2025-07-28 18:00:23 <@alexsaezm:fedora.im> !startmeeting Go SIG meeting
2025-07-28 18:00:24 <@meetbot:fedora.im> Meeting started at 2025-07-28 18:00:23 UTC
2025-07-28 18:00:24 <@meetbot:fedora.im> The Meeting name is 'Go SIG meeting'
2025-07-28 18:00:28 <@alexsaezm:fedora.im> !topic Roll Call
2025-07-28 18:00:30 <@mikelolasagasti:matrix.org> !hi
2025-07-28 18:00:32 <@zodbot:fedora.im> Mikel Olasagasti Uranga (mikelo2)
2025-07-28 18:00:41 <@alexsaezm:fedora.im> Hi! as always we'll wait a feew minutes
2025-07-28 18:01:29 <@buckaroogeek:fedora.im> !hi
2025-07-28 18:01:30 <@zodbot:fedora.im> Brad Smith (buckaroogeek)
2025-07-28 18:06:00 <@alexsaezm:fedora.im> ok, looks like we are just mikel and brad 🙂
2025-07-28 18:06:13 <@buckaroogeek:fedora.im> And alexsaezm 
2025-07-28 18:06:21 <@alexsaezm:fedora.im> I always forget about that guy
2025-07-28 18:06:37 <@mikelolasagasti:matrix.org> 🤣
2025-07-28 18:07:07 <@alexsaezm:fedora.im> Last week I was on PTO and away from the laptop, but it was a little bit chaotic. there is a bunch of things that happened, is there anything relevant we should cover first?
2025-07-28 18:07:19 <@alexsaezm:fedora.im> Last week I was on PTO and away from the laptop, but it was a bit chaotic. There are a bunch of things that happened. Is there anything relevant we should cover first?
2025-07-28 18:07:36 <@alexsaezm:fedora.im> Or we should start with the elephant in the room: the Go 1.25 mass rebuild 😄
2025-07-28 18:08:03 <@buckaroogeek:fedora.im> gotmax23: did an excellent response and follow through
2025-07-28 18:08:24 <@mikelolasagasti:matrix.org> yep, he deserves a big cookie
2025-07-28 18:08:40 <@alexsaezm:fedora.im> we need to ask for a big cookie reward
2025-07-28 18:09:08 <@alexsaezm:fedora.im> !topic Go 1.25 Mass Rebuild Status - DWARF5 Issue Resolution
2025-07-28 18:09:46 <@mikelolasagasti:matrix.org> the good thing is that most of the packages affected in the f43 mass rebuild built just fine once the hotfix landed
2025-07-28 18:10:35 <@alexsaezm:fedora.im> If I'm not missing anything, everything is fine, but the solution should be a temporary fix.
2025-07-28 18:10:54 <@alexsaezm:fedora.im> thanks for rebuilding the packages mikel 
2025-07-28 18:11:00 <@mikelolasagasti:matrix.org> source-only packages just were OK (unless not affected by orphans or other stuff) 
2025-07-28 18:11:18 <@mikelolasagasti:matrix.org> status can be checked in https://kojipkgs.fedoraproject.org/mass-rebuild/f43-need-rebuild.html and https://kojipkgs.fedoraproject.org/mass-rebuild/f43-failures.html 
2025-07-28 18:11:27 <@mikelolasagasti:matrix.org> !link https://kojipkgs.fedoraproject.org/mass-rebuild/f43-failures.html
2025-07-28 18:11:35 <@mikelolasagasti:matrix.org> !link https://kojipkgs.fedoraproject.org/mass-rebuild/f43-need-rebuild.html
2025-07-28 18:13:00 <@mikelolasagasti:matrix.org> I think we should work and encourage others to move to Go Vendor Tools to be able to retire most of those packages
2025-07-28 18:13:22 <@mikelolasagasti:matrix.org> and I think until we retire current orphans we won't be able to detect new leave packages 
2025-07-28 18:14:06 <@mikelolasagasti:matrix.org> alexsaezm: you should start orphaning you leave packages https://gitlab.com/fedora/sigs/go/package-data/-/blob/main/leaves-by-maintainer?ref_type=heads#L379
2025-07-28 18:14:45 <@alexsaezm:fedora.im> Got it, I'll start with that list this week.
2025-07-28 18:15:30 <@mikelolasagasti:matrix.org> you can use https://raw.githubusercontent.com/decathorpe/miscripts/refs/heads/master/orphan.py to automate the process
2025-07-28 18:15:55 <@mikelolasagasti:matrix.org> it doesn't handle the errors, but you can either skip the failing packages or fix the script ;)
2025-07-28 18:16:25 <@buckaroogeek:fedora.im> mikel: will the listed packages that failed need a manual rebuild or will they just work with the dwarf5 fix in place?
2025-07-28 18:16:55 <@mikelolasagasti:matrix.org> most of the packages are failing because other problems (failing tests, missing dependencies)
2025-07-28 18:17:00 <@buckaroogeek:fedora.im> I did a local build of containernetworking-plugins for rawhide and it works fine
2025-07-28 18:18:29 <@mikelolasagasti:matrix.org> then I understand you need to do manual rebuild 
2025-07-28 18:19:14 <@buckaroogeek:fedora.im> Ok. thanks. Bump the release and submit normally i suspect
2025-07-28 18:21:05 <@gotmax:matrix.org> If the last rebuild failed, you don't need to bump the release again
2025-07-28 18:21:16 <@gotmax:matrix.org> You can just submit the build
2025-07-28 18:21:56 <@mikelolasagasti:matrix.org> correct, the rebuilds are done with a commit that will bump the release
2025-07-28 18:22:11 <@mikelolasagasti:matrix.org> correct, the mass rebuilds are done with a commit that will bump the release
2025-07-28 18:23:40 <@gotmax:matrix.org> I plan to rebuild everything after Go 1.25.0 was released to make sure everything that was built with the beta got built with the GA release and also to catch anything that didn't get rebuilt at all because of the debuginfo failures
2025-07-28 18:24:12 <@gotmax:matrix.org> I filed an issue in the Go SIG tracker
2025-07-28 18:24:38 <@alexsaezm:fedora.im> I want to try a tool that a coworker has made. Do you mind if I try it first ?
2025-07-28 18:25:06 <@gotmax:matrix.org> Is that for Copr rebuilds or actual rebuilds in Koji?
2025-07-28 18:26:01 <@alexsaezm:fedora.im> In COPR. I didn't read the documentation but I don't think it supports Koji yet
2025-07-28 18:26:24 <@gotmax:matrix.org> https://gitlab.com/fedora/packager-tools/mass-prebuild ?
2025-07-28 18:26:35 <@alexsaezm:fedora.im> yeah that one
2025-07-28 18:27:41 <@gotmax:matrix.org> That is only for test rebuilds in Copr but I don't think it handles real mass rebuilds where the releases need to be bumped in dist-git and Koji builds have to be submitted
2025-07-28 18:28:03 <@gotmax:matrix.org> But I have heard good things about it for the Copr rebuild usecase
2025-07-28 18:28:36 <@alexsaezm:fedora.im> I can bother fberat with questions 🙂
2025-07-28 18:29:09 <@alexsaezm:fedora.im> but if it's only for COPR it won't solve the whole problem
2025-07-28 18:29:35 <@mikelolasagasti:matrix.org> +1 for me if the testing of this tool doesn't block the mass-rebuild with 1.25 for too long
2025-07-28 18:30:07 <@gotmax:matrix.org> !info https://pagure.io/GoSIG/go-sig/issue/62
2025-07-28 18:31:02 <@gotmax:matrix.org> MPB is built entirely around Copr so I don't think it'll work here
2025-07-28 18:31:40 <@gotmax:matrix.org> But we should definitely try it out prior to Go 1.26 to be aware of any changes there before we push it to rawhide
2025-07-28 18:32:30 <@gotmax:matrix.org> AIUI that's mass *pre*-build's purpose. It's for use *before* a mass rebuild in Rawhide
2025-07-28 18:32:34 <@alexsaezm:fedora.im> yeah, I want to try it to be sure that I can run it from time to time before finding out ourselves in something like merging to rawhide and waiting for the mass rebuild
2025-07-28 18:32:58 <@alexsaezm:fedora.im> my intention if mass prebuild works for us, is to run against go tip
2025-07-28 18:33:26 <@alexsaezm:fedora.im> Waiting for rc to be on Rawhide is getting more and more errorprone.
2025-07-28 18:33:38 <@alexsaezm:fedora.im> with very little time to handle issues
2025-07-28 18:34:23 <@alexsaezm:fedora.im> but working on go tip is quite easy, and we can even report new issues to upstream
2025-07-28 18:35:43 <@mikelolasagasti:matrix.org> gotmax23: when do you plan to retire the golang-* orphans? 
2025-07-28 18:36:13 <@gotmax:matrix.org> I did want to talk about that actually
2025-07-28 18:36:53 <@gotmax:matrix.org> I have updated the orphan packages report (https://a.gtmx.me/orphans/orphans.txt) now that FESCo has approved our exception
2025-07-28 18:37:30 <@gotmax:matrix.org> Take a look especially that the new SPECIAL NOTE paragraph at the beginning
2025-07-28 18:37:35 <@gotmax:matrix.org> Does that look okay to y'all?
2025-07-28 18:38:43 <@gotmax:matrix.org> packages will resume after a six-week grace period."
2025-07-28 18:38:43 <@gotmax:matrix.org> I need to clarify the last sentence, as it should be "Once the new guidelines are fully implemented, mass retirements of **orphaned** Golang
2025-07-28 18:40:06 <@mikelolasagasti:matrix.org> based on that list in 4-5 weeks many packages will be retired
2025-07-28 18:40:18 <@gotmax:matrix.org> packages will resume after a six-week grace period."
2025-07-28 18:40:18 <@gotmax:matrix.org> I need to clarify the last sentence, as it should be "Once the new guidelines are fully implemented, automatic retirements of 
2025-07-28 18:40:18 <@gotmax:matrix.org> orphaned Golang
2025-07-28 18:40:23 <@mikelolasagasti:matrix.org> ~end of september 
2025-07-28 18:41:45 <@gotmax:matrix.org> Assuming they're not in the exemptions list, yes
2025-07-28 18:43:18 <@mikelolasagasti:matrix.org> ok
2025-07-28 18:43:34 <@mikelolasagasti:matrix.org> most of the package will be part of F43 then, as branching happens on Aug-12
2025-07-28 18:48:21 <@alexsaezm:fedora.im> Do you think we should remove them before that happens?
2025-07-28 18:48:50 <@mikelolasagasti:matrix.org> no, just noting that
2025-07-28 18:49:39 <@mikelolasagasti:matrix.org> the only problem I see is that if a package has simple dependency chain of 4 levels, it would take ~24 weeks to be retired
2025-07-28 18:50:53 <@gotmax:matrix.org> If anyone wants to add recursive logic to the go_leaves script, they're welcome to :D
2025-07-28 18:51:20 <@gotmax:matrix.org> But once we have the new guidelines, we can at some point start mass retirements of leaves like we did a few releases ago
2025-07-28 18:51:34 <@gotmax:matrix.org> Which will speed up the process
2025-07-28 18:51:42 <@mikelolasagasti:matrix.org> agree
2025-07-28 18:53:05 <@mikelolasagasti:matrix.org> any other topics?
2025-07-28 18:53:36 <@alexsaezm:fedora.im> no, I think we covered everything relevant
2025-07-28 18:53:53 <@alexsaezm:fedora.im> done btw
2025-07-28 18:54:25 <@gotmax:matrix.org> I've been making slow but steady progress on the new guidelines
2025-07-28 18:54:48 <@gotmax:matrix.org> Nothing much to report yet, but feel free to look at the guidelines fork repository on Gitlab
2025-07-28 18:55:22 <@mikelolasagasti:matrix.org> I've been porting packages to gvt, updating specs and adding some packages to EPEL. I think today it's the first time in 2 months I sit in my desk, so I consider it a good amount of work from my side 😅
2025-07-28 18:55:56 <@mikelolasagasti:matrix.org> I'll check the guidelines when I've a few minutes 
2025-07-28 18:55:56 <@zodbot:fedora.im> gotmax23 has already given cookies to mikelo2 during the F42 timeframe
2025-07-28 18:56:25 <@mikelolasagasti:matrix.org> gotmax23: are you OK about adding packages to EPEL with the current status of gvt? It's working fine, but I'm not sure if you like the idea of having many packages using it
2025-07-28 18:56:26 <@gotmax:matrix.org> I've been meaning to make an announcement on the EPEL SIG mailing list or Matrix room about GVT being added
2025-07-28 18:56:47 <@gotmax:matrix.org> But I'm glad that you've started adopting it and that it's working :)
2025-07-28 18:56:59 <@mikelolasagasti:matrix.org> I love how simple it is
2025-07-28 18:57:05 <@gotmax:matrix.org> That's a good question :)
2025-07-28 18:57:08 <@mikelolasagasti:matrix.org> *to use it
2025-07-28 18:58:23 <@gotmax:matrix.org> At this point it's still technically in beta/preview but I guess starting to use it in a limited number of packages is okay, and I don't see any major changes at this point
2025-07-28 18:58:58 <@mikelolasagasti:matrix.org> and the other question I've is if still using `%gocheck` is a good idea for packages like `rclone` where there are a bunch of `-d path` to skip some tests until we figure out a clean way to do it with `%gotest`
2025-07-28 19:00:09 <@mikelolasagasti:matrix.org> if all beta/preview software was this good the world would be a better place 😅
2025-07-28 19:00:34 <@gotmax:matrix.org> I had some minor issues with the original CLI interface that I had at one point planned to refactor in https://gitlab.com/fedora/sigs/go/go-vendor-tools/-/issues/63 but I decided against it so after that I was more comfortable branching it for EPEL
2025-07-28 19:01:12 <@gotmax:matrix.org> I have to go so can we table the gocheck discussion for next meeting?
2025-07-28 19:01:29 <@alexsaezm:fedora.im> absolutely 
2025-07-28 19:01:31 <@gotmax:matrix.org> I have some thoughts about that also :D
2025-07-28 19:01:41 <@gotmax:matrix.org> Don't think we should start that up now
2025-07-28 19:01:45 <@mikelolasagasti:matrix.org> happy to read them 
2025-07-28 19:01:51 <@mikelolasagasti:matrix.org> no worries
2025-07-28 19:02:16 <@mikelolasagasti:matrix.org> I'll use whatever works now and change later 
2025-07-28 19:02:19 <@mikelolasagasti:matrix.org> I'll use whatever works now and change later if required
2025-07-28 19:02:51 <@gotmax:matrix.org> SGTM
2025-07-28 19:03:09 <@mikelolasagasti:matrix.org> I should leave also, but before leaving just a ping on prometheus renaming review https://bugzilla.redhat.com/show_bug.cgi?id=2383787 I would like to adhere my packages as soon as possible to the correct namign
2025-07-28 19:04:06 <@mikelolasagasti:matrix.org> leaving, have a nice rest of the week :)
2025-07-28 19:04:21 <@alexsaezm:fedora.im> thanks everyone for attending 🙂
2025-07-28 19:05:19 <@alexsaezm:fedora.im> !action Discuss gocheck status in the next meeting
2025-07-28 19:05:30 <@alexsaezm:fedora.im> !endmeeting