17:00:06 <zbyszek> #startmeeting FESCO (2022-11-15)
17:00:06 <zodbot> Meeting started Tue Nov 15 17:00:06 2022 UTC.
17:00:06 <zodbot> This meeting is logged and archived in a public location.
17:00:06 <zodbot> The chair is zbyszek. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
17:00:06 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:06 <zodbot> The meeting name has been set to 'fesco_(2022-11-15)'
17:00:06 <zbyszek> #meetingname fesco
17:00:06 <zodbot> The meeting name has been set to 'fesco'
17:00:06 <zbyszek> #chair nirik, decathorpe, zbyszek, sgallagh, mhroncok, dcantrell, music, mhayden, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor
17:00:06 <zodbot> Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku dcantrell decathorpe mhayden mhroncok music nirik sgallagh zbyszek
17:00:09 <zbyszek> #topic init process
17:00:12 <mhayden> .hi
17:00:13 <zodbot> mhayden: mhayden 'Major Hayden' <mhayden@redhat.com>
17:00:14 <zbyszek> .hello2
17:00:16 <zodbot> zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' <zbyszek@in.waw.pl>
17:00:29 <mhayden> 👋🏻
17:00:29 <nirik> morning
17:00:37 <Eighth_Doctor> .hello ngompa
17:00:38 <sgallagh> .hi
17:00:38 <zodbot> Eighth_Doctor: ngompa 'Neal Gompa' <ngompa13@gmail.com>
17:00:41 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
17:00:54 <zbyszek> So we have quorum.
17:01:14 <zbyszek> Let's dive right in.
17:01:18 <zbyszek> #topic #2883 Change: Ostree Native Container (Phase 2, stable)
17:01:18 <zbyszek> .fesco 2883
17:01:19 <zodbot> zbyszek: Issue #2883: Change: Ostree Native Container (Phase 2, stable) - fesco - Pagure.io - https://pagure.io/fesco/issue/2883
17:02:09 <nirik> I'm in favor of this, although I am not sure I understand all the implications. :)
17:02:28 <zbyszek> Same for me.
17:02:42 <sgallagh> Also, same
17:03:49 <mhayden> there's a *lot* in that change proposal
17:04:06 <sgallagh> I suggest we approve it and reserve the right to veto anything that turns out to cause a lot of problems.
17:04:27 * mhayden would love to see a visual (diagram?) or two 🤞🏻
17:04:35 <DaanDeMeyer[m]> (For the frame pointer proposal, can we delay that by one more week, we're trying to determine why Python slows down so much more than the rest of the benchmarks)
17:05:57 <zbyszek> DaanDeMeyer[m]: sure, we can do that.
17:06:25 <mhroncok> .hello churchyard
17:06:26 <zodbot> mhroncok: churchyard 'Miro Hrončok' <mhroncok@redhat.com>
17:06:46 <mhroncok> sorry, package delivery exactly when the meeting was about to start
17:07:10 <nirik> so, yeah, I guess +1 and if anything blows up that we didn't understand would blow up, revisit it... ?
17:07:28 <sgallagh> +1
17:07:50 <mhroncok> 0, wasn't able to dedicate my time looking into this yet, sorry
17:08:57 <zbyszek> 0, I'd need to re-read the proposal before voting, sorry.
17:09:33 <mhayden> 0, read it but would like to see some kind of visual representation to help me understand it better
17:10:30 <zbyszek> Hmm, I'd prefer to instead ask people to vote in the ticket.
17:10:44 <zbyszek> And if there aren't enough votes, we vote on the next meeting.
17:10:49 <mhayden> 👍🏻
17:11:48 <zbyszek> OK, any other votes?
17:12:00 <zbyszek> Oh, wait, it's 5 already.
17:12:15 <zbyszek> Eighth_Doctor didn't vote.
17:12:27 <Eighth_Doctor> 0
17:12:36 <zbyszek> With +2,4,0 it wouldn't pass.
17:12:40 * nirik has no objection to waiting more if we have questions.
17:12:49 <sgallagh> That's fine with me also
17:12:57 <Eighth_Doctor> I simply don't understand the implications of this proposal
17:12:59 <zbyszek> #agreed Everybody is to re-read the prosal and vote in the ticket.
17:13:20 <zbyszek> #info If that fails, we'll vote next week in the meeting.
17:13:32 <zbyszek> #topic #2887 move IMA RPM file-signature change to F38?
17:13:32 <zbyszek> .fesco 2887
17:13:33 <zodbot> zbyszek: Issue #2887: move IMA RPM file-signature change to F38? - fesco - Pagure.io - https://pagure.io/fesco/issue/2887
17:13:37 <Eighth_Doctor> and one of them about chaining everything on top of FCOS concerns me because I don't understand the implications of what layered products wanting to remove stuff does
17:14:14 <zbyszek> Eighth_Doctor: that kind of question is more suited for the mailing list… Maybe it's time to revive the thread.
17:14:26 <Eighth_Doctor> probably
17:14:56 <zbyszek> So… for IMA, the big question is whether something user-visible happened.
17:15:07 <nirik> I'm not sure what to do here. It looks like the ima signing might not be working, but I don't understand it well enough to know whats broken. ;(
17:15:31 <zbyszek> If all the parts that were done are in the backends and tooling and there's no user-visible change, then we can (and must) just mark the change as not-implemented.
17:15:40 <zbyszek> I.e. move it to F38 I guess.
17:15:54 <nirik> yeah, seems so...
17:15:58 <Eighth_Doctor> so we aren't doing IMA signing atm?
17:16:13 <zbyszek> And in the future, there's a lesson for us: if the change is described so sparsely that we can't even say if it happened, then we need to reject the change.
17:16:39 <Eighth_Doctor> yeah
17:16:47 <nirik> well, I think we are, but it's somehow not getting out to the end rpms in the composes...
17:16:54 <mhroncok> The "how to test" section should really have a section that allows us or bcotton_ to verify
17:17:04 <music[m]> .hello music
17:17:05 <zodbot> music[m]: music 'Benjamin Beasley' <code@musicinmybrain.net>
17:17:06 <nirik> I am not sure if thats because koji isn't doing the right thing, or sigul or what
17:17:21 <Eighth_Doctor> I guess it makes sense to defer to f38
17:17:23 <mhroncok> "You can verify that a signature has been put in place by looking at the extended attribute by running: getfattr -d -m security.ima /usr/bin/bash (change /usr/bin/bash with the file to check). "
17:17:31 <Eighth_Doctor> by virtue of the fact it gives time to fix it
17:17:36 <zbyszek> The rpms are rewritten during signing… Maybe the signatures get lost then?
17:18:04 <zbyszek> mhroncok: I tried that.
17:18:07 <nirik> koji uses detached signatures and writes out signed copies as requested.
17:18:26 <nirik> so, might be that part thats not right, or might be the detached sigs or...
17:19:15 <zbyszek> proposal: Change doesn't seem to have been implemented. It can be resubmitted for F38 with more details.
17:19:23 <nirik> +1
17:19:56 <mhayden> +1
17:20:14 <sgallagh> +1
17:20:25 <music[m]> +1
17:20:35 <zbyszek> Eighth_Doctor?
17:20:39 <mhroncok> But the change owners say it is implemented
17:20:48 <mhroncok> surely they know somethign we don't
17:21:08 <Eighth_Doctor> +1
17:21:23 <nirik> well, I don't know that we should talk about implementing... rather working. It was implemented, but it's clearly not working right.
17:21:55 <mhroncok> +1
17:21:56 <mhroncok> sigh
17:22:15 <zbyszek> #agreed Change doesn't seem to have been implemented. It can be resubmitted for F38 with more details.
17:22:19 <zbyszek> #undo
17:22:19 <zodbot> Removing item from minutes: AGREED by zbyszek at 17:22:15 : Change doesn't seem to have been implemented. It can be resubmitted for F38 with more details.
17:22:34 <zbyszek> #agreed Change doesn't seem to have been implemented. It can be resubmitted for F38 with more details (+6, 0, 0)
17:22:43 <zbyszek> #topic #2817 Change proposal: Add -fno-omit-frame-pointer to default compilation flags
17:22:46 <zbyszek> https://pagure.io/fesco/issue/2817
17:23:08 <Eighth_Doctor> I think we're pretty much set to reject this one
17:23:10 <zbyszek> So… DaanDeMeyer[m] asked for this to be moved to next week.
17:23:55 <Eighth_Doctor> this whole thing makes me wish we could do OBS-style automatic alternate publish targets with tweaks like this
17:24:11 <zbyszek> FWIW, I'm thinking that we might want to enable this for F38 with the proviso that if it turns out to cause a noticable overhead, we'll discuss reverting for F39.
17:24:17 <Eighth_Doctor> but then we also need tracked automatic rebuilding, which we also don't have... and arrgh
17:24:57 <mhroncok> See also https://lists.fedoraproject.org/archives/list/python-devel@lists.fedoraproject.org/thread/6TQYCHMX4FZLF27U5BCEC7IFV6XNBKJP/
17:26:15 <zbyszek> But I haven't read all the latest comments, so if nobody objects, I'd skip the discussion today.
17:26:24 <zbyszek> #info https://lists.fedoraproject.org/archives/list/python-devel@lists.fedoraproject.org/thread/6TQYCHMX4FZLF27U5BCEC7IFV6XNBKJP/
17:26:25 <Eighth_Doctor> here's the problem I have: are we deciding to make a Change that guarantees a 2%~10% perf loss on the possibility someone might work on the codebases to restore that?
17:27:01 <Eighth_Doctor> because unless someone is outright committing to working to restore the performance lost, I don't think this is a good idea :/
17:27:15 * nirik nods.
17:27:16 <Eighth_Doctor> but I am willing to be convinced otherwise
17:27:47 <zbyszek> Eighth_Doctor: various features that make debugging and introspection easier have a cost: minidebuginfo, -debug packages, package notes, using -O2 instead of -O3, etc.
17:28:09 <Eighth_Doctor> we use -O2 instead of -O3 because -O3 breaks packages too much
17:28:20 <Eighth_Doctor> otherwise, I think we would use it
17:28:42 <Eighth_Doctor> we already don't do super-well in benchmarks, I'd rather not make it worse :(
17:28:58 <Eighth_Doctor> and there's also downstream implications
17:29:22 <zbyszek> Eighth_Doctor: I'm willing to accept the argument that this makes benchmarking significantly more profitable and can result in much bigger savings.
17:29:22 <Eighth_Doctor> this doesn't really do a ton of good if prominent downstreams rebuild Fedora without it because the performance lost is too much to bear
17:29:39 <zbyszek> Anyway… OK to move to the next topic?
17:29:53 <nirik> +1
17:30:09 <zbyszek> Everyone's favorite!
17:30:10 <zbyszek> #topic Next week's chair
17:30:22 <music[m]> I don’t know what to make of the Python upstream recommendation. As I said in the ticket, for me 1% is probably worth it, 2-3% is hard because it is obviously worth it to some and obviously unacceptable to others depending on background and priorities. I don’t think I could consider voting for it without understanding why Python is so much worse than 2-3% because I’m afraid there are other applications or ecosystems that could
17:30:22 <music[m]> be similarly affected.
17:30:40 <mhayden> Next week is Thanksgiving week here in the USA 🦃
17:30:43 <sgallagh> I'm out next week, as I expect many others in the US will
17:30:47 * nirik will not be here next week.
17:31:19 <zbyszek> dcantrell, Eighth_Doctor are US-based too.
17:31:25 <Eighth_Doctor> zbyszek: I'm willing to accept it if I thought it would be used
17:31:25 <Eighth_Doctor> the problem is: I don't think it will
17:31:43 <Eighth_Doctor> lol
17:31:44 <Eighth_Doctor> laggodoom
17:31:52 <Eighth_Doctor> but yes Thanksgiving next week :D
17:32:06 <Eighth_Doctor> (wifi in the office is half-broken, hence this happening)
17:32:29 <zbyszek> #info Next week is Thanksgiving in US. Next meeting will be in two weeks.
17:32:38 <zbyszek> OK, volunteers to chair in two weeks?
17:32:38 <mhroncok> yay!
17:32:39 <mhroncok> DST?
17:33:07 <zbyszek> So… we agreen on 17:00 UTC.
17:33:11 <zbyszek> *agreed
17:33:25 <Eighth_Doctor> it apparently means lunchtime here
17:33:32 <zbyszek> I'm fine with moving the meeting, if people want to.
17:33:39 * Eighth_Doctor shrugs
17:33:42 <Eighth_Doctor> no strong feelings
17:34:06 <nirik> this is actually better for me
17:34:36 <zbyszek> I think people in Europe prefer that too, it's always nicer not to end at 20:00.
17:34:37 <music[m]> This is trivially worse for me but still OK.
17:34:39 <Eighth_Doctor> I just eat lunch later now I guess
17:34:54 <Eighth_Doctor> or if I'm lucky enough to get food before this
17:35:08 <Eighth_Doctor> then I can eat and watch text scroll by
17:35:12 <zbyszek> #info We will keep the meeting time at 17:00 UTC for now, which means 18:00 in Europe.
17:35:21 <zbyszek> So… volunteers for next chair?
17:35:34 <mhroncok> "in Europe" might be confusing for some
17:35:51 <zbyszek> #undo
17:35:51 <zodbot> Removing item from minutes: INFO by zbyszek at 17:35:12 : We will keep the meeting time at 17:00 UTC for now, which means 18:00 in Europe.
17:35:55 <zbyszek> #info We will keep the meeting time at 17:00 UTC for now, which means 18:00 in EU.
17:35:57 * mhroncok checks the calendar
17:35:57 <sgallagh> I'll take it in two weeks
17:36:32 <zbyszek> #action sgallagh will chair the next meeting (Nov 29th, 17:00 UTC)
17:36:38 <zbyszek> #topic Open Floor
17:36:43 <zbyszek> sgallagh: thanks!
17:37:03 <mhroncok> zbyszek: there are three timezones in the EU (don't undo it, it's fine)
17:37:04 <sgallagh> No problem
17:37:34 <zbyszek> mhroncok: Hmm, I forgot about Ireland.
17:37:40 <zbyszek> Oh, and Greece and so on.
17:37:47 <zbyszek> OK, that was stupid.
17:38:28 <zbyszek> OK, if nobody has nothing, I'll close in a minute.
17:38:57 <mhroncok> zbyszek++
17:38:57 <zodbot> mhroncok: Karma for zbyszek changed to 3 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
17:39:14 <zbyszek> See y'all in two weeks.
17:39:15 <zbyszek> #endmeeting