16:01:02 #startmeeting fpc 16:01:02 Meeting started Thu Sep 2 16:01:02 2021 UTC. 16:01:02 This meeting is logged and archived in a public location. 16:01:02 The chair is geppetto. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:02 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:01:02 The meeting name has been set to 'fpc' 16:01:02 #meetingname fpc 16:01:02 #topic Roll Call 16:01:02 The meeting name has been set to 'fpc' 16:01:11 * limburgher here 16:01:21 hi 16:01:27 #chair mhroncok 16:01:27 Current chairs: geppetto mhroncok 16:01:38 #chair limburgher 16:01:38 Current chairs: geppetto limburgher mhroncok 16:02:01 * nirik has something for open floor. 16:02:18 Hey, folks. 16:02:21 nirik: cool, we'll probably get there quickly … but might not have 5 16:02:24 #chair tibbs 16:02:24 Current chairs: geppetto limburgher mhroncok tibbs 16:02:26 Another bad day for me but I'm trying to be abount. 16:02:43 Still no air conditioning in my office.... 16:03:55 It's a beautiful day here … had been humid, but the great flood fixed that … along with turning NJ and NYC into modern atlantis 16:04:39 We also still have human rights for women, up here. 16:04:55 So like a two for one thing. 16:05:02 Noice 16:09:53 open the floor for nirik? 16:10:13 #topic Schedule 16:10:16 #info https://lists.fedoraproject.org/archives/list/packaging@lists.fedoraproject.org/message/EG5NTEC2MHBLZC5A7KFJZUC3DFQVYZNF/ 16:10:28 No sense in not talking about stuff if a few of us are here. 16:10:29 So … nothing changed this week 16:11:07 One thing is that meetbot website was down after the meeting last week, and still isn't back … so no emails about wtf happened. 16:11:15 #topic Open Floor 16:11:21 nirik: Go for it 16:11:27 Ah, someone was asking about that. 16:11:37 I think I at least have my log if it's needed. 16:11:39 I have 2 related(ish) things... first: I suggested on the devel list that we ask FPC to take over maintaining redhat-rpm-macros. Is that something you all would want to do? or do you think it's logistically not doable, etc? and second: In the last EPEL meeting we talked about some new macros that were added in Fedora, and we wondered if we could get new macros as they are added in Fedora also added in EPEL. Ideally at the same time. 16:12:08 tibbs: You want to do that? 16:12:09 geppetto: it's due to the python2->python3 conversion of zobot. The meetings are saved, just not in the place you expect them. we are working on fixing it. 16:12:39 nirik: Yeh, I figured it had to be on someone's radar … so wasn't bothering people about it 16:12:51 I mean, someone does need to "take responsibility" for that package. 16:13:08 It should at least be close to us, if not ours. 16:13:08 But I would really like to have at least Panu and Florian involved as well. 16:13:11 tibbs: Well you have, by far, the most experience with it 16:13:14 But everybody is busy. 16:13:14 THAT^ 16:13:21 hey, sorry, I'm late, I spent the last 7 hours hiking and just got back home :( 16:13:25 they want to be involved... but in a 'NO,don't do that' capacity 16:13:25 #chair decathorpe 16:13:25 Current chairs: decathorpe geppetto limburgher mhroncok tibbs 16:13:56 they don't have the cycles or interest to merge PR's about other things they don't care about. 16:14:04 nirik: tryna steal our jobs 16:14:07 anyhow, you all could comment on that thread 16:14:44 I haven't had time to read much Fedora email since the semester started. 16:14:46 nirik: I might have some cycles to get that package into shape, get rid of cruft and delegate specific things to their own maintainers 16:14:51 last weeks meeting: https://meetbot-raw.fedoraproject.org/fedora-meeting-1/2021/fedora-meeting-1.2021-08-26-16.00.html 16:14:58 nirik: will do. I think splitting off more of the domain-specific stuff into separately maintained packages would make this easier 16:15:23 however, there are things there that nobody wants to touch but removng them would have consequences 16:15:32 1) nim's autoautoautomacros 16:15:39 2) kernel srpm macros 16:15:41 If we are going to take care of that package, maybe we can fix the somewhat bizarre versioning setup while we're at it. 16:15:58 yeah, splitting might be ok, but note that adding things to the koji srpm-buildroot would possibly need doing, and more packages there the slower it is 16:16:18 Shouldn't be any need to change koji. 16:16:30 Since you can just add dependencies to pull in other macro packages. 16:16:37 people have been busy removing stuff from the buildroot, let's get busy putting more stuff back in ;) 16:16:40 for that matter 'redhat-rpm-config' is a pretty weird name these days. ;) 'fedora-rpm-macros' ? 16:16:48 It's what we do now though it might have gotten slightly out of controol. 16:17:12 maintaining the macros in a proper project outside of dist-git would be nice ... 16:17:17 And yeah, I proposed renaming and even actually had a fedora-specific package but there was outcry and flaming and someone actually had it removed from the distro. 16:17:33 Several Red Hat employees did not like it at all. 16:17:50 * nirik shrugs. Perhaps they are all gone now? 16:18:03 I have old memories of people inside RH saying they didn't want to ship something called fedora* in RHEL 16:18:06 They aren't. 16:18:07 good that this is a community governed projct so as long as those emloyes were not legal, we should be good 16:18:37 anyway, this requires time an effort and I am not sure it's worth doing 16:18:56 first things first, somebody shoul feel responsible for triaging redhat-rpm-config bugzillas and PRs 16:18:59 This is a really weird issue that probably isn't worth getting into, but it makes me wary about messing with that package and I haven't been much involved with it since that happened. 16:19:43 https://src.fedoraproject.org/rpms/fedora-rpm-macros/blob/rawhide/f/macros.fedora this thing still exists BTW 16:19:59 decathorpe: mhroncok: tibbs: Did either of you three want to take ownership and start doing things? 16:20:05 I'm happy to run interference with anyone who objects. :) 16:20:27 Yeah, we can do it. Given my current status, though, it can't just be me. 16:20:46 I am tempted, but I am afraid of burning out 16:21:10 thats another reason I thought "FPC" would be good... because it would spread the issues to a group instead of just one or two people... but dunno 16:21:29 Yeh, that's fine … I just don't want FPC to be the owner, as I fear we'll ignore it hoping someone else deals with stuff 16:21:47 One thing I'd like to make sure I understand, though: when we inevitably merge something that breaks the distro, how do we get it backed out? 16:22:02 decathorpe: That kind of leaves you ;) 16:22:14 Oh, I'll volunteer. 16:22:26 But it's not going to get better if it's just me. Especially now. 16:22:36 And obviously by being an owner doesn't mean you need to do everything … just be the person who knows wtf is going on and needs to be done 16:22:48 nirik: What's required to untag a package, since if we break this one we can't even build a replacement. 16:22:53 ? 16:22:54 tibbs: it's quite easy nowdays -- 1) request a side tag 2) tag older redhat-rpm-config in it 3) build the fixed redhat-rpm-config 4) do an update from the side tag 16:23:45 but, untagging is also an option, that requires people with special powers 16:24:07 Any special perms needed for that? I'd want good instructions for that checked into the package repo alongside the spec just in case. 16:24:21 Because I know it will happen and I know it will be 3AM. 16:24:22 tibbs: no, just packager permisions 16:24:42 technically, you need a provenpackager or maintainer of redhat-rpm-config to commit the fix 16:25:08 any releng / koji admin can untag 16:25:24 or if it becomes a problem we can grant that to some group 16:25:47 I'm just truing to cover my rear here since I know it will be me that screws it up. 16:26:07 Anyway, sounds like there's a solution. The rest is just mechanics. 16:26:18 yeah, there is 16:26:29 it happens with other stuff... much more rarely these days th 16:26:44 also can add gating test(s) to not tag it in if it breaks too badly 16:27:30 gating tests would be cool 16:27:42 Yeah, it would be nice to find some way to test that any revision can still be used to build the package itself but I don't know how that would be possible. 16:27:52 Since it's kind of a recursive operation. 16:28:25 tibbs: it is possible 16:28:39 Not something we have to talk about now, of course, unless we're out of things to talk about. 16:28:41 the CI test can rebuild redhat-rpm-config 16:28:50 the epel thing 16:30:45 BTW, I noticed that Florian is the primary maintainer of redhat-rpm-config and there are six total maintainers. 16:31:15 isn't that the same problem … getting the r-r-m updated in EPEL when it's updatd in Fedora? 16:31:20 Oh, and there are two Florians. 16:31:38 Well EPEL has its own macro package that we have historically used for "backports". 16:31:45 nirik: Or do the macros need to be added somewhere else? 16:31:47 there is no redhta-rpm-macros in epel 16:32:05 it needs somebody to maintain epel-rpm-macros and care 16:32:16 Ahh, because the real one will be in RHEL … and isn't updated 16:32:25 That used to be me as well but now it's more "community maintained". 16:32:26 my experience is that I need to hunt down people to merge my Python macro backport PRs there 16:32:33 * geppetto goes to find the snafu stamp 16:32:41 we have a epel-rpm-macros 16:32:56 and yes, it would be good to maintain that with the fedora side. ;) 16:33:04 Yes, epel-rpm-macros is the place for working around redhat-rpm-macros. 16:33:08 could the CPE epel team maintain that? 16:33:14 Now that there is no EPEL6, things get at least a bit easier. 16:33:31 I think we would want to try to bring major changes to both where possible. 16:33:33 mhroncok: So Carl will have an opinion on that 16:33:40 But he isn't here this week. 16:34:31 I don't think this is a particularly huge issue if the default thing is "try to bring changes to Fedora's redhat-rpm-macros into epel-rpm-macros if possible". 16:34:58 somebody still needs to validate and test the changes 16:35:03 do they work with ancient rpm? 16:35:17 I know of maintain it. 16:35:19 do they require other backports that were neglected before? 16:35:22 but I am low on time... 16:36:09 Who isn't low on time? 16:36:27 Anyway, I guess a bit of bug triage would do both packages good, and then just have a list of what needs doing. 16:36:38 this sort of stuff is hard and needs soembody who has time, is dedicated and understands packaging deeply 16:37:08 if we don't have that kind of person in epel, handing it over to fpc won't help 16:37:20 Well making it better just needs a little effort. Fixing all of the problems might be outside of what all of us together are able to put in. 16:38:44 I think the first step is getting away from having those packages be seen as untouchable. 16:38:50 I'd be happy with better if we can't make perfect. :) 16:39:05 I don't want to volunteer Carl … but he might have some time, or be able to poke some people who do. 16:39:40 So I'm happy to just say tibbs is the owner of r-r-m, but can't do much in the short term without help 16:39:53 I wish I could get paid to do this kind of thing instead of volunteering what time I have. 16:39:55 And we can leave the epel side for a week when Calr is here. 16:40:26 nirik: That good as a start? 16:41:33 before we end I'd also like to finally approve https://pagure.io/packaging-committee/pull-request/1066 16:42:04 well, I mean, I am happy to keep merging PR's and doing what I can, but co-maintainers welcome. I don't know dumping it on one person is good. 16:42:53 one thing I think does ot help is having too many maintainers listed 16:42:55 mhroncok: I thought we had. I will double check and +1 the ticket. 16:43:05 e.g. redhat-rpm-config has 6 maintainers listed 16:43:21 Here's the diff: https://pagure.io/fork/tstellar/packaging-committee/c/97b3d2a6e4726e131967aa4282795fde53f9a29a 16:43:30 Looks like mhroncok and decathorpe are already +1 16:43:44 I voted, yes? 16:44:30 Apparently not, till now. 16:44:55 Actually … I feel like we all voted at the meeting last week 16:45:02 And it got approved 16:45:25 Yeah, I remember that. There was one caveat which was noted and has already been addressed. 16:45:44 mhroncok: Yeh, this is ready to go … blame meetbot ;) 16:46:03 alright 16:46:17 merged 16:46:22 Ok, is there anything else for the last few minutes? 16:46:46 nirik: I'll try to figure out how much time I can dedicate to redhat-rpm-config 16:46:49 There is still a lot of triage to be done. Maybe one day I'll find the time. 16:46:56 I'll start by writing that "can build self" test 16:46:57 I'm basically in the same boat. 16:47:11 feel free to ping me anytime if I can help with it any 16:48:03 sorry, I was away for a bit. somebody wants me to become rpm macros maintainer? I didn't read the whole backlog ;) 16:48:49 decathorpe: yes, you will maintain all the macros, is that ok? 16:48:59 oh, sure 16:49:03 :) 16:49:06 decathorpe: yep, you own it all now … for epel too 16:49:14 especially for epel 16:49:20 I'll slip in some crypto miners while I'm at it >:-) 16:49:37 limburgher: off topic: https://pagure.io/releng/fedora-scm-requests/issue/36654 should be fixed now 16:49:53 brp-mine-bitcoin 16:50:27 mhroncok: looks good, thanks! 16:51:02 in all seriousness, once the basics are decided, we can set up a proper project, and I'd be fine with being one of the maintainers 16:51:14 I wish there was some way to get Panu looped in to redhat-rpm-config maintenance without overburdening him. 16:51:41 There was one recent issue where he objected to something in more than one place but it got pushed out everywhere anywah. 16:52:48 the memory limit macros? yeah ... 16:53:05 or the %cc macro that broke everything? :) 16:53:08 And I don't know if there was any discussion outside of that tickets that I can read. 16:53:31 The %cc thing was just a lack of testing and got resolved favorably in more than one place. 16:54:14 But I don't know what to do about the memory limit thing. 16:56:54 that's just sad 16:57:12 especially since Panu actually provided feedback and than the next thing got merged without him anyway 16:57:14 :( 16:57:29 well, we'll iterate somehow 16:57:36 * geppetto nods 16:57:41 bettar than gatekeeping the package and never accepting any chages 16:57:47 Will be better with tibbs not doing those merges ;) 16:57:53 Yeah, I don't know what to do about it. And there's this approved feature on it, too. 16:58:22 Certainly ignoring his objections is not the way to get Panu onboard with any improved maintenance of the package. 17:00:21 #link https://src.fedoraproject.org/rpms/redhat-rpm-config/pull-request/144 17:00:27 True, but it didn't sound like he had the time to do much work 17:00:29 [WIP] CI: Rebuild redhat-rpm-config to assert we did not break the world 17:00:41 works on my machine, but let's see what does the CI say 17:00:48 Anyway … unless there's anything we really need to talk about … we hit the one hour mark, and I should let you all go 17:01:07 mhroncok++ 17:01:07 geppetto: Karma for churchyard changed to 11 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 17:01:35 See you next week 17:01:42 #endmeeting