=== jamesh_ is now known as jamesh === pieq_ is now known as pieq === cpaelzer__ is now known as cpaelzer === pieq_ is now known as pieq [07:47] Does anyone know when GTK4 will hit Hirsute? [07:58] jo-erlend: given gtk4 is only packaged in debian experimental currently, it will not be part of hirsute, so you'll have to wait until hirsute+1 [07:59] Though the final release of GTK4 is in NEW now, so may make it into unstable soon. But I say there's no rush. [08:04] Well there is for me. I've been dying to get started. I guess I'll just switch to Debian. [08:06] what's already using gtk4 that you can't wait? just being curious [08:09] tarzeau, my program? :) [08:13] jo-erlend: url? [08:14] tarzeau, the program doesn't exist because I don't have access to GTK4 yet. I really want to get it going. [08:16] I guess I could build it, but I'm lazy :) [08:16] so it's not something that already works with gtk3 (or 2 or 1)? === pieq_ is now known as pieq [08:17] what was your reason to pick gtk and not say something else? [08:17] Only supporting GTK4 seems a bit... [08:25] tarzeau, well, I love GTK, but since my program will depend heavily on trees and grids and that's fundamentally changed in GTK4, there's no reason for me to write it in GTK3 first. [08:26] i mean you're aware there are other gui toolkits? (and i'm not refering to motif or tcl/tk) [08:26] although gtk is not very old, i wouldn't call it very modern, nor portable [08:27] tarzeaum, [08:27] tarzeau, yes there are lots of other toolkits. There are other operating systems too. But I happen to love GTK. [14:06] jo-erlend, is it ok to have gtk in a ppa? [14:08] https://launchpad.net/~costamagnagianfranco/+archive/ubuntu/locutusofborg-ppa/+packages [14:12] I don't see any reason against it IMO if it's just for testing and for yourself purposes. [15:14] ItzSwirlz, was that an answer to me? [15:14] lol [15:14] jo-erlend, I had to sync also pango-1.0 from debian, but meh, it built fine, so you are free to use it if you like [15:18] LocutusOfBorg yeah [15:20] ItzSwirlz, thanks for allowing me to do that, but I did to answer another user lol [15:21] ddstreet: rbalint: Laney: does one of you happen to know the location (and who has e.g. restart powers) of systemd ci tests running on Ubuntu infra? [15:22] I can derive between ubuntu https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic/bionic/s390x/s/systemd/20201215_153456_752ce@/log.gz and systemd-ci https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic-upstream-systemd-ci-systemd-ci/bionic/s390x/s/systemd-upstream/20201229_192335_7ee5a@/log.gz logs that there is a different [15:22] "autopkgtest-bionic-upstream-systemd-ci-systemd-ci" scope [15:22] I was wondering if there would be a counterpart to https://autopkgtest.ubuntu.com/packages/s/systemd/bionic/s390x that is about those upstream tests somewhereß [15:22] ? [15:23] there's no web UI, no, sorry :( [15:23] We (bluca and me) would want to try re-starting a job for https://github.com/systemd/systemd/pull/18095#issuecomment-752218666 [15:24] what else than web could we use to retry those without pushing a new commit - is there something hidden in the github integration that can rerun it? [15:24] Not by me, I don't have the keys - ddstreet does and maybe some upstream people [15:25] ok, thank you a lot for the info Laney! [15:25] most likely waiting for ddstreet to be around then ... [15:26] cpaelzer for restarting, it would require the github private key for the systemd project, which I don't have, so i can't restart the upstream systemd ci test runs; however keszybz does have a script that he got from mpitti that allows him to restart tests [15:27] oh that's good to know, let me ping keszybz [15:27] doh [15:27] sounds like those keys should be shared to ddstreet to me :> [15:27] at least we got the info we needed, thanks ddstreet [15:27] I hereby proposed KEY="password" [15:27] bluca: can you work with that from here? [15:28] yes I can go down the restart route and cross fingers, but if it fails again due to selinux linkage, then I'm going to be back to the start... [15:28] build failures aren't usually flaky though [15:28] ... [15:29] for context, there's a PR where the bionic s390x fails to link a binary, and we can reproduce it on the PR consistently, at least when we looked [15:29] but cannot repro it on a VM or on cpaelzer's testbed [15:29] cpaelzer bluca re: a better way to get the upstream ci test results, i use autopkgtest-manager to pull them down to my local system; it's still a bit painful (and there are LOTS of upstream test results so it takes a while) but better than hand-inputting the urls [15:29] Laney: we have build the same branch in a container (including the PPA that upstream enables) just fine multiple times and iterating on config opts [15:30] the error is: [15:30] /usr/lib/s390x-linux-gnu/libselinux.so: undefined reference to `pthread_once' [15:30] when lining the cdrom_id binary which the PR changes - but no change whatsoever to build flags, only code changes [15:30] so it's too much of a coincidence to ignore, but a complete WTF otherwise [15:31] (no selinux changes, no pthread changes) [15:31] cpaelzer: still, I doubt it's a flake but rather some difference in the way the tests are run [15:31] if it was an unrelated binary or flaky I'd shrug it, but it's the source being touched and it is consistent [15:32] Maybe you can use Canonistack to run them, see the first line in a failing log for the command used [15:32] https://wiki.ubuntu.com/ProposedMigration#Reproducing_tests_in_the_cloud [15:34] thanks - cpaelzer do you have superpowers to access that? [15:34] in the meanwhile I pinged keszybz [15:34] fingers crossed ;-) [15:35] we had 4/5 builds in a row fail like that, not just a couple - only hope at the moment is that something was temporarily broken in the image/env that day... === hellswor1 is now known as hellsworth [15:57] big nope all around: it failed again, exactly the same [15:57] https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic-upstream-systemd-ci-systemd-ci/bionic/s390x/s/systemd-upstream/20210104_155304_a4b6e@/log.gz [15:59] we are really stuck on this, and would appreciate any kind of help [16:08] linker command is identical in working vs failing case [16:09] cpaelzer: https://paste.ubuntu.com/p/4jjvRFwpmW/ <- that works for me to reproduce the issue [16:09] you'll want to fix the --keyname at least [16:10] unless you named your key in tribute to me, in which case aw shucks [16:12] awesome, I'll give it a shot after the calls here [16:12] thanks Laney [16:12] thank you both [16:27] Laney heh i live in Raleigh, tho i assume yours is named for the bike company :) [16:27] heh [16:28] well if I elide some tedious details then yes it is [16:28] they're from nottingham which is where I am [16:28] they do make great bikes [16:29] their huge factory used to cover the site where the university's computer science building now is === rharper_ is now known as rharper === ijohnson is now known as ijohnson|lunch === ijohnson|lunch is now known as ijohnson [20:51] rafaeldtinoco: Hey! Looks like the script worked for some packages, but I'm having trouble updating agordejo with is in the ubuntustudio-audio seed. Not sure why it didn't get picked-up by the script. [20:51] lemme check for u [20:52] Eickmeyer[m]: it did not pick agordejo [20:52] Weird. [20:52] let me c [20:53] ok [20:53] Eickmeyer[m]: is it really in the seed ? [20:54] my germinated does not show it (when I ran) [20:54] Weird. Lemme see... [20:54] i could find it under hirsute*universe*binary [20:56] Let me check the metas real quick, but it's saying "everything up to date" on the seed itself. [20:56] I'll query the acls by hand (and update if needed) [20:57] I'm running update on the metas, I'm not seeing it showed-up in the recommends for some reason. [21:01] Eickmeyer[m]: I added it to ubuntustudio pkgset [21:02] make sure its in the seed (for the next time I run the script) [21:02] rafaeldtinoco: Ok, thanks. :) [21:02] let me know if it works ;) [21:04] rafaeldtinoco: Yeah, the meta updater came up with "No changes found." so I don't know what's going on there. [21:05] hum [21:08] https://git.launchpad.net/~ubuntustudio-dev/ubuntu-seeds/+git/ubuntustudio/commit/?id=90dc0cfa79983b45d88c1a2a774f09010f4fc544 [21:08] you added qwinff in 12-28 [21:08] lets see that one [21:11] rafaeldtinoco: Well, I just uploaded the new version of agordejo, so that worked. Not sure about qwinff yet. [21:12] I have nothing to update there. [21:12] Hopefully this next week that won't be an issue anyhow since I'm applying for MOTU. [21:12] you will have the same issue.. the changes I had done automatically were [21:12] Would add sources to ubuntustudio: [21:12] ayatana-ido [21:12] libayatana-appindicator [21:12] libayatana-indicator [21:12] mate-optimus [21:12] python-setproctitle [21:12] sphinx [21:13] That makes sense since mate-optimus pulls that in. We added it for hirsute. It's not just a mate tool and should really be renamed, but it works for all DEs. [21:15] just ping me if you find any other package that should have been added to your pkgset and I'll add it [21:15] I have to review the pkgset script further this cycle anyway... [21:24] rafaeldtinoco: Ok, thanks for your help! [21:24] np, anytime