15:00:55 #startmeeting FESCO (2020-12-02) 15:00:55 Meeting started Wed Dec 2 15:00:55 2020 UTC. 15:00:55 This meeting is logged and archived in a public location. 15:00:55 The chair is sgallagh. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:55 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:55 The meeting name has been set to 'fesco_(2020-12-02)' 15:00:55 #meetingname fesco 15:00:55 #chair nirik, ignatenkobrain, decathorpe, zbyszek, sgallagh, mhroncok, dcantrell, cverna, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor 15:00:55 The meeting name has been set to 'fesco' 15:00:55 Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku cverna dcantrell decathorpe ignatenkobrain mhroncok nirik sgallagh zbyszek 15:00:55 #topic init process 15:01:03 .hello2 15:01:04 decathorpe: decathorpe 'Fabio Valentini' 15:01:06 .hello2 15:01:06 .hello2 15:01:07 sgallagh: sgallagh 'Stephen Gallagher' 15:01:10 bcotton: bcotton 'Ben Cotton' 15:01:11 .hello2 15:01:11 .hello ngompa 15:01:16 dcantrell: dcantrell 'David Cantrell' 15:01:17 morning 15:01:18 King_InuYasha: ngompa 'Neal Gompa' 15:01:45 I am kind of around o/ 15:01:48 morning... here to answer any questions you may have... 15:02:00 jlaw: What is the meaning of life? 15:02:08 42, of course 15:02:41 Well, I think that settles things 15:02:56 jlaw: why do fools fall in love? 15:03:30 I'll give folks two more minutes to appear. 15:03:36 :-) maybe they're not fools if they're in love? :-) 15:03:37 .hello2 15:03:37 mhroncok let me know he'd be unavailable 15:03:37 zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' 15:03:57 . 15:04:32 OK, I think we can get started. 15:04:36 Short agenda today. 15:04:46 #topic #2507 F34 Change: GNU Toolchain update (gcc 11, glibc 2.33) 15:04:46 .fesco 2507 15:04:47 sgallagh: Issue #2507: F34 Change: GNU Toolchain update (gcc 11, glibc 2.33) - fesco - Pagure.io - https://pagure.io/fesco/issue/2507 15:05:42 * nirik was +1 in ticket... I guess we just stamp this unless someone has a issue? 15:05:50 We already have +7 on the ticket, but I requested that we put it on the agenda to finalize the decision so the toolchain folks can get this in place quicker. 15:05:53 The only misgivings I have about this is that glibc 2.33 breaks stuff in weird ways in all kinds of environments :/ 15:06:03 More time in Rawhide == more time to fix issues. 15:06:25 decathorpe: What are you seeing? 15:06:47 .bug 1869030 15:06:49 decathorpe: 1869030 – glibc: Back out glibc-rhbz1869030-faccessat2-eperm.patch workaround for systemd UAPI breakage - https://bugzilla.redhat.com/1869030 15:07:10 ah, yes, that's a mess. Carlos and Florian own that issue 15:07:11 this is already in rawhide, so it doesn't change my vote, but I still think it's bad 15:07:50 Then it's in good hands in my opinion. 15:07:59 * King_InuYasha shrugs 15:08:08 i think they're still negotiating with seccomp and kernel folks on the right path forward 15:08:28 I gave it +1 because I don't expect anything to be any better with this verses the normal case 15:08:41 but as decathorpe notes, that's already in rawhide and we *have* to get it resolved 15:08:48 we've seen this sort of stuff before and really the only decision is forward and fix things 15:08:55 tons of things will break, we'll still have issues with fixing code, and so on 15:08:58 dcantrell: +1 15:09:21 yeah. which is why I still voted +1. just wanted people to be aware of possible pitfalls. 15:09:26 King_InuYasha: I'm optimistic that breaking things earlier is better 15:09:39 it is, as long as someone is paying attention 15:09:43 decathorpe: yeah, it's good to be aware of this sort of thing and to know the temporary workarounds that exist (if any) 15:10:28 decathorpe: absolutely. Jakub wants to be more conservative for gcc than we've been with glibc, but still get the gcc bits in earlier than we have in the past (Now vs Jan/Feb) 15:11:39 if someone is paying attention and we don't have the compiler continuously broken because of annobin<->gcc things, then meh, let's do this 15:11:42 Does anyone want to vote -1 or shall we certify the current vote at +7 ? 15:11:48 i'll note that it would be good to have proposals in before code lands in rawhide. change proposals can be submitted any time (we already have an approved F35 proposal) 15:12:10 King InuYasha: I think we've got the annobin vs gcc stuff sorted out starting with gcc-11 15:12:21 I can go into the details, but they're probably too low level for here 15:12:21 that's great news :-) 15:12:43 yea, it's been a huge headache 15:12:45 yay 15:12:54 :D 15:14:22 sgallagh: there seems to be no disagreement, I think we can proceed. 15:14:44 Anyone present that wants to add their +1 who didn't do so on the ticket? 15:14:44 * nirik nods 15:15:23 Nope, we all have 15:15:45 #agreed F34 Change: GNU Toolchain update (gcc 11, glibc 2.33) is accepted (+7, 0, -0) 15:15:55 We have one late addition to the agenda: 15:16:29 #topic #2501 F34 System-Wide Change: Remove nscd in favour of sssd and systemd-resolved 15:16:29 .fesco 2501 15:16:30 sgallagh: Issue #2501: F34 System-Wide Change: Remove nscd in favour of sssd and systemd-resolved - fesco - Pagure.io - https://pagure.io/fesco/issue/2501 15:17:04 arjun has updated the proposal as requested 15:17:18 I haven't read the updated proposal, doing so now 15:17:26 * nirik is still +1 15:17:33 I'm still gravy with it 15:17:43 jlaw: If you're still here, what is the toolchain's perspective on the topic? 15:18:04 my only feedback here is that the Change page itself should be renamed 15:18:11 since we're not removing nscd anymore 15:18:13 it's something we certainly want to do... arjun is working on Carlos & Florian's behalf on this issue 15:18:34 arjun "moved the needle" completely towards pure deprecation, with no plan for "removal". Should be move the needle back a bit and approve removal in F35? Or should that happen latter, separately? 15:18:44 proposal sounds good to me, changing my vote from 0 to +1 15:19:05 I would like to see a "remove nscd" proposal made for us to approve for F35 15:19:19 I'm somewhat biased, given my history with SSSD, but I'm a firm +1 to shooting nscd in the head, burying it in the woods and then burning the map to it. 15:19:21 King_InuYasha: yeah, that'd be ideal 15:19:24 if the previous proposal was split into two, one for f34 and one for f35, I'd approve it today 15:19:28 yea, if we can't get to removal this cycle, we should queue it up for the next cycle 15:19:56 King_InuYasha: agreed, s/Remove/Deprecate/ on this proposal and a new one to remove for F35 15:19:59 * nirik is fine with removal now, but sounds like others would prefer to wait more to let the info get out 15:20:11 i can relay that to arjun & carlos 15:20:13 jlaw: I think we've agreed *not* to remove it in F34, to give time for users to adapt. 15:20:27 nirik: don't get me wrong, I think we should purge it now, but if people want to slow-roll it, fine 15:20:32 yup, that's my reading from here 15:20:34 yeah. 15:20:55 basically, my ask is that I want a removal proposal ASAP for fesco to approve 15:21:08 slower than the glibc team would prefer, but I think deprecating now and removal in f35 is reasonable 15:21:18 and I'll ask arjun to get the f35 proposal in immediately 15:21:40 jlaw++ 15:21:40 zbyszek: Karma for jlaw changed to 1 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:21:44 the benefit of doing so is that we can get other distros to also start implementing the change and drop their dep on nscd 15:21:52 arjun++ 15:21:52 zbyszek: Karma for submachine changed to 1 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:21:57 which means that dropping nscd from glibc can happen _faster_ 15:22:24 * smooge gets nostalgic about all the flame wars and this is killing Linux when nscd appeared. I should see who many of those are now posting hate emails about it going away 15:22:33 haha 15:22:42 smooge++ 15:22:42 bcotton: Karma for smooge changed to 5 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:22:43 smooge: part of me is absolutely expecting a few threads like that 15:22:54 smooge++ 15:22:54 Conan_Kudo: Karma for smooge changed to 6 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:23:05 I *absolutely* expect it 15:23:20 I mean pulseaudio is now the bee's knees once you say you are going to replace it 15:23:20 I'm happy to write a separate proposal. Actually, maybe I will move the current one to DeprecateNSCD, and edit this back to a removal. 15:23:24 How does that sound? 15:23:24 I am glad I don't have to killall nscd anymore. ;) 15:23:24 but that is sort of the point. marking it deprecated will at least give those users an opportunity to be heard 15:23:26 OK, let's get a final vote: 15:23:26 Proposal: Accept the Change as it stands with today's updates 15:23:33 ah, didn't know you were here arjun 15:23:34 and not be surprised when nscd is actually removed 15:23:43 jlaw: Hi :) 15:24:18 We should probably nudge the PR team to emphasize this impending removal too. 15:24:31 * sgallagh looks pointedly in bcotton's direction 15:24:33 arjun: yeah, I think that'd be better, because the page name is "RemoveNSCD", and that's a bit confusing. 15:24:34 so perhaps we just wait and approve both next week? 15:24:41 arjun: perfect! 15:24:45 unless there is a hurry? 15:25:00 nirik: I'd rather start he communication of this rolling, personally. 15:25:14 well, make the new deprecate page, and mark that approved 15:25:16 that's easy enough 15:25:16 If we know we're going to approve it next week anyway, why wait? 15:25:30 and we can also pre-approve the removal page, knowing it's just the other half 15:25:38 Yes 15:25:40 since we had the original combined proposal 15:25:42 so we know what we are approving exactly, but ok... whatever. 15:25:53 Revised Proposal: Approve deprecation for F34 and removal for F35 15:26:03 sgallagh: +1 15:26:05 sgallagh: +1 15:26:08 +1 15:26:12 +1 15:26:13 +1 15:26:23 +1 15:27:11 #agreed Deprecation of nscd is approved for F34, removal of nscd is approved for F35 (+6, 0, -0) 15:27:15 For my part, I will (a) create a new DeprecateNSCD proposal, for F35. And (b) edit back the current page to be a removal proposal, but for F35. 15:27:26 Sorry, Deprecate for F34. 15:27:35 and update the links on the fesco ticket 15:27:37 sounds good 15:27:41 and probably change the fesco ticket title 15:27:48 #action arjun will split the Change proposal for F34 and F35 and update the FESCo ticket with the new links. 15:28:32 bcotton: Once those appear, go ahead and proceed with the post-acceptance process. 15:28:39 sgallagh: ack 15:28:42 Thanks 15:28:54 i'll still post on devel-announce for transparency 15:28:57 ack 15:29:06 #topic Next week's chair 15:29:41 * sgallagh takes the potato off the grill, lobs it. 15:29:51 as always, if there's a toolchain issue that's not getting the attention it needs, don't hesitate to reach out to me directly and I'll start chasing folks down 15:30:00 law@redhat.com 15:30:45 jlaw++ 15:30:45 sgallagh: Karma for jlaw changed to 2 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:30:57 Is next week after election results? (That's me deflecting the veggie spinning down from above.) 15:30:58 jlaw++ 15:30:58 Conan_Kudo: Karma for jlaw changed to 3 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:31:00 He's not kidding, I've seen his Mandalorian armor 15:31:11 zbyszek: yes, it is 15:31:27 Oh yeah, I need to go vote... 15:31:33 O.o 15:32:00 So it's either ignatenkobrain (who's not here) or King_InuYasha or cverna 15:32:31 next week is a heavy week for me, I'm not even sure if I'll be here or not 15:32:51 (it's my last week of work before PTO to the end of the year) 15:33:36 but if cverna can't, I can do it 15:33:57 I am on PTO starting Dec 7th and returning Jan 4th. Similar to mhroncok, I will check fesco mail and tickets weekly but otherwise can't guarantee I will be around for meetings. 15:34:01 I'll take it again if that's the only real choice. 15:34:08 * bcotton can be a non-member chair if needed 15:34:20 oh right, bcotton exists :) 15:34:36 sorry catching up 15:34:48 King_InuYasha: I'm not convinced 15:34:54 I can chair next week 15:34:57 Oh, thanks 15:35:01 excellent! 15:35:08 #action cverna will chair the next meeting 15:35:14 #topic Open Floor 15:35:23 * bcotton has two things for open floor 15:35:25 * sgallagh peers into the shark tank 15:35:34 bcotton: You have the floor 15:36:11 #info Elections voting ends at 2359 UTC on Thursday: https://elections.fedoraproject.org 15:36:14 after that I wonder if we shouldn't figure out which meetings we are skipping due to holidays. Or should we just wait until after elections? 15:36:15 also 15:36:43 nirik: that'd be probably more effective to do with the new cast 15:36:56 * nirik nods, ok 15:36:57 yeah, I'd say let's wait for next Thursday 15:36:57 Yes, especially if we end up moving the meeting time 15:37:05 err Wednesday 15:37:06 iirc miro set up the labels on the FESCo repo initially, but are there any objections to replacing the F* labels with pagure milestones instead? this way we can keep some of the cruft hidden 15:37:22 i don't know if anyone is using the labels for anything, so i don't want to break it 15:37:36 I think milestones make sense for that 15:37:38 but milestones seem like a better fit for tracking release-specific issues 15:37:40 since they are timelines 15:37:48 and we can attach dates and stuff to close milestones 15:37:48 Yeah, that seems reasonable. 15:38:22 Yeah, milestones are reasonable for this. 15:38:31 sure. although... 15:38:50 it's sometimes hard to use fedora release dates as milestones... since we have alternates/etc 15:39:14 well, no, we wouldn't use that for our milestones 15:39:19 we'd use the change deadline 15:39:24 that's pretty much fixed 15:39:30 That's a good idea 15:39:34 nirik: i'm just thinking of making "f34" a milestone instead of a label. no dates for pagure's purposes 15:39:52 bcotton: I think having the change deadline as a date makes sense 15:39:56 sure, thats fine 15:40:09 (to bcotton ) 15:40:25 if we're not using dates, there's no meaningful reason to switch from labels 15:41:21 King_InuYasha: yes there is. we don't end up with lots of unused labels clogging the ui. and since labels are used for other purposes, that's an advantage 15:41:36 hmm, I guess I hadn't thought of that 15:41:36 also, i think with closed milestones, they don't appear in the drop down (but I could be wrong) 15:41:55 bcotton: Will you convert the existing closed tickets to the milestones? 15:41:59 i mean this is 95% percent for my benefit if i'm being honest 15:42:02 For search purposes 15:42:37 bcotton: at least on Fedora Workstation, they show up here: https://pagure.io/fedora-workstation/roadmap?status=all 15:42:46 sgallagh: that was going to be my next question. i can definitely do that for previously labeled issues (which goes back a few releases). do we want to go back beyond that? (through all 2900+ issues) 15:42:57 bcotton: and they show up in the search filter as inactive milestones 15:43:23 as long as you don't _delete_ the milestone, it's still searchable that way 15:43:30 bcotton: Just the ones that have existing labels is fine. 15:43:51 e.g. https://pagure.io/fedora-workstation/issues?milestone=Fedora+26&status=Closed 15:44:11 sgallagh: ack 15:45:47 King_InuYasha: right, i mean it inactive milestones won't show up in the selection when editing issue metadata so it's cleaner 15:46:25 Ok, anything else for open floor? 15:46:31 so it sounds like there's a general consensus that moving F* from labels to milestones, including updating closed issues is good? 15:46:38 yep 15:47:23 okay, i'll open an issue to show my work :-) 15:48:46 Great 15:48:58 EOF 15:49:03 I’ll close the meeting in two minutes if there is nothing else. 15:49:13 Thanks sgallagh 15:49:26 thanks sgallagh 15:49:35 thanks sgallagh 15:50:21 sgallagh++ 15:50:21 decathorpe: Karma for sgallagh changed to 1 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 15:50:29 good luck with the elections everybody 15:51:04 Thanks for coming, everyone. 15:51:06 thanks all 15:51:09 #endmeeting