Thread: Commitfest 2022-03 Starts Now
The final commitfest of this release begins now. Whereas most commitfests are about getting feedback to authors so they can advance the patch -- this one is about actually committing patches to wrap up the release. Please when reviewing patches try to push yourself to make the difficult call about whether a patch will actually be feasible to commit in this commitfest. If not be up front about it and say so because we need to focus energy in this commitfest on patches that are committable in Postgres 15. There are a lot of great patches in the commitfest that would really be great to get committed! -- greg
So it's 8 days into the commitfest. So far 3 patches have been committed: * parse/analyze API refactoring by Peter Eisentraut * FUNCAPI tuplestore helper function by Melanie Plagemen committed by Michael Paquier * Typo in pgbench messages by Kawamoto Masay committed by Tatsuo Ishii (There was also a flurry of five commits from Tom on Feb 28 including one that needed to be backpatched subsequently) In addition 5 patches have been moved moved to a future commitfest, withdrawn, or rejected. So that has removed 8 patches from the commitfest. There are now 189 "Needs Review" and 24 "Ready for Committer" patches. Of those 24 "Ready for Committer" patches only 3 actually have committers assigned andonly 6 have had any emails since the beginning of the commitfest. Is there anything I can do to get committers assigned to these patches? Should I do a round-robin style assignment for any of these?
On 3/9/22 13:38, Greg Stark wrote: > > Is there anything I can do to get committers assigned to these > patches? Should I do a round-robin style assignment for any of these? I don't think this is a good idea. Committers pick the patches they are going to commit. What prefer to do is bump any committers that have been involved in a patch thread to see if they are willing to commit it. Regards, -David
On Wed, 9 Mar 2022 at 15:44, David Steele <david@pgmasters.net> wrote: > > On 3/9/22 13:38, Greg Stark wrote: > Should I do a round-robin style assignment for any of these? > > I don't think this is a good idea. Committers pick the patches they are > going to commit. > > What prefer to do is bump any committers that have been involved in a > patch thread to see if they are willing to commit it. Well yes, I suppose that's probably what I had in mind despite calling it that. But I've been skimming the set of "Ready for Committer" patches and I'm a bit down on them. Many of them seem to mostly have gotten feedback from committers already and the type of feedback that leads me to think it's ready for commit. I suppose people mark patches "Ready for Committer" when the level of feedback they require is more in depth or more design feedback that they think requires a committer even if it's not ready for commit. So I'm going to go through the patches and ask the committers who have already commented if they think the patch is on track to be committed this release or should be pushed to the next commitfest. -- greg
On Wed, 9 Mar 2022 at 16:46, Greg Stark <stark@mit.edu> wrote: > Many of them seem to mostly have gotten > feedback from committers already and the type of feedback that leads > me to think it's ready for commit. Er. I meant *not* the type of feedback that leads me to think it's ready for commit. I mostly see patches with early design feedback or even feedback questioning whether they're good ideas at all. -- greg