18:59 #startmeeting Developer Membership Board Meeting 18:59 Meeting started Mon Jun 15 18:59:58 2020 UTC. The chair is rafaeldtinoco. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 18:59 18:59 Available commands: action commands idea info link nick 19:00 o/ 19:00 #chair rafaeldtinoco 19:00 Current chairs: rafaeldtinoco 19:00 #topic Review of previous action items 19:00 ddstreet: should i start with your items ? 19:01 sure, i think most are done 19:01 alright 19:01 ddstreet follow up on packageset owning team for ubuntu-mate (carried over) 19:01 done, team's created and pkgset is moved over to using it 19:02 I have a question regarding the team creations to the end 19:02 moving on 19:02 ddstreet to handle PPU rights of tux{math,paint,type},x11vnc for alkisg ahead of edubuntu packageset removal in the future 19:02 done 19:02 ddstreet send out announcement for oSoMoN addition to mozilla packageset (rolled over) 19:02 yep sent, done 19:02 ok.. now my stuff 19:02 one addition first 19:02 sure 19:02 briefly 19:03 as alkisg should now have all the upload rights needed from the edubuntu pkgset, we should add a action item to remove it, unless there is any other reason for keeping it 19:04 what happened to the seeds, do u know ? 19:04 i can add filters to pkgset scripts, but they are basically creating the sets coming from existing seeds 19:04 no, i haven't looked at it other than adding alkisg's upload pkgs 19:04 ok.. let me add an action to myself here 19:05 The packageset scripts have a list of autogenerated packagesets, don't they? 19:05 rbasak: yep, I can simply filter out edubuntu 19:05 im wondering about the seeds existance 19:05 should we care ? 19:05 I thought was an inclusion list rather than an exclusion list? 19:06 its a shell script with an array currently 19:06 But yeah, we could be good citizens and remove the seed too 19:06 and then germinate runs through all generated files 19:06 ok, for the seed removal we will need an aa 19:06 #action rafaeldtinoco to check edubuntu seed <-> pkgset relationship (generation) and if edubuntu pkgsets can be dropped 19:06 * meetingology rafaeldtinoco to check edubuntu seed <-> pkgset relationship (generation) and if edubuntu pkgsets can be dropped 19:06 and we should probably manually remove edubuntu from groovy, right? but leave it untouched from focal and earlier? 19:06 That might be more of a release team or AA thing, yeah 19:06 if you all agree ^ 19:07 Presumably the image generation machinery needs disconnecting if it isn't already 19:07 +1 19:07 hum. true 19:07 +1 19:07 it will be a good exercise to understand in deep the correlation 19:07 Yes - I'd love to learn from you when you're done figuring it out :) 19:07 ddstreet: yes (for the groovy removal part) 19:08 yep the pkgsets changes emails are about the re-start once im done 19:08 let me move to my items 19:08 and we can catch up after 19:08 rafaeldtinoco to follow up with TB / packageset tooling on ubuntu-studio-uploaders being added to ubuntustudio packageset for upload rights 19:08 > Archive Upload Rights for ubuntu-studio-uploaders: archive 'primary', package set 'ubuntustudio' in groovy 19:08 done 19:08 rafaeldtinoco to handle request "Adding lubuntu-update-notifier to the Lubuntu package set" (carried over) 19:08 > Archive Upload Rights for lubuntu-dev: archive 'primary', package set 'lubuntu' in groovy 19:08 done 19:09 rafaeldtinoco to complete documentation on existing packageset tooling, and discuss the current (large) delta in the packagesets that need adjusted (rolled over) 19:09 > Work in progress, move to long term please. 19:09 #action rafaeldtinoco to move pkgset tooling documentation to long term 19:09 * meetingology rafaeldtinoco to move pkgset tooling documentation to long term 19:09 rafaeldtinoco to handle getting gnome-shell into ubuntu desktop exceptions list 19:09 > Archive Upload Rights for ubuntu-desktop: archive 'primary', package set 'ubuntu-desktop' in groovy 19:09 done 19:09 rafaeldtinoco to handle "Refresh the packagesets" mailing list item (https://lists.ubuntu.com/archives/devel-permissions/2020-April/001486.html) 19:09 done ^ 19:09 @ddstreet lets talk about this one after I finish my items 19:09 per your previous email 19:10 ack 19:10 rafaeldtinoco to look into mailing list item "Including yaru-theme in the ubuntu-desktop set" (https://lists.ubuntu.com/archives/devel-permissions/2020-April/001487.html) 19:10 > Archive Upload Rights for ubuntu-desktop: archive 'primary', package set 'ubuntu-desktop' in groovy 19:10 done 19:10 rafaeldtinoco to look into mailing list item "Include fprint packages in the ubuntu-desktop set" (https://lists.ubuntu.com/archives/devel-permissions/2020-April/001488.html) 19:10 > Add libfprint to exceptions for ubuntu-desktop pkgset 19:10 committed 19:10 > $ edit-acl add -S {groovy,focal,eoan,bionic} -s libfprint -P ubuntu-desktop -t upload 19:10 done 19:10 rafaeldtinoco to handle "Packages to Remove from ubuntustudio packageset" (https://lists.ubuntu.com/archives/devel-permissions/2020-May/001497.html) 19:10 > Done with the pkgset tool re-run for Groovy. 19:10 rafaeldtinoco to handle "Applying packageset exceptions transitively" (https://lists.ubuntu.com/archives/devel-permissions/2020-May/001501.html) 19:11 > Work in progress, move to long term please. 19:11 #action rafaeldtinoco to document exceptions inclusion for DMB team 19:11 * meetingology rafaeldtinoco to document exceptions inclusion for DMB team 19:11 ok 19:11 we have no applications today 19:11 should we discuss anything here 19:11 or at the end ? 19:12 i guess just the mailing list items? 19:12 #topic Outstanding mailing list requests to assign 19:12 jackd2 for ubuntustudio packageset (Erich Eickmeyer) 19:12 ddstreet: ^ 19:13 rafaeldtinoco so, i don't really understand your email reply, but i think that's because i don't understand the tooling yet 19:14 let me clarify for you then 19:14 in single terms 19:14 simple 19:14 all pkgset inclusion/exclusion is done by the tooling.. for -devel only 19:14 based on the existing seeds for the -devel version. 19:15 if there is a pkg that is in the wrong pkgset (server instead of ubuntustudio, lets say) 19:15 we have one file to be updated to mark it as an exception 19:15 the tooling will be ran 2 or 3 times a week and send an email with the changes 19:16 if no complains, the changes will be applied (to devel, based on seed, again) 19:16 so, packages can only be in a single packageset? 19:16 good question 19:17 there are seeds that extend other seeds 19:17 AIUI, there are various edge cases that are handled poorly or have surprising results 19:17 rbasak: +1 on what you said 19:17 the logic is a big python batch with lots of edge cases 19:17 When we've tried to clean these up in the past, different DMB members have had different opinions on what the answer should be 19:17 that part is the one I still have to work on 19:18 so with this example, was it wrong for me to add jackd2 to ubuntustudio packageset? 19:18 because that would break some other packageset? 19:18 so with that example 19:18 I think for now we have to just keep adding exceptions case-by-case, and longer term it'd be nice to specify what the behaviour should be and then start adjusting the tooling to head towards that goal. 19:18 for all previous versions .. doing by hand was correct (because we are not automatically updating those). 19:18 for groovy .. the correct thing was to put it as an exception.. so instead of "server" it would be "ubuntustudio" in the next automatic generation changes 19:19 an email would be sent to us saying that it was about to add the source package (jackd2) to ubuntustudio 19:19 we would not complain and the changes would be committed 19:19 rbasak: yes 19:19 i want to put what we have "in production" 19:19 and read the batch better 19:19 and document what it does (with germinate output) 19:19 and we can maybe discuss 19:19 so our packageset list doesn't include 'hidden' packagesets like 'server'? https://people.canonical.com/~ubuntu-archive/packagesets/groovy/ 19:20 or do you mean 'ubuntu-server' 19:20 ddstreet: a consequence of manually adjusting an automatic packageset is that the next time the tooling is run, it might try to undo what you did, and then the person running it has to either go with it or code up an exception. Sometimes that's hard to do. The alternatives are to fix the tooling before making the adjustment, or to leave the uploader stuck without the packageset change. None of 19:20 these options are good :-/ 19:20 ddstreet: https://paste.ubuntu.com/p/H4GBJFYpNK/ 19:20 right, i was clear that my addition of jackd2 might be undone later, which is why i added the follow-up in the agenda to see if the seed update was all that's needed 19:21 the jackd2 case I'll re-check 19:21 actually 19:21 I replied to you on that already 19:21 right ? 19:21 (just checking) 19:21 ddstreet: yeah and to be clear I'm fine with your approach. 19:21 ok, maybe it's just me who doesn't see the link between the seeds and packagesets, i'm hoping to get a clearer picture once the tooling is documented... 19:22 maybe i should just read the existing tooling now 19:22 well, not *now*, but i mean before refactoring 19:22 #action rafaeldtinoco add jackd2 as an exception (from ubuntu-server to audio-plugins perhaps) 19:22 * meetingology rafaeldtinoco add jackd2 as an exception (from ubuntu-server to audio-plugins perhaps) 19:22 In the case that a package is only seeded from one place, it'll end up in the corresponding automatic packageset if there is one. 19:22 If it's seeded from more than one place, then the tooling has various logic to try and resolve it. 19:22 That's the extent of my knowledge :) 19:23 yes 19:23 rbasak: said correctly 19:23 (and when I say seed, I probably mean germinate - it doesn't matter whether the package is directly or indirectly seeded) 19:23 there are some decisions made to see where the package will stay (and now it even reads the pkg dependencies) 19:23 this part was made by iain in last change 19:24 im about to enter that batch code now =) 19:24 ddstreet: let me document something "quick" for the next meeting 19:24 "what to do instead of edit-acl for -devel" 19:24 - guaranteeing seed is good 19:25 - editing exceptions file and commiting 19:25 - waiting pkgset auto re-generation 19:25 type of thing 19:25 +1 would definitely help me 19:25 #action rafaeldtinoco to create, for now, a small "what-to-do" for pkgset changes in -devel 19:25 * meetingology rafaeldtinoco to create, for now, a small "what-to-do" for pkgset changes in -devel 19:25 #action rafaeldtinoco to put pkgset tooling to automatically update pkgsets (crontab) 19:25 * meetingology rafaeldtinoco to put pkgset tooling to automatically update pkgsets (crontab) 19:25 ok 19:26 that was it on my side for this #) 19:26 YEAHHHHH 19:26 =) 19:26 ddstreet: should I move on to TB bugs ? 19:26 sure 19:26 #topic Open TB bugs 19:26 The idea is that flavour developers get to maintain their own stuff, btw, and 'their own stuff' is defined as the things in their flavour, i.e. what they seed 19:27 * Laney runs 19:27 Laney: +1 and dont go to far 19:27 Packageset modification for ~ubuntu-mate-uploaders 19:27 @ddstreet fix released ^ 19:27 yep 19:27 though 19:28 I have a question about this 19:28 just for my own info, should the devel release have been done differently? 19:28 (or any of it) 19:28 devel should be done automatically (through seed updates -> pkgset changes OR by exceptions file being changed) 19:29 ah for this change 19:29 no 19:29 sorry, this is just a permission change 19:29 right ? from person to the team 19:29 i think just uploader change is totally separate from the tooling, right? 19:29 yep 19:29 that is the exact question I have 19:29 who defined we would start creating the teams ? 19:29 i mean, is there anything we should follow ? 19:30 (start creating dev groups with dmb permissions) 19:30 there's some docs in the kb 19:30 https://wiki.ubuntu.com/DeveloperMembershipBoard/KnowledgeBase 19:30 because this change gives DMB the power to add/remove people from pkgsets 19:31 yep 19:31 should we do this ad-hoc like you're doing ? 19:31 or try to guarantee all pkgsets have their respective devel group ? 19:32 * rbasak doesn't follow 19:32 rbasak: last changes coming from ddstreet 19:32 like this 19:32 https://bugs.launchpad.net/ubuntu-community/+bug/1881666 19:32 Launchpad bug 1881666 in ubuntu-community "[TB/DMB] Packageset modification for ~ubuntu-mate-uploaders" [Undecided,Fix released] 19:32 changing the ubuntu-mate packageset uploader from an individual to a team 19:33 note that this was the last 'flavor packageset' that *was not* set to team-uploader 19:33 so now, all the flavor packagesets have team uploaders 19:33 ah was it ? 19:33 The main thing is that we own the team, and nobody else apart from the DMB can change its membership 19:33 ok.. so my question is dead 19:33 That seems correct 19:33 rbasak so team members can't add other members? 19:33 There's also the stuff about ~ubuntu-dev vs. ~ubuntu-upladers which is covered in the docs I think? 19:33 Correct 19:34 That can be done with a delegation, but the team must explicitly have that 19:34 we should make sure not to add anyone to an uploader team as 'administrator' though, right? 19:34 Right 19:34 unless they have a delegation 19:34 ack 19:34 hum. 19:34 that is new to me, will have to read 19:35 rbasak rafaeldtinoco note that i did update the dmb kb with some clarification steps 19:35 ddstreet: tku, ill read it carefully 19:35 https://wiki.ubuntu.com/DeveloperMembershipBoard/KnowledgeBase?action=diff&rev2=33&rev1=27 19:35 or just check the 'info' tab to select diffs 19:35 cool 19:35 i think everything i added is correct, but please do correct anything as needed 19:35 Delegations documented here: https://wiki.ubuntu.com/UbuntuDevelopers/TeamDelegation 19:36 alright, got homework on this 19:36 anything else on this discussion ? 19:36 we should link the teamdelegation from our kb page 19:36 if it isn't already 19:37 #action rafaeldtinoco link teamdelegation from dmb kb page when reading ddstreet updates 19:37 * meetingology rafaeldtinoco link teamdelegation from dmb kb page when reading ddstreet updates 19:37 ill do it since I'll read it again 19:37 thanks 19:37 #topic PPU addition for ~alkisg 19:37 https://bugs.launchpad.net/ubuntu-community/+bug/1881653 19:37 Launchpad bug 1881653 in ubuntu-community "[TB/DMB] PPU addition for ~alkisg" [Undecided,Fix released] 19:37 fix released also ^ 19:38 yep 19:38 do I have to check groovy for this one ? 19:38 ah it was permission only 19:38 nm 19:38 ppu, ok moving on 19:39 right, and next one is ppu also i think 19:39 #topic PPU addition for ~rcj 19:39 https://bugs.launchpad.net/ubuntu-community/+bug/1866679 19:39 Launchpad bug 1866679 in ubuntu-community "[TB/DMB] PPU addition for ~rcj" [Undecided,Fix released] 19:39 fix released also 19:39 yep, and while the tb bug doesn't list groovy, i am pretty sure he already had ppu for groovy 19:39 cool 19:40 #topic Select a chair for the next meeting 19:40 rbasak: can u chair the next one ? 19:40 you're the next on the list 19:41 I can change it later if you can't 19:41 #topic Any other business 19:41 ok.. 19:41 any other stuff to be discussed ? 19:42 nothing else from me 19:42 rafaeldtinoco: sure 19:42 rbasak: tku 19:42 ok.. im calling it in ... 5 19:42 4 19:42 3 19:42 Just as an fyi ... I'll need to catch up on the current todo, but I'm back after a few weeks of absence. 19:43 slashd: welcome back =) 19:43 slashd: you will be chairing after rbasak 19:43 glad to have you back slashd :) 19:43 fyio 19:43 3 19:43 rafaeldtinoco: no problem for the chairing order 19:43 2 19:43 1 19:43 #endmeeting