15:03 <juliank> #startmeeting Weekly Ubuntu Foundations team
15:03 <meetingology> Meeting started at 15:03:07 UTC.  The chair is juliank.  Information about MeetBot at https://wiki.ubuntu.com/meetingology
15:03 <meetingology> Available commands: action, commands, idea, info, link, nick
15:03 <juliank> #topic Lightning rounds
15:03 <juliank> #link https://discourse.ubuntu.com/t/foundations-team-updates-thursday-2024-06-13/
15:03 <juliank> brb :D
15:05 <bdmurray> bdrung: Mantic isn't EoL yet!
15:06 <adrien> dbungert: I've seen launchpad reporting that a package is ready in my PPA and a build triggered after that would still us the previous version :/
15:06 <bdrung> bdmurray, let me check my clock. wait, let me check my timezone... ;)
15:06 <adrien> rmadison is a good idea; do you have a script around it to automate the wait?
15:07 <dbungert> rmadison script to trigger - no but that sounds useful
15:07 <bdrung> adrien, the build needs to be published before it can be used
15:08 <bdmurray> I wonder if bryce's PPA tools has something
15:08 <adrien> not that I know of; actually I was using ppa-dev-tools to wait on publishing which is the only wait I know of
15:08 <adrien> bdrung: yes, but in practice you need to wait for more than that
15:09 <vorlon> launchpad's definition of "published" is not the same as downloadable from the archive
15:09 <vorlon> so you can't use the api to answer this
15:09 <adrien> dbungert: ok, I don't have a need right now but I'll do that next time I need it, we'll see when that happens (I'm usually not fond of rmadison due to how slow it is :/ )
15:10 <adrien> at some point I thought I would chdist apt and poll for changes :D
15:10 <bdmurray> Its a perl script if anybody is feeling ambitious
15:11 <adrien> :scared:
15:14 <juliank> moving on
15:14 <juliank> #topic Release incoming bugs
15:14 <juliank> #link http://reports.qa.ubuntu.com/reports/rls-mgr/rls-oo-incoming-bug-tasks.html#foundations-bugs
15:14 <juliank> I think we talked about bug 1969394 last week
15:15 <juliank> Ah shoot I see now, the comment in there is from 2022
15:15 <vorlon> dbungert: "rebuild clicks" -> ubuntu-build
15:15 <juliank> How do we go about approaching this, this needs desktop design work
15:16 <juliank> But update-manager is our package
15:16 <bdrung> IMO this is a ticket for the desktop team
15:17 <juliank> I left a comment asking
15:17 <vorlon> the desktop team is not the design team
15:17 <vorlon> is it update-manager that needs to change?
15:18 <juliank> I think you also need software-properties-gtk change
15:18 <dbungert> vorlon: thanks!
15:18 <vorlon> dbungert: I posted about ubuntu-build on ubuntu-devel and nobody replied, so I'll be changing it soon to make --batch implicit and the only interface
15:18 <juliank> You need a separate button for livepatch and that needs then work to start software-properties with the livepatch tab directly?
15:18 <vorlon> (the non "batch" interface is blech)
15:19 <juliank> In any case we can circle back on this with desktop and/or design
15:19 <juliank> Let's see what they have to say
15:19 <juliank> bug 2069041
15:20 <juliank> This is a bit unfortunate
15:21 <juliank> enr0n: What does the daemon-reload do?
15:21 <enr0n> juliank: re-runs the generators
15:21 <vorlon> enr0n: you commented that a different command is necessary from the user to restart because you have to do a daemon-reload.  Is there no supported way that we could automatically retrigger the generator on a file change?
15:21 <juliank> enr0n: ah
15:21 <juliank> We could trigger the generator
15:21 <juliank> But we don't want to autoreload
15:21 <juliank> But you'd get a warning
15:21 <juliank> You'd need a unit to active on the path changed
15:22 <enr0n> I think the answer is basically no, there's not currently a way to make that automagic
15:22 <enr0n> You could do some things with path units etc., but all would involve having an ExecStart=systemctl daemon-reload somewhere, which I don't think is a good idea
15:22 <schopin> a good first step could be to mention that in the default config comments for the Ports option?
15:22 <tobhe> ^ thought the same
15:22 <vorlon> I mean, I can think of a way to do it, you can trigger units based on inotify events?
15:23 <juliank> enr0n: I guess my point is with the path unit; you then get a warning that the daemon-reload is needed from systemd
15:23 <enr0n> vorlon: Right, it's not a question of how to trigger the event, it
15:23 <vorlon> but yeah if you would just be wrapping daemon-reload, then eew
15:23 <enr0n> the problem is that you would be triggering daemon-reload everytime the config changed, without the admin knowing
15:23 <vorlon> so, in that case we should still prioritize the documentation change
15:24 <juliank> You can _manually_ trigger the generator, it writes the openssh.socket to the generator dir, and then systemd complains
15:24 <vorlon> comment in the default config file + added to the discourse post
15:24 <enr0n> Manually triggering the generator does not help. Still needs to be loaded into the manager
15:24 <enr0n> So yes, documentation change is best here.
15:24 <juliank> Yes the point aggain would have been it told you it's outdated
15:25 <enr0n> Okay, I will assign this to myself
15:25 <juliank> You know, systemd prints a message when you restart a unit that has changed on disk without reloading :)
15:25 <juliank> but config documentation probably is nicer in any case
15:26 <juliank> #link http://reports.qa.ubuntu.com/reports/rls-mgr/rls-nn-incoming-bug-tasks.html#foundations-bugs
15:26 <enr0n> juliank: yes, I know.
15:26 <juliank> bug 2066995
15:26 <juliank> I guess this is in progress already, bdrung
15:27 <bdrung> yes, we have one fix but that might not be enough
15:27 <schopin> This is a compound issue, we fixed one bug but still don't know what causes the loop.
15:27 <juliank> hmm
15:27 <vorlon> yeah fwiw a chmod of /var/crash hasn't made the problem go away
15:27 <juliank> Well are we to a point where we can analyse it in a pulse?
15:28 <schopin> vorlon: it hasn't? Interesting.
15:28 <bdrung> vorlon, did you also chown the group afterwards?
15:29 <vorlon> the owning group is whoopsie
15:29 <juliank> Can we not try to debug it here?
15:29 <bdrung> for all files in /var/crash?
15:29 <juliank> I'll tag it
15:29 <bdrung> we need to find the reason since there were one other user reporting the same
15:29 <bdrung> so tagging is fine
15:30 <juliank> bug 2068933
15:30 <bdrung> you can tag that as well. that should be easy to fix
15:30 <juliank> doing so
15:30 <juliank> OK that's it for incoming bugs, mm only has the desktop ones, and jj, ff are empty
15:30 <juliank> #topic Team proposed-migration report
15:31 <juliank> #link https://ubuntu-archive-team.ubuntu.com/proposed-migration/update_excuses_by_team.html#foundations-bugs
15:31 <juliank> vorlon:
15:31 * juliank is preparing for apt 2.9.5 which will fix the migration but can't seem to land the final feature missing :D
15:31 * juliank also needs to re sortout fwupd-efi
15:32 <vorlon> juliank: fwupd-efi, we needed a sourceful upload, were you still taking that? ok
15:32 <vorlon> pyopenssl vs python-oauth2client
15:32 <adrien> needs a new package: https://gitlab.com/adrien-n/python-google-api-core
15:32 <vorlon> adrien: are you still working through this?  making progress?
15:32 <adrien> well, https://salsa.debian.org/adrien-n/python-google-api-core/
15:32 <vorlon> ah
15:33 <vorlon> does that just need sponsorship to oracular?
15:33 <adrien> various packages had not been updated for four years so there's a lot of debt
15:33 <adrien> I was thinking going through debian as that's the advised way AFAIK
15:33 <adrien> (since it's not packaged yet)
15:34 <adrien> and there was an ITP in debian (opened january 2023, not materialized yet)
15:34 <juliank> You already broke python-googleapi so let's go jfu it
15:34 <juliank> (to Ubuntu)
15:34 <vorlon> with my AA and release team hats I'm going to say this is a case you should upload directly to Ubuntu
15:34 <juliank> :D
15:35 <juliank> Yes, python3-googleapi in oracular release pocket is broken without it
15:35 <vorlon> unless python-googleapi is removeable instead
15:35 <adrien> good, let's do that then; I'll need minor adjustments to the changelog and maybe control files but I don't expect much
15:35 <vorlon> can I get a volunteer to sponsor adrien's package to the NEW queue?
15:35 * juliank need it for reading emails
15:35 * juliank volunteers
15:35 <vorlon> thanks
15:35 <adrien> well, I don't know how long it would take to fix by going through debian
15:35 <vorlon> gcc-13: dbungert has this in progress
15:35 <adrien> maintainer of python-googleapi has not been particularly fast on this
15:36 <vorlon> adrien: yes and that's why we shouldn't wait, for something that is on our todo list :)
15:36 <schopin> juliank: is that needed for lieer?
15:36 <adrien> yes
15:36 <vorlon> libsub-override-perl... hang on, what's happening here, didn't I see libsub-prototype-perl on component-mismatches for demotion?
15:36 <adrien> vorlon: let me see what I can prepare before EOD
15:36 <vorlon> hmm no it has always been in universe
15:37 <vorlon> cpete: so is the MIR for this still in progress?
15:37 <cpete_backup> vorlon: MIR is ready for review
15:37 <cpete_backup> I can take something else
15:38 <vorlon> is there someone who can move that MIR forward? I guess slyon isn't here in the meeting
15:39 <vorlon> cpete_backup: right, libsub-override-perl is listed on https://ubuntu-archive-team.ubuntu.com/component-mismatches-proposed.html for demotion
15:39 <vorlon> so... if I demote it then no MIR is needed
15:40 <cpete_backup> ...ah. TIL to check that
15:40 <vorlon> apt: juliank?
15:40 <vorlon> danilogondolfo: are you still working on git?
15:40 <vorlon> ah you mentioned ikiwiki-hosting in your report so I guess so :)
15:40 <danilogondolfo> vorlon, yes, I'm testing fixes for both regressions
15:42 <vorlon> ginggs: you had systemd last week, I still see a lot of autopkgtest regressions that persisted on retry, do you know what the story is here?
15:42 <vorlon> and does enr0n want to look at these
15:42 <vorlon> bdrung: dracut vs systemd, any progress on that one specifically?
15:42 <enr0n> vorlon: I am planning to upload systemd today, so yes I have this
15:42 <ginggs> there are a lot of "out of disk space" errors, and enr0n is planning another upload
15:43 <vorlon> ok
15:43 <bdrung> vorlon, yes, new dracut merged, found and addressed one bug, but still more to fix
15:43 <vorlon> zhsj: and I see you're still working on fakeroot
15:43 <zhsj> yes
15:43 <vorlon> libxml-libxml-perl is a desktop-owned lib transition
15:43 <vorlon> and openjdk-21 is just waiting for test results
15:44 <vorlon> so nothing new to assign out right now!
15:44 <vorlon> juliank:
15:44 <juliank> #topic AOB
15:44 <juliank> I just realized I wrote "to dumb", I meant to dump, but I will argue I meant "to dump in a dumb way" as a word play
15:44 <juliank> (https://salsa.debian.org/apt-team/apt/-/merge_requests/354 ftr)
15:47 <vorlon> mentioned in my status report and by email, but I'm OoO for 3 weeks starting Monday
15:47 <juliank> I think I'll just release apt 2.9.5 now, plenty of bug fixes, just no new features :(
15:47 <vorlon> so get in your last-minute panicked requests today
15:47 <bdmurray> I'll be OoO next week and tomorrow
15:55 <juliank> #endmeeting