On Tue, Oct 30, 2018 at 11:41 PM Junio C Hamano <gits...@pobox.com> wrote: > > Stefan Beller <sbel...@google.com> writes: > > > I also picked up the patch for pending semantic patches, as the > > first patch, can I have your sign off, please? > > I find this step quite lacking. > > What was posted would have been perfectly fine as a "how about doing > it this way" weatherbaloon patch, but as a part of real series, it > needs to explain to the developers what the distinctions between two > classes are, and who is to use the cocci patch at what point in the > development cycle, etc., in an accompanying documentation update.
if only we had documentation [as found via "git grep coccicheck"] that I could update ... I'd totally do that. I guess this asks for documentation to begin with, now? > So can we have both sign-off and write-up (perhaps cut&paste from > the original e-mail discussion)? I'll see where to put the docs; I assumed commit messages are enough. 63f0a758a0 (add coccicheck make target, 2016-09-15) is what I found nice. > > t/helper/test-repository: celebrate independence from the_repository > > It seems that this topic is full of commits with overly long title. yep. > > git range-diff origin/sb/more-repo-in-api... > > [...] # rebased to origin/master > > I offhand do not recall what happened during these 100+ pacthes. > DId we acquire something significantly new that would have > conflicted with this new round of the topic? I do not mind at all > seeing that a series gets adjusted to the updated codebase, but I do > dislike seeing it done without explanation---an unexplained rebase > to a new base is a wasted opportunity to learn what areas of the > codebase are so hot that multiple and separate topics would want to > touch them. >From the point of view of these large scale refactorings, all of the code is hot, e.g. the patch that was present in the RFC "apply all semantic patches" would clash with nearly any topic. As I do not carry that patch any more, I do not recall any conflicts that needed to be resolved. Thanks.
