[05:26] Good morning [07:09] Sweetshark, ping, wrt https://bugs.launchpad.net/ubuntu/+source/libreoffice/+bug/1200277 . Where are the desktop files linked. I dont see this in debian/rules . [07:10] Launchpad bug 1200277 in libreoffice (Ubuntu) "[LibreOffice] - libreoffice-writer.desktop when drag/drop to desktop, 100% broken. " [Undecided,New] [09:03] hey [09:08] good morning desktopers! [09:11] today is GTK 3.10 to trusty day [09:11] larsu, be ready ;-) [09:11] hah [09:11] yues [09:11] yesterday was mesa 10 day :D [09:12] did you finally managed to get that one to build? [09:12] 10.0.0-0ubuntu1 built on all platforms, but tjaalton wanted llvmpipe for armhf :P [09:13] it can wait [09:13] since it doesn't even build right now :) [09:13] it does [09:14] 10.0.0-1ubuntu2 release (main) [09:14] ah [09:14] hey seb128 mlankhorst tjaalton [09:15] Laney, howday, happy tuesday [09:15] grumpy tuesday for me [09:15] howdy even [09:15] ;-) [09:15] but hey [09:15] oh? [09:15] do you need to build webkit again? :p [09:15] haha nothing work related [09:15] need to return some christmas present I bought and now I can't find the receipt [09:15] -> angry [09:15] oh :/ [09:16] oh well [09:16] * seb128 is not going to comment since he has no organisation whatsoever, for papers, receipts, expense claims, etc [09:16] haha [09:16] I know where to find stuff on my hdd though ;-) [09:16] I'm going to call the shop and see if they will take it back anyway, but not too hopeful [09:17] tracker <3 [09:17] ebay to the rescue if you can't return I guess... [09:17] jibel: I need youuuuuuuuuuuuuu [09:17] jibel: (pango1.0 blocks on the old notify-osd) [09:22] Laney, looking [09:23] thanks [09:28] Laney, fixed, next britney run should promote it (or next next) [09:28] jibel: thanks, what was the bug? [09:29] did you fix the cause or the symptom? :-) [09:30] Laney, I fixed the symptom, there was no result for the latest version of the tuple pango/notify-osd [09:31] mmm, would be nice to eliminate the root cause ideally [09:31] Laney, I verified that they both passed and forced a result. another way to fix that is to restart the test of notifi-osd [09:31] y [09:31] investigating [09:44] niiice [09:45] the new webkit patch seems to make s-c happy [09:45] autopkgtest for libreoffice 1:4.1.3-0ubuntu2: FAIL (Jenkins: public, private) [09:45] Not considered [09:45] aaaaaaaaaaaaaa [09:45] Laney, I'm going to upload webkit again to trusty unless you have a reason I shouldn't [09:45] yes I do [09:45] please wait for this stuff to migrate first [09:45] I figured out that might be the case [09:45] should just be one round [09:45] which is why I asked [09:45] sure [09:45] after forcing this [09:47] shrug [09:48] we have a stack of stuff doing the buggy g_source_remove() [09:49] larsu, desrt: I blame you for those ;-) [09:49] joke aside [09:49] what sort of bugs do that lead to? leaks/extra wakeups? [09:49] seb128: me?! [09:50] larsu, congrats, you have been classified in the "glib maintainers" group ;-) [09:50] * larsu runs as fast as his legs can take him [09:51] larsu, joke aside do you know if it's fine to just ignore those or if we better fix them this cycle (e.g what sort of side effect those bugs have) [09:51] what else has it? [09:52] software-center [09:52] yelp [09:52] just run into those [09:52] by trying to run some webkit rdepends clients [09:52] seb128: the removing wrong source ID stuff? [09:52] so I guess we are going to find others [09:52] desrt, hey, happy mid-night? [09:52] desrt, yes [09:52] it's almost 5am here! [09:53] * desrt is keeping pitti hours [09:53] desrt: oh, so that's "good morning" for you? or good night? :-) [09:53] you are almost ready to move to Germany ;-) [09:53] seb128: I can't think of a side effect other than "it's ugly". I'm sure desrt can [09:53] so here's the story: [09:54] back in the good old days we allows removing non-existent sources [09:54] mostly because it was pretty harmless -- if the source didn't exist, we could just return FALSE... no harm done [09:54] but we started having apps that used sources _a lot_ [09:55] which means that we had to start recycling source IDs [09:55] which means that the source ID that you used to have may no longer belong to you -- it may have been handed out again [09:55] so removing a non-existent source ID now that we recycle them means that you could be removing someone else's source [09:55] which is obviously very very bad [09:55] wait ... which app used more than 2^32 sources? [09:55] so we aim to catch people who are doing this [09:55] larsu: not hard if you have any kind of server [09:56] ok [09:56] desrt: didn't think we had many of those [09:56] so in practice for e.g yelp it's not likely to be an issue [09:56] still a wrong usage and good to fix [09:56] right [09:57] I'm just going to file upstream bugs about those for now [09:57] seb128: please do [09:57] try to get a few fixed on the way (like the notify-osd one, thanks larsu and Laney) [09:57] might not be too hard to give patches for, if that one's anything to go by [09:57] right [09:57] well you have to understand the logic of the code first [09:58] it's for sure not days, but can easily be an hour [09:58] * seb128 is trying to wrap other things before eoy, so not likely to pick on those [09:58] starting by GTK 3.10 to trusty today [09:58] * seb128 goes back to do that [09:58] awesome! http://blogs.gnome.org/mclasen/2013/12/09/a-terminal-surprise/ [10:00] that is neat [10:00] we just need to unblock new g-t versions in ubuntu :P [10:01] what are they blocked on? [10:01] umm, let me try to remember [10:01] ugh, we're still on 3.6 :-/ [10:02] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=706065 [10:02] Debian bug 706065 in libvte-2.90-common "libvte-2.90-common: /etc/profile.d/vte.sh is not sourced by interactive shells" [Normal,Open] [10:03] seb128: did that webkit patch get reviewed upstream yet? [10:04] ah thanks [10:05] I guess we could do what they did [10:05] but no doubt people would complain [10:06] Laney, Company wrote it, kov test built it and asked me to test if it was fixing our issue [10:06] Laney, so I guess "yes" [10:07] okay, I checked the bug and didn't see any new patch there is all ;-) [10:07] Laney, we discussed that on IRC yesterday, I guess kov is going to update the bug today when I come back saying it works [10:07] nod [10:16] seb128: can we remove the old webkit source? [10:16] Laney, it was building the same binaries right? (e.g nothing that is still in use there) [10:17] yeah: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=727015 [10:17] Debian bug 727015 in ftp.debian.org "RM: webkit -- ROM; renamed to webkitgtk" [Normal,Open] [10:18] Laney, ok, I'm going to have a look once I'm done with my webkit upload and GTK 3.10 [10:19] neat [10:44] ritz: seems (from extracts of my launchpad spam) you found the part about /usr/share/application symlinks yourself? [10:58] Sweetshark, yes [10:58] I could be wrong though [10:58] still building, to test [11:09] Laney, new GTK uploaded, do you think we should block in proposed for today just to lead CI tests and stuff run against it before it hit the archive? [11:17] larsu, https://launchpad.net/ubuntu/+source/gtk+3.0/3.10.6-0ubuntu1 [11:17] \o/ [11:17] ;-) [11:20] larsu, btw, do you know if Marek does IRC? could we give him a nudge to review https://bugzilla.gnome.org/show_bug.cgi?id=692931 ? [11:20] Gnome bug 692931 in printing "The list of printers should be searchable/sortable" [Enhancement,Unconfirmed] [11:20] larsu, it's the small patch from attente we discussed some time ago, if you remember [11:21] ah right! ya, I'll ping him [11:21] thanks [11:21] if I remember his nick :D [11:21] though I wonder why that compare function has all his pango crazyness [11:22] seb128: you mean http://ci.ubuntu.com/smokeng/trusty/desktop/ ? ;-) [11:23] Laney, well, I was more meaning e.g if it creates issues for unity7 or compiz CI/autopilot tests [11:25] oh, well, feel free if you like: https://lists.ubuntu.com/archives/ubuntu-devel-announce/2013-October/001068.html [11:25] when are those tests run usually? [11:26] not sure nowadays [11:26] didrocks, ^ [11:26] Laney, thanks for url, I have forgotten about that one [11:27] seb128: see #gtk. mkasik tries to have a look next week [11:27] larsu, thanks! [11:29] seb128: sorry, in a hangout [11:29] I don't know about the smoke tests for dekstop TBH [11:29] didrocks, man, you live in hangouts nowadays :/ [11:31] seems so… [11:31] hrm [11:31] why can't I launch u-s-s in another language? [11:33] Laney, what's the error? [11:33] no error, just doesn't get translated [11:33] Laney, unset LANG LANGUAGE [11:33] I can see that it does read the right mo file [11:33] locale says de_DE [11:34] Laney, do the unset [11:34] "Konten" is translated but nothing else [11:34] did [11:34] then export LC_ALL=de_DE.UTF-8 if needed [11:34] yeah, it is [11:35] Laney, weeeird [11:35] Konten comes from the online accounts translation domain [11:36] LANG= LANGUAGE= LC_ALL=de_DE.UTF-8 strace -e open system-settings 2>&1 | grep system-settings.mo [11:36] open("/usr/share//locale/de/LC_MESSAGES/ubuntu-system-settings.mo", O_RDONLY) = 19 [11:36] displaying all translated [11:37] Laney, do you have language-pack-de installed (e.g the locale generated) (I guess otherwise the online account translation would probably not work) [11:38] yep I see that [11:38] the strace [11:38] other things like apt output are translated [11:38] bah [11:39] it opens the same file? [11:39] yea but mine doesn't have the // [11:39] locale -a lists de_DE? [11:39] yep [11:39] what do you mean the //? [11:40] in your string from strace [11:40] oh [11:40] Laney, msgunfmt /usr/share//locale/de/LC_MESSAGES/ubuntu-system-settings.mo [11:40] ? [11:40] looks fine [11:41] also: didn't know that program, nice [11:41] Laney, can you pastebin the output of the command I gave you [11:41] ;-) [11:42] http://paste.ubuntu.com/6550634/ [11:42] Laney, sorry, I meant the strace one [11:43] oh [11:43] ok [11:43] just want to check that one [11:43] if the output looks fine I'm going to blame it on qt5.2 (are you still on that, or did you downgrade after testing the other bug fixes?) [11:43] http://paste.ubuntu.com/6550646/ [11:43] I only took that in a chroot [11:43] I think ... [11:44] Laney, that strace has no successfull open [11:44] open("/usr/local/share//locale/de/LC_MESSAGES/ubuntu-system-settings.mo", O_RDONLY) = -1 ENOENT (No such file or directory) [11:44] open("/usr/share/locale/de/LC_MESSAGES/ubuntu-system-settings.mo", O_RDONLY) = 12? [11:44] shrug, was looking at the bottom :p [11:45] maybe it forgets it there [11:46] right [11:46] I don't understand why it tries to load the translations again [11:49] heh [11:49] I symlinked it into /usr/local/... and now it works [11:49] what is this [11:52] do you have a plugin in /local or something? [11:52] seems not from the strace [11:53] Laney, can you pastebin your "env"? [11:56] seb128: http://paste.ubuntu.com/6550708/ [11:57] hum, nothing weird that I can [11:57] Laney, I don't get it, your strace doesn't try to load the translations from the langpack location in the first loading blob [11:58] then it does another one later than loads from local and langpacks [11:58] dunno :/ [11:58] is that specific to settings? e.g can you load other qt5 apps translated fine? what about gtk? [11:58] what's a translated qt app? [11:58] ubuntuone-control-panel is ok [12:00] & rhythmbox [12:06] seb128: do you have a /usr/local/share/locale/de/LC_MESSAGES directory on your system? [12:07] if I remove that it works [12:09] Laney, no [12:09] I don't sudo make install stuff [12:09] my /usr/local is clean [12:10] ok, well can you try making it for a test? [12:10] yeah, I confirm it makes translations not work [12:11] interesting [12:11] libc bug? [12:11] in the fallback path to load translations [12:11] well other translations still work [12:11] I guess qt or u-s-s [12:11] don't insult u-s-s :p [12:11] tbh that does hardly anything with translations [12:12] I guess you can try gallery-app or notes-app to see if that's qt5/uitk issue [12:12] I even commented out the initTr and it didn't change anything [12:12] true [12:12] yeah confirmed [12:12] it breaks the gallery too [12:12] k [12:12] "good" [12:12] toolkit or qt issue [12:12] I guess next test would be to try a qt5 app not using our toolkit [12:13] * Laney tries to find one [12:17] I'm blaming the UITK just from reading the code [12:18] that would be my bet as well (just from gut feeling though) [12:18] just file a bug there and let them reassing to qt if needed? [12:18] BAH [12:18] let me show you the block in question [12:19] Jenkins Failure - trusty-adt-gtk-3.0 26 [12:19] you'll believe me too [12:19] you bastard [12:19] oh no :( [12:19] you=jenkins [12:19] why so much hate [12:19] https://jenkins.qa.ubuntu.com/job/trusty-adt-gtk-3.0/26/ARCH=amd64,label=adt/console [12:19] UnicodeEncodeError: 'ascii' codec can't encode character u'\u2018' in position 68: ordinal not in range(128) [12:19] https://bazaar.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/trunk/view/head:/modules/Ubuntu/Components/plugin/i18n.cpp#L107 [12:20] is it because of that warning? [12:20] I bet it is [12:20] not sure [12:20] I'm trying locally [12:20] stderr usually makes them fail [12:20] unless you explicitly allow that [12:20] and the upstream tests passed [12:20] @uitk: no cookie [12:20] seb128: Error: "uitk:" is not a valid command. [12:21] but yeah, that's buggy [12:21] the "build", "run" [12:22] * seb128 reads http://packaging.ubuntu.com/html/auto-pkg-test.html [12:22] I don't exerce those stuff often enough [12:22] yeah that stuff is nice [12:24] jibel, pitti: is [12:24] " File "/usr/bin/adt-run", line 159, in psummary [12:24] print >>summary_stream, m [12:24] UnicodeEncodeError: 'ascii' codec can't encode character u'\u2018' in position 68: ordinal not in range(128)" [12:25] something known/worth reporting a bug about (where)? [12:30] seb128: yes, please file a bug about it with a reproducer [12:31] sounds like a non-ASCII test case name in an ASCII environment [12:31] (LANG=C or so) [12:31] pitti, ok, against what? [12:31] pitti, that's https://jenkins.qa.ubuntu.com/job/trusty-adt-gtk-3.0/26/ARCH=amd64,label=adt/console [12:31] seb128: autopkgtest [12:31] pitti, gtk 3.10 upload to trusty [12:31] pitti, I'm trying locally now [12:31] nice [12:33] seb128: so, no other regressions found except for the weird dialog boxes? (missing label break) [12:34] pitti, none that didn't get fixed (larsu spent a week poking overlay-scrollbar to find a way to make them work with the new GTK rendering behaviour), some theme issues and other small stuff, larsu got them all resolved before upload [12:36] pitti, should that work? [12:36] $ ./bin/run-adt-test -r trusty -a i386 file:///tmp/gtk+3.0-3.10.6 gtk+3.0 [12:36] 2013-12-10 13:35:50: Info: Using default VM configuration file [12:36] 2013-12-10 13:35:50: Info: Cleaning up [12:36] $ [12:36] work/do something [12:36] * seb128 goes back to read http://packaging.ubuntu.com/html/auto-pkg-test.html [12:37] did you prepare-testbed first? [12:37] i don't think you need the package name at the end there [12:37] seb128: yes, that looks fine [12:37] Laney: you do, otherwise it builds the package again [12:37] oh, I never include it :-o [12:37] well, unless you actually want that; could be [12:37] I mostly write new tests for existing packages in which case I don't want to build/install the debs [12:38] oh, I didn't prepare-testbed no, I did it a while ago and I was unsure if that was needed again, but I was not on trusty [12:38] * seb128 does it [12:38] seb128: right, you need to build a VM first [12:38] I though I had one [12:38] seb128: alternatively you can run the test on your local desktop, which is much faster [12:38] but if I had it was saucy anyway [12:38] seb128: and I guess for gtk that shoudl be fine [12:38] just using adt-run? [12:38] seb128: in the source tree, try "sudo adt-run -B ./ --- adt-virt-null" ? [12:40] bah [12:40] not enough random it says! [12:40] * seb128 creates entropie [12:41] yeah, it started ;-) [12:42] so yeah, I can reproduce the issue [12:42] it doesn't like the deprecation warning [12:46] pitti, danke, filed https://bugs.launchpad.net/ubuntu/+source/autopkgtest/+bug/1259529 and I'm fixing the GTK test to not throw a deprecation warning [12:46] Launchpad bug 1259529 in autopkgtest (Ubuntu) "UnicodeEncodeError in adt-run psummary call" [Undecided,New] [12:46] seb128: thanks, will look at it [12:47] thanks [12:49] seb128: meeting reminder reminder btw [12:50] Laney, oh, thanks [12:50] Laney, done [12:50] ta === greyback is now known as greyback|lunch === mjohnson15_2 is now known as mjohnson15 === Ursinha-afk is now known as Ursinha === Ursinha is now known as Guest48115 === Guest48115 is now known as Ursula === mjohnson15_2 is now known as mjohnson15 === jdstrand_ is now known as jdstrand === greyback|lunch is now known as greyback [15:30] larsu: is there any particular reason that the sound-indicator uses "audio-volume-low-zero-panel" and "audio-volume-high-panel" for the icons next to the volume-scale instead of symbolic icons? [15:31] ochosi: the only reason I know of is that the symbolic ones didn't exist back then and humanity doesn't have them [15:31] oh, it's meeting time! [15:32] Aiee! [15:32] o/ [15:33] qengho, Sweetshark, mlankhorst, Laney, tkamppeter_, desrt, attente, larsu: hey, it's meeting time! [15:33] larsu: same goes for nm-applet seemingly and the volume-icon in the panel itself also isn't symbolic, would you consider a patch for this? [15:33] seb128: you're late. i'm disappointed. [15:33] G'day maaatu [15:33] desrt, yeah, just back from exercice [15:34] nice weather today, I took the opportunity to go while there was still sun out there ;-) [15:34] ya... pretty nice here, too [15:34] * Sweetshark was excercising a lot in the last weeks *cough* cough* === alan_g is now known as alan_g|tea [15:34] * desrt cringes [15:34] Sweetshark, excercising your pillow by napping are you? ;-) [15:34] well, I've excercised a lot. just not monday and weekend. :P [15:34] ok, let's get started [15:34] O_O [15:34] qengho, hey [15:35] hey guys! [15:35] * chromium-browser 31.0.1650.63 done and given to #security. [15:35] * Automated workflow for updating PPA. [15:35] * To do: 1) supportable "hal"-like support for Flash. 2) more apparmor testing b [15:35] efore profile move to Cr pkg. 3) better autopkgtests to test webapps functionali [15:35] ty. [15:36] EOF [15:37] qengho, so, chromium-browser is stucked in trusty-proposed since 2013-11-05 ... when is that changing? I though we agreed that should be fixed last week one way or another? [15:38] we need that resolved and not at the bottom of a list... [15:38] seb128: Yes, I've been trying to get the same package out since Friday morning. [15:38] what is blocking? can I help? [15:39] is the issue a real one? [15:39] or should we discuss overwriting the autopkgtest result and publish it? [15:39] (if that's a buggy test that's going to be fixed soon) [15:40] woohoo, gtk 3.10 lands in trusty!! [15:40] seb128: The only thing lacking is time of #security folks. [15:40] seb128: If you want to sponsor trusty specally, I can point you to the package. [15:41] yes please [15:41] you should probably go through normal sponsoring for trusty [15:41] no need of security there [15:41] they have enough to do I guess [15:42] chrisccoulson, or is there a reason you prefer to sponsor to current serie as well, e.g to have all updates out synced? [15:42] well, anyway, let's discuss that after the meeting [15:42] qengho, thanks [15:42] Sweetshark, hey [15:42] heya [15:42] drudged through the verification of precise (done), raring (rescinded) and saucy (done) SRUs [15:42] made libreoffice l10n not only buildable separately, but also installable: see http://nabble.documentfoundation.org/Packing-from-instdir-also-lets-kill-scp2-td4084340.html#a4087375 === alan_g|tea is now known as alan_g [15:42] squashed and proposed on the release branch: https://gerrit.libreoffice.org/#/c/6982/ [15:42] now tweaking the ubuntu packaging to use this to build a learn an mean libreoffice-l10n (WiP) [15:42] fixed LibreOffice 4.1 series on trusty, b0rked by mdds update :/ (incedentally also needed by the saucy SRU) [15:42] survived two or three heart attacks given by the suggestion to completely tear out Java out of LibreOffice for the LTS no less. [15:42] EOF [15:43] the java discussion is still only, right? [15:43] upstream compiz seems to be abandoned and has been removed from debian... are there plans to maintain it, or to migrate unity off of it? [15:43] thanks for fixing on saucy ;-) [15:44] psusi, hey, we are in a meeting, but basically both, we are maintaining compiz for some years and we are moving away from it with unity8 [15:44] seb128: I need to get an update on the java thing. I dont think dropping will fly and dont see it happening. [15:44] seb128: so, not too stressed there anymore. === tkamppeter_ is now known as tkamppeter [15:45] oh, for the -l10n source package-foo I would need someone reviewing that at some point in time. As you can imagine its not debian-policy-by-the-book as the whole package isnt. [15:45] ok, feel free to ping me about that [15:45] it might be an after holidays stuff though [15:45] Sweetshark, thanks [15:45] seb128: willdo [15:45] mlankhorst, hey [15:45] - mesa 10 upload [15:45] - xorg-server 1.14.5rc1 upload for bug #1238410 [15:45] - triaged and fixed #1224254 with tseliot [15:45] - triaged and fixed related bug #1259561 with him too [15:45] - commit dri-image series [15:45] Launchpad bug 1238410 in X.Org X server "Inconsistent cursor visibility with cursor plugin enabled" [Medium,In progress] https://launchpad.net/bugs/1238410 [15:45] Launchpad bug 1259561 in xorg-server (Ubuntu) "config timestamps not updated on hotplugging a output with hybrid graphics" [High,In progress] https://launchpad.net/bugs/1259561 [15:45] - prepare and upload pixman sru's for lts-saucy now it's no longer blocked by a CVE [15:45] NUL [15:46] would be NULL, i guess [15:46] mlankhorst, there is also a bug marked against gnome for this [15:46] wrt cursor visibility [15:46] ritz: yeah [15:46] https://bugzilla.gnome.org/show_bug.cgi?id=706229 [15:46] Gnome bug 706229 in libgnome-desktop "mouse cursor occasionally invisible after logging in" [Normal,Resolved: fixed] [15:47] ritz: I'm waiting for 1.14.5rc1 to be renamed to 1.14.5, then I'll sru it and start the rename upload [15:47] ritz, read https://bugs.launchpad.net/ubuntu/+source/gnome-settings-daemon/+bug/1238410 [15:47] aah, sweet, thanks [15:47] Launchpad bug 1238410 in X.Org X server "Inconsistent cursor visibility with cursor plugin enabled" [Medium,In progress] [15:47] ritz, the GNOME bug you pointed is specific to GNOME 3.10 we don't have [15:47] it's also pointed in that launchpad bug [15:47] aah, my bad. [15:48] mlankhorst, did you finally get the trusty stuff to build/migrate to the release today? [15:48] desrt: NUL as in the ASCII char, not a null pointer. ;) [15:48] seb128: erm it was already migrated on monday [15:48] mlankhorst: it was a list [15:48] a double NUL terminated list [15:48] ;D [15:49] poor empty string [15:49] cannot possibly be in your list [15:49] everyone hates that string anyway [15:49] mlankhorst, ok, good, I just saw you did a couple of uploads for other archs/depends [15:49] don't get me started... [15:49] lol [15:49] mlankhorst, thanks [15:49] Laney, hey [15:49] meee [15:50] • Some Debian merges / updates [15:50] • Handle gnome-online-accounts transition; undo libgdata's no-goa-on-armhf hack which isn't necessary any more. [15:50] • Final batch of time whacking my head against webkitgtk - finally it succeeded on all architectures; thanks to hallyn for making qemu-arm64 available and to Debian for partch (powerpc porterbox), which were both invaluable. [15:50] • harfbuzz transition [15:50] ∘ pango1.0 uncovered notify-osd had an autopkgtest failure. Was a real bug (double g_source_remove()). Fixed that. Yay for tests discovering real bugs & leading to fixes. [15:50] • Look at use of upstart / XDG autostart for indicator launching; give some feedback on how that can be improved within the current scheme (c.f. gnome-settings-daemon's use of XDG autostart) (no trolling here, no sir) [15:50] • Discussions / experimentation around using libicu for u-s-s's locale selector. Seems like it should work. [15:50] • When running u-s-s's new autopilot tests, discover that translations are broken in some cases. Investigate, blame the uitk, file a bug & work around in u-s-s. [15:50] D O N E [15:50] usual question: status of u-s-d/u-c-c? [15:50] robert is doing that thing [15:50] desrt, robert_ancell's topic [15:51] I think it got put into some ppa [15:51] desrt, I think I already told you last week :p [15:51] he needs to be in our meeting! [15:51] yeah, agreed! [15:51] call him RIGHT NOW [15:51] seb128: ya... i kinda assume that you and/or Laney will be landing it when the time comes, though [15:51] dunno, he's got the same powers as us [15:51] will be reviewing / testing though, but haven't done that yet [15:52] right, I was sort of waiting for robert_ancell to do his call for testing [15:52] he said he would have it in a ppa soon and ask for review/testing then [15:52] I need to stay longer some evening to catch him [15:52] or try email maybe ;-) [15:52] pfft [15:52] Laney, thanks [15:52] * Laney just found http://www.ubuntugeek.com/ubuntu-control-center-ucc-simple-tool-for-ubuntu-administration.html [15:53] tkamppeter, hey [15:53] O_O [15:53] lol [15:53] those people are gonna be pissed [15:53] ah, here we are [15:53] he called it unity-control-center [15:53] Laney, desrt, others: https://launchpad.net/~ubuntu-desktop/+archive/unity-control-center [15:53] https://launchpad.net/~ubuntu-desktop/+archive/unity-control-center [15:53] bah [15:53] snap :p [15:53] that.... makes a good deal of sense [15:54] quite some other sources, included outdated one [15:54] since he's renaming the binary [15:54] he needs to patch stuff calling g-c-c to try both [15:54] I'm still not convinced we shouldn't have g-c-c being a wrapper calling one or the other [15:54] can't we do some alternatives thing? [15:54] well at least as a transitional thing [15:54] we can [15:55] or just a symlink in case g-c-c proper is not installed [15:55] robert believes it's bitter to just bite the rename [15:55] can we paint the bikeshed green? [15:55] I like green [15:55] * desrt starts up gcalctool to run the math on that one [15:55] ...oh wait [15:55] Welcome to LibreOffice Calc [15:55] well, even if g-c-c is installed, you want to call u-c-c under unity [15:56] since g-c-c is going to depends on shell runtime for stuff [15:56] I agree with seb128, there's quite a few things that invoke the binary [15:56] but it should warn, so that we can fix these [15:56] larsu, the list is in the ppa [15:56] awesome [15:56] so i'll put a patch in glib that checks in g_spawn if you're trying to run something called "gnome-control-center" and patch it over to unity-control-center if XDG_CURRENT_DESKTOP is set accordingly [15:56] it's from previous discussion with robert_ancell/a query on http://codesearch.debian.net/ [15:57] desrt: that's nasty [15:57] +99999 [15:57] * desrt has trouble with the 'not friday' thing [15:57] lol [15:57] make g-c-c depend on u-c-c and do the check itself [15:57] ...the final insult [15:57] ok, tkamppeter didn't pong and we sidetracked [15:57] desrt, your turn :p [15:58] spent most of last week continuing the dconf work [15:58] talked to some people who are doing odd things there (like gdm) and got a better idea of requirements and did some designing [15:58] turns out it's a lot of work... i started on it already [15:58] maybe will be done this week, or maybe next... but definitely by end of year, which is what i promised the phone guys [15:59] i also spent a good deal of time trying to improve the portability story of glib... i helped the openbsd/freebsd guys to get jhbuild rigs going on their platforms and spent some time filing bugs for fixing portability issues that have crept in just in the last little while (since there are no regular builds on !linux) [16:00] that's about it [16:00] now you have a debian account you can get shells on all the cool systems [16:00] haha [16:00] Laney: all using debian... [16:00] yep, the best userspace [16:00] my goal is to test userspaces that are not the best :) [16:01] besides... i thought ubuntu was the best? [16:01] even this one has issues :P [16:01] ahem. who's next? [16:01] cool, when we need an AIX build of LibreOffice, Ill hint the red hat guys at desrt! [16:01] :) [16:01] * desrt has a hard time caring about a non-free OS that's not windows or mac [16:02] that sounds overly specific :) [16:02] maybe... but it's very close to what almost everyone actually does care about [16:02] so is "linux, osx, and windows" [16:03] maybe [16:03] but glib's project description has 'portability' in the name [16:03] so.... [16:03] seems sliding offtopic for the meeting [16:03] ya, not saying it's bad, just being pedantic [16:03] desrt, thanks ;-) [16:03] attente, hey [16:03] hey seb128 [16:03] really slow week... u-s-s language changes with thanks to Laney and GunnarHJ, continuing work on compiz to get gnome-key-grabber working, thinking about how to fix the modifier-tweaks branch so that it'll be accepted upstream by MC Return [16:03] we can't really make credible claims there if our portability story is "we will patch bugs when you tell us .0 doesn't build,... 6 weeks after we released it." [16:05] attente, ? [16:05] seb128, yes [16:05] seb128: implying that he started with ? [16:05] * desrt frowns [16:05] it's the ending of the end! [16:05] lol [16:06] we're having hard times with EOFs today [16:06] oh god, don't leave us [16:06] [16:06] mlankhorst: XML: yr doin it wrong [16:06] attente, I didn't follow much the compiz changes, just read the mp comments last week, do you think you are figuring it out? [16:06] [16:06] :D [16:07] attente, is MC Return being overpicky? (I saw some concerns about ccsm and corner cases) [16:07] desrt: ah of course, it's [16:07] ;> [16:07] seb128, making it work in a way that's acceptable to him might not be that easy... [16:08] Laney: you better close that. [16:08] * desrt watches larsu twitch [16:08] * larsu is getting a bit anxious indeed [16:08] seb128, one thing i thought we could try is falling back on xkb grp [16:08] attente, ok, let's chat a bit after the meeting about what are his current concerns, maybe we can just do without resolving some of those [16:08] :D [16:08] seb128, ok [16:08] attente, thanks [16:08] larsu, your turn [16:08] larsu can't go yet... [16:09] ?> [16:09] don't have that much to report this week either [16:09] okay [16:09] quite a few distractions and discussions [16:09] for example, the fix that makes gsettings-qt work should be in 5.2 [16:09] qt 5.2, that is [16:09] but mainly I've been working on porting evince to GActionGroup, so that we can have a menu for it in unity [16:10] instead of the appmenu and cog [16:10] :D [16:10] \o/ [16:10] that progresses nicely, I think I'll be done with it in a couple of days [16:10] larsu: sorry that these dconf changes got better than i thought.... and also sorry that i got distracted over the weekend :) [16:10] *bigger [16:10] the porting part will hopefully be accepted upstream, I already talked to chpe about it [16:11] desrt: no worries, I'm using g_application_add_accel now [16:11] it's not so bad if you do it in a loop [16:11] in a loop? [16:11] ...maybe strapped to an xml parser [16:11] bah [16:11] no [16:11] ...and if you write that, let me know so that i can merge it into gtk :) [16:11] seb128: eof btw [16:11] larsu, thanks [16:12] ok, my turn [16:12] desrt: let's talk about this outside of the meeting [16:12] desrt: also, the context-menu part [16:12] larsu: NO, you can't put a btw after a eof, it's a violation of syntax! [16:12] mlankhorst: watch me! [16:12] * kept chassing e.u.c issues with some success: [16:12] - got a software-properties fix uploaded (enforced a depends to a new enough version) [16:12] - write a telepathy-mission-control fix for a segfault, got it merged upstream/uploaded to trusty [16:12] - got branches up for review for issues in update-manager/apturl/software-properties [16:12] - debugged a bunch of other ones/opened bugs with some debug infos,hint for people knowing the code better [16:12] - collected debug info for software-center/webkit issue, got some upstream traction and a patch that seems to be working, thanks Company! [16:12] -> seems like we got most of the 30 from saucy handled, I'm going to keep watching the list and try to clean some bugs every week but do a bit less there [16:12] [16:12] * backported a telepathy-glib fix for incorrect g_source_remove use [16:12] * filed a bug about similar issue for notify-osd, fixed by larsu and Laney, thanks! (some others to file) [16:12] * some desktop updates and sponsoring [16:12] * some extra testing for GTK 3.10, looks good, uploaded to trusty, thanks again larsu! [16:12] * got the ubuntu-system-settings autopilot branch from QA merged [16:12] * some u-s-s code reviews [16:12] None [16:13] None? [16:13] python [16:13] but uh.... python doesn't do that [16:13] ya [16:13] you kept up on your promise to have more detailed reports [16:13] * larsu got pinged twice in that report [16:13] Laney, ;-) [16:14] larsu, those are good pings, they come with icecream for you next team we see each others [16:14] ice cream pings! [16:14] did you fix the gtk+3.0 tests? [16:14] my favorite kind [16:14] also: any sprint news? [16:14] @gtk-test: yes [16:14] seb128: Error: "gtk-test:" is not a valid command. [16:14] I just need to commit/upload [16:14] I decided to do that after exercice and I went back just for the meeting [16:14] neat [16:14] doing to do that in a bit [16:14] gtk tests? [16:15] * desrt would sure love to know about the sprint [16:15] autopkgtest failed on deprecations [16:15] more -Werror, or what? [16:15] ah, I thought it had to do with my patch still [16:15] the autopkgtest failed because the small build example uses a GtkStick [16:15] GtkStock [16:15] stderr → fail [16:15] ... [16:15] i have nothing to say about that, on account of what my mother taught me [16:15] unless you allow it [16:15] lol [16:16] well, small/stupid issue, going to be fixed [16:16] -> desktop week [16:16] confirm that it's going to happen in february [16:16] it's likely we can get London the week after fosdem [16:16] though not confirmed yet [16:16] nice [16:17] awesome [16:17] I want the hotel with the weird alarms and funky lighting again [16:17] citizen|M? [16:17] ya [16:17] otherwise we would have option for U.S later in the month (Orlando, Austin, ...) where it's warmer, but I don't think we care much about the warm? [16:17] spacecapsuleprefabhotel [16:17] they had nice breakfast, too [16:17] I also made an argument for not making people deal with cross atlantic twice and tz changes twice [16:17] big beds, all good [16:17] yeah london seems sensible [16:17] since some are in London for another sprint before fosdem [16:17] and there is fosdem [16:17] seb128: london, please [16:18] london cold is just fine.... and FOSDEM!! [16:18] desrt, right, I'm with you, looks like it should work out ;-) [16:18] ok [16:18] oh, on another note [16:18] eoy holidays start soon for some of us [16:18] end of week for me at least (didrocks as well I think) [16:18] nice [16:18] you got three weeks? [16:18] !! [16:18] Laney, do you want to lead the meeting next week if you are still around? [16:19] will be and can do [16:19] Laney, yeah, 3 full weeks [16:19] \o/ [16:19] I used like 2 this year [16:19] seb128: I'm off from Friday too [16:19] Laney, thanks [16:19] davmor2, enjoy the holidays ;-) [16:19] well, anyway [16:19] seb128, sorry, I missed my call, here is my parts: [16:20] if you need me to review stuff/give some guidance/land something, please ping me in the next days [16:20] - cups-filters/Poppler: Finished fixes for Landscape output. [16:20] - cups-filters: PPD-less printing on PostScript printers [16:20] - cups-filters: Several bug fixes and clean-ups. [16:20] - Bugs [16:20] tkamppeter, hey, thanks [16:20] ok [16:20] seb128: Thanks you too [16:20] on that note, end of meeting! [16:20] seb128: any idea when we'll 'know for sure' about london? [16:21] desrt, this week [16:21] great. thanks [16:21] yw [16:21] desrt: so.... accels [16:22] larsu: will be part of the action description xml, overridable by some user config [16:22] either a file in homedir keyed by appid, or something in dconf [16:22] dconf sounds better to me tbh [16:22] org.gtk.Accelerators [16:22] and then what? [16:22] a gigantic dictionary keyed by appid and then keyed again by detailed action name? [16:23] this is where I got stuck in my thinking [16:23] no [16:23] gsettings list! [16:23] remember: gtk doesn't have dconf -- it only has GSettings [16:23] right :) [16:23] is that never going to arrive? [16:24] meh. [16:24] alternatively we could have a file in .config [16:24] wouldn't even need to be xml [16:24] for easy editing and all [16:24] ya [16:24] was thinking keyfile [16:24] "app.action=Ctrl+Q" ? [16:24] yes [16:24] Q more likely, actually [16:24] ya, whatever the format for gdk is [16:25] although it may be interesting to change the way we format accels due to the clash with xml and the fact that we use xml freaking everywhere [16:25] attente: there's a climbing centre 16 minutes from blue fin (re: our conversation in Oak) ;-) [16:25] * desrt would be up for climbing [16:25] Laney, ha, looking forward to it :) [16:25] awesome [16:32] Anyone here know how I can disable suspend? [16:32] jpds, on lid close or? [16:34] seb128: Completely. [16:35] jpds, you mean lockdown? forbid user to access it? [16:36] seb128: I never want the system to suspend. [16:36] jpds, when does it suspend? we only do that on lid close and low battery afaik and you can disable/change both from system-settings->power [16:37] seb128: Well, right now there's the option in indicator-session that I can't seem to disable with PolicyKit. [16:37] jpds, you made to make so the "suspend" item in the indicator doesn't do what it says? [16:37] (Though I can enable hibernate fine). [16:37] you want to make* [16:38] seb128: I want to remove it. [16:38] that's a different question from preventing suspend [16:38] if you would describe the usecase I could maybe help you [16:38] is that in context of vendor/admin lockdown? [16:38] is your machine having a bug with suspend? [16:40] jpds, what I want to know is if that's like asking to take the "mute" item out of sound menu, a weird request just because you want to tweak your UI, or if there is a real usecase behind [16:47] larsu: sorry for bothering you in advance of the meeting, wasn't aware it was going to commence. are you open to patches wrt using symbolic icons in the indicators? [16:56] seb128, what are the current plans for gtk 3.10? [16:58] it's uploaded [16:59] mterry, I'm fixing the autopkgtest if that's the question? [16:59] seb128, ah, no. It was a bigger picture question than that. Sounds like we're close to landing it [16:59] mterry, otherwise in case you just didn't pay attention to trusty-change/our meeting, we uploaded it to trusty today [16:59] seb128, yup, that's the bit I missed :) [17:00] k [17:00] seb128, awesome work [17:00] larsu, ^ that credits mostly goes to you [17:00] :) [17:00] bah [17:00] mterry, thanks larsu with an icecream next time you see him [17:00] why does nautilus not remember the "show hidden files" setting? [17:00] he did most of the patches porting, including the menus stuff and scrollbars === tvoss is now known as tvoss|test === tvoss|test is now known as tvoss|dinner [17:01] Laney, we have bug reports about that, since you can reproduce ... is that the preference setting or the by folder value? [17:01] Laney, https://bugs.launchpad.net/ubuntu/+source/nautilus/+bug/1247221 [17:01] Launchpad bug 1247221 in nautilus (Ubuntu) "Nautilus doesn't remember setting for showing hidden files correctly" [Low,Confirmed] [17:01] the latter one [17:02] oh, right, I can confirm that [17:08] seb128, did you get a chance to look at my fixes to my uss branch? [17:08] kenvandine, not yet, let me have a look [17:08] thx [17:08] yw [17:09] sorry for not getting to it earlier, it has been another crazy day here ;-) [17:09] no worries, i've switched context to content-hub for the day :) [17:10] Laney, do you want to debug nautilus (or at least upstream a bug report)? I saw the issue flying by my launchpad bug emails a few time, would be good to have fixed for the lts [17:10] yeah i'll put it on the list [17:11] thanks [17:11] it might be worth testing with nautilus 3.10 if it's fixed [17:11] yup [17:11] I'm pondering updating that one [17:11] they didn't do too much work [17:11] and they share the sidebar the fileselector now [17:11] that might be a nice consistency improvement for the LTS [17:12] work->changes (e.g 3.8->3.10 is not crazy or controversial iirc) [17:13] guess we took that on with 3.8 in the first place ... [17:13] "that"? [17:14] the crazy changes [17:14] let me check NEWS/git log again [17:16] https://git.gnome.org/browse/nautilus/log/?h=gnome-3-10&qt=grep&q=places-sidebar [17:16] no, that's only in 3.10 (the commit at the top) [17:16] oh sorry, I meant other crazy changes [17:16] e.g. nautilus controversy [17:16] right [17:17] sorry I misread you [17:18] nod [17:21] robru: did you consolidate the copyright data as I suggested or is it following the same format as before? [17:37] ochosi: [sorry, was out for a walk] not if you break unity in the process :P [17:37] ochosi: so if you rename icons in humanity I'm fine with it [17:37] larsu: hehe, well i mean it seems an easy enough and obvious fix [17:37] basically humanity needs some additional icons [17:37] and then the indicators should take advantage of that [17:38] yep, I totally agree [17:38] for some reason, indicator-power already does accept symbolic icons [17:38] it's just that noone got around to doing it [17:38] (which is why i got curious about the rest) [17:38] I think that's because it's used by xfce and elementary [17:38] and we had a feature request for that [17:38] I'd be happy if all the -panel ones go away [17:38] actually in xubuntu we will only be able to use it in 14.04 for the first time [17:39] our panel only now supports gtk3 indicators [17:39] hm, might have been elementary then (or some other project) [17:39] yeah, i suppose elementary [17:39] would you be the right person to ask for review if a merge-request were done? [17:40] on the request please put indicator-applet-developers, so that other subscribed people see the changes as well [17:40] but feel free to ping me when you have it up, I'll review it [17:40] ok, thanks [17:40] oh, for the icons I don't know [17:40] hm, does it even make sense? [17:41] i mean chances are you'll switch to a new icon-theme in 14.04 [17:41] at least if tiheum keeps up the good work ;) [17:41] we can't break unity icons, not even during development [17:41] I'm fine with merging once we've switched icon themes though :) [17:41] so i guess ideally the new icon-theme will contain the symbolic icons [17:42] definitely, yes [17:42] and then you'll probably do the change automatically [17:42] ya [17:42] finally no more -dark and -bright themes... [17:42] i mean icon-themes [17:43] not sure if that'll work for all icons [17:43] bu tyeah, would be awesome [17:43] *but [17:43] with gtk3, there's not much reason not to, unless your icon-theme is incomplete, i think [17:49] hmm [17:49] the u-s-s autopilot tests assume the battery panel is always visible [17:49] but that is false [17:51] ya, not visible for some user settings and when there's no battery installed [17:52] yeah I wonder how we should handle that in the tests [17:52] is it legit to mock upower in autopilot tests? [17:53] why wouldn't it be? [17:54] thought you might be expected to use the real one in the session you are running autopilot in === m_conley_away is now known as m_conley [17:54] ah right. I don't know what the policy is there [17:54] or if we even have one :) === alan_g is now known as alan_g|EOD === m_conley is now known as m_conley_away === m_conley_away is now known as m_conley [18:03] climbing time, happy evening all === Ursula is now known as Ursinha === mjohnson15_2 is now known as mjohnson15 === m_conley is now known as m_conley_away [20:04] attente, do you know of any reason why we couldn't move /usr/share/dbus-1/services/org.freedesktop.IBus.service from gnome-settings-daemon to IBus? Are there other desktops that might be using it differently? [20:12] robert_ancell, i'm not really sure, can't think of a reason why that would cause problems [20:13] attente, I wonder why upstream didn't do it? Perhaps the ibus project is not responsive? [20:13] I need to shift it so we can have both a gnome-settings-daemon and a unity-settings-daemon [20:16] hmm.. could it just be that the ibus devs didn't think it needed to be dbus activated? [20:18] the commit where that's added says that gnome-shell wants to replace the existing ibus-daemon with special arguments [20:19] robert_ancell, specifically to disable the panel [20:20] attente, right, but it still seems kind of wrong to not have that in the ibus package. It's gnome-settings-daemon kind of hijacking the ibus config there [20:21] i can only speculate, but i guess it's just because ibus-daemon should be running from the start of the session [20:22] and i guess g-s-d is just adding that service file very specifically to overwrite the arguments to ibus-daemon [20:25] the other option is to make a special package with just that config e.g. gnome-settings-daemon-ibus-config [20:25] That would be the safest, but a bit weird [20:25] But I guess that matches the weirdness of the reality [20:26] hmmm.. i wonder if we even need that in unity, being able to dbus activate ibus [20:27] attente, oh, that would be even better. :) [20:27] attente, shall I make the packages without it and you can tell me if it's a problem? [20:28] bschaefer, do you know if ibus-daemon needs to be dbus activated under unity? [20:28] robert_ancell, i think that should be ok for time being [20:28] hmm if you wan it to work with nux, IIRC. As nux talks with the ibus daemon, which i think it does it through dbus [20:29] but isn't ibus-daemon already running in the session by virtue of im-config? [20:31] attente, yeah, quickly looking at what nux does. It uses ibus api which then ends up using dbus commands to send messages to the daemon (quickly lookig) [20:31] looking* [20:32] i guess it's easy enough to just check what happens when the file is gone [20:32] brb [20:36] robert_ancell, so, i moved the file around, and IM with ibus is still working [20:36] attente, ok, I'll just not have it in u-s-d [20:36] sounds like it should be ok, unless bschaefer has something to say about it [20:37] * bschaefer reads backlog [20:37] bschaefer, i just tried removing the service file :) [20:38] attente, sorry, i miss read your first statement :), no ibus doesn't need to be dbus started [20:38] when you start the daemon it should be doing that on its own [20:38] ok, so that is just a gnome-shell thing, fantastic [20:38] like when its started with upstart [20:39] robert_ancell, hey, check with happyaron as well about ibus stuff, he's maintaining that stack for us === m_conley_away is now known as m_conley