19:00:03 #startmeeting FESCO (2021-08-23) 19:00:03 Meeting started Mon Aug 23 19:00:03 2021 UTC. 19:00:03 This meeting is logged and archived in a public location. 19:00:03 The chair is defolos. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:03 Useful Commands: #action #agreed #halp #info #idea #link #topic. 19:00:03 The meeting name has been set to 'fesco_(2021-08-23)' 19:00:15 #meetingname fesco 19:00:15 The meeting name has been set to 'fesco' 19:00:28 #chair nirik, decathorpe, zbyszek, sgallagh, mhroncok, dcantrell, defolos, mboddu, Conan_Kudo, Pharaoh_Atem, Son_Goku, King_InuYasha, Sir_Gallantmon, Eighth_Doctor 19:00:28 Current chairs: Conan_Kudo Eighth_Doctor King_InuYasha Pharaoh_Atem Sir_Gallantmon Son_Goku dcantrell decathorpe defolos mboddu mhroncok nirik sgallagh zbyszek 19:00:35 .hello ngompa 19:00:36 Eighth_Doctor: ngompa 'Neal Gompa' 19:00:39 .hello2 19:00:40 dcantrell: dcantrell 'David Cantrell' 19:00:41 .hello2 19:00:42 .hello2 19:00:43 decathorpe: decathorpe 'Fabio Valentini' 19:00:46 .hi 19:00:46 defolos: defolos 'Dan ČermΓ‘k' 19:00:49 dustymabe: dustymabe 'Dusty Mabe' 19:00:55 .hello churchyard 19:00:56 mhroncok: churchyard 'Miro Hrončok' 19:00:59 .hello sgallagh 19:01:00 sgallagh: sgallagh 'Stephen Gallagher' 19:01:03 morning 19:01:09 .hello2 19:01:10 .hello kevin 19:01:10 bcotton: bcotton 'Ben Cotton' 19:01:13 nirik: kevin 'Kevin Fenzi' 19:01:17 #topic init process 19:01:24 I'm here for any votes that need me, but I'm not completely following along. 19:01:45 no worries sgallagh 19:02:01 who's missing? 19:02:08 Please feel free to ping me if any topics clearly need my attention :) 19:02:20 Mohan? 19:02:28 please speak up if you are missing 19:02:44 xD 19:02:48 mboddu: 19:03:10 he is on vacation this week 19:03:11 well, let's start with the light stuff 19:03:46 #topic #2659 Arbitration request: Crypto policy prevents VPN connections 19:04:01 Just for completenes, since it's still tagged with meeting 19:04:14 the authors wrote in the ticket that they don't need further guidance 19:04:37 anyone got anything to add to that ticket or are we all happy with the current state? 19:05:02 I wouldn't say "happy" but meh 19:05:05 nirik: ack, Mohan is considered excused πŸ˜‰ 19:05:42 I'm "meh" about it too 19:05:52 I'm not particularly happy about the situation 19:05:57 but we can live with it 19:06:36 yeah. 19:07:02 so untag from meeting and close the ticket? 19:07:33 or just close... 19:07:51 any objections? 19:07:58 none from me 19:08:37 no objections. 19:08:50 #agree ticket can be closed 19:09:28 hm, was that right? 19:09:37 didn't get a reply from zodbot 19:09:46 zodbot doesn't reply to #agree/#agreed 19:09:56 ah, ok 19:09:59 or most other # commands 19:10:02 #topic #2661 Add fedora-third-party-refresh.service to 90-default.preset for rpm-ostree 19:10:21 also for completeness here, there's +6 on the ticket and no 0 or -1 19:10:27 so if you don't like it scream 19:10:37 otherwise we can approve it 19:10:53 .fesco 2661 19:10:55 defolos: An error has occurred and has been logged. Please contact this bot's administrator for more information. 19:11:28 O.o 19:11:58 did I vote on this ticket? if not, count me as +1 ... 19:12:31 Fabio Valentini (decathorpe): you did not 19:12:39 consider me 0, I was not following the tickets this week much due to life 19:12:49 sorry about the ticket thing. :( 19:12:59 this seems to have enough support already, so no need for me to go and read it properly 19:13:04 zodbot moved to python3 and there's some rough edges. ;( 19:13:19 nirik: wow :) 19:13:26 nirik: no worries 19:14:41 ok, so Fabio Valentini (decathorpe) is in favor, so this can be considered approved 19:15:09 #agree no objections, will be approved 19:15:20 #action defolos to close the ticket as approved 19:15:51 #topic #2662 MinGW debug symbols location change 19:16:18 there hasn't been any comments on that ticket 19:16:31 afaik this is a deferred change proposal from f35 19:16:51 I'm fine with this change 19:16:51 shall we vote on that here or in the ticket? 19:17:01 since nobody voted in the ticket, we might as well do it here 19:17:39 https://pagure.io/fesco/issue/2662 19:18:16 I must confess that I have no idea about mingw 19:18:16 +1 here 19:18:46 same reason as before, consider me 0 (with inclination to approve if my vote will be needed) 19:19:12 I'm fine with this change, +1 from me 19:19:15 +1, I assume smani knows what they're doing 19:19:24 +1 19:19:27 and the proposal sounds fine 19:20:00 sgallagh: anything from your side? 19:20:34 +1 19:20:51 I'll throw a +1 as well 19:21:15 that makes it a +6 and a 0 19:21:23 was that enough to approve it? 19:22:16 defolos: yes 19:22:33 defolos: 5 or more +1s make it approved here 19:22:58 #agree DECISION (+6, 1, 0) 19:23:19 #topic #2663 Start signing repodata 19:23:37 dmach asked here about starting to sign repodata 19:23:44 imho that sounds like a good idea 19:24:02 but I'm not sure if there is anything for us to do here before the infra is in place 19:24:21 nirik: you might have more insight into this? 19:24:50 sorry, got distracted, lets see. 19:25:03 IMHO this shouldn't be a fesco ticket. 19:25:22 not sure what to say here other than "it would be nice to have this, but needs people to implement the missing bits" - and FESCo can't force people to do this. 19:25:26 it's up to interested parties to coordinate/get this setup, then submit a change proposal like anything else? 19:25:43 * nirik nods 19:25:56 I agree, this should be submitted as a Change with interested parties volunteering to do the work on it 19:26:18 * bcotton goes mobile to get the kids off of the school bus 19:26:26 or, ideally, with the implementation already done and the change was about enabling it by default. 19:26:42 That being said, if we have a strong opinion that it should NOT be done, we should chime in before making people write up a Change proposal 19:26:53 IMHO it's not at all ready to enable by default 19:27:12 well, then making it optional would be a first step? 19:27:13 ok, anyone has strong objections against having this enabled if the infra is put into place? 19:27:13 I mean we should/can sign the repodata, but we need a while after that for dnf to work well with that setup 19:27:50 I agre with "coordinate/get this setup, then submit a change proposal" 19:27:57 there's serveral horrible interface issues with dnf and signed repodata. One of the bugs was a release blocker a while back... until thats fixed we should not enable it by default. 19:28:27 nirik: would you mind to follow up with dmach on this topic? 19:28:52 defolos: well, sure, I think the work needed is known... just needs work doing it. 19:30:29 so I see the concensus as "get this setup and then send in a change proposal"? 19:30:29 defolos: That's what it looks like to me 19:30:35 #agree proposal sounds reasonable, should be coordinated properly and then resubmitted as a change proposal 19:31:07 and close the ticket I assume? 19:32:13 yep 19:32:28 #action defolos to comment on the ticket and close it 19:32:48 #topic #2624 Ban bots from submitting non-scratch builds 19:33:09 Fabio Valentini (decathorpe): there has been some movement there recently, anything that needs to be done her? 19:33:14 s/her/here 19:36:23 hm, I guess we lost Fabio 19:36:23 I think the last few comments were on some new issue... 19:36:23 but related. 19:36:23 I don't think there's anything else we need to do there just yet 19:36:23 it looks like lsm5 is taking care of it 19:36:23 I agree 19:36:23 nope I'm here 19:36:23 but it should be fine ... 19:36:23 yeah, I think we talked about having some policy... 19:36:27 ok, so no action needed at the moment 19:36:35 "document-it" 19:36:44 #agree no action needed at the moment besides documenting it 19:36:54 * nirik fixed the ticket/url thingie.. 19:37:10 and now for the elephant in the room 19:37:29 #topic #2660 F35 incomplete changes: Code complete (testable) deadline 19:37:41 .fesco 2660 19:37:42 nirik: Issue #2660: F35 incomplete changes: Code complete (testable) deadline - fesco - Pagure.io - https://pagure.io/fesco/issue/2660 19:37:56 nirik++ 19:37:57 defolos: Karma for kevin changed to 48 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 19:38:06 thanks nirik! 19:38:12 #topic GNU toolchain update 19:38:37 tracking bug rhbz#1982744 has been set to ON_QA, so this appears to be done properly 19:39:31 anyone knows of anything that would speak against that? 19:39:49 .bug 1982744 19:39:51 defolos: 1982744 – GNU Toolchain update (gcc 11, glibc 2.34, binutils 2.37, gdb 10.2) - https://bugzilla.redhat.com/1982744 19:40:09 I think it's done yeah 19:41:04 it looks done to me as well 19:41:10 so let's consider that one done then 19:41:16 any objections? 19:41:54 none from me 19:42:01 * nirik has none 19:42:07 Ben Cotton (he/him/his): apparently this room doesn't have a public MXID 19:42:19 nor is it published like the other f-m rooms 19:42:24 I can't link to it to anyone 19:42:38 #agree the change appears to be done 19:42:52 #topic CompilerPolicy Change 19:43:21 #info the devel thread resulted in the suggestion for FPC to co-maintain redhat-rpm-macros 19:43:39 fpc#1066 is however still open 19:43:56 how shall we proceed here? 19:44:10 the meeting is on Thursday 19:44:19 Conan Kudo: mhroncok Fabio Valentini (decathorpe) you're in the FPC, right? 19:44:20 I won't be there but I voted +1 in the ticket 19:44:26 yes 19:44:28 well, %cc is broken, so as is, it is not acceptable 19:44:29 defolos: yes 19:44:41 it is tagged with meeting and should get approved on that meeting, if there is quorum 19:44:42 Fabio Valentini (decathorpe): I already merged the fix to use `%build_cc` 19:44:52 though rpm upstream is fixing the macro name restrictions 19:45:04 oh, good. then then fpc PR needs to be updated? 19:45:09 yes 19:45:15 that's why I didn't merge it yet 19:45:18 marvelous 19:45:59 but I guess this will not get in before beta freeze then? 19:46:06 (that's tomorrow) 19:46:30 fpc docs are independent of implementation (which is in already) 19:46:35 the guidelines don't need to follow beta freezes 19:46:39 so I wouldn't worry about it 19:46:49 they should follow rawhide anyway 19:46:55 yup 19:47:12 ok, so this is done then? 19:47:26 (for all practical purposes) 19:47:30 yes 19:48:05 #info implementation is in place, documentation will be updated soon 19:48:18 #agree implementation considered done 19:48:31 Conan Kudo: room should have a public address now 19:48:35 #topic Golang 1.17 19:48:41 Ben Cotton (he/him/his): thanks! 19:49:07 soooo, apparently we got a build of 1.17 in rawhide and a ton of build failures in the copr 19:49:31 Hi! I'm one of the proposal owners :) 19:49:33 :( 19:49:41 alexsaezm: can you tell us the status? 19:51:02 I made a COPR project with all of the packages in rawhide and Go 1.17 (which now is on rawhide, but it wasn't when I made the initial build): https://copr.fedorainfracloud.org/coprs/alexsaezm/go-rawhide/builds/ 19:51:09 from ~1800 packages, 228 failed 19:51:36 a lot of them have failing bugs filled 19:51:48 For context: https://copr.fedorainfracloud.org/coprs/alexsaezm/go-rawhide/builds/ 19:52:22 Some of the packages have an easy fix (example: tinygo, it depends on LLVM 12) but others have weird issues 19:52:42 like golang-github-prometheus 19:52:43 thats quite a lot of failures. ;( 19:52:48 some of those failed builds have newer succeeded ones 19:52:52 e.g. weldr-client 19:52:53 so, too late for beta freeze, right? 19:53:03 Eighth_Doctor, yes, that is true 19:53:20 Didn't check the numbers now that some maintainers fixed some issues 19:53:42 I can tell from the errors that they are not related with the update in most of the cases but I can't guarantee it 19:53:47 well, also we have branched, so f35 has diverged. 19:54:58 can we get this in after release? 19:55:03 are any of the 228 faiilures critical path? 19:55:06 or is that too disruptive 19:56:24 lot of the failing packages are dependencies (based on the naming ) but tbh I don't know how to evaluate the impact 19:57:01 perhaps we could get adamw to test against openqa and see if it would break composes in rawhide or not... 19:57:23 I don't see any critical path packages in https://copr.fedorainfracloud.org/coprs/alexsaezm/go-rawhide/packages/ 19:57:33 err I don't see any failing critical path packages 19:57:35 if it doesn't, I'd say land in rawhide asap... but I think f35 ship has sailed and the change should move to f36 19:57:41 nirik: this is rather easier if there's an update or at least something in koji 19:57:51 nirik, 1.17 is on rawhide since past week :) 19:57:56 i could bodge it up for a copr repo, maybe, though, i guess 19:57:59 kinda busy this week 19:58:00 adamw: yeah, I guess a side tag would be good at this point? 19:58:11 since we have to do that anyhow... 19:58:49 but it's too late for f35, or do we want to risk it? 19:58:56 it's too late for f35 19:59:26 anyone objecting? 19:59:43 sgallagh: ? 20:00:05 * Eighth_Doctor shrugs 20:00:11 I'd prefer to see it in F35 20:00:18 but ehh 20:00:39 looking 20:01:01 Eighth_Doctor: me too, but beta freeze is tomorrow :( 20:01:15 Too late for F35, but it could go in as a module... 20:01:25 he said the M-word! 20:01:26 *too late to be default for F35, I should say 20:01:27 why can't these just go out as updates for F35? 20:01:35 πŸ˜‰ 20:01:53 dcantrell: It changes the default Go compiler after Beta Freeze 20:01:57 Which is generally a bad idea 20:02:13 and it would require a huge rebuild 20:02:21 if Daniel Mach's modularity improvements are available, it might be workable as a module 20:02:28 with chances for even more breakage 20:02:34 the problem with language stack modules right now is we can't ever get rid of them 20:03:05 Let's not get sidetracked by the module argument. 20:03:06 then I say this stack has missed F35 if it's that critical 20:03:28 I think we're mostly in agreement that it's too late to be the default Go 20:03:45 (in F35, of course) 20:03:54 so let's settle on "defer to f36, change owners can consider adding a module for f35"? 20:04:02 f36 will be along soon... no need to panic. ;) 20:04:16 not sure how to do the module path, but I'll check it :) 20:04:38 * nirik is not in favor of adding modules, but you're welcome to do as you like there. 20:04:50 ask sgallagh, he does it for nodejs 20:04:53 alexsaezm: It might be worthwhile for the compiler tools, but probably NOT for all the packages. 20:04:59 #agree defer to F35, too late for beta freeze 20:05:06 eehhh? 20:05:12 don't you mean defer to F36? 20:05:19 :) 20:05:30 someone is counting with an early production run Pentium, I see 20:05:32 #undo 20:05:32 Removing item from minutes: AGREED by defolos at 20:04:59 : defer to F35, too late for beta freeze 20:05:48 #agree defer to F36, too late for beta freeze 20:05:52 thanks Conan Kudo 20:05:57 Conan Kudo++ 20:06:06 #topic LLVM13 20:06:07 ha 20:06:11 tstellar: around? 20:06:27 anyone knows the state of this? 20:06:53 I've seen some chatter around breakages due to an incomplete side tag merge or something related to llvm13 20:07:03 I think it's done... 20:07:08 yeah I think I saw something like that over the weekend 20:07:16 there was one package in the side tag that was mistagged somehow.. it was fixed saturday? 20:07:26 yep 20:07:34 I see llvm 13.0.0~rc1 in f35: https://koji.fedoraproject.org/koji/buildinfo?buildID=1814364 20:07:58 so I'd say this is probably done 20:08:02 yep. 20:08:27 Eighth_Doctor: rc1? 20:08:32 not a final release? 20:08:40 final release doesn't happen for a while 20:08:48 it's usually fairly close to final freeze 20:09:26 historically, we land the RCs as early as we can, since that's after ABI freeze for an LLVM release 20:09:38 and either wind up shipping the final release just before or just after GA 20:09:41 defolos, I pinged him just right now 20:09:54 not sure if he is available 20:10:09 LLVM is pretty out of sync of Fedora, so it's a little on the painful side to depend on 20:10:16 but we need it for Mesa 20:10:20 "Continue packaging newer release candidates into rawhide and f35 until the final release is complete (Late September 2021)" 20:10:27 ah 20:10:29 this was part of the change 20:10:34 #info https://bugzilla.redhat.com/show_bug.cgi?id=1982397#c2 20:10:37 https://fedoraproject.org/wiki/Changes/LLVM-13 20:10:54 #info all LLVM13 packages have been tagged into f35 20:11:03 (from the ticket) 20:11:08 so this is done then 20:11:21 agree? 20:11:29 * nirik nods. I think so yes. 20:12:08 alexsaezm: Thanks, he clarrified it in the bugzilla tracker bug 20:12:29 #agree done according to the tracker bug 20:12:45 #topic Introduce Module obsoletes and EOL 20:12:52 nothing happened there 20:12:58 Daniel Mach is here to talk about that 20:13:05 I guess that mcurlej is probably on vacation 20:13:18 Daniel Mach: mach set a needinfo on the bugzilla bug 20:13:29 but didn't get a reply yet 20:13:40 .bug 1834844 20:13:42 defolos: 1834844 – Introduce module Obsoletes and EOL - https://bugzilla.redhat.com/1834844 20:14:29 well, it's pretty clear this ain't landing in f35 20:14:39 question would be: defer or reject? 20:14:48 well, all the tech has landed 20:14:52 just not the docs 20:15:08 and the policies 20:15:27 right, but neither of those two are actually tied to Fedora release criteria 20:16:49 any opinions about this besides Neal? 20:17:18 I don't see it done. there has been no update whatsoever about this anywhere 20:17:27 as an observer, it appears ver not done to me 20:17:31 *very 20:18:01 OTOH it doesn't really matter how do we call it, if it is already done :D 20:18:26 I'd suggest to defer this and not reject as Conan Kudo says that the infra is in place 20:18:36 defolos: +1 20:19:09 I can query the status on the internal modularity meeting if it is not cancelled 20:19:27 mhroncok: that would be great 20:19:42 it's on Wednesday 20:19:58 #action mhroncok to ask internally about the status 20:20:27 unless anyone objects I'll write down that we agree on deferin 20:20:37 sgallagh: ☝️ 20:20:45 sounds good 20:21:32 defolos: how many more items on the agenda? 20:21:43 that's more or less the last one 20:21:54 #agree defer to f36 20:22:11 #topic Enhanced Inscript as default Indic IM 20:22:13 just ftr, this is done 20:22:27 #info tracker bug is on ON_QA, change owner replied that this is done 20:22:27 in that case, I'll disappear now, the dogs need to be fed 20:22:30 see you later 20:22:37 πŸ‘‹ 20:22:46 #topic open floor 20:23:06 who's not asleep yet and got something to add? 20:23:20 πŸ‘‹ - i'm going to be trying to come to the fesco meetings in the future to help represent Fedora CoreOS 20:23:30 we have some initiatives and ideas we'd like to run through the group 20:23:43 will open tickets and such, when appropriate. Just wanted to say hey. 20:24:01 sounds good, you're always welcome here 20:24:01 wrt repodata signing, there's also an open ticket to improve gpgcheck settings in dnf for this: https://bugzilla.redhat.com/show_bug.cgi?id=1433592 20:25:14 * defolos is already on CC 20:25:40 ah, forgot one thing 20:25:41 #topic Next week's chair 20:25:47 voluntereers? 20:25:54 * volunteers? 20:26:21 I guess I can do it 20:27:16 #action ngompa will chair next meeting 20:27:21 thanks neal! 20:27:50 I'll wait until 20:30 UTC and then close the meeting 20:28:03 (unless something happens of course) 20:28:32 Ben Cotton (he/him/his): what's the process for the changes that we decided to defer to F36? 20:29:28 defolos: just note the decision and i'll take care of the paperwork 20:29:38 * bcotton is a full-service FPgM 20:29:54 nice 20:29:59 thanks Ben Cotton (he/him/his) 20:30:11 bcotton: will you clean my windshield too while you pump gas? 20:30:23 dustymabe: yes! 20:30:37 i've been missing out 20:30:44 do you guys live even remotely close by πŸ˜›? 20:31:00 FSVO "close" 20:31:30 can't do that here... :) 20:31:37 on an astronomical scale πŸ˜‰ 20:31:57 well, as this is OT now 20:31:57 #endmeeting