14:00:13 <dcantrell> #startmeeting FESCO (2021-03-24) 14:00:13 <zodbot> Meeting started Wed Mar 24 14:00:13 2021 UTC. 14:00:13 <zodbot> This meeting is logged and archived in a public location. 14:00:13 <zodbot> The chair is dcantrell. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:13 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:00:13 <zodbot> The meeting name has been set to 'fesco_(2021-03-24)' 14:00:13 <dcantrell> #meetingname fesco 14:00:14 <dcantrell> #chair nirik, ignatenkobrain, decathorpe, zbyszek, sgallagh, mhroncok, dcantrell, cverna, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor 14:00:14 <zodbot> The meeting name has been set to 'fesco' 14:00:14 <zodbot> Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku cverna dcantrell decathorpe ignatenkobrain mhroncok nirik sgallagh zbyszek 14:00:16 <dcantrell> #topic init process 14:00:26 <mhroncok> .hello churchyard 14:00:27 <zodbot> mhroncok: churchyard 'Miro Hrončok' <mhroncok@redhat.com> 14:00:28 <Eighth_Doctor> .hello ngompa 14:00:30 <zodbot> Eighth_Doctor: ngompa 'Neal Gompa' <ngompa13@gmail.com> 14:00:32 <dcantrell> .hello2 14:00:33 <zodbot> dcantrell: dcantrell 'David Cantrell' <dcantrell@redhat.com> 14:00:40 <dcantrell> hi, Miro. hi, Neal. 14:01:30 <zbyszek> .hello2 14:01:31 <zodbot> zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' <zbyszek@in.waw.pl> 14:02:03 <dcantrell> hi, Zbigniew 14:02:14 <mhroncok> hi, David 14:02:21 <nirik> morning 14:02:38 <dcantrell> hi 14:02:49 <dcantrell> and that's 5? 14:02:51 * dcantrell counts 14:02:53 <zbyszek> Hi David, hi everyone. 14:03:36 <dcantrell> I forget, do we need 5 or 6 to proceed? 14:03:38 * decathorpe 's ghost hovers around 14:03:40 <zbyszek> 5 14:03:50 <dcantrell> ok, and there's 6 anyway. very good 14:04:09 <zbyszek> We should also talk about the meeting time. 14:04:10 <dcantrell> Alright everyone, I think this will be short today. Hopefully. Thanks for joining. let's get going 14:04:21 <dcantrell> zbyszek: let's do that after the two topics here 14:04:36 <dcantrell> #2582 F35 Change: rpmautospec - removing release and changelog fields from spec files 14:04:39 <dcantrell> .fesco 2582 14:04:40 <zodbot> dcantrell: Issue #2582: F35 Change: rpmautospec - removing release and changelog fields from spec files - fesco - Pagure.io - https://pagure.io/fesco/issue/2582 14:04:42 <sgallagh> Oops, I have this on my calendar in an hour. 14:04:44 <sgallagh> I'm here now 14:04:44 * cverna waves 14:04:51 <dcantrell> welcome, sgallagh and cverna 14:05:15 <cverna> sgallagh: same here, calendar playing tricks on me :) 14:06:08 * Eighth_Doctor sighs 14:06:19 <dcantrell> for 2582, we've got feedback from zbyszek and Eighth_Doctor. not sure if you two have anything else to add 14:06:51 <zbyszek> I'm typing in the ticket now. 14:07:22 <dcantrell> ok, I will wait. I'm also ok with this one just continuing in the ticket. zbyszek was the final vote I believe 14:07:34 <zbyszek> I think I understand where Eighth_Doctor is coming from, but I don't share his view. 14:07:55 <sgallagh> I haven't had time to review the latest chatter. 14:08:18 <Eighth_Doctor> I used to have zbyszek's position, then I spent a couple of years in openSUSE where it doesn't work that way 14:08:35 <Eighth_Doctor> and it's really freeing and makes it easier to do bigger things 14:08:47 <mhroncok> I haven't had time to read thie either 14:09:02 <Eighth_Doctor> that part of the openSUSE workflow is what I want to bring into Fedora, because I think it'll make things much better 14:09:27 <dcantrell> ok, I think then maybe we should all continue this one outside the meeting and followup in the ticket 14:09:34 <Eighth_Doctor> honestly, my opinion is that if that isn't part of the goal, I'm inclined to reject the whole proposal 14:09:37 <dcantrell> it feels like we're converging on a decision but not everyone is caught up right now 14:09:53 <Eighth_Doctor> because tbh, it's not worth it for changelog merge conflicts 14:10:02 <zbyszek> Eighth_Doctor: but the thing you want to make easier is something that our guidelines say shoudln't be done. So if you want to convince me that it is a good thing, please start with the motiviation. 14:10:52 <Eighth_Doctor> zbyszek: well, I don't think soname bumps should happen in stable branches without exceptional need, I'm referring to rawhide 14:10:54 <Eighth_Doctor> remember that everything uses bodhi now 14:10:55 <mhroncok> note that if we make it work ina way that requires the commit, we can later revisit this 14:11:06 <Eighth_Doctor> there's no reasonable way to make this work with commits 14:11:10 <mhroncok> but if we build multiple builds from single commit from the start, it will be problematic to revisit that 14:11:34 <Eighth_Doctor> because if we have to make commits, someone has to have write privileges 14:11:57 <mhroncok> that is not necessarily a bad thing 14:11:59 <Eighth_Doctor> and also the commit sequence doesn't imply the order of merging either 14:12:07 <zbyszek> Without a commit, there is no commit message, and this is an anti-feature for me. 14:12:08 <Eighth_Doctor> yes it is 14:12:15 <mhroncok> zbyszek: +1 14:12:41 <Eighth_Doctor> if you want commits and commit messages for every build 14:12:48 <Eighth_Doctor> there's no point in doing rpmautospec at all 14:13:01 <mhroncok> Eighth_Doctor: not true 14:13:07 <mhroncok> the point is: no more bloody conflicts 14:13:17 <Eighth_Doctor> wrong 14:13:18 <zbyszek> Eighth_Doctor: that's an overstatement. There is a lot of value in having that, even if the particular feature you care about, which we don't currently use in Fedora, is not supported. 14:13:21 <mhroncok> you can have more PRs open at the same time and marge them in any order 14:13:41 <mhroncok> Eighth_Doctor: wrong how? 14:13:47 <zbyszek> Exactly. And trivially cherry-pick patches between branches. 14:13:56 <mhroncok> or between distros 14:14:25 <Eighth_Doctor> how many people actually have that problem? 14:14:32 <Eighth_Doctor> because I don't 14:14:35 <zbyszek> "that problem"? 14:14:46 <mhroncok> Eighth_Doctor: I don't know how many 14:14:48 <Eighth_Doctor> I can probably bet that closer to 80% of packagers maintain their packages with a single effective branch 14:15:11 <mhroncok> Eighth_Doctor: e.g. entire pytohn-maint has this problem 14:15:35 <mhroncok> Eighth_Doctor: packagers that maintain their packages with a single effective branch don't need a solution 14:15:38 <Eighth_Doctor> yeah, because your packages are different per branch, which is rare 14:15:46 <mhroncok> packagers that acually follow the update policy need the solution 14:16:32 <zbyszek> Eighth_Doctor: the packages that are multi-branch are often very important packages with a lot of commits. So optimizing for them, even if it's a small chunk of the overall package count, is still worth it. 14:16:50 <Eighth_Doctor> uh huh, but the existence of changelog files that must exist in the repo means that you're still screwed 14:16:52 <mhroncok> let's have a look at https://fedoraproject.org/wiki/Changes/rpmautospec#Benefit_to_Fedora 14:17:00 <mhroncok> Eighth_Doctor: no 14:17:05 <Eighth_Doctor> and if you have to edit the changelog, you're back to square one again 14:17:22 <Eighth_Doctor> because we don't have a git-data based way to do it 14:17:32 <mhroncok> the benefit does not tell: "anyone can rebuild anything anytime without write permissions" 14:18:06 <Eighth_Doctor> because rpmautospec itself doesn't do it, that would be a follow-on automation change 14:18:07 <mhroncok> Eighth_Doctor: as long as you keep the changelog-edit commits separate, you don't have a problem 14:18:27 <Eighth_Doctor> yeah, that's not going to happen 14:18:37 <Eighth_Doctor> nobody does that now with changelogs in spec files 14:18:42 <Eighth_Doctor> you could do that if you wanted, but you don't 14:18:54 <mhroncok> Eighth_Doctor: for editing the old ones? sure I do that 14:19:08 <Eighth_Doctor> dude, you have to regenerate that file every time you need to edit it 14:19:08 <mhroncok> Eighth_Doctor: why would I edit an old changelog entyr together with other changes in the spec file? 14:19:13 <dcantrell> the point is valid, that would be more self-discipline than anything 14:19:16 <Eighth_Doctor> otherwise you lose changelog entries 14:19:33 <mhroncok> yes, how often do we need to do this? 14:19:37 <Eighth_Doctor> the changelog part of the system is the poorly designed part 14:19:56 <Eighth_Doctor> mhroncok: how often do people actually maintain packages with differing branches? 14:20:04 <mhroncok> oftne 14:20:06 <mhroncok> *often 14:20:10 <mhroncok> big, complicated stuff 14:20:10 <Eighth_Doctor> the point is, if you want to solve the 1% cases, you really need to justify it 14:20:17 <Eighth_Doctor> and you really aren't selling me on it 14:20:22 <mhroncok> Eighth_Doctor: rigth back at you, really :D 14:20:32 <dcantrell> ok guys, let's keep it civil :) 14:20:37 <mhroncok> Eighth_Doctor: both of us wants to solve "their" problem 14:20:49 <mhroncok> dcantrell: sorry, I thought the :D makes it less horrible 14:20:49 <zbyszek> Eighth_Doctor: but the packages with differing branches are the ones we are solving here. Mono-branch packages don't need this and don't care. 14:21:08 <dcantrell> mhroncok: it did, but I just wanted to restate the obvious. tone is difficult in irc 14:21:22 <Eighth_Doctor> zbyszek: literally the reason for %autorelease is for non-commit rebuilds 14:21:29 <mhroncok> Eighth_Doctor: apologies, no harm intended 14:21:36 <mhroncok> Eighth_Doctor: no 14:21:38 <Eighth_Doctor> yes 14:21:39 <zbyszek> Eighth_Doctor: not for me 14:21:50 <mhroncok> Eighth_Doctor: the reason for %autorelease is for non-conflicitng commits 14:21:53 <Eighth_Doctor> well, that's why myself, pingou, and nils designed it 14:22:17 <Eighth_Doctor> it came out of the conversation at flock in 2019 14:22:19 <mhroncok> clearly said in https://fedoraproject.org/wiki/Changes/rpmautospec#Benefit_to_Fedora 14:22:41 <mhroncok> The Release and %changelog fields are the two most conflicting fields in RPM spec files. They impact most pull requests if they involve updating the package or if the package is updated/rebuilt while pull-request are being reviewed. 14:23:11 <Eighth_Doctor> it's clearly omitted is what it is :( 14:23:40 <mhroncok> let me wrap it up 14:23:46 <Eighth_Doctor> it was even mentioned in the nest talk last year :( 14:23:47 <mhroncok> Eighth_Doctor wants one thing 14:23:56 <mhroncok> me, decathorpe, zbyszek want another thing 14:24:18 <mhroncok> Eighth_Doctor says our thing is not worth doing alone and let's not even approve it if the other thing is not possible 14:24:25 <zbyszek> Also, there's the following issue: this scheme is opt-in. If we were to accomodate what Eighth_Doctor wants, we would need to require all packages to follow the new scheme. This is another reason not to do that. 14:24:54 <mhroncok> we say let's have our thing first and if the other thing really is needed, it can be discussed and added later 14:25:20 <Eighth_Doctor> well if I know all of you are going to vote it down, why I would I bother to write the other half to make it possible? 14:25:37 <zbyszek> Yeah, I think we should consider Neal's approach separately. Maybe it's a good thing in some scenarios, but we shouldn't mix it. 14:25:58 <dcantrell> heads up, we're almost at 15 minutes on this topic. 14:26:00 <mhroncok> Eighth_Doctor: not necessarily if you show how it's better than not having it 14:26:10 <dcantrell> (2 minutes away by my stopwatch) 14:26:19 <mhroncok> Eighth_Doctor: I've even voted +1 for the proposal as is, becasue I disagree about this, but I don't want to block progress 14:26:46 * Eighth_Doctor shrugs 14:26:53 <mhroncok> but let's not block the benefits of %autorel(ease) and %autochangelog just because one other thing would not be included 14:27:05 <zbyszek> Yeah, I'm ready to move on too. I'll vote in the ticket if that's ok. 14:27:42 <Eighth_Doctor> if we're going to do this, my condition is that we need a more uniform workflow for changelogs 14:27:52 <Eighth_Doctor> the current proposed workflow is awful 14:28:27 <dcantrell> ok, that's 15 minutes. we'll continue in the ticket 14:28:45 <mhroncok> if everybody keeps projecting their own little disagreements here and there, we will never have this 14:29:08 * mhroncok stops 14:29:24 <dcantrell> #action everyone read and add additional comments to the ticket; hopefully voting takes place there 14:29:30 <dcantrell> #2589 Disallow "Block Un-Signed commits" hook in dist-git 14:29:30 <dcantrell> .fesco 2589 14:29:31 <zodbot> dcantrell: Issue #2589: Disallow "Block Un-Signed commits" hook in dist-git - fesco - Pagure.io - https://pagure.io/fesco/issue/2589 14:29:57 <dcantrell> this one I left on the agenda, but I'm prepared to vote +1 now 14:29:58 <Eighth_Doctor> I think we pretty much agree that we should do this 14:30:18 <dcantrell> thanks for the feedback after my comment 14:30:22 <Eighth_Doctor> aside from dcantrell everyone voted +1 14:30:25 <mhroncok> dcantrell: I appreciate the "we should have asked trousers mainatainers why they do this in the first place" thought 14:30:53 <dcantrell> mhroncok: thanks. for future items, I would like to make sure we do try to get input from the example package in question 14:31:03 <mhroncok> dcantrell: but OTOH I cannot even imagine a reason they'd give that I'd consider and say: yes, sure, let's keep this 14:31:22 <dcantrell> mhroncok: true, but I never want to assume :) 14:31:38 <dcantrell> so, uh vote here or in the ticket? 14:31:54 <mhroncok> we need to vote here 14:31:58 <zbyszek> dcantrell: more people voted in the ticket. If you vote there too, it'll be +9. 14:31:59 <mhroncok> ticket cannot pass if it has -1 14:32:22 <dcantrell> alright 14:32:31 <mhroncok> dcantrell: unless you want to withdraw that vote 14:32:42 <zbyszek> mhroncok: it's already been superseded 14:32:43 <dcantrell> I just voted +1 in the comment thread 14:32:51 <dcantrell> or did while talking here 14:32:54 <mhroncok> oh 14:33:27 <mhroncok> let's call it approved +9 here, so we can announce and close? 14:33:40 <zbyszek> ack 14:33:44 <nirik> also should file a infra/pagure ticket to do it. ;) 14:33:51 <dcantrell> #agreed (+9,0,-0) 14:34:04 <mhroncok> #action mhroncok to open infra ticket to do it 14:34:37 <dcantrell> #topic Next week's chair 14:35:06 <zbyszek> I can. 14:35:13 <dcantrell> thanks, zbyszek 14:35:24 <dcantrell> #action zbyszek will chair next meeting 14:35:30 <dcantrell> #topic Open Floor 14:35:44 <zbyszek> Meeting time? 14:35:54 <mhroncok> impossible 14:35:58 <dcantrell> heh 14:36:05 <nirik> Just a reminder that the new account system is rolling out. Thanks in advance for everyone's patience as we deploy. 14:36:35 <dcantrell> nirik: thank you. I look forward to the improvements 14:36:49 <dcantrell> zbyszek: did you want to change the meeting time or discuss changing it? 14:37:06 <zbyszek> We had a poll, the results are here https://whenisgood.net/raqpfhx/results/wjshcxp 14:37:21 <zbyszek> Myself, I'm OK with the current time. 14:37:51 <Eighth_Doctor> I'm pretty much stuck at the current time :/ 14:38:14 <Eighth_Doctor> nirik, dcantrell: also a packaged version of the new account system has landed in Fedora for other people to use for their own deployments :) 14:38:15 <dcantrell> looking at the results, I don't really see a better option. everything appears to be not ideal 14:38:30 <dcantrell> Eighth_Doctor: nice 14:39:10 <mhroncok> we won't have decathorpe and ignatenkobrain 14:39:18 <mhroncok> ^ if we keep it now 14:39:18 <cverna> current time is Ok for me, not ideal but OK :) 14:39:29 <mhroncok> and if we move it, we loose different people 14:39:51 <mhroncok> "ideal" time for this meeting is like an unicorn 14:40:11 <mhroncok> all we need is a time that is mostly OK 14:40:29 <dcantrell> reading the results on this url is difficult 14:40:38 <dcantrell> trying to find the best options, that is 14:40:52 <zbyszek> dcantrell: look for the one with least red dots 14:41:09 <zbyszek> also you can click on a person to exclude their vote. 14:41:19 <zbyszek> People are sorted by the number of slots they entered. 14:41:36 <zbyszek> And if you're logged in, the times are in your own timezone. 14:42:02 <sgallagh> Not so 14:42:10 <sgallagh> I'm logged in and they're definitely in UTC 14:42:20 <sgallagh> They're in my TZ when I edit my replies, though 14:42:21 <zbyszek> Yeah, UTC. Sorry, I was confused by DST. 14:42:21 <dcantrell> wow, so Eighth_Doctor has the most restrictive time options here in the results 14:42:46 <Eighth_Doctor> yep 14:43:05 <Eighth_Doctor> over the past half year, so many meetings have built up in my calendar that I have no wiggle room 14:43:07 <sgallagh> I can probably make that Monday slot where Neal and I are the only exceptions 14:43:20 <Eighth_Doctor> I had to drop out of EPEL meetings because it moved from Friday to Tuesday 14:43:48 <sgallagh> I won't be able to chair, but it's a conflict with a meeting I can probably dual-attend. 14:43:48 <Eighth_Doctor> I can maybe show up on the Tuesday slot where it overlaps with the OKD WG meeting 14:43:58 <Eighth_Doctor> no chairing, like sgallagh said 14:44:08 <Eighth_Doctor> but I could attend since it's an IRC meeting and the other is a video meeting 14:44:15 <Eighth_Doctor> but being double-booked sucks 14:44:28 <dcantrell> ok, so moving to Mon at 16:00 UTC could possibly work for all of us? 14:44:30 <zbyszek> Dunno, if nobody is too unhappy with the current time, let's just keep it. 14:44:43 <Eighth_Doctor> I literally can't do Monday 14:44:44 <Eighth_Doctor> ever 14:45:05 <Eighth_Doctor> that's my $dayjob weekly scrum meeting 14:45:08 <Eighth_Doctor> 11am US/ET on Monday and Friday 14:45:11 <zbyszek> dcantrell: 16 UTC monday is pretty bad for me. 14:45:15 <dcantrell> I'm beginning to think you are only in meetings :) 14:45:16 <nirik> Eighth_Doctor: it moved to wed afternoon. ;) 14:45:33 <sgallagh> zbyszek: This time doesn't permit decathorpe or ignatenkobrain to attend. 14:45:47 <sgallagh> The Monday time would "only" exclude Eighth_Doctor 14:45:55 <Eighth_Doctor> yep 14:46:03 <sgallagh> From a purely utilitarian standpoint, that would seem to be preferable. (Sorry, Neal) 14:46:04 <mhroncok> we could have both 14:46:10 <sgallagh> .fire mhroncok 14:46:10 <zodbot> adamw fires mhroncok 14:46:16 <Eighth_Doctor> if you want to move it to Monday, I'll just never be able to show up, which if you're okay with that, sure 14:46:18 <mhroncok> sgallagh: I mean, every toher week 14:46:25 <mhroncok> *other 14:46:34 <zbyszek> sgallagh: Hmm, I see Can't make it: David Cantrell, Neal Gompa, Stephen Gallagher, cverna, nirik, zbyszek 14:46:37 <sgallagh> That seems complicated, but it's an option. 14:46:43 <Eighth_Doctor> the Tuesday one I can at least sanely double book because work doesn't really care (as it's a community timeslot anyway) 14:46:47 <mhroncok> Eighth_Doctor would never meet decathorpe, but otherwise... 14:47:05 <sgallagh> zbyszek: ? 14:47:34 <sgallagh> zbyszek: Monday at 1500 UTC on the results page is free for everyone but Neal and I 14:47:49 <sgallagh> But I can probably make it work for me (albeit double-booked) 14:48:12 <zbyszek> sgallagh: Oh, 15:00, not 16:00 UTC. 14:48:25 <dcantrell> oh, I goofed on that. 15, not 16 14:48:40 <Eighth_Doctor> 5pm UTC on Tuesday excludes only cverna 14:48:49 <Eighth_Doctor> as I can double-book for that 14:49:12 <zbyszek> I'd vote for Tuesday over Monday. 14:49:28 <Eighth_Doctor> that's 1pm EDT, and that collides with OKD WG, but this is an IRC meeting, so I can be half-present on weeks when that's going 14:51:01 * nirik is fine with 17UTC tue 14:51:07 <dcantrell> so is there a proposed meeting time slot move? 14:51:19 * dcantrell looks at zbyszek, Eighth_Doctor, sgallagh 14:51:46 <zbyszek> dcantrell: 17UTC tue 14:52:32 <dcantrell> zbyszek: alright, I'm fine with that. everyone else? 14:52:35 <sgallagh> Looks like that only blocks cverna on WhenIsGood. 14:52:38 <sgallagh> I'm fine with it. 14:53:13 <nirik> cverna: any chance you could move something and make that? 14:55:19 <dcantrell> cverna: ? 14:55:21 <cverna> let me look 14:55:58 <cverna> 17UTC is kind of late for me, but don't block on me 14:56:15 <cverna> so that's fine for me 14:56:20 <dcantrell> thanks 14:56:31 <dcantrell> zbyszek: can I action you to update the meeting time in the docs? 14:56:38 <zbyszek> yes 14:57:00 <dcantrell> #action zbyszek will update the FESCo meeting times to 17UTC on Tuesdays 14:57:03 <dcantrell> thanks 14:57:07 <dcantrell> anyone else for open floor? 14:57:22 <mhroncok> yes 14:57:30 <mhroncok> zbyszek will chair next meeting 14:57:34 <mhroncok> even on Tuesday? 14:57:51 <zbyszek> Reluctantly 14:58:07 <Eighth_Doctor> are we still in f-m-2? 14:58:35 <dcantrell> Eighth_Doctor: we should verify, zbyszek can you check that as well? 14:59:05 <zbyszek> dcantrell: I'm not sure if I know how. 14:59:22 <dcantrell> when stuck, we can always #action bcotton 14:59:42 <dcantrell> #action bcotton will help us verify irc meeting room we can use on Tuesdays at 17UTC 14:59:50 <zbyszek> https://apps.fedoraproject.org/calendar/location/fedora-meeting-2%40irc.freenode.net/ 14:59:57 <zbyszek> 17tue seems free 15:00:03 <dcantrell> fantastic 15:00:13 <dcantrell> alright, anyone else for open floor? 15:00:16 <mhroncok> so is #fedora-meeting 15:00:21 <mhroncok> let's move there? 15:00:27 <Eighth_Doctor> sure, I suppose 15:00:33 <dcantrell> I'm fine with either room 15:00:34 <Eighth_Doctor> might as well 15:00:51 <zbyszek> Yeah, let's move back to #fedora-meeting. 15:00:57 <mhroncok> 90 more people idle in #fedora-meeting :) 15:01:01 <mhroncok> (ATM) 15:01:14 <mhroncok> that's it from me 15:01:24 <mhroncok> thanks all 15:01:30 <dcantrell> thanks, anyone else? 15:01:32 <mhroncok> Eighth_Doctor: nic arguing with you :) 15:01:36 <mhroncok> *nice 15:01:49 <mhroncok> let's repeat that some day :P 15:01:53 * Eighth_Doctor sighs 15:02:25 <dcantrell> thanks all! 15:02:28 <dcantrell> #endmeeting