On Fri, Jul 15, 2022 at 03:23:29PM +0400, Pavel Borisov wrote: > On Fri, 15 Jul 2022 at 15:20, Pavel Borisov <pashkin.elfe@gmail.com> wrote: > > Hi, hackers! > > v42 stopped applying, so we rebased it to v43. Attached is a GitHub link, > > but I see Cfbot hasn't become green. Apparently, it hasn't seen changes in > > GitHub link relative to v42 attached as a patch. > > Github link is as follows: > https://github.com/ziva777/postgres/tree/64xid-cf > Maybe this will enable CFbot to see it.
My suggestion was a bit ambiguous, sorry for the confusion.
The "Git link" is just an annotation - cfbot doesn't try to do anything with it[0] (but cirrusci will run the same checks under your github account). What I meant was that it doesn't seem imporant to send rebases multiple times per week if there's been no other changes. Anyone is still able to review the patch. It's possible to build it by applying the patch to a checkout of the postgres tree at the time the patch was sent. Or by using the git link. Or by checking out cfbot's branch here, from the last time it *did* apply. https://github.com/postgresql-cfbot/postgresql/tree/commitfest/38/3594
[0] (I think cfbot *could* try to read the git link, and apply patches that it finds there, but that's an idea that hasn't been proposed or discussed. It'd need to know which branch to use, and it'd need to know when to use the git link and when to use the most-recent email attachments).
Hi, Justin!
I can agree with you that sending rebased patches too often can be a little annoying. On the other hand, otherwise, it's just red in Cfbot. I suppose it's much easier and more comfortable to review the patches that at least apply cleanly and pass all tests. So if Cfbot is red for a long time I feel we need to send a rebased patchset anyway.
I'll try to not doing this too often but frankly, I don't see a better alternative at the moment.
Anyway, big thanks for your advice and attention to this thread!