17:00:10 #startmeeting FESCO (2023-05-02) 17:00:10 Meeting started Tue May 2 17:00:10 2023 UTC. 17:00:10 This meeting is logged and archived in a public location. 17:00:10 The chair is mhayden. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 17:00:10 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:00:10 The meeting name has been set to 'fesco_(2023-05-02)' 17:00:10 Nee, thirds! 17:00:15 #meetingname fesco 17:00:15 The meeting name has been set to 'fesco' 17:00:27 #chair nirik, decathorpe, zbyszek, sgallagh, mhroncok, dcantrell, music, mhayden, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor 17:00:27 Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku dcantrell decathorpe mhayden mhroncok music nirik sgallagh zbyszek 17:00:27 #topic init process 17:00:29 .hi 17:00:30 sgallagh: sgallagh 'Stephen Gallagher' 17:00:38 .hello2 17:00:39 mhayden: mhayden 'Major Hayden' 17:00:51 .hello zbyszek 17:00:52 zbyszek[m]: zbyszek 'Zbigniew Jędrzejewski-Szmek' 17:01:00 oh man, i got my matrix/irc thing working properly and .hello2 works 🎉 17:01:03 hi fed500 17:01:10 .hello2 17:01:11 Eighth_Doctor: Sorry, but user 'Eighth_Doctor' does not exist 17:01:17 .hello ngompa 17:01:18 Eighth_Doctor: ngompa 'Neal Gompa' 17:02:05 Conan Kudo: i can help with your matrix/irc name sync stuff later if you're interested (if your problem is the same as mine) 17:02:05 morning 17:02:15 .hello fed500 17:02:16 bkm: fed500 'None' 17:02:25 .hi 17:02:26 neil: neil 'Neil Hanlon' 17:02:38 .hi 17:02:39 decathorpe: decathorpe 'Fabio Valentini' 17:03:09 i think we have quorum now if i counted properly, but i'll give it another minute 17:03:26 some fun home repairs happening today so my brain is scattered 17:03:50 Must be nice to have an excuse... 17:04:06 sgallagh: you have kids 17:04:25 that is your goto excuse til they are 50 17:04:37 * neil is waiting to hear from his realtor about a bid on a house .. 17:04:46 .hello music 17:04:47 music[m]: music 'Benjamin Beasley' 17:04:53 sending good luck your way, neil 17:05:01 thanks mhayden.. we need it lol 17:05:07 okay, y'all, let's get this party started 17:05:17 #topic #2984 Change: Remove webkit2gtk-4.0 API Version 17:05:27 .fesco 2984 17:05:28 mhayden: Issue #2984: Change: Remove webkit2gtk-4.0 API Version - fesco - Pagure.io - https://pagure.io/fesco/issue/2984 17:05:46 this one might not be a big deal, but i added it since it touches some important packages 17:06:06 if it's not a big concern, we can vote/ask questions in the ticket and move on with the meeting 17:06:58 kudos to Michael Catanzaro for explaining the change well in the proposal 👏 17:07:00 Eh, nobody cares about Emacs anyway... 17:07:03 Michael Catanzaro++ 17:07:12 well, packages have either 1. already been ported to the new API, 2. have already been fixed, or 3. were already removed since they used libsoup-2.4 ... 17:07:15 .hello catanzaro 17:07:16 MichaelCatanzaro: catanzaro 'Michael Catanzaro' 17:07:22 in ticket should be fine IMHO. I guess this means the end for midori. oh well. 17:07:36 nirik: No not really 17:07:43 Our Midori is different from upstream Midori 17:07:57 We have an old version; the new one is another Chromium reskin 17:07:57 no? 17:08:07 The "new" one... is also old. 17:08:17 * sgallagh is not thrilled with the new browser monoculture... 17:08:19 do you have a link for the chromium one? 17:08:31 decathorpe: libsoup2 was not retired: https://src.fedoraproject.org/rpms/libsoup 17:08:40 https://fedoraproject.org/wiki/Changes/libsoup_3:_Part_Two 17:08:55 it wasn't, but any packages that pull in both libsoup-3 and libsoup-2.4 via their dependency tree are b0rked 17:09:17 nirik: https://github.com/midori-browser/core/commits/master, no commits in 3 years, I guess the "new" one is indeed also old... alas 17:10:05 yep. Thats not a chromium thing tho... 17:10:15 nirik: Indeed, it uses WebKitGTK O_O 17:10:16 anyhow, thats a side track. ;) 17:10:18 What do I know, then. Huh. 17:10:46 * nirik is the maintainer of the fedora midori packages. ;) It was cool back in the day... 17:12:00 do we have any concerns on the proposal that aren't addressed yet? or actions? 17:13:31 * nirik doesn't. 17:14:13 alrighty. i guess we can move on here. if folks can have a look through the ticket and vote / ask questions, that would be great 👍 17:14:13 nope. just general "meh, if necessary, fine" 17:14:28 There are some other packages in the “intersection” list that I recognize as ”significant,” with large user bases and/or active upstreams, e.g. exaile, quodlibet/exfalso, gthumb. I wish I knew how gnarly their path to staying in the distribution was. In some cases it might just be that the spec file is out of date. But we need to keep chipping away at the libsoup2 transition, so we can’t just keep everything around 17:14:28 indefinitely either. 17:14:30 thanks for the proposal Michael Catanzaro 17:15:01 next item up... 17:15:06 #topic #2982 Create an email list policy 17:15:09 .fesco 2982 17:15:10 mhayden: Issue #2982: Create an email list policy - fesco - Pagure.io - https://pagure.io/fesco/issue/2982 17:15:15 * mhayden slowly backs away 😄 17:15:28 * nirik adds a shot to his coffee. 17:15:29 Difficult topic 17:15:41 Stephen Gallagher made a good point to focus on the change being requested here 17:15:47 PRobably dead list policy seems reasonable to start with 17:16:14 so, I'll start by saying fesco should not make such a policy for all lists. If they want to for the technical side of things, ok... 17:16:27 I've heard rumblings that this is a Council responsibility rather than a FESCo one 17:16:36 mhayden: I think you're conflating two issues, one that hasn't had a formal request made to FESCo yet 17:16:48 it has actually. see the next ticket. 17:16:54 Ah, sorry. 17:17:06 sgallagh: totally possible that i am getting it backwards 17:17:11 Ok for council to consider it 17:17:20 IMHO we should close this one, and focus on https://pagure.io/fesco/issue/2989 17:17:24 Regarding the mailing list policy, I propose "Any list that hasn't seen a message for the last four Fedora release cycles will be archived" 17:17:26 Though good to separate from Discourse discussion 17:17:42 if making a "Inactive Mailing List Policy" is Council territory then I'd say let's close this as NOTOURBUSINESS 17:17:49 Proposal on archiving seems reasonable to me 17:17:53 But yeah, that could well be COuncil territory 17:18:32 I don't think it's council territory any more than it is ours. Are council members somehow experts on dead mail lists? 17:18:40 I'll take the AI to open a Council ticket for discussing a list retirement policy. 17:18:46 Can investigate and send to council. Probably infra will need to give input 17:18:57 > "Any list that hasn't seen a message for the last four Fedora release cycles will be archived" 17:18:57 Sounds reasonable. 17:19:01 well, a number of lists fall under mindshare... 17:19:01 #action Stephen Gallagher to open a Council ticket to discuss a list retirement policy 17:19:01 zbyszek: Council == strategic stuff, FESCo = tactical stuff 17:19:06 thank you, Stephen Gallagher 17:19:06 why is fesco deciding what happens to them? 17:19:25 s/=/==/ 17:19:28 I'm not a fan of the idea of using Discussion for Change discussion 17:19:33 but if there are teams that want to shift their discussion from lists to forums, that's their prerogative 17:19:44 but yes, we must separate "your mailing list is dormant and needs to be deleted" and "let's go to discourse" 😉 17:20:11 to be clear, I am fine with closing inactive lists... although it's work infra has to do, while just ignoring them is no work. 17:20:12 Conan Kudo: that's a separate topic. Let's discuss the topic at hand. 17:20:16 well, we don't need to delete lists, we just close them 17:20:36 and we probably shouldn't delete lists since they can contain historical context 17:20:46 archive not delete 17:20:48 right, Conan Kudo -- i used the wrong words earlier 17:20:59 ignoring them is still work. I have spent a lot of time 'cleaning' the ignored ones 17:21:02 we don't. But 'closing' them means having to set a bunch of things in mailman that are anoying. 17:21:30 mhayden: Regardless of whose decision it is, I think the dormant-list idea is the only part of https://pagure.io/fesco/issue/2982 that isn’t premature at this point. 17:21:37 does marking a mailing list as "closed" at least reduce the time spent on triaging held messages? 17:21:47 music: agreed 17:22:00 nirik, smooge: so this is actually important. Is it less work to keep the list or to close it? 17:22:07 well... it depends. 17:22:23 held messages go to the list owners/moderators. 17:22:43 I am not convinced we have any left that moderate non subscribed emails, but I suppsoe we might 17:22:53 we do not. 17:23:07 IMHO it's less work to just leave it alone... 17:23:14 most of the people I contacted about non-subscribed email told me that they had set up filters to go to trash a long time ago 17:23:21 but if everyone wants things 'closed' we can do it, it's just more work 17:23:27 * nirik nods 17:23:31 Can it be automated? 17:23:57 sure. But someone has to make the scripting. 17:24:01 so a lot of the issues to make it easier to do any of these things requires mailman3, hyperkitty and other tools updated 17:24:40 My view (again, perhaps not my decision) is that the only reason to close/archive inactive lists is if the people maintaining the mailing lists say it will reduce maintenance burden. 17:24:44 which is underway still... and may finish. In fact it's all in fedora now, we could look at that. 17:26:13 indeed 17:26:19 but otherwise I agree with nirik. It is a lot of work with limited resources which will require a lot of dedicated focus to accomplish 17:26:27 I'm not sure what else to say here... or what fesco would like to do. 17:26:57 Ok. Fine to leave as is, and revisit once Discourse discussion is done if needed. 17:27:13 I don't think that inactive lists matter for users. We have a hundred somewhat active lists anyway, so users have to use search anyway, so the extra 580 inactive lists don't make much of a difference. 17:27:23 The issue was raised thera are may inactive lists 17:27:26 .hi 17:27:27 salimma: salimma 'Michel Alexandre Salim' 17:27:54 hey Michel Alexandre Salim 🎩 17:27:56 how about i update the ticket to say Stephen Gallagher will bring it up with Council (the inactive mailing list part) and that a move to Discourse really isn't a choice for us to make? 17:28:10 as someone who admins a couple of inactive lists, I would not mind shifting them to Discourse. maybe let list owners opt in 17:28:31 IIRC SIGs used to be required to have lists, right? 17:28:37 yes 17:28:46 though I think SIGs have generally not requested them as of late 17:28:47 packager sigs MUST have lists... and do. 17:28:59 you 100% need somewhere to cc bugs. :) 17:29:08 right 17:29:16 that's why both the asahi and budgie sigs have them 17:29:20 can that work with Discourse? that will be interesting 17:29:36 Possibly... but not sure yet. 17:30:02 I guess... I would say we should let things progress and see where we are down the road... 17:30:48 nirik++ 17:30:54 if we want the council to weigh in on inactive lists, sure, ok... 17:31:56 Michel Alexandre Salim 🎩: you and I should sync up on the steps left to get mm3 stack into epel9 17:32:01 since I think everything is now in Rawhide 17:32:21 I think epel9 just needs django? 17:32:22 yeah, let's take that offline after this meeting 17:32:35 it's mostly django related IIRC 17:32:42 nirik: I don’t, particularly. But if somebody thinks it’s important, especially if there is a concrete benefit to doing something about them, fine. Otherwise I’d rather see people spend time on things that are directly helpful. 17:32:57 okay, i feel like we've talked through the mailing list ticket pretty well thus far ;) 17:33:08 mailman3 still needs unretiring, I think it builds fine now but it is quite hairy, so if Conan Kudo can update that I'd feel more comfortable with it 17:33:19 Michel Alexandre Salim 🎩: sure, let's sync and get that done 17:33:29 I think mm3 package itself is actually in a weird releng state right now 17:33:36 which is why I haven't fixed it before 17:33:48 * nirik happy to help sort it out of band too 17:34:08 #action mhayden to update #2982 with some updates from this meeting 17:34:16 For housekeeping, I think we should close the ticket as "not desired, at least for now". That seems be the outcome of the discussion here. 17:34:28 agreed 17:34:30 +1 17:34:36 +1 17:34:38 1 17:34:42 + 17:34:56 Ok 17:34:56 +1 17:34:59 nice... markdown + for the win. ;) +1 17:35:04 +1 17:35:12 nirik: very creative 17:35:18 okay, moving right along... 17:35:26 #topic Next week's chair 🪑 17:35:58 So, did we want to discuss https://pagure.io/fesco/issue/2989 ? or did it come in too late and we should punt to next week? 17:36:10 anyone want to chair next week's meeting? 17:36:18 nirik: oh, open floor still coming up 17:37:15 punt 17:37:16 I don't really want to talk about it right now 17:37:43 I can do it. I don't have anything too much on fire next week. 17:37:54 zbyszek: thank you much 17:38:07 #action zbyszek to chair the next meeting 🪑 17:38:12 zbyszek++ 17:38:15 #topic Open Floor 17:38:27 nirik: want to bring up 2989? 17:38:28 .fesco 2989 17:38:29 mhayden: Issue #2989: Proposal to adjust Changes Policy to use Fedora Discussion instead of the devel list - fesco - Pagure.io - https://pagure.io/fesco/issue/2989 17:38:30 * nirik is fine with punt too... as treebeard says "don't be hasty!" 17:38:57 punt +1 17:39:00 I am curious what the general thought is... but we can punt to next week, thats fine with me 17:39:05 punt +100 17:39:40 just vote with all your IRC nicks ;) 17:39:40 nirik: I don't personally want to do it, I think the way forums work make it unappealing for that discussion 17:39:42 🙀 17:39:42 * zbyszek[m] wonders if we're waiting for any new information, because otherwise the punt is not useful. 17:40:14 I'm -1 to making any changes in the immediate future until the loss of the Fedora PgM has time to settle 17:40:17 and some of the "features" of discourse make it even more unattractive for long discussions 17:40:22 zbyszek: I think "I cannot productively discuss this right now" is a valid reason to punt to be honest 17:40:26 I'm fine to try it out with Changes discussion... I'm happy to help anyone who wants to interact with it via email (I have been doing this for a while). If it doesn't work out, we can go back. 17:40:51 yeah. Agree with everyone there. ;) 17:40:51 Now Neal started a discussion… 17:41:27 but honestly, the number one reason I don't want to do it is because Discourse allows mod/site side topic splitting, which I personally consider extremely rude 17:41:40 shoving people out of a discussion is awful 17:41:44 nirik: Ok, general thoughts: I’m in Matt’s “open to the idea” category at this point. I haven’t used Discourse. I see the motivations for migrating, and I see some of the concerns too. I suspect switching things to Discourse might make my life a little worse, but I’m not certain of that. If migrations are going to happen, the Changes process might be an OK place to start eating our own dog food. 17:42:11 between that and the automatic topic shutdown after 30 days and some of the other things on discourse, I definitely think it's inappropriate for technical discussions 17:42:42 You can set the 'auto close' thing per thread (or whatever they call them) I am pretty sure. 17:42:42 Eighth_Doctor: Do you know if the topic shutdown is configurable? I would find that harmful too. 17:42:54 music[m]: the few times I've interacted, it doesn't appear to be 17:43:17 I think discourse is perfectly fine for user support and things of that nature, but development discussions suck on discourse 17:43:41 I wonder if the Python folks have done any retrospective after switching to Discourse for development 17:43:57 It does seem editable per thread. 17:44:01 and making it more difficult for connecting outside and inside parties is just unappealing to me 17:44:06 Michel Alexandre Salim 🎩: there was a mail from Miro on fedora-devel. 17:44:21 ah thanks. I need to catch up on that thread 17:44:22 (again) 17:44:49 nirik: What about a site-wide default? 17:45:12 Not sure, but I don't see any autoclose at all on ask threads... 17:45:31 I also don't think discourse is going to help anything for increasing contributors 17:46:10 our problems are much more fundamental than the medium of longform async discussion 17:46:17 I think there's a lot of folks these days that don't bother with mailing lists... and it would be sure nice to get them involved. 17:47:02 nirik: i'm in the "lost with mailing lists" boat. i go back to work with my team, then return to the list, and an entire novel has been written since i left 😅 17:47:14 MIro's message about Python: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/message/GMK75TUS5LA4Z7F3GWAFYTH4TUKAPAJC/ 17:48:01 I also think that discourse has the potential to make "connecting inside and outside parties" actually easier. It should make it easier for people to move from being pure users who ask questions, to participate in discussions, and then to get more involved. 17:48:58 I started paying more attention to discussion.fp.o over the last few weeks and I'm surprised how many user topics there are e.g. about systemd, that I would never have been involved in, and that would never make it to fedora-devel. 17:49:00 The email interface has it's issues for sure, but it does mostly work if you are used to interacting via email... 17:49:41 Yeah, the mail interface is usable, in particular when just reading discussions. 17:49:48 Lists on sourcehut are really great. 17:50:00 zbyszek: do we really want that? devel discussions already tend to go off into tangents and side-topics that don't really make following the actual discussion easier 17:50:09 However, mostly go, so difficult to package and update 17:50:26 bkm: sourcehut is written in Python 17:50:39 Email part is mostly go 17:50:44 anyway, doesn't matter 17:50:59 the bigger thing is that the forums culture is naturally very different 17:51:27 for good or ill, we are probably going to lose a significant chunk of our contributor base 17:51:32 I remember when web forums (mostly) died. Now they are fashionable again. 17:51:35 if we kill mailing lists 17:52:00 Fabio Valentini: I'm not sure why discourse would be worse in that regard. At least there there's the possiblity of splitting out subtopics and other measures to make the discussion easier to follow. 17:52:21 zbyszek: which is a bad feature 17:52:32 because you effectively shunt away people 17:52:46 it's not isolating things, there's a link right there saying "XYZ posts have been split off to topic: something else" 17:52:59 i think it is a culture issue as you said forum people don't see it as shunting 17:53:12 right 17:53:14 I don't see it as shunting either, but oh well. 17:53:32 as soon as you've separated a discussion, you've split the audience 17:53:36 that's a shunt 17:53:38 Conan Kudo: you wrote that before also. I understand the worry, but I think that it won't be a problem for us. I assume our admins will do things in a reasonable way. I don't think we need to assume hostile use or overpolicing. 17:54:11 zbyszek: I don't want to get into it, but I have decent reason to believe the goal is to use that capability a lot more than you think 17:54:18 I'd be less worried about this if we could make FESCo members admins of those topics 17:54:33 (s/topic/whatever the correct term is for Discourse/) 17:54:35 well, take the discourse thread on devel... there was a subthread about the discourse email interface... and because it wasn't a seperate thread, people asked the same questions multiple times because they didn't want to read the entire thread 17:54:42 one of the private pitches I've gotten for discourse is specifically around that capability 17:54:51 we're approaching :55 minutes past -- i need to run at the top of the hour :| 17:54:52 Fabio Valentini: I think that might indeed be possible... 17:55:06 I need to go too... 17:55:12 we are finishing up a plugin to sync account system groups perms. 17:55:31 ie, 'fesco' could be automatically granted perms for #devel tag or something. 17:55:31 and we got into discussing something I didn't want to talk about today because I don't really have a good headspace for this right now 17:55:41 and this is my fault 17:55:42 sorry 17:55:56 I think we should delegate administrations of topics a lot in general, because otherwise it'll be way too much for a small group of people. 17:55:58 no problems. :) lets stop. 17:56:00 not your fault Neal, you did say you wanted to punt :) 17:56:37 okay, y'all -- thanks to everyone for showing up and participating! 👏 17:56:56 i hope you all have an awesome rest of the week 17:57:01 Thanks for the discussion 17:57:25 thanks mhayden! 17:57:33 anytime! 17:57:42 see you folks around! I'm not voting anyway so it doesn't matter, but will be at conferences in Vancouver next week 17:57:44 i'll go ahead and close us out 17:57:45 #endmeeting