17:00:18 #startmeeting FESCO (2024-01-04) 17:00:18 Meeting started Thu Jan 4 17:00:18 2024 UTC. 17:00:18 This meeting is logged and archived in a public location. 17:00:18 The chair is sgallagh. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 17:00:18 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:00:18 The meeting name has been set to 'fesco_(2024-01-04)' 17:00:18 #meetingname fesco 17:00:18 The meeting name has been set to 'fesco' 17:00:18 #chair nirik, jistone, zbyszek, sgallagh, humaton, dcantrell, mhayden, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor, tstellar 17:00:18 Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku dcantrell humaton jistone mhayden nirik sgallagh tstellar zbyszek 17:00:32 .hello ngompa 17:00:34 #topic Init Process 17:00:34 Son_Goku: ngompa 'Neal Gompa' 17:00:35 .hi 17:00:36 sgallagh: sgallagh 'Stephen Gallagher' 17:00:43 .hello2 17:00:44 dcantrell: dcantrell 'David Cantrell' 17:00:46 .hi 17:00:47 jistone: jistone 'Josh Stone' 17:00:59 * decathorpe watching from the cheap seats 17:02:29 * sgallagh waits to see if we achieve quorum 17:03:03 .hello2 17:03:04 zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' 17:03:33 I'm in the car, so I'll bit a bit slow to respond. I should be home in about 10 minutes. 17:03:55 zbyszek: Don't text and drive, please 17:04:23 .hello humaton 17:04:24 jednorozec: humaton 'Tomáš Hrčka' 17:04:36 .hello tstellar 17:04:37 tstellar: tstellar 'Tom Stellard' 17:04:59 Well, we now have quorum, so we'll begin. 17:05:19 First order of business: Welcome jednorozec and jistone to FESCo! 17:05:39 thanks! 17:05:40 welcome! 17:05:59 thanks 17:06:08 Congratulations on your election. Your official rubber stamps would be in the mail if they existed and people still used mail. 17:06:29 people still use mail 17:06:54 I know because of my very realistic portrayal! 17:07:01 Second order of business: A salute to decathorpe and mhroncok, both of whom opted not to run for FESCo again. Thank you for your service. 17:07:22 * nirik is still on vacation, but happens to be at the keyboard... :) 17:07:24 * dcantrell applauds 17:07:32 * decathorpe bows 17:07:33 thank you, decathorpe and mhroncok 17:07:35 congratulations to decathorpe and mhroncok 17:07:43 .hello salimma 17:07:43 yeah! many thanks! 17:07:44 michel-slm: salimma 'Michel Lind' 17:07:52 sorry 'thank you' would have been more appropriate 17:07:59 * michel-slm just realized this is still on IRC 17:08:10 smooge: "Congratulations on your graduation", maybe? 17:08:22 congrats to the new members and the emeritus ones :D 17:08:25 smooge: "congratulations, you got out in time" is perfectly fine 17:08:56 I promise I was trying not to start off the year with my usual snarkiness 17:09:24 thank you both for everything :) 17:09:26 #topic #3101 Change: Remove OpenSSL Compat 17:09:30 .fesco 3101 17:09:31 sgallagh: Issue #3101: Change: Remove OpenSSL Compat - fesco - Pagure.io - https://pagure.io/fesco/issue/3101 17:10:45 When last we left this topic, we were punting until we had a clear plan in place for Python 3.6. 17:11:19 It still seems unlikely to me that python3.6 will be ported to OpenSSL 3. 17:11:22 has anyone *tried* to make an openssl3 patch? 17:11:30 Last I looked, the "plan" was to move Python 3.6 over to OpenSSL 3, but that doesn't seem to be happening 17:11:33 I applied the python2.7 patch, and it went ok, I think 17:11:34 not that I have seen 17:11:43 well, I don't know that it's realistic to expect anyone to have worked on this over their holidays. 17:11:49 ++ 17:12:11 nirik: Absolutely fair, but it's been a month of no motion 17:12:27 openssl maintainers want the compat package removed, so I don't think it's fair to keep this around since no one is going to work on it 17:12:39 I'm not keen on punting the decision much further. I'd like to propose the following: 17:13:00 sgallagh: going from how things are in various groups in Red Hat.. expecting motion from Nov 24->Jan 7 is betting against the odds 17:13:21 Proposal: The Change is approved, and Python 3.6 must either migrate to OpenSSL 3 prior to Beta Freeze or be dropped from Fedora at that time. 17:13:58 I think they can either migrate or drop the dependency on openssl-compat. just disable the ssl modules that require it 17:14:06 either way, the decision is for the python maintainers 17:14:27 * nirik nods. 17:14:43 What if the python maintainers took over maintainence of openssl-compat is that an acceptable solution too? 17:14:55 they've already indicated they can't take on that work 17:14:57 tstellar: Frankly, I'd say "no" 17:15:07 tstellar, I dont think so the team is understaffed 17:15:20 They don't want to and they're not (on the whole) subject matter experts 17:16:09 sgallagh: can we rephrase your proposal a bit to say Python 3.6 must either migrate to OpenSSL 3 prior to Beta freeze or drop the dependency on openssl-compat? 17:16:31 dcantrell: I'm not sure that's realistic; SSL is pretty baked-in, isn't it? 17:16:39 so I think many of the reasons that python3.6 is kept is to use ssl items 17:16:50 per this discussion, it's a possible path for them: https://discussion.fedoraproject.org/t/f40-change-proposal-removing-openssl-1-1-package-system-wide/92899/4 17:17:07 Ok, so in that case, I'm +1 sgallagh's proposal. It seems like no one wants to maintain openssl-compat any more, and we can't really force people to do the work. 17:17:34 I'll amend my proposal slightly 17:17:47 count me a +1 to the proposal, afk for just a bit 17:17:49 actually, that isn't true 17:18:00 tstellar: jonathanspw offered to maintain it last time we discussed it 17:18:09 Proposal: The Change is approved, any package still depending on openssl-compat at Beta Freeze will be dropped from Fedora at that time. 17:18:17 Less python-specific this way as well. 17:18:35 sgallagh: -1... I'd rather ask that they orphan rather than retire it 17:19:12 given the packages that still depend on it, the fact it's technically maintained through the end of RHEL 8 EOL, a packager could maintain it by syncing RHEL 8 to Fedora's compat package 17:19:46 I'm kind of wary of community-maintained OpenSSL, if only because a non-corporate entity is less likely to get timely notification of CVEs 17:19:47 do we *want* anyone to be adhoc'ing a security-sensitive package like that? 17:20:17 given that bascially only an old version of python still depends on it ... I guess the potential damage is very limited as well 17:20:57 I feel the same way as sgallagh. If it were another package I think orpahning instead of retiring would be fine, but I don't think we should plan to ship a release with components that depend on an umaintained crypto library. 17:21:01 well, at least in fedora provided packages... ;) There may be 3rd parties using it. 17:21:22 3rd parties can reintroduce the package themselves 17:21:24 it's... python36 only in the Fedora repos, but I bet there are many third-party / internal packages that still depend on it too, right 17:21:25 jistone: you said that you applied the patch and it worked… Can you expand a bit on this? 17:21:29 nirik: Third parties are more likely to bundle it than use ours 17:21:52 jistone: link to a patch or pr? 17:21:58 sure, they could do all kinds of things. 17:22:00 I got it far enough to build locally -- trying a mockbuild to see how it goes in full 17:22:05 so yeah if it's maintained by volunteers... it really needs to keep up with the RHEL8 openssl at least 17:22:20 I haven't posted anywhere yet, but I will if this passes 17:22:43 jistone: tham\nks 17:23:00 So this sounds doable. 17:23:12 +1 to sgallagh's proposal 17:23:14 The fedora one is not 100% the same as the thel8 one too.... the env is going to be different somewhat and we have compat-*, so you can't just copy the spec... so there could be subtle bugs that creep in 17:23:40 yeah, I can be +1 to sgallagh's proposal too 17:24:15 +1 from me 17:24:19 +1 17:24:33 I think it's much safer to drop it than to pretend that it's mainained. Openssl is a too juicy target with too many cves to wing it. 17:25:24 I count (+5, 0, -1) so far, so this will pass. dcantrell, tstellar: do you want to weigh in? 17:25:35 * dcantrell is back 17:25:36 +1 17:25:47 sgallagh: I'm +1 17:26:11 #agreed The Change is approved, any package still depending on openssl-compat at Beta Freeze will be dropped from Fedora at that time. (+7, 0, -1) 17:26:47 #topic New Meeting Time 17:27:11 #link https://whenisgood.net/agyhckd/results/sxn8wpk 17:27:41 At present, we have one available slot shown, but it's at a time where I suspect jednorozec would not prefer. 17:28:07 basically this slot, but monday? 17:29:01 this slot on Monday overlaps with blocker review 17:29:04 so the timES i SEE ARE MY LOCAL TIME OR utc? 17:29:04 No, but I just realize WhenIsGood isn't displaying timezones properly. 17:29:13 huh 17:29:25 Which means probably all of the responses are wrong... 17:29:25 new keyborad sorry 17:29:28 I think that's showing in Eastern 17:29:32 mine is right :P 17:29:42 Right, I submitted it in Eastern US time, which timezones enabled. 17:29:43 it did let me set my timezone when I responded 17:30:04 So the current slot looks like 3pm EST (three hours later than the current slot) 17:30:24 But if folks thought that was UTC, everything is likely borked 17:30:33 I hate timezones. 17:30:33 people can update their responses right now 17:30:41 whenisgood lets you do that 17:30:44 I submitted my response in my timezone. 17:31:13 3hrs later is 9 ish over here :/ 17:31:37 Right, I'm looking at the responses and they actually do look right for their resp. timezones. 17:32:00 But as I said, I suspect the currently-available slot is bad for jednorozec 17:33:13 Am I the only one in Europe? 17:33:22 I could try and move my standup and do 1 hour eariler... 17:33:23 No, zbyszek is also 17:33:35 his responses seems like he respondet in CET 17:33:45 but didnt change the tz 17:33:56 His TZ is marked as "Warsaw" 17:33:57 Yes, and the one time is 21 CET, i.e. 20 UTC. 17:34:09 I see 17:34:16 It's late, but I'll try to do it. 17:34:28 At least I won't have other things to do at time usually. 17:34:30 if we had a time machine, we wouldn't have this problem 17:34:38 lol 17:34:44 dcantrell: No, we'd have many, many problems instead 17:34:48 remember: you only need to invent a timemachine _once_ :) 17:34:57 violating causality for convenience 17:35:06 time machines go only forward we have science to back that up now :D 17:35:23 jistone: we can't all be Hermione Granger, I guess :) 17:35:36 Or The Doctor... 17:35:38 anyway 17:35:43 mavity 17:35:44 ok lets do monday 20UTC 17:35:59 I'm still trying to figure out how all of FESCo could fit in a DeLorean. Those were 2 seaters 17:36:13 The next, least-bad option is two hours earlier, but Neal declined that one. 17:36:14 * Son_Goku is listening to a Doctor Who theme remix right now :P 17:36:27 also my standup should be done by 19:30utc... so we could do that to be 30min eariler for eu folks? 17:36:33 @Son_Goku Is that set in stone, or could we make that available? 17:36:42 sgallagh: 12pm is blocker review and 1pm is kde sig meeting 17:36:58 it's not totally set in stone, I could double up on kde sig and this 17:37:18 but we should not overlap with blocker review 17:37:22 19:30 UTC would work for me too. 17:37:37 half hour earlier is fine-ish 17:37:43 I need to step away for a few minutes, but I'm +1 to either time. 17:37:44 I can drop out halfway in kde sig meeting 17:38:01 or at least be quiet in a video meeting :P 17:38:11 Isn't blocker review at 1700 UTC? 17:38:39 yes, that's 12pm EST 17:39:00 KDE SIG meeting is pegged to US/Eastern, which is currently 18:00 UTC / 13:00 EST 17:39:24 will the meeting be anchored on UTC, or shift with EST/EDT? 17:39:34 (not a problem for me either way) 17:39:51 Right, I was asking about 1800 UTC, which I realize collides with KDE. I was wondering if that could be doable to make life easier for the European folks 17:39:59 jistone: Traditionally we track UTC 17:40:29 I think we can go with 19:30UTC and see how it goes 17:40:51 OK, if that works for you. 17:40:56 I'm fine with it too 17:41:35 #info The new FESCo meeting time will be at 1930 UTC, beginning on 2024-01-16 17:41:40 monday is also nice because freeze events are tuesdays... 17:41:48 yup 17:41:49 yup 17:42:17 Since Monday is just two working days away, I vote we skip a formal meeting next week but I'll send out an announcement-only message. 17:42:36 +1 17:42:40 +1 17:43:02 we can process async voted stuff on Monday regardless 17:43:10 That's what I meant. 17:43:23 #topic Next Week's Chair 17:43:35 Who wants to chair the 2024-01-16 meeting? 17:44:02 #action sgallagh to send out the voting announcements on 2024-01-08 17:44:21 Whoops, I mean the 2024-01-15 meeting 17:44:39 (Forgot I had my calendar set to Monday as first day of the week) 17:44:48 heathen :P 17:44:59 !hi 17:45:00 .hi 17:45:02 jonathanspw: jonathanspw 'Jonathan Wright' 17:46:28 For the benefit of our new members, FESCo policy has us rotating the chair each week. The protocols and processes for chairing the meeting are recorded at https://fedoraproject.org/wiki/FESCo_meeting_process 17:46:45 also, what about finally switching to matrix? 17:46:57 I was going to bring that up in Open Floor 17:47:04 okay 17:47:22 rotating makes sense, but I'd rather sit through a few before driving :) 17:47:59 jistone: Thanks for volunteering! 😈 17:48:21 (I kid, of course) 17:48:43 jistone: If you want, I can chair and you can shadow me. 17:49:05 tstellar: That sounds like a great idea. 17:49:05 sure, if you're volunteering 17:49:40 #action tstellar will chair the 2024-01-15 meeting and offers to mentor jistone at the same time 17:49:46 Thank you both 17:49:52 #topic Open Floor 17:50:00 So... Matrix? 17:50:07 yes please 17:50:07 yes please 17:50:28 +1 17:50:29 I'm all for it, assuming we have a meetbot that supports it. 17:50:46 Will that necessitate a change to the meeting protocol? 17:50:54 most of the commands are the same 17:50:59 just ! instead of # 17:51:01 The only thing really missing is the email to meetingminutes, and I guess people don't find that too important... 17:51:11 well and no chair functionality 17:51:20 (almost all other meetings have already moved) 17:51:22 that is slightly irritating 17:51:28 yeah 17:51:44 Does that mean that any participant can use ! commands? 17:51:56 yep 17:51:57 I'm not sure that's too much of a problem in practice. 17:51:59 I think from a syntax sugar perspective, the only thing we might be missing is !fesco wrapper for "!pagureissue fesco" 17:52:59 Does uploading of logs to meetbot work? 17:53:01 yes 17:53:02 yes 17:53:11 So yeah, I think we should switch. 17:53:34 nothing was ready when I brought this up a while ago (re: moving to matrix), but it feels usable enough now to move 17:54:44 FWIW I'll probably miss the meeting on the 15th. Ski time. 17:54:54 ryanlerch did a great job porting everything over 17:55:15 OK, so I'm not hearing any dissent, so I suggest we skip a formal vote and just switch to Matrix. 17:55:28 :D 17:55:31 One thing I didn't check... what meeting channel will we use for the new time. 17:55:41 we should be good to use fedora-meeting 17:55:46 #meeting:fedoraproject.org 17:56:48 Works for me 17:57:15 #action sgallagh to update the FESCo Meeting Process with new bat-time, new bat-location 17:57:30 :D 17:58:06 tstellar: Since this will be the first Matrix meeting, probably worth calling that out loudly when sending the meeting announcement 17:58:20 sgallagh: OK 17:58:41 Any other topics for this week? 17:58:57 We should also probably put a notice in the IRC channel at the time of the meeting to redirect any confused folks. 18:00:09 Agreed 18:01:09 sgallagh: please wrap this up. 18:01:17 Sounds good to me. 18:01:22 Thank you for coming, folks! 18:01:26 #endmeeting