=== micahg_ is now known as micahg [06:44] Good morning [06:44] hi pitti! [06:45] wie war dein Wochenende? [06:45] hey larsu! gut, wir waren wieder in Dresden [06:46] und bei Dir? [06:47] auch schön. hab gestern ein bisschen mit poettering über fsckd gesprochen :) [06:47] (es war mal wieder open source meetup in berlin) [06:49] ah, nett! [06:50] good morning desktopers [06:50] hey larsu pitti [06:50] hi seb128! [06:51] Howdie all! [06:51] hey tmpRAOF [06:52] “vestigial” is an excellent word that I like to use in commit messages! [06:52] Hey seb128! [07:08] good morning [07:09] bonjour didrocks, ça va ? [07:14] pitti: ça va bien, et toi ? [07:14] didrocks: ça va aussi ! le week-end à Dresden était bien [07:16] bonjour didrocks! [07:18] hey larsu [07:45] pitti: seems lennart answered (and mention the fsckd patch as well) [07:45] :( [07:46] didrocks: I saw, I'll answer today [07:50] didrocks-fsckd vs the universe ;-) [07:50] seems so :/ [07:51] * didrocks back to autopkgtests… why lightdm isn't starting into my test bed [07:55] seb128: especially the previous email stated that we tried without it… [07:55] anyway [08:15] didrocks: missing "needs-recommends"? [08:16] didrocks: lightdm without recommends doesn't start; needs Xorg, mesa, and stuff [08:18] pitti: yeah, I just validated that theory ;) [08:18] thanks for confirming! [08:19] ok, so now plymouth… [08:23] btw, I had a weird transition issue yesterday where I ended up with gdm after a dist-upgrade [08:23] removing it left my system in an unbootable state (with systemd) [08:23] reconfiguring lightdm made it work [08:23] as did booting with upstart [08:23] not sure if this helps anyone... [08:26] when you removed gdm, you should had have a prompt with which asked you which dm to choose, right? [08:27] on install, yes; on removal, not sure if there shold be a dialog if there's only one DM left [08:27] (what would you choose?) [08:28] yeah, but it should have set /e/X/d-d-m anyway to the remaining one [08:35] morning [08:41] didrocks, pitti: didn't have that [08:41] but lightdm was definitely still installed [08:43] would have been interested to check /etc/X11/default-display-manager if you reproduce it (adding on my list) [08:43] * didrocks has the test writing to stderr, but not when I try to reproduce it of course :/ [08:44] hey all [08:44] seb128, can you test baobab (with titlebar patch) on ppa:ubuntu-gnome-packaging/staging [08:45] hey darkxst [08:45] darkxst, sure [08:46] evening darkxst [08:46] seb128, upstream look like they will take the patch but want to use XDG_CURRENT_DESKTOP==Unity, rather the gtk-shell-uses-headers [08:46] hey didrocks [08:47] darkxst, going to make e.g xfce sad [08:47] wait, what?! [08:47] we have a gtk-shell-uses-headers [08:47] ? [08:47] larsu, gtk-dialogs-use-headers even [08:47] https://bugzilla.gnome.org/show_bug.cgi?id=745346 [08:47] Gnome bug 745346 in general "Use traditional title bars on Unity" [Normal,New] [08:48] darkxst: oh, that. Don't use that for header bars [08:48] others have [08:49] it's wrong, though === dbarth is now known as dbarth-afk [08:50] also, a toolbar with a title? [08:50] I'd get rid of the toolbar entirely, replacing the menu button with a menu bar [08:51] and merging the title into the title bar [08:56] darkxst: mind chaning the patch to simply hiding the headerbar when on !GNOME? [09:01] larsu. what about the "Back" button then though? [09:01] darkxst: ugh, right... [09:01] * larsu hates toolbars with only one or two buttons [09:05] morning [09:05] hi Laney! [09:06] it's a Laney! [09:06] what's up? [09:06] right on time for some headerbar discussion [09:06] how was your weekend? [09:06] hey Laney [09:06] larsu, Laney, speaking about headerbar, gnome-log-viewer has some in utopic/vivid [09:07] * Laney sinks [09:07] seb128: and no menubar anymore. Why do we update those apps before we port them? [09:08] larsu, we don't, that one came with a direct sync from Debian [09:09] what's gnome-log-viewer? [09:09] the log viewer app from GNOME [09:09] yeah, I know, the name is not obvious :p [09:09] I mean which package [09:10] seb128, gnome-logs? [09:10] because [09:10] laney@raleigh> apt-file search gnome-log-viewer ~ [09:10] * larsu was struggling to explain it without using "gnome", "log", and "view" [09:10] laney@raleigh> ~ [09:10] which is systemd only? [09:10] no [09:10] the old thing we install for ever [09:10] gnome-system-log? [09:10] or how is it called [09:10] ya, gnome-system-log [09:10] just type "log" in your dash [09:10] oh this thing [09:11] bug #1376565 [09:11] bug 1376565 in gnome-system-log (Ubuntu) "Utopic system log wrong design" [Medium,Confirmed] https://launchpad.net/bugs/1376565 [09:11] "wrong" *giggle* [09:11] I would almost be of the opinion to don't bother, keep the headerbar for those sort of non important applications [09:11] also, the bug uses "should". We should warn people when doing that [09:12] but we would need to have them themed to look at least a bit similar to our theme [09:12] and to have unity handle them correctly of course, which is still not the case... Trevinho! [09:13] seb128: I tend to agree. Maybe style the close buttons better, like mitya57 started in https://code.launchpad.net/~mitya57/ubuntu-themes/window-buttons/+merge/250526 [09:13] larsu, I never understood why not using "should" in bug description, I seem to remember mpt doesn't like that but it doesn't make sense to me [09:13] description or title [09:13] seb128: because it implies a solution [09:13] well, "nautilus should not crash" [09:13] doesn't imply a solution [09:13] a bug report should do at least one thing: describe a problem [09:13] just state a fact [09:13] and optionally describe possible ways of solving the problem [09:14] no it doesn't [09:14] cf ^ [09:14] seb128: not true in all cases. It crashes by design when running out of memory, not finding .ui or .gschema files, etc. [09:14] I'm +1 on theming as mitigation [09:14] "should not crash" is just not true in the general case [09:15] "nautilus should not hit segfault due to invalid use of memory" [09:15] ? [09:15] hm, I wonder why I didn't get an email for mitya57's comments - I'm subscribed according to the right sidebar [09:16] seb128: it should, though [09:16] seb128: you might be right for some cases of course, but generally we should (!) avoid 'should' [09:16] larsu, ok, what about "indicator-session should not format your drive when installed"? ;-) [09:16] thou shalt not double free [09:17] seb128: rephrase as "indicator-session formats my hard-drive when I install it" [09:17] larsu, well, I find them both as correct [09:17] but yeah, in those very obvious cases the difference isn't large [09:17] anyway [09:17] :) [09:17] can we theme the gtk decoration to be closer from ours? [09:18] like thiner bar, round buttons, colored close? [09:18] seb128: yes. I've played around with the branch a bit this morning and the things you describe are possible [09:18] seb128: thinner bar will be hard because of the subtitle widget [09:18] it's already quite crammed [09:19] hum, k [09:19] but we really, really need Trevinho to land frame extents before we can do this [09:19] right [09:19] because we also need to add client-side shadows [09:19] well, theming would put us a step closer [09:19] and it doesn't conflict with the unity work [09:19] and there's a spacing problem which I indend to fix in gtk in some way (see that MR) [09:19] seb128: indeed [09:20] nice [09:22] seb128: "alternatives.log — updated yesterday 12:00" would be better for us and give us slighly thinner title bars (note that we're still bound by the buttons on the right side) [09:22] seb128: but we can't change that from the theme sadly [09:22] gtk-has-thin-toolbar [09:23] yeah... [09:23] no... [09:23] we're getting into weird territory [09:24] hahaha [09:24] ↑ poe's law [09:25] Laney: I was sincere.. [09:25] about what? [09:25] not liking adding another xsetting. seriously, app authors will have to test all kinds of combinations [09:26] we already have a few of those, I'd like to keep that number as low as possible [09:26] yes it was me who wasn't being serious [09:26] that is a stupid thing to have a setting for [09:27] oh. oops :) [09:28] we can probably just live with it [09:28] hey Laney [09:28] make the font smaller and dim the label [09:29] which Adwaita seems to be doing despite the wider bar [09:29] maybe we can do something similar for baobab? [09:29] hi darkxst [09:30] did I start a theme war ;) [09:30] rather you rekindled it ;) [09:31] it's a good war [09:31] noble and just [09:31] Game of thrones vs Ubuntu ? [09:37] the new gnome-shell sass theme is a little rough around the edges right now [09:40] MATE is not fully compatible with glib 2.43 [09:40] We've created the required patches. [09:40] Tested in Debian Jessie and Ubuntu 15.04. [09:40] Debian have denied our unblock request to patch Debian, due to freeze for Jessie final. [09:40] Therefore I need to get these patches applied in Ubuntu directly. [09:41] How do I progress that? [09:41] Laney, Your thoughts? ^^^^^^ [09:41] flexiondotorg: patches to what? [09:41] flexiondotorg: file a bug with a pointer to the patches (or a branch) [09:42] I think you should put them in the sponsor queue for normal processing [09:42] Laney, 5 MATE components that require a patch. [09:42] Laney, All 5 packages have been synced directly from Debian in the past. [09:43] flexiondotorg: do you know who is maintaining those packages in ubuntu? [09:43] Sounds like you want to make them diverge [09:43] larsu, Well I am on the upstream Debian maintainers team. But so far, I'm not aware of anyone maintain them in Ubuntu. [09:44] larsu, But it should probably be me considering I'm the lead for Ubuntu MATE. [09:44] Laney, Thought so. [09:44] Laney, No idea where to start on this though. [09:44] File bug, attach debdiff, subscribe ubuntu-sponsors [09:44] Laney, Thanks. [09:44] Or: upload into unstable and don't care that it's not going to go into jessie, sync [09:47] Laney, We've been told to not do that. Not even to experimental 😞 [09:47] ... [09:48] Laney, I'll start preparing debdiffs. [09:48] Up to you and your sponsor I guess. If experimental's empty then I would use it personally. [09:48] 'kay, diff away [09:50] Laney, We (The Debian maintainers for MATE) want to upload to experimental. But we have been expressly told not to do so. [09:50] By one of your sponsors? [09:50] We filed an unblock request. The DDs on our team have been told to not upload these patches packages to anywhere. [09:51] restricting uploads to debian experimental seems weird [09:51] We've spent to weekend trying to get some other DDs to help fight our corner. But no. [09:51] are you sure that's what happened? [09:51] I think it is political, because these patches "fix" Ubuntu 15.04. [09:52] But without them, Jessie will work. [09:52] Technically, out implementation for gsettings was incorrect in some places. [09:52] I don't know who would have the authority to tell you not to upload to experimental [09:52] So, these patches actually make MATE technically correct. [09:53] Laney, I can dig up the unblock. [09:53] Please [10:01] Laney, https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=779340 [10:01] Debian bug 779340 in release.debian.org "unblock (pre-approval): marco/1.8.2+dfsg1-5" [Normal,Open] [10:08] flexiondotorg: I don't think it's right to interpret anything in there as preventing an upload to experimental [10:09] Laney, I believe our DD has had other emails. I am no privvy to them. [10:09] I'm still pushing that end. [10:09] But, I'll prepare debdiffs. [10:09] As you wish. I think that sunweaver might have the wrong end of the stick on this oen. [10:11] Laney, Possibly. [10:11] Laney, I'll preapre the debdiffs so I'm ready to go either way. [10:12] I don't want to diverge. [10:52] Laney, https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=779340#24 [10:52] Debian bug 779340 in release.debian.org "unblock (pre-approval): marco/1.8.2+dfsg1-5" [Normal,Open] [10:52] what about it? [10:55] Laney, I'm agreeing with your reading. [10:55] There is nothing there to suggest we couldn't upload to experimental. [10:56] Indeed. mehdi's on the release team so he's commenting on jessie matters, and then suggesting some things you might do to help the release happen sooner so you can get your stuff into testing again. [10:57] Nothing which stops you uploading to exp (or even sid, really, but that might get in the way of future fixes) [10:57] Laney, Well my take is glib 2.43.x is in experimental. Therefore, if we upload MATE fixes to experiemental we are "fixing" the issue that resides there. [10:57] Yes, that's fine [10:57] So if you can convince sunweaver, more's the better, otherwise toss the patches in the queue [11:08] Laney, I'm making progress in Debian land 😃 [11:08] \o/ [11:21] Laney, Would you be able to upload some updated package for Ubuntu MATE? [11:21] *packages [11:21] flexiondotorg: I'll be patch piloting at some point soon - toss them in the queue and I will look then [11:22] Laney, OK. New words to me there. "patch piloting"? What queue? [11:28] flexiondotorg: 1. https://wiki.ubuntu.com/UbuntuDevelopment/CodeReviews?action=show&redirect=PatchPilot#Patch_Pilots [11:28] 2. http://reqorts.qa.ubuntu.com/reports/sponsoring/ (subscribe ubuntu-sponsors) [11:39] Laney, I'm not clear how I add something to the queue. [11:40] flexiondotorg: Take your bug and add ubuntu-sponsors as a subscriber [11:50] Laney, Understood. Sorry for delayed replay. [12:01] Laney, do I raise the bugs against 'ubuntu-mate'? [12:01] flexiondotorg: against the source package you want updating [12:01] Laney, Thanks. === MacSlow is now known as MacSlow|lunch [12:45] hum, those using vivid (and systemd maybe?), is picking another user from the indicator-session bringing you to the greeter to enter the password for that user [12:46] ? [12:46] didrocks, Laney, pitti ^ [12:46] it just locks unity for me [12:49] guest ok or a real user? [12:50] real user [12:51] guest opens directly a session [12:51] it's not sending you to the greeter to auth [12:51] yeah just wondering where the problem is [12:51] * Laney makes a user [12:54] this password strength checker is mean [12:55] seb128: once my current tests finshes, I can give it a try [12:55] seb128: don't want to mess up with it :) [12:55] * Laney is also waiting for a glib build to finish :P [12:55] well, let's wait first for Laney's test [12:55] ahah [12:55] nearly done [12:55] dh_shlibdeps [12:56] Status: successful [12:56] Laney: here, the testbed is installing 300M of deps, not nice turnaround fix/run timing [12:56] * Laney clicks the big bad button [12:56] brb, session restart [12:56] seb128: I'm at the greeter [12:56] with the user selected [12:56] Laney, ok, so it works for you :-( [12:56] Laney, upstart or systemd? [12:56] systemd [12:56] bah [12:56] I wonder why it doesn't work here [12:57] sorry! [12:57] no worry [12:57] needs more debugging I guess [12:57] I'm unsure if that "bring to greeter" is lightdm's doing? [12:57] but brb first [12:57] dm-tool iirc [12:57] do some playing there I guess [12:57] need to restart that session, click screwed it [12:58] still got those borders in the greeter [12:58] * Laney files a bug [12:58] oh, someone at the door, brb [12:59] didrocks you didn't debug more that click/schroot/ecryptfs/homedir mount issue we looked at in Bruxelles? [12:59] seb128: that usually works for me, hang on [13:00] seb128: just tried it again, works fine [13:00] pitti, danke [13:00] seb128: didn't get the time yet to get there [13:00] seb128: but I have an ecryptfs directory working here [13:01] the other issue was click/schroot, and I didn't start looking at the sdk issues yet [13:01] (the non cleaning up part) [13:03] didrocks, k, that issue is starting to really annoy me [13:03] I wanted to try to publish a click, but I can't install a framework due to that [13:04] Laney, I've files the bugs. Is it appropriate for me to assign importance to them for 'Ubuntu'? [13:04] seb128: did you talk about it with the sdk guys? [13:04] let's see if mvo replies to my pings, I'm a bit stucked meanwhile :-/ [13:04] didrocks, define "sdk" :-) [13:04] didrocks, zoltan&co bounce those issues to click, which is mvo's afaik [13:05] seb128: well, click is part of the sdk to me, but yeah, fundation's for that part of the sdk :) [13:10] flexiondotorg: No real benefit unless you want to for your own purposes [13:11] didrocks: did you get to look at v2 patch for transient presets? === alan_g is now known as alan_g|lunch [13:18] Laney, I've added the bugs to ubuntu-amte and set the priority there. [13:19] k [13:28] xnox: didn't yet === alan_g|lunch is now known as alan_g [14:09] bah [14:10] mvo, same issue when running qtcreator from another user [14:10] it still tries to unmount the homedir from my user [14:10] even more annoying, it does manage to unmount the ecvryptfs mount [14:11] I've to run again ecryptfs-mount-private to get back to my user dir [14:14] seb128: meh :/ === MacSlow|lunch is now known as MacSlow [14:42] good morning, all [14:43] hey desrt, had a good w.e? [14:43] not really -- sick again :( [14:43] :-( [14:44] do you feel better today? [14:44] not really :p [14:44] get some rest! [14:44] * desrt has been working, but has annoying cough that won't go away [14:44] worse is that i'm going to a sleater-kinney concert tonight [14:44] i hope it's loud, 'cause i don't want to be annoying everyone around me with my cough :p [15:01] wooot [15:01] got it to work [15:02] mvo, it's not your fault, seems to be https://bugs.launchpad.net/ubuntu/+source/ecryptfs-utils/+bug/769595 [15:02] Launchpad bug 769595 in schroot (Ubuntu) "Encrypted home not mountable under chroot" [High,Triaged] [15:02] seb128: puhhh, sounds like I got lucky this time :) [15:03] mvo, I'm unsure if click could help though, since it's an issue with the default fstab [15:03] though that the click config uses rbind, which something suggested in that bug [15:06] seb128: fyi: libreoffice 4.4.1 has been in the ppa for more than a week without any issues, so we should bump to that version on vivid this week. i'll prepare the package. [15:06] Sweet5hark, great, give me a ping when you are ready for sponsoring [15:06] seb128: willdo [15:09] Laney, when are you being a pilot or whatever? [15:09] in a minute [15:09] usually people expect to wait a bit of time for their patches to be sponsored [15:09] OK. Just seeing if I have time to push some more stuff today. [15:10] There's people looking at the queue every day [15:10] Laney, Cool. [15:22] k [15:22] mvo, so yeah, it's somewhat a click bug too, same bug as stock schroot [15:23] changing [15:23] /home /home none rw,rbind 0 0 [15:23] to [15:23] /home/user /home/user none rw,bind 0 0 [15:23] resolves the issue with ecryptfs [15:24] didrocks, ^ just as a fyi [15:36] seb128: oh, I think there was a branch for this actually at some point [15:38] mvo, https://code.launchpad.net/~mvo/click/lp1319790-chroot-fstab/+merge/219728 ? === czajkowski is now known as laurac [15:46] seb128: yeah [15:47] seb128: nice catch! === rickspencer3_ is now known as rickspencer3 [15:53] seb128: Here's my plan for the Chromium bug reports at errors.u.c . Tell me if it makes sense to you? ... [15:53] didrocks, mvo, opened https://bugs.launchpad.net/ubuntu/+source/click/+bug/1427264 about it (just as a fyi) [15:53] Launchpad bug 1427264 in click (Ubuntu) "using ecryptfs, creating frameworks fail to bind mount issues" [High,New] [15:53] qengho, k [15:57] seb128: First finding is that the GPU vendors in reports roughly match the population's hardware. This suggests it's wide-ranging, and not a particular hardware feature that causes it. Lots of Intel, AMD, some nvidia, and even "vmware". The GPU watchdog timeout is 10 seconds. That's a long time, but not so long that I'm sure it's an actual hang. Very busy CPU could cause it too. Some reports mention business. [15:58] seb128: So, I plan to 1) submit a apport bugpattern that scans the dmesg for specific GPU-related messages and route them to gpu-specific bug reports. [15:59] seb128: 2) extend the chromium watchdog timeout by a bit. If it's really a hang, then another 10 seconds won't help, and if it's not, the user gets a machine back without a watchdog killing something. [16:01] qengho, increasing the timeout makes sense to me yeah, having an apport script to collect more info as well [16:06] seb128: In addition to load, what additional info should the apport hook collect? Already getting "hardware", df, drm, dmesg, lspci. [16:07] qengho, those seem enouygh to me [16:07] seb128: thanks for listening! [16:07] yw! [16:08] thanks for investigating the issue [16:08] :) === greyback__ is now known as greyback [16:28] larsu, http://paste.ubuntu.com/10503326/ ... bug in libmessaging-menu or in the user of the lib? (g_variant_unref: assertion 'value->ref_count > 0' failed) === laurac is now known as czajkowski [16:30] seb128: good question :) My first guess would be in the library [16:31] but we've never seen this before... [16:31] larsu, or we never paid attention to the warnings in logs :-) [16:33] seb128: indeed. I'll have a look. Do you have a bug? [16:33] larsu, no, I'm asking where to file it :-) [16:34] seb128: file it for i-messages please. I'll reassign if necessary [16:36] Laney, saw you upload ubuntu-mate-artwork. Thanks. [16:37] Laney, Do you want me to ad LP: #xxx references to the changelog in the other packages? [16:37] Yes please [16:37] larsu, https://bugs.launchpad.net/ubuntu/+source/indicator-messages/+bug/1427286 [16:37] Launchpad bug 1427286 in indicator-messages (Ubuntu) "g_variant_unref warning displayed when calling back after a missed call " [Undecided,New] [16:38] seb128: thanks. The bt doesn't correspond to trunk - do you know which version it is of? [16:38] rtm 248?! [16:39] * larsu wonders how to find out which i-messages version is shipped on there [16:39] larsu, https://launchpad.net/ubuntu-rtm/+source/indicator-messages/13.10.1+14.10.20141007-0ubuntu1 [16:40] seb128: thanks! [16:40] larsu, dget https://launchpad.net/ubuntu-rtm/+source/indicator-messages/13.10.1+14.10.20141007-0ubuntu1 [16:40] ups [16:40] larsu, dget https://launchpad.net/ubuntu-rtm/+archive/primary/+files/indicator-messages_13.10.1%2B14.10.20141007-0ubuntu1.dsc [16:40] still having firefox often not taking my right click/copy url orders :-/ [16:40] larsu, yw [16:42] Laney, Will do. [16:44] seb128: turns out it is an issue in telephony-service... [16:44] * larsu reassigns [16:45] larsu, not surprised, thanks! [16:46] seb128: floating refs ftw! [16:46] heh [16:47] seb128: should I ping someone about it? [16:47] I described the problem and how to fix it [16:47] larsu, no, we can do that in some time if nobody responds [16:47] ok [16:48] but seems like tiagosh looks at those bugs [16:48] so no need to be nagging (yet) [16:48] larsu, danke [17:36] My IRC client adds (I think) duplicate entries to the messaging menu. Anyone have a tip of where to start to debug what's wrong? http://i.imgur.com/7TscLqn.png [17:39] neat icons... [17:40] :) [17:43] Laney, All done. [17:43] qengho: in d-feet, navigate to smuxi on the session bus and call "ListSources" on the com.canonical.indicator.messages.application interface (should be on at least one object) [17:43] qengho: I have to run now, but feel free to file a bug with the ouput of that call [17:43] and ping me with it [17:44] larsu: Thanks. [18:11] see you tomorrow guys! [18:11] me too [18:11] bye! === alan_g is now known as alan_g|EOD [18:16]