16:27 #startmeeting Developer Membership Board 16:27 Meeting started at 16:27:32 UTC. The chair is rbasak. Information about MeetBot at https://wiki.ubuntu.com/meetingology 16:27 Available commands: action, commands, idea, info, link, nick 16:27 #topic Package Set/Per Package Uploader Applications 16:27 #subtopic PPU application for s390-tools by Frank Heimes at DMB meeting 2024-09-16 (will also join meeting on 2024-09-02, just in case there is remaining time, if not it's also fine). 16:28 fheimes: o/ 16:28 hi \o 16:28 shall I introduce myself quickly? 16:28 Sure, thanks. 16:29 Ok, I'm Frank (fheimes) and work at Canonical's Partner Engineering department (formerly known as hardware enablement) and the Partner I take care of is IBM, with their Z platform (s390x) and their IBM Power family (ppc64el) 16:30 Prior to my time at Canonical I worked at IBM's R&D lab in Germany in the area of Linux, esp. Linux on s390x. 16:30 So, I take care about s390x hw enablement duties, I do quite some work in the s390x space, hence me applying for PPU for s390-tools (package set, consists of two packages). 16:30 rbasak: sorry for lateness. (had an IRL emergency problem) 16:31 OK, questions for fheimes then please. 16:31 Where can the release schedule be found, and on what date was the freeze for Oracular? 16:32 that can be found on discourse, there are release schedule pages for all Ubuntu releases, let me quickly find it ... 16:32 https://discourse.ubuntu.com/t/oracular-oriole-release-schedule 16:33 FF was on Aug 15th 16:33 fheimes: OK. And if it's after feature freeze, Oracular has 2.34.0-0ubuntu2 and s390x-tools upstream release 2.34.1, what do you need to check to verify that an upload to Oracular will not violate feature freeze? 16:33 today was beta and hwe freeze 16:34 well, there are several things (I guess) 16:34 one thing to check is if the new release is a bug fix release only 16:35 (there is a high chance with a x.y.1) 16:35 if it incl. new features, that are urgently needed and it would violate the FF, then a FFe can be thought of 16:36 if a new version incl. new features AND bug fixes and some fixes are needed, then they need to be cherry-picked -- the package can no longer be simply version bumped after the FF 16:37 I hope I covered what you were looking for? 16:37 Yes thanks. Typing question next question... 16:37 If you upload and it gets "stuck in proposed", who is expected to resolve the situation? 16:38 well, it's primarily me 16:39 OK. And how would you approach that - for example where's the documentation for dealing with this situation? 16:39 I'll should check and shepared the package through the process and should see when it gets stuck, but I will also get a notification in case it's stuck for a certain amount of time 16:40 I would first of all look at launchpad, trying to find out why it's stuck (FTBFS, missing approval) 16:41 britney is another source, actually the update excuses page, like: 16:41 https://ubuntu-archive-team.ubuntu.com/proposed-migration/oracular/update_excuses.html#s390-tools 16:42 Do you get dep8 results for s390-tools in practice? 16:42 there are many docs where someone can find more info on what to do, for example in case of an SRU: https://people.canonical.com/~ubuntu-archive/pending-sru.html 16:42 I see the packaging doesn't ship tests (I think?) but maybe you get rdep dep8 tests triggered? 16:43 @rbasak: the status on that is what I can see here: https://ubuntu-archive-team.ubuntu.com/proposed-migration/oracular/update_excuses.html#s390-tools 16:44 There's nothing there at the moment though because it isn't pending migration right now 16:44 and I can often find directly the reason there (failt to build, missing approval etc) 16:44 yes, that's correct atm there is nothing there (was just an example in case its stuck) 16:44 question: what if it's not there (update_excuses.html) and the package has still not migrated? 16:45 where will you look? 16:45 or what will you do? 16:45 and failed tests of dependent packages can be identified and if needed, retriggered (in different ways) or tests fixed (if needed) 16:46 this is a good doc on this: https://wiki.ubuntu.com/ProposedMigration#How_to_re-run_autopkgtests_with_dependencies_on_other_packages_in_the_proposed_pocket that I sometimes refer to 16:47 Thanks. I have no further questions! 16:47 Does anyone else have any further questions for fheimes? 16:47 my question^ 16:47 fheimes, have you considered adding dep8 tests to s390-tools? 16:48 so far I got cases where things fauiled to migrate (https://people.canonical.com/~ubuntu-archive/proposed-migration/) via LP itself, and update excuses 16:48 ok, gotcha 16:48 fine, it's okay 16:48 Sorry I thought he was answering it in his subsequent messages 16:49 OK, are we all ready to vote? 16:49 yes, I have considered that - there is a local version that I havebut it's difficult to trigger that automatically, since special hw access is needed, that is no always possible in KVM 16:49 yes, ready to vote 16:49 i am ready to vote as well 16:50 #vote Grant fheimes PPU for s390-tools and s390-tools-signed 16:50 Please vote on: Grant fheimes PPU for s390-tools and s390-tools-signed 16:50 Public votes can be registered by saying +1, -1 or +0 in channel (for private voting, private message me with 'vote +1|-1|+0 #channelname') 16:50 +1 16:50 +1 received from rbasak 16:50 +1 16:50 +1 received from bdrung 16:50 (for example test on s390x hw cryptography can only be run if the hw got passed throught to a KVM instance, and ensuring that in PS is barely possible ...) 16:50 Sorry I was just checking the application page for the exact set of package names, and then started pondering about the -signed PPA business. 16:50 oh ... 16:50 But I think it's probably irrelevant for initial PPU. 16:50 +1; Frank has a lot of experience with the package and I think I trust him to be an excellent in taking care of the package. 16:50 +1; Frank has a lot of experience with the package and I think I trust him to be an excellent in taking care of the package. received from utkarsh2102 16:50 +1 16:50 +1 received from bdmurray 16:51 omg, bdmurray, hiiiiiiiiiiiiiiii! 16:51 teward: are you available to vote? 16:51 yes i am 1 minute 16:51 rbasak: we already have 4 votes 16:51 +1 16:51 +1 received from teward 16:51 ah nice 16:51 perfect! 16:51 Yeah but people should have the opportunity to register if they dissent :) 16:51 rbasak: voted. sorry but like 2 minutes ago I get a call from CEO at DAYJOB 16:52 and that's one of those "Always Answer" cases 16:52 (and it's nice when they also +1) 16:52 #endvote 16:52 Voting ended on: Grant fheimes PPU for s390-tools and s390-tools-signed 16:52 Votes for: 5, Votes against: 0, Abstentions: 0 16:52 Motion carried 16:52 fheimes: hey, congratulations! <3 16:52 Congratulations fheimes! 16:52 congrats fheimes :) 16:52 rbasak: can you add his acl and I can take care of the announcement? 16:52 Any volunteers for the action items to enact this? 16:52 congrats fheimes 16:52 ack 16:52 coolio 16:52 #action utkarsh2102 to announce fheimes' successful application 16:52 * meetingology utkarsh2102 to announce fheimes' successful application 16:52 Many thx ! I'm excited and happy ! 16:52 #action rbasak to add fheimes PPU ACL 16:52 * meetingology rbasak to add fheimes PPU ACL 16:53 congratulations 16:53 I know we are running out of time but can we look at aciba, too? 16:53 I have a hard stop 16:53 fheimes, I missed to ask one very important question since you stated that you live in the Schwäbische Alb: What is a Muggeseggele? 16:53 teward, bdrung, bdmurray? 16:53 i have no objection 16:53 ;-) 16:53 no objections 16:53 rbasak: but are you okay with cloud-init PPU? 16:53 oh, hey pleia2 ! 16:53 I'm OK if you'd like to continue anyway. If you can please ask the usual questions then I'll check back later and vote if needed. 16:54 On the matter of cloud-init PPU, I believe Chad already has PPU for cloud-init? 16:54 Chad and James have 16:54 i think he does 16:54 yes 16:54 @bdrung I'm living here, but didn't grew up here, but nevertheless it is a very small unit ;-) 16:55 I'd like to see that ML thread concluded with DMB consensus on the topic, but maybe it's best to follow existing precedent for cloud-init in the meantime 16:55 Seeing as aciba has already been let down 16:55 i don't understand what that means 16:55 And because it'd be a separate matter as to whether cloud-init is considered "core". 16:55 fheimes, I was born in Baden-Württemberg, but grew up in Berlin. 16:56 are you okay with us continuing? 16:56 I think it is probably "core", but cloud-init has been working well enough with PPU recently, especially as any changes have to effectively pass an SRU reviewer anyway. 16:56 are you okay with us continuing> yes 16:56 great! 16:57 #chair utkarsh2102 bdrung bdmurray teward 16:57 Current chairs: bdmurray, bdrung, rbasak, teward, utkarsh2102 16:57 I need to step away imminently so someone else will need to take over chairing 16:57 bdrung: are you okay to chair? 16:58 okay 16:58 #subtopic PPU application for cloud-init by Alberto Contreras at DMB meeting 2024-09-16 (as 2024-09-02 did not happen). 16:58 \o/ 16:59 o/ 16:59 #link https://wiki.ubuntu.com/AlbertoContreras/DeveloperPerPackageUploadApplication 16:59 aciba, do you quickly want to introduce yourself? 16:59 yes 16:59 I am Alberto Contreras, I have been working for Canonical during the last ~2.5 years as part of the Canonical Public Cloud (CPC) team focused on cloud-init on cloud-specific features / fixes. 17:00 I would like to apply to get upload permissions for cloud-init. 17:00 More about myself and the application: https://wiki.ubuntu.com/AlbertoContreras/DeveloperPerPackageUploadApplication 17:01 Thanks. Questions for aciba please. 17:04 aciba, you list LP #2045582 as first entry under areas of work. What went well there and what did you learn there? 17:05 aciba: question - after you answer the above, what happens if the verification of the SRU fails? what do we do then? 17:06 the sru process, having the package in -proposed allowed other developers to test that new version and detect a couple of bugs prior to releasing it, iirc 17:07 so we detected a couple of bugs and proposed a new version of the package with those bugs fixed to be SRUed 17:09 thanks. perfect transition to utkarsh2102's question 17:10 if the verification fails in a valid way, we need to hold that version and fix it 17:10 proposing a new version including the fix or fixes 17:10 as we did in #2045582 17:10 we need to add the tags as verification failed to the lp bug 17:11 perfect, thanks 17:11 aciba: what freezes do you need to keep in mind before uploading? 17:12 feature freeze, documentation freeze, beta freeze, final freeze, they are doucmented here: https://discourse.ubuntu.com/t/oracular-oriole-release-schedule/36460 17:12 great 17:12 aciba, when you upload a package to the archive, which steps are taken until the package hits the release pocket? 17:13 typically: unnapproved queue > -proposed > -updates 17:14 those steps are outlined here: https://canonical-sru-docs.readthedocs-hosted.com/en/latest/howto/standard/ 17:14 aciba, and in case of uploads to oracular? 17:15 it depends on the time of the release cycle we are at 17:16 prior to feature freeze, it goes directly to the release pockets 17:16 after feature freeze but prio to the release, it requires a ffe and goes directly to the release pocket after the ffe has cleared 17:17 before release it follows the normal SRU process 17:17 that is: unnapproved queue > -proposed > -updates 17:18 did you mean s/before release/after release/? 17:18 yes, sorry 17:19 aciba, there is a step in between the upload and release pocket on the development series. 17:20 it goes to the unpproved queue, but it gets automatically accepted under certain circumstances, 17:21 aciba, and what about autopkgtest? 17:21 they are run when the package hits -proposed, every reverse dependency's dep8 test on the package get executed 17:22 on can see the results in the update excuses page 17:22 and the SRU drive must monitor and take care of them 17:22 s/drive/driver/ 17:23 aciba, are autopkgtest only run for SRU uploads and not for development release uploads? 17:24 the rdeps ones yes, I belive the ones associated with the package itself do run on every build 17:27 package uploads to the development release and up in -proposed first. 17:28 s/and/end/ 17:28 then yes, autopkgtest run for development release uploads 17:29 Do you need a FFe for every upload after the feature freeze (and before the release)? 17:30 no, for example, if there is a new upstream version that contains only bug fixes, then a new upload containing that does not require and ffe 17:30 before the release, the SRU process holds 17:30 s/before/after/? 17:30 s/after/before/ again 17:31 sorry I am a bit nervous 17:31 thanks for being honest here (written text does not convey things like that) 17:32 exceptions to the FFe process can be found here: https://wiki.ubuntu.com/FeatureFreeze and https://wiki.ubuntu.com/FreezeExceptionProcess 17:33 Any questions from my DMB fellows? 17:35 not from me 17:35 none here 17:35 aciba, You list https://github.com/canonical/cloud-init/commit/57bb32388 as one example. Anything one should be aware of when (re)moving files in /etc? 17:38 it should not changer behavior, it should not remove config files created by the user 17:39 it should aviod user intervention 17:39 s/aviod/avoid 17:41 does that satisfy your question? 17:41 aciba, if upstream decides to drop a file in /etc can the package just remove the file in /etc on upgrade? 17:43 drop the file how? 17:43 at runtime? 17:44 bdrung 17:44 drop = delete / not ship it in the file in the release tarball 17:44 yes, the package maint scripts can delete a config file when it is not need anymore 17:45 but it depends in a case by case basis I think 17:45 aciba, and in case the file was modified by the user? 17:45 we should minimize user disruption 17:46 and if the user modified the file, the maint scripts shouldn't remove the file 17:46 perfect thanks. i was aiming for the helpers from https://manpages.ubuntu.com/manpages/noble/en/man1/dpkg-maintscript-helper.1.html to be used to handle that. 17:47 I am done with my questions. 17:47 Everyone ready for voting? 17:47 yes 17:47 teward, bdmurray? 17:47 I have no questions. 17:48 bdrung: i think we should proceed 17:48 it's wayyyyy over time 17:49 teward will follow up 17:49 no questions 17:49 there :) 17:49 #vote Grant Alberto Contreras (aciba) PPU for cloud-init 17:49 Please vote on: Grant Alberto Contreras (aciba) PPU for cloud-init 17:49 Public votes can be registered by saying +1, -1 or +0 in channel (for private voting, private message me with 'vote +1|-1|+0 #channelname') 17:49 (you already asked if we had more questions and I didn't) 17:51 writing a review.. 17:53 +1; I think Alberto has quite a lot of experience with cloud-init & its internals. I think they'd be a great addition as an uploader of the package. Yes, there were some answers which weren't completely clear but that's okay, I think those are smaller and can be learned on the fly. Alberto, I'd ask you to exercise your upload rights with caution & 17:53 +1; I think Alberto has quite a lot of experience with cloud-init & its internals. I think they'd be a great addition as an uploader of the package. Yes, there were some answers which weren't completely clear but that's okay, I think those are smaller and can be learned on the fly. Alberto, I'd ask you to exercise your upload rights with caution & received from utkarsh2102 17:53 please don't hesitate to ask before uploading or whenever in doubt. More reviews, better the upload is. So please keep in mind all those things when you upload. ;) 17:54 bdrung, bdmurray, teward? 17:55 +1 I concur what utkarsh2102 wrote 17:55 +1 I concur what utkarsh2102 wrote received from bdrung 17:55 +! 17:55 oops 17:55 +1 17:55 +1 received from teward 17:55 +1 17:55 +1 received from bdmurray 17:56 #endvote 17:56 Voting ended on: Grant Alberto Contreras (aciba) PPU for cloud-init 17:56 Votes for: 4, Votes against: 0, Abstentions: 0 17:56 Motion carried 17:56 yay! 17:56 aciba: congratulations! <3 17:56 wow, many thanks, I will take care and ask when in doubt 17:56 aciba: super! thank you! 17:57 aciba, congrats. asking when in doubt is the correct mindset. nobody knows everything. 17:57 #action rbasak to add aciba to the right place & utkarsh2102 to announce the successful application 17:57 * meetingology rbasak to add aciba to the right place & utkarsh2102 to announce the successful application 17:57 ok, I'm going to call if off. It was a pretty long meeting :) 17:57 thank you so much for chairing, bdrung. 17:57 much appreciated, have a good day o/ 17:57 exactly 17:57 you were great! o/ 17:58 #endmeeting