[01:35] good morning [03:39] Oops. Morning callmepk [03:48] Hi duflu [05:19] good morning desktoppers [05:20] hi oSoMoN [05:25] Hi oSoMoN [05:25] And hi seb128 [05:25] Back later... [05:26] hey duflu, and see you later :) [05:26] hey callmepk, duflu [05:26] salut seb128 [05:26] lut oSoMoN, hey callmepk [05:26] how are you? [05:26] I'm good, had a good and uninterrupted night of sleep \o/ [05:26] how are you yourself? [05:32] I'm good, up too early to my taste though :-) [05:32] but it's friday, hopefully more sleep tomorrow [05:33] hi seb128 [05:34] hi all [05:36] salut jibel [05:40] bonjour oSoMoN [05:47] lut jibel, comment ça va ? [05:52] seb128, ça va bien, merci. [06:05] Hi jibel [06:14] morning duflu [06:22] good morning [06:22] Hi didrocks [06:24] hey duflu [06:32] lut didrocks [06:39] salut seb128, ça va ? [06:40] didrocks, ouais, un peu fatigué, réveillé avant 7h par le petit qui a fait un cauchemard :p [06:40] et toi ? [06:41] seb128: réveillé à 5h, mais sinon, ça va [06:41] tu 'gagnes' :p [06:42] façon de parler [06:53] question for the python people on the channel [06:53] https://git.launchpad.net/software-properties/tree/softwareproperties/gtk/LivepatchPage.py#n38 [06:54] why is the [06:54] STRING = _()... [06:54] then later [06:54] https://git.launchpad.net/software-properties/tree/softwareproperties/gtk/LivepatchPage.py#n276 [06:54] print(STRING) [06:54] not working? [06:55] (there is an parathensis error on l42 which makes it use the wrong string, it should give the version without the substitution to gettext, but fixing that isn't enough) [07:05] jamesh, ^ seems like the kind of problem you might have a clue about? [07:06] * jamesh looks [07:09] what error are you getting? [07:10] I don't see any print() calls in that code [07:17] it looks correct, how do you run this code? [07:17] good morning callmepk duflu oSoMoN seb128 jibel didrocks jamesh [07:18] Hi marcustomlinson [07:18] happy friday :) [07:19] Bonjour marcustomlinson [07:20] hey marcustomlinson [07:20] seb128, oops got a crash of g-c-c when I clicked on "connect to" [07:20] jamesh, jibel, sorry it was maybe not clear [07:20] https://git.launchpad.net/software-properties/tree/softwareproperties/gtk/LivepatchPage.py#n276 [07:20] error_message=self.GENERIC_ERR_MSG [07:21] that displays in english and doesn't fetch the gettext translation [07:21] if I replace it by [07:21] error_message=_('Canonical Livepatch has experienced an internal error.' [07:21] ' Please refer to {} for further information.').format(COMMON_ISSUE_URL) [07:21] it works [07:21] (which is the same string as the const) [07:22] g-c-c crash https://errors.ubuntu.com/bucket/?id=/usr/bin/gnome-control-center%3A11%3Acc_panel_get_title_widget%3Aactivate_panel%3Aset_active_panel_from_id%3Alaunch_panel_activated%3Ag_closure_invoke [07:22] it's bug #1873272 [07:22] bug 1873272 in software-properties (Ubuntu) "Error in livepatch is not translated" [Low,Incomplete] https://launchpad.net/bugs/1873272 [07:23] jibel, the gcc bug is bug #1862553 which we discussed/assigned during the team meeting this week [07:23] bug 1862553 in gnome-control-center (Ubuntu) "gnome-control-center crashed with SIGSEGV in cc_panel_get_title_widget()" [High,In progress] https://launchpad.net/bugs/1862553 [07:23] the livepatch one [07:23] seb128: the code as written would try to look up "Canonical Livepatch has experienced an internal error. Please refer to https://wiki.ubuntu.com/Kernel/Livepatch#CommonIssues for further information." in the message catalogue, which doesn't match any tagged string constant [07:24] so you'd get the english version [07:24] moving the format() call outside of _() means you'd be looking up the string without the URL substituted [07:24] jamesh, see what I wrote earlier [07:24] (there is an parathensis error on l42 which makes it use the wrong string, it should give the version without the substitution to gettext, but fixing that isn't enough) [07:24] also I fixed that in the paste I just did [07:25] seb128: is there a translation in the message catalogue? [07:25] moving a ) before the .format [07:25] yes [07:25] as said, if I do [07:25] if I replace it by [07:25] error_message=_('Canonical Livepatch has experienced an internal error.' [07:25] ' Please refer to {} for further information.').format(COMMON_ISSUE_URL) [07:25] it works [07:25] (which is the same string as the const) [07:25] I've the UI displayed in french in the GUI then [07:26] changing self._trigger_ui_update(skip=True, error_message=self.GENERIC_ERR_MSG) I meant [07:26] there is something about using the const that is wrong [07:26] seb128: the code as written in that branch is wrong [07:26] I wonder if that's because it does the query at object creation and the domain is not set or something? [07:26] seb128: your change is what it should be doing [07:27] salut didrocks, good morning marcustomlinson, hey jamesh [07:29] jamesh, the change works, I don't understand why having it the way it's now isn't equivalent [07:29] is that because it does the call to gettext too early with the variable assignement at the beginning of the class? [07:30] seb128: the _() call serves two purposes: (1) it is something that the xgettext tool recognises to extract string constants from source code to create the message catalogue [07:30] right [07:30] and (2) at runtime convert an english string to the equivalent native language string [07:30] right :) [07:31] as written, for (1) the call recognises the string "... Please refer to {} for further information" [07:31] but for (2) it is looking up the string "... Please refer to http://... for further information" [07:31] sorry I'm not being clear [07:31] because there is a format() call within the _() invocation [07:31] let me try a pastebin [07:32] and that doesn't match anything in the catalogue [07:33] salut oSoMoN, ça va ? [07:33] jamesh, https://paste.ubuntu.com/p/vbmFxZtYDk/ [07:34] jamesh, the wrong parameter/string being passed to gettext I fixed and is ortogonal, it's just not enough to have the code working [07:34] jamesh, the pastebin is a summary of the thing that is confusing me [07:35] jamesh, I guess in the first case it might do the translate fetch when creating the object which might be before the gettext init/domain is set? [07:35] where in the second case it does it when the function is called? [07:35] seb128: ohh. This would be an ordering issue with the bindtextdomain call [07:36] so in the first case the fetch is not done at the call time? [07:36] The class statement is executed before gettext is configured [07:36] right, that's what I guess [07:36] so there is no message catalogue to look up [07:37] right [07:38] what would be the 'right' fix? [07:38] Fixing this would either involve delaying import of the module until after gettext is set up, or not using module or class level attributes that are evaluated at module import time [07:38] let me try if moving the constants defintion to the init is enough [07:38] if not I will go for the simple fix [07:39] that would be enough [07:39] ok, let me do that then [07:39] jamesh, thanks! and sorry for not being clear with the question, the {} / ) issue which added on top did make it even more confusing [07:40] jibel, 'connect to', where was that? you were were setting up livepatch or hit the issue from another workflow? [07:40] seb128: an alternative would be to edit https://git.launchpad.net/software-properties/tree/software-properties-gtk -- move softwareproperties import to below the textdomain() call [07:40] or move the textdomain() call up higher [07:41] right [07:42] that feels unclear having some code moved before the imports though [07:42] I will go for just moving the const to the init [07:42] jamesh, thanks again! [07:42] seb128: Are you sure that's the only problem string though? [07:43] you are never sure, we didn't get other reports and there is no other gettext calls made in a class directly from what I can see [07:44] you would just move the gettext/textdomain() init before the import? [07:47] didrocks, is there a particul process for yaru-theme uploads? Trevinho asking me to sponsor his fix from https://github.com/ubuntu/yaru/commits/master , I can just do the normal 'manual' package build/dput right? [07:48] seb128: just ensure that you push the tag to the upstream repo, but then yes. gbp builpackage -S + dput [07:48] didrocks, thx [07:49] the tag is already pushed by Trevinho [07:49] so just upload to ubuntu [07:49] didrocks, thx [08:02] hey [08:03] morning Laney [08:07] hey Laney [08:08] hey Laney, marcustomlinson, happy friday! [08:14] Hi Laney [08:15] moin didrocks marcustomlinson seb128 duflu [08:15] happy final freeze friday [08:15] fff [08:16] indeed! [08:17] or fffff [08:17] f³ / f⁵ [08:22] Alliteration for human beings [08:24] * Laney waits impatiently for ISOs [08:30] weeee my wish is granted [08:32] :) [08:32] popey, https://bugs.mojang.com/browse/MCL-13512 should probably be reopened, it would be best if they moved to a non years-old-legacy Depends still [09:34] marcustomlinson: looks like you got some replies on the atk bug! [09:36] yeah good stuff [09:55] I rebooted [09:55] My desktop locks up. [09:56] I see black screen, mouse pointer, cannot get to VT or anything. [09:58] xnox, nvidia? fractional scaling? [09:58] And I was about to EOW :/ [09:58] Should be non-nvidia but I do have it. And have highdpi infinity screen. [09:59] I am in emergency target now, let me check what's up [10:00] infinity sounds very high [10:00] Meh, it's 2015 infinity [10:01] Old kernel booted fine [10:01] * duflu nods [10:02] Anyway, let me apply all updates, if still broken something is really and. [10:02] *bad [10:03] Covering my eyes now [11:02] seb128 looks like debian are on it [11:36] ok that was disturbing [11:37] in the middle of a dist-upgrade and everything died [11:37] a whole load of [11:37] Apr 17 12:32:39 raleigh systemd[1]: wpa_supplicant.service: Unexpected error response from GetNameOwner(): Connection terminated [11:37] Apr 17 12:32:39 raleigh systemd[1]: udisks2.service: Unexpected error response from GetNameOwner(): Connection terminated [11:37] from various things [11:37] which sounds like the system bus went away? [11:59] Laney, bug #1871538 [11:59] bug 1871538 in dbus (Ubuntu Focal) "dbus timeout-ed during an upgrade, taking services down including gdm" [High,Incomplete] https://launchpad.net/bugs/1871538 [11:59] disturbing bug, also unsure how to debug it :/ [12:00] popey, right, I asked because I was suprised to see a comment on the Debian BTS that the upstream request to use a non transitional package had been closed [12:00] good they followed up now though [12:00] sounds right [12:00] any idea what happened? [12:00] no :-( [12:00] I did some dbus reloads and systemd re-execs in a loop, didn't trigger it [12:01] I didn't see an obvious clue in the journal [12:01] no [12:01] it looks like they restarted ok [12:07] ah [12:07] maybe [12:08] yeah [12:08] so if you reload the dbus configuration and the systemd one at the same time, this happens [12:21] meh that only worked once [12:22] oho, ok, put them both in a loop and then you'll probably get it [12:23] systemd bug? [12:23] dunno [12:24] that seems likely, if it's losing the connection to the system bus when it shouldn't [12:29] I should stop doing this on my host [12:51] seb128: do you remember why we upgraded atk to 2.35? Did something else require it? [12:52] kenvandine and I are wondering if we could temporarily downgrade to 2.34 again for focal until that segfault issue is resolved [12:52] marcustomlinson, we follow GNOME every cycle [12:52] it's a GNOME component [12:53] version downgrades suck, can't you just revert the one problematic commit? [12:53] We worry that any patch to fix this issue wouldn't have enough time for testing [12:53] I would still revert the commit [12:53] rather than play weird 3.35.is.3.34 update [12:53] I hate those versions [12:53] since it's a well identified commit [12:54] it is? [12:54] I mean it may well be, I'll look [12:54] I didn't think it was [12:56] 'Of particular interest is the introduction of the atk_socket_component_real_get_extents() function, and the NEWS entry: "Make AtkSocket get_extents return parent extents by default (MR!23)"." [12:56] is a misleading comment then [12:56] it sounded like !23 was the obvious candidate [12:56] sorry, though I think you are right [12:56] https://gitlab.gnome.org/GNOME/atk/-/commit/71ecedd9f1d12ad74f660a7b2827e5574ca7e510 [12:56] agreed [12:57] I would start by trying to revert that one [12:57] I'm going to add that as a release blocker on our cool new wiki post [12:58] this may take a while I've not touched this project before [12:58] I mean the packaging [12:59] what branch do I work on? focal-proposed? [12:59] marcustomlinson, apt-get source atk1.0 [12:59] there is no ubuntu branch [13:00] also I would ask on the upstream bug if they think it's safe to revert [13:00] I'd make a branch if it were me [13:00] I've no real knowledge of the a11y stack, I'm not sure the feature/version you want to revert isn't providing something that at-spi or other component might use [13:01] (I would probably personnally feel better try to go the fix way and try the patch draft upstream provided) [13:01] at least start by getting feedback on that, if it works we might have a proper fix tested by monday [13:02] marcustomlinson, if you want to follow Laney's lead, the vcs in Debian is https://salsa.debian.org/gnome-team/atk [13:03] unsure if you have an account on salsa yet [13:03] might be a good opportunity to get one otherwise :) [13:04] wait a second, I'll just make the branch, that'll be easier [13:04] thx [13:09] atk ubuntu/master 75979e0 Sebastien Bacher debian/ (5 files in 2 dirs) * Import Debian changes 2.35.1-1ubuntu1 * https://deb.li/LW7I [13:09] atk ubuntu/master 472cb96 Iain Lane debian/ control control.in gbp.conf * control, gbp.conf: Update for ubuntu/master branch * https://deb.li/3qqVA [13:10] there [13:11] bah, why did we fail to update/merge 2.36? [13:11] arg, version was still have 2.34 as the stable serie so it's in the yellow section :( [13:12] annoying [13:12] can be SRUed though [13:12] Samuel says reverting that commit will be fine [13:12] great [13:12] this one: https://gitlab.gnome.org/GNOME/atk/-/commit/71ecedd9f1d12ad74f660a7b2827e5574ca7e510 [13:12] I suggest doing that and then testing the proposed patch, giving feedback on that and hopefully later we can replace the revert with a fix [13:13] If I confirm it to work could I just ask you Laney to do the revert and publish etc [13:13] I'm sorry, sounds awfully lazy but it will certainly take much longer for me [13:14] I'll test this now [13:14] seb128, bug 1873454 is surely something to fix for the release. It's a g-c-c crash when a user tries to register his U1 account in software-properties to use livepatch. [13:15] bug 1873454 in gnome-control-center (Ubuntu) "/usr/bin/gnome-control-center:11:cc_panel_get_title_widget:activate_panel:set_active_panel_from_id:launch_panel_activated:g_closure_invoke" [High,Confirmed] https://launchpad.net/bugs/1873454 [13:15] that patch is not a patch for this issue [13:15] it's a "it's sorta kinda like this" [13:16] yeah sure [13:16] lunch now [13:26] seb128: I have an account on salsa but it's appended with "-guest" [13:26] don't remember why [13:30] oh right coz I'm not a Debian Developer [13:35] jibel, right, we said that during the weekly meeting [13:35] marcustomlinson, correct :) [13:37] jibel, did you see my comment/question this morning? [13:38] jibel, https://irclogs.ubuntu.com/2020/04/17/%23ubuntu-desktop.html#t07:22 [13:39] seb128, no, I missed it sorry [13:39] np [13:39] but your description seems to confirm it's a dup [13:40] also Robert proposed https://gitlab.gnome.org/GNOME/gnome-control-center/-/merge_requests/746 earlier today but he didn''t make it clear if that was a fix for this issue, I'm going to test that next [13:41] yes, it's the same case than comment 8. I duplicated the report created from errors.u.c [13:42] thx [13:56] Laney: reverting that commit resolves the crash without any regression as far as I can see - took the screenreader for a spin on a couple other apps too. [13:56] The commit is: 4b5ed8630c5ae9ef37884c4199da8463b7e5a127 [14:01] Laney: and for the changelog, the bug is (LP: #1870508) [14:01] Launchpad bug 1870508 in ubiquity (Ubuntu) "ubiquity crashed with SIGSEGV in g_type_check_instance_is_a()" [High,Confirmed] https://launchpad.net/bugs/1870508 [14:01] that's the only changes needed [14:13] mutter debian/master Marco Trevisan * [open] merge request !60: d/p: Fallback to closed laptop lid configuration if no other available * https://deb.li/Nf8d [14:15] seb128: thanks for pointing out my own reference to the problematic commit (facepalm). I blame lockdown [14:15] marcustomlinson, no problem, glad that you figured it out, nice to hit friday afternoon with problems solved :) [14:15] digging out what the heck was happening fried my brain [14:16] mutter debian/master Marco Trevisan * [update] merge request !60: d/p: Fallback to closed laptop lid configuration if no other available * https://deb.li/Nf8d [14:20] mutter Marco Trevisan 155899 * commented commit 6661980 * https://deb.li/ig0tV [14:30] atk ubuntu/master 3138b37 Iain Lane debian/patches/ atk_requires_glib series Revert-atksocket-make-get_extents-return-parent-extents-b.patch * Revert "atksocket: make get_extents return parent extents by default" * https://deb.li/3zfce [14:30] atk ubuntu/master 9466ac0 Iain Lane debian/changelog * Update changelog * https://deb.li/33HhI [14:30] atk ubuntu/master 558773e Iain Lane debian/changelog * Finalise changelog * https://deb.li/iO9yA [14:38] thanks Laney [14:40] kenvandine: Can we talk about https://gitlab.gnome.org/Community/Ubuntu/gtk-common-themes/-/issues/13 ? It's directly affecting Ubuntu Studio at this point. [14:41] np marcustomlinson [14:41] would appreciate it if you could check out any patches that get proposed [14:41] sure [14:41] atk signed tags 1976303 Iain Lane ubuntu/2.35.1-1ubuntu2 * atk1.0 Debian release 2.35.1-1ubuntu2 * https://deb.li/M0Ig [14:42] Eickmeyer: we really wanted to hold off on adding more themes because the auto theme installation work is coming real soon [14:42] kenvandine: Unfortunately, because of the lack of that theme, people are installing snap-store and it's unusable. [14:43] i see [14:43] https://bugs.launchpad.net/snap-store/+bug/1867417 [14:43] Ubuntu bug 1867417 in snap-store "snap-store launches without theme (transparent) when using non standard themes" [Undecided,New] [14:43] Eickmeyer: i'd be interested in seeing how much bigger that would make the snap [14:44] Eickmeyer: any chance you could prepare a PR adding it? It should be trivial to follow one of the other examples [14:44] Well, it's the default theme we've been carrying in Ubuntu Studio for a year now. [14:44] I could look into it, but I honestly lack the know-how. [14:44] understood [14:45] Eickmeyer: can you provide a link to the git source for the theme? [14:46] kenvandine: https://github.com/nana-4/materia-theme [14:47] Eickmeyer: i can probably take a stab at it tonight [14:47] kenvandine: I'd appreciate it. :) [15:17] morning folks [15:21] hey hellsworth, how are you? happy friday! [15:22] i am ok.. tired from a baby up for a few hours again teething but coffee is slowly waking me up :) [15:22] happy friday indeed [15:22] hey hellsworth [15:23] hey marcustomlinson how's it going? [15:23] hellsworth: meh. also tired :P [15:23] most importantly.. how's Isla :) [15:25] she's doing great :) her sleep is starting to stretch at night which is helping a bit [15:25] oh horray for that! [15:26] how's your little one doing? [15:27] she has teeth coming in.. so last night she was up from 1:30-5. it was rough. wants to be held for hours. and then she finally went to sleep and woke up at the usual 7:30 [15:27] i'll be glad when all the teeth are in [15:27] and when daycare is open again [15:27] i dream of daycare.. [15:27] ok yeah... [15:28] /s/ok/oh [15:31] hey hellsworth :) [15:32] hi there didrocks [15:32] good luck with the teeth [15:32] i'll take all the luck i can get :) [15:32] heh [15:32] spoiler alert: it takes time though :p [15:32] i know.. i asked the pediatrician when it will stop and she looked at me and kind of laughed and said not for a long time [15:33] roh :) [16:12] Eickmeyer: can you test a channel of gtk-common-themes for me? [16:13] kenvandine: Sure. What channel? [16:13] Eickmeyer: snap refresh --channel-=latest/stable/materia gtk-common-themes [16:13] I'll get on it. :) [16:13] Eickmeyer: snap refresh --channel=latest/stable/materia gtk-common-themes [16:13] typo the first time :) [16:13] Heh [16:14] Eickmeyer: i added all the variants for both gtk3 and gtk2 [16:15] Eickmeyer: if snap-store has a background process running you'll need to kill it first [16:15] killall snap-store [16:16] kenvandine: Looks good! [16:16] great [16:18] igordhossegor: I see that someone arranged with a redirect for bitlocker and rst after all. Do you know how that happened? I would like to know, because I fear that the bzr branch we use for maintaining help.ubuntu.com is no longer in sync with what's actually on the web server. [16:18] GunnarHJ i asked IS yesterday [16:18] for the redirect before we talked [16:19] i can ask them to remove the server redirect and then you can do the one you want [16:21] GunnarHj i asked IS to remove the redirects [16:21] after that it should work fine [16:21] (for how you are going to do it - with html pages and the redirects you can make) [16:22] igordhossegor: Ok.. I don't have any own preferences. But I pushed the files to the branch yesterday, and it ought to work without redirects. My only concern atm is that some IS guy bypassed the branch. [16:24] kenvandine: I have confirmation from others that it fixes the transparency issue. [16:25] Eickmeyer: ok, i'll have that in the candidate channel today and probably stable by monday [16:25] Perfect. Thanks! :) [16:45] :( adding gamemode after final freeze [16:53] indeed :-/ [17:17] good night all, have a great we [17:26] have a good one marcustomlinson [18:31] Happy Friday all ! [18:31] Trying again - on 20.04 gnome desktop - how to show top bar on all desktops with multi monitors ? Thanks ! [19:10] Eickmeyer: it's in candidate now [19:28] kenvandine: Thanks! [20:13] have a good week-end everyone! [20:14] you too!