2025-03-04 17:00:07 <@humaton:fedora.im> !startmeeting FESCO (2025-03-04)

2025-03-04 17:00:08 <@meetbot:fedora.im> Meeting started at 2025-03-04 17:00:07 UTC
2025-03-04 17:00:08 <@meetbot:fedora.im> The Meeting name is 'FESCO (2025-03-04)'
2025-03-04 17:00:17 <@humaton:fedora.im> !meetingname fesco
2025-03-04 17:00:17 <@humaton:fedora.im> Chairs: @conan_kudo:matrix.org, @ngompa:fedora.im, @nirik:matrix.scrye.com, @humaton:fedora.im, @zbyszek:fedora.im, @sgallagh:fedora.im, @fale:fale.io, @dcantrell:fedora.im, @decathorpe:fedora.im, @salimma:fedora.im
2025-03-04 17:00:17 <@humaton:fedora.im> !topic Init Process
2025-03-04 17:00:18 <@meetbot:fedora.im> The Meeting Name is now fesco
2025-03-04 17:00:26 <@conan_kudo:matrix.org> !hi
2025-03-04 17:00:28 <@zodbot:fedora.im> Neal Gompa (ngompa) - he / him / his
2025-03-04 17:00:55 <@decathorpe:fedora.im> !hi
2025-03-04 17:00:57 <@zodbot:fedora.im> Fabio Valentini (decathorpe) - he / him / his
2025-03-04 17:01:21 <@zbyszek:fedora.im> !hi
2025-03-04 17:01:22 <@zodbot:fedora.im> Zbigniew Jędrzejewski-Szmek (zbyszek)
2025-03-04 17:01:26 <@zbyszek:fedora.im> BTW, was the agenda sent out?
2025-03-04 17:01:44 <@humaton:fedora.im> oh
2025-03-04 17:01:49 <@nirik:matrix.scrye.com> morning
2025-03-04 17:02:43 <@humaton:fedora.im> I was out sick yesterday and forgot about it
2025-03-04 17:04:15 <@humaton:fedora.im> ok there is 5 of us
2025-03-04 17:04:23 <@humaton:fedora.im> lets start
2025-03-04 17:04:45 <@humaton:fedora.im> !topic 3347 RFC: Schedule checkpoint for landing major toolchain upgrades
2025-03-04 17:04:48 <@humaton:fedora.im> !fesco 3347
2025-03-04 17:04:49 <@zodbot:fedora.im> **fesco #3347** (https://pagure.io/fesco/issue/3347):**RFC: Schedule checkpoint for landing major toolchain upgrades**
2025-03-04 17:04:49 <@zodbot:fedora.im> 
2025-03-04 17:04:49 <@zodbot:fedora.im> ● **Assignee:** Not Assigned
2025-03-04 17:04:49 <@zodbot:fedora.im> ● **Last Updated:** 2 hours ago
2025-03-04 17:04:49 <@zodbot:fedora.im> ● **Opened:** a month ago by decathorpe
2025-03-04 17:05:02 <@zbyszek:fedora.im> Please don't forget to announce the changes that you marked as announced in the ticket today…
2025-03-04 17:05:33 <@humaton:fedora.im> oh yeah I plan to do that in the email after the meeting, that is why they are still opened
2025-03-04 17:05:57 <@humaton:fedora.im> back to the ticket
2025-03-04 17:06:12 <@humaton:fedora.im> there is a proposal there and 2 out of the 5 present voted
2025-03-04 17:06:25 <@zbyszek:fedora.im> Conan Kudo and I were +1
2025-03-04 17:07:09 <@humaton:fedora.im> 
2025-03-04 17:07:09 <@humaton:fedora.im> !topic Proposal: Major toolchain upgrades must land at least 7 days before the scheduled start of the mass rebuild.
2025-03-04 17:07:19 <@decathorpe:fedora.im> (for the record, "one week before the start of the mass rebuild" might also mean "mass rebuild is postponed if necessary" :D
2025-03-04 17:07:31 <@conan_kudo:matrix.org> that's why it was written that way, yes
2025-03-04 17:07:49 <@nirik:matrix.scrye.com> well, the proposal had more than that too tho right?
2025-03-04 17:07:51 <@zbyszek:fedora.im> Yeah, probably. We'd probably vote on this.
2025-03-04 17:08:07 <@conan_kudo:matrix.org> both solutions are valid, but it is something we'd deal with on a case-by-case basis
2025-03-04 17:08:12 <@decathorpe:fedora.im> 
2025-03-04 17:08:12 <@decathorpe:fedora.im> the proposal literally has only
2025-03-04 17:08:12 <@decathorpe:fedora.im> > Major toolchain upgrades must land at least 7 days before the scheduled start of the mass rebuild.
2025-03-04 17:08:39 <@decathorpe:fedora.im> we might want to drop "scheduled" so it's clear that both endpoints of those 7 days can move.
2025-03-04 17:09:03 <@conan_kudo:matrix.org> well shifting the start of the mass build is a fesco decision
2025-03-04 17:09:10 <@nirik:matrix.scrye.com> yeah, but... tools folks were asking about the mass prebuild. I guess if thats not part of the proposal no one cares? ;)
2025-03-04 17:09:26 <@conan_kudo:matrix.org> MPB doesn't matter from our perspective
2025-03-04 17:09:35 <@conan_kudo:matrix.org> it basically doesn't exist for policymaking purposes
2025-03-04 17:09:49 <@nirik:matrix.scrye.com> I think thats a odd way to look at it, but ok.
2025-03-04 17:10:17 <@conan_kudo:matrix.org> the problem is when things land in fedora, and mpb doesn't affect that, you can run mpb things whenever you want
2025-03-04 17:10:46 <@nirik:matrix.scrye.com> sure, but it makes the thing less surprising when it lands.
2025-03-04 17:11:02 <@nirik:matrix.scrye.com> since all those bugs/etc are already filed/found/known, or at least some of them
2025-03-04 17:11:15 <@decathorpe:fedora.im> fwiw there apparently was a "mass prebuild" for GCC 15 in December (?) but they just didn't act on the results or publish them at all 
2025-03-04 17:11:25 <@zbyszek:fedora.im> Yeah, there was.
2025-03-04 17:11:37 <@decathorpe:fedora.im> so saying a "mass prebuild" needs to happen means nothing 
2025-03-04 17:11:38 <@zbyszek:fedora.im> They didn't have enough time to analyze the results properly and report.
2025-03-04 17:12:02 <@conan_kudo:matrix.org> and I don't think we should be making policy around mpb anyway
2025-03-04 17:12:18 <@nirik:matrix.scrye.com> ok, fair, we can leave that to them...
2025-03-04 17:12:24 <@conan_kudo:matrix.org> in an ideal world, mpb runs are happening regularly against gcc and llvm snapshots like python does with every milestone for new python releases
2025-03-04 17:12:31 <@zbyszek:fedora.im> Yeah, we shouldn't do micromanagment. 
2025-03-04 17:12:31 <@conan_kudo:matrix.org> but that's not really up to us
2025-03-04 17:14:15 <@nirik:matrix.scrye.com> ok, anyhow, I guess I am +1 to the proposal. Hopefully it won't cause tools folks too much pain...
2025-03-04 17:14:58 <@decathorpe:fedora.im> ideally, less pain than for the F42 cycle :)
2025-03-04 17:15:23 <@decathorpe:fedora.im> having to fix compiler bugs under time pressure doesn't sound like fun
2025-03-04 17:15:51 <@nirik:matrix.scrye.com> sure, but having to also rush out a release because it needs to land earlier also doesn't sound fun.
2025-03-04 17:16:24 <@humaton:fedora.im> ok I count 4+1 including me and Fabio Valentini is also +1 since he is the proponent on the ticket
2025-03-04 17:16:27 <@decathorpe:fedora.im> it's **at most** like 5-6 days earlier in rawhide 
2025-03-04 17:16:40 <@decathorpe:fedora.im> (yes, I'm also +1 to myself, fwiw)
2025-03-04 17:16:48 <@humaton:fedora.im> !agreed APPROVED(5,0,0)
2025-03-04 17:17:25 <@decathorpe:fedora.im> can we !action Aoife to add this to the schedule? 😄
2025-03-04 17:17:46 <@humaton:fedora.im> hmm
2025-03-04 17:17:51 <@conan_kudo:matrix.org> !action Aoife Moloney will add the new milestone to the F42 key tasks schedule
2025-03-04 17:17:56 <@conan_kudo:matrix.org> blech
2025-03-04 17:17:59 <@conan_kudo:matrix.org> I wish we had undo
2025-03-04 17:18:01 <@humaton:fedora.im> thanks
2025-03-04 17:18:06 <@conan_kudo:matrix.org> !action Aoife Moloney will add the new milestone to the F43 key tasks schedule
2025-03-04 17:18:23 <@humaton:fedora.im> !topic 3332 Incompatible update policy violations wrt libgit2-1.9.0 updates
2025-03-04 17:18:26 <@humaton:fedora.im> !fesco 3332
2025-03-04 17:18:28 <@zodbot:fedora.im> ● **Assignee:** Not Assigned
2025-03-04 17:18:28 <@zodbot:fedora.im> **fesco #3332** (https://pagure.io/fesco/issue/3332):**Incompatible update policy violations wrt libgit2-1.9.0 updates**
2025-03-04 17:18:28 <@zodbot:fedora.im> 
2025-03-04 17:18:28 <@zodbot:fedora.im> ● **Opened:** a month ago by yselkowitz
2025-03-04 17:18:28 <@zodbot:fedora.im> ● **Last Updated:** 2 hours ago
2025-03-04 17:18:55 <@humaton:fedora.im> there is a proposal by zbyszek https://pagure.io/fesco/issue/3332#comment-957814
2025-03-04 17:19:24 <@decathorpe:fedora.im> note that this proposal wouldn't apply to libgit2 :)
2025-03-04 17:20:01 <@decathorpe:fedora.im> because dependent packages *do* need to adjust their BRs to keep building against the same libgit2 version
2025-03-04 17:20:29 <@zbyszek:fedora.im> Hmm, that depends on how the deps are declared, no?
2025-03-04 17:20:54 <@decathorpe:fedora.im> Most just do `BuildRequires: libgit2-devel` or similar unversioned `pkgconfig(...)`
2025-03-04 17:21:08 <@zbyszek:fedora.im> If they are not say '`Requires: pkgconf(libgit) < 1.9.0`' than they would be OK?
2025-03-04 17:21:12 <@conan_kudo:matrix.org> I'm more concerned if the flatpak buildroot isn't inheriting updates properly
2025-03-04 17:21:28 <@decathorpe:fedora.im> you would need to use something like `(pkgconfig(libgit2) >= 1.4 with pkgconfig(libgit2) < 1.5)`
2025-03-04 17:21:35 <@conan_kudo:matrix.org> this breakage shouldn't have happened in the first place
2025-03-04 17:21:42 <@conan_kudo:matrix.org> because by all rights the packager did the right thing
2025-03-04 17:21:54 <@decathorpe:fedora.im> they should, but basically no package does this 
2025-03-04 17:21:56 <@zbyszek:fedora.im> Yeah, so I think that's fine. If they dependent packages need to be updated to work with the new version, and they maintainers know this, then some coordination is necessary.
2025-03-04 17:23:12 <@zbyszek:fedora.im> In the proposal "Advance notification is not required when **no changes or rebuilds in dependent packages are necessary**", emphasis added.
2025-03-04 17:23:22 <@decathorpe:fedora.im> I am aware
2025-03-04 17:23:36 <@decathorpe:fedora.im> I just wanted to explicitly say that this makes this not apply to the libgit2 update in question ;)
2025-03-04 17:23:40 <@decathorpe:fedora.im> as such, I'm +1 to the proposal 
2025-03-04 17:24:02 <@conan_kudo:matrix.org> +1 as well
2025-03-04 17:24:20 <@humaton:fedora.im> but yeah what about the libgit2
2025-03-04 17:25:33 <@decathorpe:fedora.im> give a "fesco-themed virtual slap on the wrist"? not sure we can do more 
2025-03-04 17:25:43 <@nirik:matrix.scrye.com> +1 to the proposal, seems reasonable
2025-03-04 17:26:07 <@humaton:fedora.im> ok lets do the formal vote 
2025-03-04 17:26:12 <@humaton:fedora.im> !topic Proposal: Changes to packaging that maintain compatibility can happen at any time, though it is recommended to avoid bigger changes in stable releases. Advance notification is not required when no changes or rebuilds in dependent packages are necessary. When considering compatibility, maintainers must take all into account all the ways that packages are used, i.e. not just rpm builds, but also ostree and flatpaks deployments.
2025-03-04 17:26:32 <@conan_kudo:matrix.org> +1
2025-03-04 17:26:37 <@decathorpe:fedora.im> +1
2025-03-04 17:26:38 <@zbyszek:fedora.im> +1
2025-03-04 17:28:43 <@humaton:fedora.im> nirik ?
2025-03-04 17:29:07 <@humaton:fedora.im> huh I commented on wrong ticket prematurely :D
2025-03-04 17:29:43 <@nirik:matrix.scrye.com> I was +1 before, still am. ;)
2025-03-04 17:30:00 <@humaton:fedora.im> !agreed APPROVED (5,0,0)
2025-03-04 17:31:52 <@humaton:fedora.im> and that is all I had 
2025-03-04 17:32:01 <@humaton:fedora.im> there si couple of fasttrack tickets
2025-03-04 17:32:25 <@humaton:fedora.im> one has reached +7 with my vote and one is missing few https://pagure.io/fesco/issue/3369
2025-03-04 17:33:00 <@zbyszek:fedora.im> Those are for CVEs, so I think it'd be nice to vote on them quickly.
2025-03-04 17:33:08 <@zbyszek:fedora.im> We're pretty slow already.
2025-03-04 17:33:52 <@humaton:fedora.im> !topic Next week's chair
2025-03-04 17:34:05 <@humaton:fedora.im> !action Fale  will chair next meeting
2025-03-04 17:34:16 <@humaton:fedora.im> !topic Open Floor
2025-03-04 17:35:03 <@humaton:fedora.im> anything for open floor?
2025-03-04 17:35:58 <@conan_kudo:matrix.org> nothing from me
2025-03-04 17:36:58 <@decathorpe:fedora.im> nope
2025-03-04 17:37:21 <@humaton:fedora.im> thank you all and see you around!
2025-03-04 17:37:24 <@humaton:fedora.im> !endmeeting