19:07:00 #startmeeting 19:07:00 Meeting started Wed Jan 27 19:07:00 2021 UTC. The chair is elbrus. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:07:00 Useful Commands: #action #agreed #help #info #idea #link #topic. 19:07:15 #topic Admin 19:07:25 #info Previous minutes: http://meetbot.debian.net/debian-release/2020/debian-release.2020-12-23-18.58.html 19:07:52 #info ginggs had an action for a patch to britney 19:08:02 * elbrus hasn't seen it ;) 19:08:14 no, sorry :) 19:08:17 #info ginggs had an action to think about a proposal for blocking binNEW in unstable 19:09:13 ginggs: ^? 19:09:17 nothing 19:09:21 #info elbrus had an action to continue the arch qual by mail 19:09:29 done and mail sent afterwards 19:09:39 #info elbrus had an action to list (build-)essentials 19:09:49 done and added to the freeze_policy 19:10:17 #info ginggs had an action to come up with key_package threshold impact 19:10:48 ginggs: did you do anything with the info ivodd added to this chat after the meeting? 19:10:49 i did come up some numbers, and ivodd came up with others 19:11:14 there was a hint to dive into meta packages... 19:11:31 and I think you mentioned a good reason later in private 19:11:46 but I can't straight remember it 19:11:57 i only added it to this meeting's agenda for further discussion 19:12:08 right, lets' have it there then 19:12:14 #topic Transitions 19:12:22 not much left :) 19:12:23 adsb: thanks for the reminder, i totally forgot about it. I'll do that asap 19:12:35 Sebastinas: ^^ 19:12:45 openmpi and pytest are done! 19:13:03 \o/ 19:13:09 quicker than I expected 19:13:28 Sebastinas: you had libxmlb (#981078) 19:14:37 Ah, yes. 19:15:23 But I think it's too late. At first I thought it might be beneficial to do, but that's all stuff we can't easily remove if something goes wrong 19:16:15 any of the more experienced members around? 19:16:23 to judge? 19:16:36 ivodd: around? 19:17:34 Sebastinas: maybe try to catch one of the others if your unsure 19:18:15 are there other things on the transitions front? 19:18:45 * elbrus spotted https://release.debian.org/transitions/html/libgdk-pixbuf-2.0-0.html 19:18:45 Not that I am aware of, but I've also been rather busy @ work this month. 19:19:01 I've added a tracker for that, but that's nothing we need to process. 19:19:15 right, forgot about that "soft transition" 19:19:35 so, looks all good for bullseye 19:19:47 #topic Current state of bullseye 19:20:05 as mentioned a couple of days ago, we have 19:20:06 https://release.debian.org/transitions/html/libgdk-pixbuf-2.0-0.html 19:20:11 arch 19:20:20 https://wiki.debian.org/Teams/ReleaseTeam/ReleaseCheckList/BullseyeCheckList 19:20:21 that 19:20:59 all the tasks in "Before Freeze" can be picked up or tagged if already done 19:21:08 shall we go through them? 19:21:15 or is that a bit much? 19:21:25 I think keys were requested 19:22:48 I suggest when people pick something up, they tag it at the end of the line with [ ] 19:23:42 anybody up to some of those tasks? if it's just me, this release is going to be slowish... 19:24:18 which brings me to the bugs... 19:24:19 https://udd.debian.org/dev/bugs.cgi?release=bullseye_and_sid&merged=ign&rtbullseye-will-remove=ign&rtbullseye-can-defer=ign&autoremovals=ign&deferred=ign&fnewer=ign&fnewerval=30&rc=1&cpopcon=1&chints=1&ctags=1&cclaimed=1&cdeferred=1&caffected=1&crttags=1&sortby=last_modified&sorto=asc&#results 19:24:27 115 on that list now 19:24:40 but quite a few old ones 19:24:55 that needs triaging and $somebody to work on it 19:25:02 we miss BSPs 19:25:07 * elbrus guesses 19:26:37 anybody, comments? (feels like these presentations nowadays just talking to your own laptop and hoping that others are listening and watching) 19:27:17 Sebastinas: you added: - chromium (#972134) 19:28:17 I don't know exactly what you wanted to say on it 19:28:39 Right. chromium came as potentially unsupportable in stable. 19:28:48 I guess that has been resolved. 19:29:11 but as far as I'm conserned, it's the security team that has to be convinced it's maintained by enough people 19:29:52 they expressed their conserns 19:30:15 ACK 19:30:26 I haven't followed the activity in the bug, but I don't think they are convinced by a one time upload (for now) 19:30:35 carnil, jmm_: ^ unless you have more concerns, I guess it will migrate before the freeze 19:30:48 by others than the "normal" maintainer, who did an upload today 19:30:56 I guess it will not 19:31:07 but is still open and RC 19:31:40 and security team should in my opinion agree on closing or downgrading 19:32:08 the bug is closed afaict? it's not fixed, but it's closed. 19:32:14 aha, wait, bug is closed 19:32:28 by latest upload 19:32:44 (#960786 can probably be downgraded or -ignored given builds happen on the buildds) 19:33:02 (from elbrus's UUD list) 19:33:26 nthykier: thanks for your note 19:33:37 nthykier: do we block binary uploads on stable? 19:34:01 if not then that's a recipe to have it break on a security update 19:34:32 jcristau: if not, bullseye might be a good release to start on doing that! ;) 19:35:37 maybe, but until/unless that happens i'm not sure it's advisable to ship with that bug. 19:35:57 Michael Gilbert was gone for a while, but seems back and he closed the task, given that two new people stepped up in helping out with Chromium that seems good enough for now 19:36:09 jmm_: thanks 19:37:13 Sebastinas: ginggs: do you intend to join triaging of open RC bugs? 19:38:02 the further into the freeze, the better view we need to have on the remaining bugs 19:38:17 can we organize a bsp for that? 19:38:18 to judge when we can release 19:38:45 do you want to organise a bsp? (that would be awesome) 19:38:53 or do you mean to work through them now? 19:39:15 I don't mean *we* fix them, just know which ones can be ignored or need to be fixed 19:39:43 now (but not in this meeting) 19:40:04 poking maintainers can sometimes have great results 19:40:12 and sometimes not by the way 19:40:21 I'll have more time again starting from mid February can then take a loook at some of them. 19:40:34 Sebastinas: any help is welcome 19:41:04 there's a couple of bts tags only we are allowed to set 19:41:11 and we have some usertags as well 19:42:12 for discussion on meetings or maybe even a sprint I used private usertags in the past to group stuff I looked into but didn't feel comfortable yet in official tagging 19:42:58 other item we need to handle is: 19:43:06 #topic Release Notes 19:43:31 last time I did most of it from our side, before that nthykier did most I think 19:43:46 anybody is free to help of course 19:44:13 there's ideas to convert it to MarkDown 19:44:24 does anybody have experience doing that? 19:44:43 I think that needs to happen soon or after the release 19:44:55 but the current format was difficult to get right 19:45:10 and adds burden to adding sections or chapters 19:45:37 I would love if some people would at least subscribe to the bts for those bugs 19:45:45 and/or to the salsa repo 19:46:10 https://bugs.debian.org/release-notes 19:46:27 Is https://salsa.debian.org/ddp-team/release-notes the salsa repo? 19:46:28 https://salsa.debian.org/ddp-team/release-notes/ 19:46:31 yes 19:47:37 I'm willing to work on it again this time, but I like help 19:48:17 i can help you 19:48:22 \o/ 19:49:32 I can't promise anything meaningful, but can at least get subscribed to the right things. Maybe that'll push me to help... 19:49:47 \o/\o/ 19:50:13 #topic popcon threshold for key packages (again) 19:50:18 ginggs: ^ 19:51:32 so (again) this is just shall we increase the popcon threshold (currently 5%) for key packages 19:51:49 what was your argument again? 19:51:53 ivodd provided some numbers, which i'll paste here 19:52:07 except a tiny bit less packages on the list? 19:52:31 currently, there are 6493 key source packages (popcon threshold at 5%) raising that would give: at 10% 6147 key source packages, at 20% 6025 key source packages, at 100% 5869 key source packages 19:52:56 so besides the obvious benefit of autoremoval 19:53:42 dependencies of packages that are going to be autoremove are warned in the bug tracker 19:54:06 but this doesn't happen for key packages 19:55:02 so, more exposure of RC bugs impacting popular packages 19:55:08 where RC bugs matter most? 19:55:31 yes 19:55:43 did you check which packages would be demoted? 19:56:08 when ivodd wrote those numbers above, it wasn't clear to me whether he was for or against adjusting the threshold 19:56:42 it's trivial to do that excersize if you know how the script runs 19:56:55 I think he just wanted to provide the numbers 19:57:11 i didn't check which packages would be affected 19:57:25 I think the original treshold was picked by belly feeling during introduction 19:57:44 personally I'm OK with raising I think 19:57:46 so i think you and i would be happy to increase the threshold, or do away with it completly 19:58:03 but does anybody feel it should stay at 5%? 19:58:05 but I also think that ivodd had an idea to reduce the set even further 19:58:20 which you might want to persue too 19:59:24 raising to 100% is effectively removing it 20:00:10 so, most packages would come because of standard, important or required 20:01:15 and d-i 20:02:26 #agreed raise popcon threshold to 100% 20:02:33 <_rene_> as a data point: even LO is only ~44%. While we (including me) don't need it that much it's different for our users 20:02:41 <_rene_> firefox also is ~44% 20:02:56 <_rene_> sorry, firefox-esr 20:03:11 #topic AOB 20:03:16 anybody? 20:03:37 (we're close to planned closure time, much better than last time) 20:04:16 From the topics discussed last time: llvm-toolchain-10 is gone from testing. 20:04:26 * elbrus noticed, thanks 20:04:33 some openjdk-* versions too 20:04:44 and others are requested to be removed (from unstable) 20:05:25 yes, only 11 16 and 17 left 20:05:40 in bullseye 20:05:59 and discussion about 17 is ongoing 20:06:57 ok 20:07:05 #topic Next meeting 20:07:12 #info Next meeting is 24th February at 19:00 UTC (import into your calendar via https://release.debian.org/release-calendar.ics) 20:07:25 ^ includes the freeze dates too, so you don't forget 20:07:36 #endmeeting