I see two problems that this solves. 1) Gets the history of approver out of GitHub and into our log (no need to look back at the PR) 2) Gives proper attribution for the single person responsible for committing/approving the code (untangle all the "Reviewers:")
And yea, I think it can be totally automated which takes the burden off the PR author. -David A On Wed, Apr 2, 2025 at 12:47 AM Ismael Juma <m...@ismaeljuma.com> wrote: > What's the burden if it's automated? I agree we should keep it simple for > anything that has to be done manually. > > Ismael > > On Tue, Apr 1, 2025 at 5:45 PM Matthias J. Sax <mj...@apache.org> wrote: > > > Hi, > > > > I understand the problem of a long "Reviewers" line and line break, so > > breaking it down into multiple lines, one per reviewer make sense to me. > > > > I don't see much value in using all these other trailers personally, and > > all the question about "when to add whom to which trailer" kinda backs > > up that it might be unnecessary complex and confusing > > > > - committer/contributor who left a comment > > - committer/contributor who hit "approve" > > - committer who merged > > - committer/contribute who co-authored > > > > Of course we can track this all very fine grained, but what it the use > > of this information? In the end, if I really want to know who "approved" > > a PR, I can go back to github and look it up. And I cannot remember that > > I would have needed to look this up often. > > > > Would be great to get some clarification why such fine grained tracking > > is useful. It puts a lot of burden on committers to fill this in > > correctly (even if we try to automate it to the best extend possible). > > > > What actual problem are we solving? > > > > > > -Matthias > > > > On 3/28/25 7:25 AM, Chia-Ping Tsai wrote: > > > hi David > > > > > >> David Arthur <mum...@gmail.com> 於 2025年3月27日 晚上9:50 寫道： > > >> > > >> Commented-by: left any comment on the PR (any contributor) > > >> Reviewed-by: did a full review on the PR (any contributor) > > >> Approved-by: committer(s) who approved the PR > > > > > > If a committer leaves the comment without approve, he/she should be > > included only by “Commented-by”, right? > > > > > > For another, is there a tool which can collect the name for > > “commented-by” automatically? > > > > > > Best, > > > Chia-Ping > > > > > > > > > > > -- David Arthur
