[08:22] <didrocks> good morning
[08:23] <duflu> Morning didrocks
[08:24] <didrocks> hey duflu
[08:31] <marcustomlinson> morning desktoppers
[08:33] <didrocks> hey marcustomlinson
[08:33] <marcustomlinson> \o
[08:33] <marcustomlinson> how you didrocks?
[08:34] <duflu> Morning marcustomlinson
[08:35] <didrocks> marcustomlinson: feeling better, thanks! And you?
[08:36] <marcustomlinson> didrocks: yeah not bad thanks, as good as one can be on Monday. Got over my cold thankfully
[08:36] <marcustomlinson> hey duflu, how's life?
[08:37] <didrocks> that’s what week-ends are for: getting sick and then better for a week of work :p
[08:37] <didrocks> (same rule applies to holidays)
[08:37] <duflu> marcustomlinson, alright I guess. Just reinstalling focal after breaking it :/
[08:37] <seb128> lut didrocks, comment ça va ?
[08:37] <seb128> hey marcustomlinson, how are you?
[08:38] <marcustomlinson> duflu: bleh
[08:38] <marcustomlinson> seb128: I'm well thanks, cold over so that good. you?
[08:39] <didrocks> seb128: mieux, mais le week-enhd n’a pas été très fun :( et toi, ce FOSDEM ?
[08:41] <seb128> didrocks, sympa de voir du monde, mais sinon l'évenement moi c'est bof, trop de monde/bousculade, pas moyen d'accéders aux présentations, etc, comme d'hab :)
[08:41] <marcustomlinson> hmm, can a package in main Build-Depends on a package in universe?
[08:41] <seb128> marcustomlinson, I'm fine, a bit tired from fosdem w.e/got another small cold (on friday, so not due to the w.e) and having a sick child at home today
[08:41] <seb128> marcustomlinson, yes
[08:42] <didrocks> seb128: le FOSDEM quoi :p
[08:42] <seb128> marcustomlinson, it it doesn't result on a runtime depends
[08:42] <seb128> didrocks, oui :p
[08:44] <marcustomlinson> I'm trying to understand: "libreoffice-core/amd64 unsatisfiable Depends: libqrcodegencpp1 (>= 1.2.1)"
[08:44] <marcustomlinson> https://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#nspr
[08:44] <marcustomlinson> libqrcodegencpp-dev was added as a Build-Depends
[08:44] <didrocks> marcustomlinson: libqrcodegencpp1 is in universe
[08:44] <didrocks> while libreoffice-corelibreoffice-core is in main
[08:45] <didrocks> and so, you can’t dep from a package in main on something in universe
[08:45] <marcustomlinson> but I only added a Build-Depends
[08:45] <didrocks> the binary package links against it
[08:45] <marcustomlinson> hmm
[08:45] <seb128> marcustomlinson, if you use a lib you depends on it :)
[08:45] <didrocks> if you ldd the binaries/libs in libreoffice-core
[08:46] <seb128> otherwise ldd would be sad
[08:46] <didrocks> you probably have a dep on it
[08:46] <seb128> what didrocks is saying :)
[08:46] <marcustomlinson> yeah makes sense now
[08:46] <marcustomlinson> sigh
[08:52] <marcustomlinson> sorry to hear about your child seb128 :(
[08:59] <seb128> marcustomlinson, thx, it's minor, just a small stomach bug, he's better today but lack energy so he's staying here and resting
[09:04] <Laney> moin
[09:04] <marcustomlinson> hey Laney
[09:04] <didrocks> hey Laney
[09:13] <Laney> hey marcustomlinson didrocks
[09:14] <Laney> didrocks: mised you this weekend :(
[09:14] <Laney> missed*
[09:15] <duflu> Morning Laney
[09:16] <didrocks> Laney: I missed you as well :(
[09:29] <seb128> hey Laney, how are you? not too tired from the w.e/getting back late yesterday?
[09:35] <Laney> hey seb128
[09:35] <Laney> I'm ok, but yeah probably could have done with more resting time
[09:36] <Laney> what about you? ok trip back?
[09:38] <seb128> yeah, trip back was easy (and train is the best :)
[09:38] <seb128> I'm a bit tired from the w.e but it's fine, was worth it :)
[09:44] <Laney> :>
[09:48] <Laney> how did g-i migrate? :(
[09:50]  * duflu nods
[09:50] <Laney> http://autopkgtest.ubuntu.com/packages/gobject-introspection/focal/amd64
[09:50] <Laney> That top test run there
[09:50] <Laney> Steve re-triggered the test and asked it to use everything from focal-proposed
[09:50] <Laney> So it got the new python3-defaults, which made the test pass
[09:50] <Laney> And then it migrated *without* that new package because that's what happens
[09:50]  * duflu nods again
[09:50] <Laney> result: broken package in focal
[09:51] <duflu> bug 1861556
[09:51] <Laney> right
[09:51] <Laney> I wanted to know how that happened
[09:51] <Laney> because it was indeed caught by the testing
[10:35] <Wimpress> Morning desktopers o/
[10:35] <Wimpress> Looks like some of you had a good weekend :-)
[10:45] <marcustomlinson> morning Wimpress
[10:49] <didrocks> hey Wimpress
[11:00] <seb128> Laney, that's an annoying proposed-migration issue than adding triggers on proposed packages doesn't create migration-bindings with those :/
[11:01] <Laney> Sure, but that is how it is I'm afraid
[11:01] <Laney> and I don't know about you but I don't even know how you'd do that with all-proposed really
[11:05] <Laney> parse the logs or something?
[11:05] <Laney> make autopkgtest output the dpkg state at the end and synthesise dependencies on everything in there that is in proposed
[11:06] <seb128> all-proposed should be banned imho
[11:06] <seb128> it's a lazy way to not have to figure out what packages you need and try with those...
[11:08] <Laney> don't particularly disagree
[11:08] <Laney> people who run complex transitions find it to be useful
[11:08] <seb128> Laney, do you plan to point out to Steve that this retry with all-proposed was not a good idea and has been leading to a regression in focal?
[11:08] <Laney> probably if it wasn't available then better tooling would have been developed
[11:08] <seb128> right
[11:32] <Laney> ah, how rude, hi Wimpress
[11:55] <Laney> Wonder if this is going to work...
[11:56] <Laney> I hacked this silo to build without focal-proposed: https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/3909/+packages
[11:56] <Laney> will copy it to the archive now and hopefully it'll migrate and fix the g-i in focal
[11:58] <GunnarHj> Hey seb128!
[11:58] <GunnarHj> seb128: As regards bug #1861481, do you think we can special case the live session and stop lightdm from setting LANGUAGE there? If so, do you know of a proper way to test whether we are in a live session?
[12:19] <seb128> GunnarHj, hey, I don't know offhand how to test that we are on the a live session sorry
[12:20] <seb128> I will have a look
[12:20] <seb128> GunnarHj, I'm unsure to understand the logic at all there though, why do we limit the locales to be used to he availability of langpacks?
[12:28] <doko> seb128: this is a severe issue in gobject-introspection, the very same will happen in debian without any all-proposed=1. fix the dependencies
[12:28] <GunnarHj> seb128: Well, the reason was to create a sensible list of options for setting language via the GUI, where each option really represents a language. I.e. as opposed to listing every generated locale, which might result in a lot of non-sensible options for e.g. Englisth, Spanish, or Arabic. But we didn't consider the live session at the time...
[12:31] <GunnarHj> seb128: A hackish way to test for live session might be to check if the user is "ubuntu".
[12:56] <seb128> GunnarHj, let me poke a bit, using the username seems suboptimal, it's likely than 'ubuntu' is being picked for test installs, public accounts, etc
[12:57] <GunnarHj> seb128: Yeah, agreed. Would be great if you could figure out a more proper way.
[12:57] <seb128> doko, right, L_aney uploaded a fix for that, still the tests caught the regressions and someone decided it was a good idea to overide those results by trying with all-proposed, which resulted in the buggy version to migrate to focal
[12:57] <seb128> doko, I think we can agree that force migrating broken component isn't right?
[13:00] <doko> seb128: I'll reserve the next math/science migrations for you, and then you can decide again, ok?
[13:03] <seb128> doko, sounds good
[13:03] <seb128> better than force migrating buggy components :-)
[15:07] <ricotz> hey desktoppers
[15:08] <ricotz> seb128, would you have time for exiv2 https://people.ubuntu.com/~ricotz/sponsoring/ and syncing gexiv2 ?
[15:28] <seb128> ricotz, hum, that includes a soname change/transition, we should perhaps wait for the python-default one to clear off first?
[15:29] <seb128> marcustomlinson, so back to your component mismatch, you either need to rebuild libreoffice without that lib or to MIR it ... MIR might take a bit, I would recommend to unblock by removing the build-depends for now (I guess it's an optional one right?)
[15:29] <marcustomlinson> seb128: yes I've been on it
[15:29] <seb128> marcustomlinson, also usually it's a good habit to list new (build-)depends in the changelog, https://launchpad.net/ubuntu/+source/libreoffice/1:6.4.0-0ubuntu1 seems a bit minimalist
[15:30] <marcustomlinson> it's an import from Debian, all changelog entries from Debian imported too
[15:31] <ricotz> seb128, yeah, sure
[15:32] <seb128> marcustomlinson, whoever do the dpkg-buildpackage -S should include -v<previous changelog version which was in ubuntu> so the .changes includes those changelog entries you imported as well
[15:33] <seb128> marcustomlinson, e.g -v6.3.4-0ubuntu1
[15:33] <marcustomlinson> ok thanks
[15:33] <seb128> thank *you*
[15:33] <seb128> sorry for nitpicking, I was just trying to find the reference to the new depends :p
[15:34] <seb128> I didn't see it in the debian changelog on salsa and the diff on launchpad is 200M+ which firefox didn't really enjoy trying to display :)
[15:35] <marcustomlinson> seb128: the dep is optional, I'm just building in a ppa to make sure I didn't break anything before i do another uploade
[15:38] <Laney> if it's merging with Debian too, the usual practice is to append 'ubuntu1' to their version string
[15:38] <Laney> so 1:6.4.0-1 plus ubuntu changes -> 1:6.4.0-1ubuntu1
[15:39] <Laney> here endeth the packaging seminar
[15:48] <hellsworth> good morning desktopers!
[15:49] <didrocks> hey hellsworth
[15:50] <hellsworth> hi didrocks :)
[15:50] <Laney> ahoy hellsworth
[17:40] <hellsworth> marcustomlinson: ricotz did something happen to my write access to https://code.launchpad.net/~libreoffice/+git/libreoffice-snap ?
[17:41] <hellsworth> it seems that i can no longer use git+ssh://hellsworth@git.launchpad.net/~libreoffice/+git/libreoffice-snap to fetch or push
[17:42] <hellsworth> https://paste.ubuntu.com/p/CsWrN355Tr/
[17:43] <marcustomlinson> hellsworth: don't know, you're still an admin member
[17:44] <hellsworth> the only thing i can think of is that i added an ssh key to my launchpad and then cloned this repo from that system (which cloned fine) but now i can't fetch/push from any system
[17:45] <seb128> git remote -v ?
[17:45] <marcustomlinson> hellsworth: ok so I've lost the ability to push too
[17:45] <seb128> could be a launchpad problem, ask on #launchpad?
[17:45] <hellsworth> ah ok thanks for testing that marcustomlinson
[17:46] <hellsworth> i'll go ask in #launchpad :)
[17:46] <marcustomlinson> hellsworth: Laney just asked
[17:46] <hellsworth> ah ok i just joined #launchpad so missed it
[17:49] <marcustomlinson> seb128: this is known right? https://launchpadlibrarian.net/463362182/buildlog_ubuntu-focal-amd64.libreoffice_1%3A6.4.0-0ubuntu2_BUILDING.txt.gz
[17:49] <marcustomlinson> Depends: gobject-introspection (>= 1.32.0) but it is not going to be installed
[17:49] <seb128> I didn't poke at it, it's a bit weird
[17:49] <marcustomlinson> was some discussion about g-i broken this morning
[17:49] <seb128> Laney, ^ did building that update out of proposed make it uninstallable in proposed?
[17:49] <marcustomlinson> oh sorry, misread your response
[17:50] <seb128> the update also got bitten gcc b0rkage on s390x
[17:50] <seb128> which failed all the autopkgtests
[17:50] <seb128> marcustomlinson, but basically not your/libreoffice's problem
[17:50] <marcustomlinson> thanks
[17:53] <seb128> it's installable for me with proposed enabled though, weird
[18:51] <Laney> Wouldn't have expected it to
[18:51] <Laney> Poor old s390x though, that's annoying
[18:54] <Laney> have to retry those with the newer gcc-9 or?
[19:27] <Laney> looks like that works
[19:27] <Laney> arm64 too
[19:29] <seb128> great
[19:30] <seb128> do you have any smart way/script for helping with such situations?
[19:30] <Laney> maybe stop doing those retries please seb128
[19:30] <seb128> or just going through clicky click
[19:30] <Laney> retry-autopkgtest-regressions
[19:30] <seb128> Laney, yeah, sorry, I did like 5 and noticed someone did
[19:30] <seb128> and closed my tabs/browser
[19:31] <Laney> thx
[19:31] <Laney> they need the extra trigger on gcc-9
[19:31] <seb128> right, I also noticed the recent green results had that
[19:31] <seb128> makes sense
[19:31] <seb128> thx also for the r-a-r hint, I will check out this one :)
[19:32] <Laney> there's a hint in the --help about how to use it with scripts
[19:36] <seb128> thx
[19:36] <seb128> on that note, dinner time, have a nice evening desktopers!
[19:38] <seb128> marcustomlinson, the build-depends not installable for libreoffice looks like it could have to do with the boost changes xnox started uploading
[19:39] <Laney> it's likely g-i is not installable in focal-proposed too since it depends on python3 << 3.8 and python3 will be 3.8
[19:39] <Laney> will try to get it moved over and re-uploaded tonight
[19:43] <seb128> right
[19:48] <marcustomlinson> Ok cool thanks seb128 and Laney
[19:53] <xnox> i should be up to libreoffice soon.
[19:53] <xnox> boost deps should be mostly okish, since both old and new boost exist in the release pocket
[19:54] <Laney> looks like g-i according to chdist anyway
[19:54] <Laney>  gobject-introspection : Depends: python3 (< 3.8) but 3.8.0-3 is to be installed
[19:55] <Laney> maybe I should consider force-skiptest if it is heading towards the right conclusion after excuses refreshes next time
[19:56] <xnox> Laney:  your build of gobject-introspection looks borked on arm64....
[19:56] <xnox> Laney:  maybe we should just build it in focal-proposed proper and deal with migrating it?
[19:56] <doko> it's so conviniant if you can convince yourself ...
[19:57] <Laney> please :(
[19:57] <Laney> i'm trying to help you unbreak focal-proposed, can leave it if you want
[19:57] <Laney> xnox: just the known gcc-9 borkage I think
[19:57] <doko> is desktop not subscribed to gobject-introspection by intent?
[19:58] <xnox> Laney:  ah, ok!
[19:58] <xnox> in that case it is alright to go in.
[19:58] <xnox> Laney:  i think it is best to force migrate it; and get a rebuild against 3.8 in focal-proposed sooner.
[20:05] <ricotz> marcustomlinson, hi, it would make sense to do a more meaningful upload of libreoffice (given its buildtime)
[20:10] <ricotz> marcustomlinson, pushed the proper way using internal qrcodegen
[20:13] <doko> why not using the external one?
[20:16] <ricotz> not my decision
[20:17] <ricotz> there are still several library which would require a MIR and therefore the internal ones are used