14:00:24 #startmeeting FESCo (2020-07-29) 14:00:24 Meeting started Wed Jul 29 14:00:24 2020 UTC. 14:00:24 This meeting is logged and archived in a public location. 14:00:24 The chair is decathorpe. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:24 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:00:24 The meeting name has been set to 'fesco_(2020-07-29)' 14:00:29 #meetingname fesco 14:00:29 The meeting name has been set to 'fesco' 14:00:31 .hello2 14:00:33 zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' 14:00:35 .hello2 14:00:36 dcantrell: dcantrell 'David Cantrell' 14:00:37 #chair nirik, ignatenkobrain, decathorpe, zbyszek, sgallagh, mhroncok, dcantrell, cverna, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor 14:00:37 Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku cverna dcantrell decathorpe ignatenkobrain mhroncok nirik sgallagh zbyszek 14:00:52 #topic Init Process 14:00:53 .hello2 14:00:54 sgallagh: sgallagh 'Stephen Gallagher' 14:01:10 morning. 14:01:14 .hello kevin 14:01:15 nirik: kevin 'Kevin Fenzi' 14:02:33 .hello2 14:02:34 bcotton: bcotton 'Ben Cotton' 14:03:42 .hello ngompa 14:03:43 King_InuYasha: ngompa 'Neal Gompa' 14:03:53 hey folks, sorry I'm a bit late 14:04:11 no problem. with you we have quorum :) 14:04:30 I'll give the others another minute, then we'll start 14:05:34 #link https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/NHIKO64AWRW72LEOD3LMJ5E6FJT62CED/ Schedule 14:06:01 #topic #2446 What to do with the 33 packagers/watchers who do not have a valid bugzilla account? 14:06:09 .fesco 2446 14:06:10 decathorpe: Issue #2446: What to do with the 33 packagers/watchers who do not have a valid bugzilla account? - fesco - Pagure.io - https://pagure.io/fesco/issue/2446 14:06:59 I commented with a three-pronged proposal: https://pagure.io/fesco/issue/2446#comment-665491 14:07:02 * zbyszek voted for proposal 3 in the ticket 14:07:26 the voting has started though I'm a bit confused what things were voted on, so I brought it up today 14:08:11 just read through it, looks reasonable. added my +1 14:08:18 zbyszek: I don't think these are contradictory proposals, I think they're one proposal with individual steps 14:08:34 I'm in favor of 2 and 3... but not sure I like 1. 14:08:54 sgallagh: yes. that was my intent. I guess I was not specific enough 14:08:56 Yeah, same as nirik here 14:09:11 what's the downside to step 1? 14:09:15 or part 1? 14:09:24 1 means if someone shows back up in a few weeks they have to ask to be re-added to all the stuff they want to maintain again. 14:09:29 Exactly. 14:09:38 hello, sorry for being late 14:09:47 hi Miro! 14:09:49 I would actually call that a benefit to part 1 14:09:56 I think keeping them as co-maintainers doesn't hurt and makes it easier to pick up again. 14:09:59 hi Miro 14:10:10 It means they'll only ask to be re-added to the ones they actually care about 14:10:23 or just decide it's too much red tape and go do something else? 14:10:25 And we'll have a more realistic view of what packages are adequately maintained 14:10:27 nirik: yes, that is a "potential downside" though if a packager is really voted to be non-responsive then I'm not sure keeping them as co-maintainer is actually a good idea, but that's a separate discussion 14:10:53 some people go non-responsive too simply because they haven't had luck handing packages off to new maintainers 14:11:02 I'm good with going with all three proposals together +1 14:11:09 and I've said as much in the ticket 14:11:11 we don't have a good exit path for people who no longer want to maintain packages or who do but lack time 14:11:13 just now 14:11:35 well, you can orphan your packages anytime? 14:11:48 nirik: If that's too much red tape, chances are that's a good indicator of how much care they'd put into their packages. 14:12:16 orphaning your package requires admitting that you can't maintain them anymore, which is a hard thing to do for a lot of people :-) 14:12:39 Question: Do we want to discuss this detail now, or should I modify my "proposal 1" slightly to remove the controversial bit? 14:12:41 nirik: true, but in some cases if packages are still used but no one has claimed it off the orphaned packages list, then the last person to touch it still gets contacted and there is a certain amount of anxiety that comes along with that 14:13:20 I don't think thats fair... say you added a stack of 30 packages that are needed for the 1 you care about... you get busy and find out that you were marked unresposive... you might come back and work on the one you care about, but you might not want to go through talking to all the new maintainerrs to be re-added to all those deps... 14:13:30 decathorpe: personally, I don't think we need to do that. the non-responsive policy is understood and I think this is an ok action when someone is non-responsive 14:13:49 decathorpe: how would you change it? 14:14:14 when somebody is nonrepsonsive, they should be removed from all their packages 14:14:27 mhroncok: +1 14:14:34 it reduces the amount of stale obsolete info 14:14:39 nirik: that's valid, but are you really maintaining the 30 packages you added or just the one? do you have the time to do all that work? 14:14:47 look, this thing has 14 co-mainters... but all of them are gone 14:14:49 mhroncok: +1 14:14:59 it's hard to say in hypotheticals. ;) 14:15:02 we have packages like that today, and it's pretty hard to get that reassigned 14:15:03 Yeah, I guess that's a good point. 14:15:10 e.g. some of the sip stack packages 14:15:25 Modified Proposal 1: When a packager is declared non-responsive, also remove them from all packages as watchers, in addition to orphaning packages where they have "main admin" / "owner" role. 14:15:30 This skips the "controversial part" for now, which we can discuss separately (since it's a modification of the Non-responsive maintainer policy) 14:15:32 +1 14:15:39 +1 14:15:51 I think if you added pkgs but were unresponsive for some and got removed, it's an indicator that help is needed. Not that you are failing. And as a project we should step in and clear out their ownership and spread the work out. 14:15:58 so... did we agree to remove people from retired packages? 14:16:20 nirik: we did 14:16:25 nirik: yes 14:16:25 decathorpe: +1 to modified proposal 14:16:41 decathorpe: +1 to this and please let me know what was the removed controvrsial bit 14:17:06 mhroncok: don't remove nonresponsive packagers as comaintainers for now, move to different discussion 14:17:45 I'm +1 to the revised Proposal 1, but I'd honestly *prefer* the original one. 14:18:25 oh, so it removes them as watchers but not commit/admin/ticket? 14:18:35 I change my vote to 0 14:18:45 sorry, I've missed that bit 14:18:57 my vote is to wipe them from ACLs and watchers 14:19:07 I have to say that mhroncok's arguments convinced me, and I'd vote for the original proposal too. 14:19:21 alright, so I withdraw modified Proposal 1 14:19:30 I'd prefer the original step 1, but am +1 on both the original or the revised one 14:19:56 I'm not in favor, but won't stand in everyone's way if you all feel thats the way to go... :) 14:20:14 Ok, can I have votes for the original P1+2+3 (as stated in the ticket)? IRC / my internet connection seems to be wonky today and messages seem to arrive out of order :( 14:20:32 +1 to original proposal in its entirety 14:20:40 +1 to original proposal in its entirety 14:20:52 -1 for 1, +1 2 and 3. ;) 14:21:01 +1/+1/+1 14:21:14 +1/+1/+1 14:21:14 decathorpe: I think it's using TCP, so we must be just talking out of order (as usual) :) 14:21:42 +1/+1/+1 14:21:51 Or decathorpe is using IRCCloud via Internet Explorer :-P 14:22:19 oh that may be it! :D 14:22:28 thanks for the votes 14:22:54 P1: (+5, 0, -1) ---- P2+P3: (+6, 0, -0) 14:23:30 #agree APPROVED (P1: (+5, 0, -1); P2+P3: (+6, 0, -0)) 14:23:52 pingou will be happy 14:24:43 #topic Next week's chair 14:24:45 nirik: one thing about "I'm out and loosing all my packages", we're only considering accounts that are invalid in bugzilla 14:24:52 I'm happy to chair next week 14:25:07 * pingou shouldn't be in two meetings at ones, so may have missed something 14:25:12 pingou: right now yes, but not moving forward. ;) 14:25:18 nirik: ah :( 14:25:32 moving forward it means when someone is unresponsive they are removed from everything. 14:25:47 I'd prefer we only remove people that somehow causes problem 14:25:49 but thats ok, we will see if it causes any issues. 14:25:59 well, we can always keep this in mind when voting non non-responsive maintainer tickets. 14:26:06 if it causes a problem, we can revisit this 14:26:07 people MIA that don't cause problem, well... don't cause problems 14:26:11 nirik: yeah, it *feels* a bit scary, but we should try if it works better than the current rules. 14:26:19 regarding next-weeks-chair: I think last week mhroncok volunteered for next week? 14:26:30 oh, that's certainly fine :) 14:26:34 pingou: I disagree 14:26:35 decathorpe: I honestly don'T recall 14:26:55 pingou: they cause outdated information 14:26:55 People MIA still cause problems because it provides a false sense of the level of maintenance a package has. 14:26:58 sgallagh: eager to hear more, but I'll ping you out of the meeting 14:27:02 Maybe it'll encourage poeple to shed packages more quickly, without going through the whole non-resp maintainer procedure. 14:27:10 ok, then the question is: any volunteeres? 14:27:23 I can chair next week 14:27:30 zbyszek: I hope that this will be the outcome, yes 14:27:42 King_InuYasha: thanks 14:27:51 #action ngompa will chair next week's meeting 14:28:05 #topic Open Floor 14:28:12 o/ 14:28:15 no idea if you already talked baout this, but to do about all the stalled changes ticktes? 14:28:22 *what to do 14:28:33 We should treat each one individually. 14:28:54 zbyszek: when is the point that we say: rejected 14:29:27 there are official deadlines for submission, and code ready. do we only kill the chnages when they ar enot approved until beta freeze? or sooner 14:29:58 depends on the scope of the proposal, i'd say 14:29:58 mhroncok: I think that for simple changes we can reject them at the last meeting before beta freeze. 14:30:14 fine by me 14:30:22 I just wanted to hear other opinions 14:30:34 * zbyszek looks for the F33 schedule 14:31:48 Tue 2020-08-25 is change checkpoint and beta freeze 14:32:16 ack 14:32:46 So the fesco meeting 3 weeks from now is the last one before that. 14:33:08 do we have any non-trivial changes left to examine? 14:34:15 PostgreSQL 13 doesn't look too good. 14:34:43 the forge macros is still waiting 14:35:05 panovotn updated the PostgreSQL 13 proposal https://fedoraproject.org/w/index.php?title=Changes%2FPostgreSQL_13&type=revision&diff=583493&oldid=582401 14:35:18 2409 (compiler policy change) is also stalled 14:35:25 nirik: there was some "discussion" on #fedora-devel aboutt them 14:36:04 yeah, but note there are 2 changes... the forge macros (33) and the autorelease stuff on top of that (34) 14:36:23 bcotton: nice 14:36:27 bcotton: oh, good, this does look better. though there's still no info as to who will actually do the necessary rebuilds. 14:36:28 at least the pgsql 13 one, the changes to the proposal look fine to me 14:36:39 I am assuming the change author is going to do the rebuilds 14:37:22 they're not a provenpackager. 14:37:35 umm well then 14:37:39 that's a problem 14:37:42 Also "revert changes" is not very ... verbose. 14:37:52 someone is going to need to help this person :( 14:39:07 so ... volunteer from FESCo to help with Change proposal and come up with more detailed rebuild plan and contingency mechanism? 14:39:45 I've tried the last time several times, they didn't communicate :( 14:39:47 anyone want to help shepherd this through? 14:40:03 * King_InuYasha is personally a little swamped 14:40:11 * sgallagh is in the same boat 14:40:30 I'd *love* to have pgsql 13 in F33, but I don't think I have bandwidth to help out 14:40:52 * nirik is idle and has lots of free time to... naaa... kidding. I'm busy too. :) 14:41:19 I guess I can try if no one else wants to... or we could ask them to line up provenpackagers on the list? 14:41:32 nirik: that'd be great :) 14:41:55 it'd be nice to see panovotn say something on-list at all, though :( 14:42:24 that is the root of the problem 14:42:42 yeah :( communication could be better (s/better/existent) 14:42:49 * mhroncok would be happy to help, we are on the same subsystem in RH, but there is no way 14:43:36 wait, what? 14:43:40 that seems broken 14:44:06 iirc, SSTs are supposed to be able to talk to each other relatively easily 14:44:17 s/SSTs/SST members/ 14:44:22 mhroncok: No way to do what? 14:44:31 * sgallagh is guessing that sentence was accidentally shortened. 14:45:53 sgallagh: no way to talk to them 14:46:21 sgallagh: as an example, there are no responses on the mailing list thread about this from the change owner 14:46:42 That... feels like something that should likely be escalated in-house. 14:46:49 and direct emails also get no answer? 14:47:06 let me rephrase 14:47:19 trough Fedora comminty channels, mailing list, fesco tickets and bugzilla, this didn't work 14:47:25 in any case yeah... should try and educate them on how to better communicate for these sort of changes? 14:47:35 I can volunteer my Fedora time to help, but I don't wish to do things interanally 14:48:11 is this person new to Fedora things? 14:49:03 no idea, there is no wiki page etc. 14:49:21 They have been in RH for a while... 14:49:43 that doesn't mean anything, unfortunately 14:49:54 the person who maintains libseccomp in RHEL, for example, does nothing in Fedora 14:50:10 same for pkgconf 14:50:13 well, that seems ... suboptimal 14:50:25 it happens a lot more these days 14:50:26 https://src.fedoraproject.org/rpms/postgresql/commits/master suggest 2 years of postgresql maintanance 14:50:26 that has kind of gotten worse over time too 14:50:39 unfortunately 14:51:16 it's something I've learned to expect more these days, sadly :( 14:51:27 :( 14:51:39 but at least this person *does* stuff in Fedora, admittedly quietly 14:52:01 mhroncok is right, we should communicate on this psql issue via fedora channels rather than internally 14:52:06 so, what do we want to do here? I can try mailing them, but if we know thats not going to work... what do we do? just reject the change? 14:52:16 there is a contingency 14:52:23 there is code 100% complete deadline 14:52:49 we approve it and we let them now that we have provenpackagers who can help if reached throu Fedora channels 14:53:11 ok. we should also note it should be done in a side tag. 14:53:54 nirik: I wanted to say that this is obvious... but probably better to explicitly state that 14:54:00 Proposal 14:54:02 ... 14:54:47 yeah, I don't see it mentioned on the change page. 14:55:13 Change is APPROVED assuming the update will be delivered via a side tag. If provenpackager powers are required, please ask for help via Fedora channels. 14:55:40 mhroncok: +1 14:55:45 mhroncok: I'm good with that. +1 14:55:45 mhroncok: +1 14:56:59 mhroncok: +1 14:57:25 mhroncok: I don't think the change is realistic as written. There's a list of ~100 packages that need to be rebuilt in the side tag, and no communication with proven packagers has happened to date. 14:57:46 zbyszek: would you like to ask them again to adapt the proposal? 14:58:04 Yes. 14:58:08 ok 14:58:26 I'll write a list of concrete questions in the ticket that I think need to be answered in the change page. 14:58:27 we still have time... 14:59:20 #action zbyszek to comment in PostgreSQL ticket with questions / missing details in Change Proposal 15:00:06 If there's anything else, speak up now, or I will close the meeting in a few minutes :) 15:00:12 Yes... 15:00:25 .fesco 2416 15:00:26 zbyszek: Issue #2416: Update 3rd party repo policy - fesco - Pagure.io - https://pagure.io/fesco/issue/2416 15:00:48 I made a proposal to approve https://pagure.io/fesco/issue/34 15:00:59 #topic #2416 Update 3rd party repo policy 15:01:26 decathorpe: i also have a thing when the current topic is done 15:01:35 There weren't any comments for a week, so I thought everyone is happy... 15:01:37 bcotton: noted 15:01:46 But maybe not. Either way, please vote or comment. 15:01:47 zbyszek: +1 15:01:59 King_InuYasha: in the ticket please, I don't think we should handle this here. 15:02:02 oh 15:02:12 issue 34 is kinda old though 15:02:44 I see a fresh question from aday in PR#34, regarding flathub 15:02:45 oh, that was supposed to fesco-docs/pull-request/34 15:02:48 King_InuYasha: that is pagure UX fail 15:03:08 King_InuYasha: please fix it :D 15:03:12 haha 15:03:19 it's because fesco-docs are not in fesco repo 15:03:21 * decathorpe wonders if that's the reason why the other proposal was worded so overly-generic :( 15:03:42 cross repo references aren't really a thing yet 15:03:52 King_InuYasha: I edited the proposal to contain the full link now. 15:03:57 zbyszek: cool, thanks 15:04:32 #action everybody will read tickets and proposed docs changes and vote in ticket 15:04:38 decathorpe: oh, right. I'll answer in the ticket. 15:04:53 zbyszek: thanks! 15:05:16 can we hand over the mic to bcotton? 15:05:28 (actually I saw the notification for this earlier, but there's so many tickets and prs that I couldn't find it... I'll keep a tab open now.) 15:05:41 yes please 15:05:57 * decathorpe hands mic to bcotton 15:06:11 #info Council is voting on an actual process for promoting deliverables to Edition status https://fedoraproject.org/wiki/Editions/Promotion_Process 15:06:18 spoiler alert: it will likely pass 15:06:44 the main reason i'm bringing it up here is because in the next few days i'll be submitting an INCREDIBLY LATE F33 system-wide change proposal for promoting IoT to edition 15:07:00 the reason i'm doing this is because most of the work was done before f32, so it's mostly a paperwork issue at this point 15:07:07 so. if anyone objects, now is the time to let me know 15:07:30 or if anyone thinks we should shortcut the process in any way, also let me know 15:07:31 EOF 15:08:11 * nirik replied to the council list thread on it, but didn't see any replies. ;) 15:08:16 meh, seems fine to me 15:08:31 seems fine to me too. System Wide Change seems the appropriate channel 15:08:32 yeah, looks reasonable 15:08:34 though it's going to be interesting to see if we'll get more editions out of it 15:08:50 iirc, one of the reasons this wasn't a real process was to *prevent* editions from spawning 15:09:34 well, there are criteria that must be met 15:09:36 nirik: oh, poop. i meant to reply. i'll do that now anyway 15:09:56 i.e. the Plasma Desktop product was rejected years ago on the basis that they didn't want such an edition 15:10:03 s/i.e./e.g./ 15:10:15 King_InuYasha: yeah, it's not about "i can make anything an edition" so much as "here's what that means" 15:10:30 hmm 15:10:32 because we discovered when promoting IoT that there were a bunch of unchecked boxes 15:10:37 and i made adam very sad 15:10:41 well yeah, because there were no boxes to check :D 15:10:48 exactly! 15:10:50 it was at the whim of the Council back then 15:11:06 which is what sort of upset a few KDE SIG folks, since they wanted to become a proper WG then 15:11:11 meh, old history 15:11:23 but relevant when considering having an actual process for making editions 15:11:29 bcotton: Is there a complementary "demotion" policy as well? 15:11:34 whimsy would look incredibly specious 15:11:39 For example, discontinuing Editions that are on life-support? 15:11:48 we still technically have a Cloud Edition 15:11:52 sgallagh: asking for a friend? 15:11:57 bcotton: other than "this looks fine at first glance", is there anything we should discuss right now? 15:11:57 which is being revived mind you, but has been dead for a few years 15:11:58 My best friend! 15:12:21 there is not, in part because it's basically "stop calling this an edition", but that's something we can look at later 15:12:37 decathorpe: nope. just wanted to make sure everyone was aware it's coming and _why_ it's coming so unbelievably late 15:12:48 bcotton: I'd like to see some marketing clarifications for Editions as a followup, but the process lgtm 15:12:57 where "it" is "the iot promotion proposal" 15:13:06 bcotton: you are still doing the spin liveness check every release right? 15:13:19 bcotton: alright thanks, I just don't want this meeting to run any longer than necessary :) 15:13:34 technically, Fedora has five editions, but only three are marketed 15:13:37 zbyszek: correct 15:13:50 bcotton: could we just include the editions in that too? 15:13:52 with IoT being promoted, that'd be six editions 15:14:04 King_InuYasha: good item for the council-discuss list 15:14:06 * dcantrell hears The Count from Sesame Street 15:14:23 I'm prepared to propose Server to drop back to being a Spin 15:14:24 bcotton: I'll bring it up sometime later 15:14:43 sgallagh: why? 15:14:47 zbyszek: maybe. there's more to an edition than "someone is still there", but it'd be good to at least know if something is dead 15:14:56 Because I can't keep running it solo and still pretend it's an "Edition" 15:15:18 we should drop the Server edition and create a Fedora Enterprise edition 15:15:21 sgallagh: sorry about that... I'd certainly be happy to help with Server WG stuff, I just didn't know what to do 15:15:22 hahaha 15:15:28 dcantrell: oh boy, hehe 15:15:34 Proposal: End meeting if there's no other topics to discuss? 15:15:43 +1 to end 15:15:45 +1 15:15:47 ack 15:15:49 ack 15:15:52 see you later 15:15:57 decathorpe++ 15:16:03 thanks, decathorpe++ 15:16:08 great. thanks everybody! have a nice week. 15:16:11 Thanks decathorpe 15:16:15 decathorpe++ 15:16:15 sgallagh: Karma for decathorpe changed to 13 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:16:19 thanks decathorpe 15:16:25 #endmeeting