Datasets:
Does opting out for one repository ensure that all forks of this repository are also excluded?
I am collecting the information on the file attribution to repositories in The Stack, and I've noticed a large amount of forks. Some of them collecting even more files than the original repository.
For example, here you can see that the original tensorflow repository has only 4 files attributed to it:
I suspect the issue here is in the deduplication process. If deduplication does not differentiate between a fork and an original repository, the files from the original repository can be removed in the process.
Another example, no files from https://github.com/sindresorhus/awesome are present in The Stack, however there are a bunch of forks of this repository:
For reference, here is the code to collect these stats: https://gist.github.com/ViktorooReps/29170f5cdc6b0038b4dbf24affe1aa32
This naturally poses the question:
If no care is given to differentiate forks from the original repository, does it mean that the opt-out request does not extend to the repository forks?