2021.01.20 20:41 "[Tiff] MergeRequest for discussion code style", by Thomas Bernard
Based on unprofessional comments in the merge request from a subset of the libtiff members, I am bowing out of trying to contribute to libtiff. I responded to a couple of the comments for those that do care just for completeness. My goal was to figure out what changes might be acceptable. I do not appreciate getting comments that start with "WTF", calling me OCD, etc. This statement is key:
changing everything just for the pleasure is just making losing other people time.
I thought there was some great information discovery so far.
The discussion you induced is indeed interesting.
I'm sorry if I and some other libtiff contributors offended you. As you may know most (all ?) of us contribute to libtiff on our free time, which is sparse.
libtiff is a huge project with a 33 years history and is used almost everywhere.
I think it was well explained how focusing on code style changes may take other peoples time: just imagine the work of someone maintaining a patch set against libtiff!
Fixing some bugs was my first motivation to contribute to libtiff, so how do you think I may feel when reviewing changes that introduce a bug without me seeing a clear benefit to the changes?
Apart from adding new features, I think work that will bring the more value to the project is to improve test code coverage, improve CI, etc. Because of its popularity, libtiff is a target of choice for exploits, we may want to avoid them...
Just an example: does some one has some time to include the fuzzer to gitlab-ci as Roger suggested? https://gitlab.com/libtiff/libtiff/-/merge_requests/151
You may also notice there are quite a lot of open bug https://gitlab.com/libtiff/libtiff/-/issues I would guess it is because time to fix them is difficult to find.