[00:00] jbicha: no. i wish GNOME would accept patches I submit on bugzilla, before I show up to guadec ;-) otherwise i'd be one grumpy attendant. [00:02] several GNOME modules are mostly unmaintained [00:02] larsu: darkxst: the ubiquity-panel, the way i test is, is stop lightdm & kill -9 X, and then from tty1: sudo start ubiquity, after changing stuff in-place in a VM..... not sure if this helps =) the highlight is old about ubiquity [00:04] jbicha: yeah, my point is other GNOME dev's that say "lgtm, why is this not merged?" should just commit. In Debian/Ubuntu we have NMU and/or strong team maintainership, yet the GNOME core people don't see this as a problem. [00:04] GNOME upload procedures are rather different than Ubuntu's [00:05] it's relatively easy to get git commit rights for everything GNOME so GNOME maintainers assume you have rights unless you tell them otherwise [00:06] on the other hand, if you're not a maintainer you can get yelled at for not getting a commit reviewed on bugzilla first [00:08] it's much much harder to get commit/upload rights for Debian or Ubuntu [00:13] xnox: so the patches you're waiting on are for gnome-sudoku and gnome-power-manager? those are good examples of mostly unmaintained modules [00:14] look in the .doap file to find out who the maintainer is and manually ping them on irc or by email [00:16] jbicha: that's not what gnome love / gnome goals say to do. [00:16] jbicha: i write a patch for gnome, not a pull request to individual person, maybe it's all should be up on github then instead. [00:18] gnome-power-manager is no longer core GNOME (note that Ubuntu GNOME doesn't ship it either) [00:24] nearly all of gnome-games went unmaintained when they split them into separate modules [00:24] GNOME will never go full-on github because github isn't FOSS === Ursinha-afk is now known as Ursinha [00:25] it's likely that the new "GNOME Software" app won't include support for paying money for proprietary apps === Ursinha is now known as Ursinha-afk [01:16] attente: the indicator-keyboard tests still fail === Ursinha-afk is now known as Ursinha [02:19] hello all, any ecrypt pros here? [02:46] jbicha, is there an updated log? or is it the same as before? [02:49] attente: I don't know http://paste.ubuntu.com/5923873/ also failed on the PPA builders but by default it doesn't give a nice log [02:52] jbicha, thanks, yeah, that's a different problem [03:00] and ibus-client-clutter needs fixing to build against ibus 1.5 http://paste.ubuntu.com/5923902/ === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [04:13] Good morning [04:24] good morning! [04:30] Good morning [04:35] good afternoon! [04:35] pitti, hey! [04:35] attente: this is correct, right? https://launchpad.net/ubuntu/+source/ibus-client-clutter/0.0+git20090728.a936bacf-5ubuntu1 [04:36] pitti, would you have any idea why apt-get doesnt seem to pick up ddebs from ppa? [04:36] darkxst: I never tried that; do they appear in the Packages.gz index? === thumper is now known as thumper--afk [04:40] pitti, nope [04:40] pitti, although apport-retrace seems to find them mostly [04:40] darkxst: and a-retrace doesn't just take them from ddebs.u.c.? [04:40] apport just uses apt [04:41] well, python-apt, but that still just uses the normal package indexes [04:41] pitti, gnome3 ppa [04:41] those would not be on ddebs.u.c [04:42] ooh [04:42] http://ppa.launchpad.net/gnome3-team/gnome3/ubuntu/dists/saucy/main/debug/binary-amd64/Packages [04:42] darkxst: ^ they are there, in a /debug sub-hierarchy [04:42] so you need a "main/debug" series apt line in addition [04:46] pitti, right! thanks [07:07] Morning! [07:08] sil2100: hello! [07:17] Stacks seem to be a bit lazy since some days - I remember times when I woke up and had to publish/fix manually stuff, since I already had all stack results [07:17] But now? I wake up and testing and building steps are still ongoing [07:22] Morning didrocks ! [07:22] didrocks: stacks still not published, they're building and checking right now! [07:22] didrocks: how's IoM? ;) [07:23] hey sil2100! [07:23] sil2100: foggy :p [07:23] sil2100: stacks are under control? why they are not ready? [07:23] like ATI machine down? [07:24] Not sure, some stacks are still in the building stage (like settings), and check jobs are queued up [07:24] I guess everything slowed down a bit after we enabled additional check jobs, like unity8 [07:24] sil2100: shouldn't [07:24] sil2100, ATI was donw [07:24] sil2100: the tests are running on both machines, right? [07:24] down [07:24] didrocks: hah, something's wrong with ATI [07:25] ok, thanks jibel :) [07:25] jibel: is it fixed now? [07:25] sil2100: ensure you kill all the -head jobs please that are duplicated [07:25] if it's down since saturday, we don't want to rebuild the same stacks 3 times [07:25] sil2100, I'm trying to bring it back but there is something strange [07:25] jibel: since I see the ATI log of the current job and it doesn't seem to move [07:25] sil2100, since 3 or 4 days it always goes down when platform tests start [07:26] hmm [07:26] didrocks, I restarted it saturday and yesterday [07:26] robert_ancell: https://launchpad.net/~ubuntu-unity/+archive/next/+packages?field.name_filter=unity-api&field.status_filter=published&field.series_filter= [07:27] jibel: urgh, so something wrong with the upstart job? [07:27] jibel: I wonder what's wrong with the platform tests ;/ Could you give me a ping once ATI is unblocked again? [07:27] didrocks, it wouldn't freeze the whole machine [07:28] didrocks, it is a hardlock, machine is up, but ssh is dead and no console [07:28] sil2100, it's up [07:28] bonjour didrocks et jibel, comment allez-vous ? [07:28] jibel: but could it be that something again is wrong? Since: [07:28] http://10.97.0.1:8080/job/autopilot-saucy-daily_release/665/label=autopilot-ati/console [07:28] Bonjour pitti . Ça va bien et toi ? [07:29] Oh, FAIL [07:29] jibel: argh, not fine :/ [07:29] jibel: je vais mieux, il ne fait plus autant chaude [07:29] salut pitti! [07:29] sil2100, FATAL: Unable to delete script file /tmp/hudson1344923958933719703.sh means that jenkins lost contact with its slave [07:29] sil2100: please publish the settings stack ASAP ;) [07:29] jibel: "chaud" [07:29] didrocks: will do, waiting for it to finish building though! [07:30] armhf is still building [07:34] didrocks, sil2100 , there is really something wrong, jenkins slave just died [07:36] sil2100, you can add unity8 to the list of tests to restart [07:42] jibel: ok! [07:43] didrocks: so, I assume you guys ACK the packaging changes for settings ;)? [07:43] didrocks: publishing1 [07:44] sil2100: yep, ack! :) [07:44] sil2100: thanks [07:45] jibel: do you think that re-running tests for platform can cause the death of ATI machine again? [07:45] jibel: since I re-ran it in the morning, so I guess it's queued up already [07:46] pstolowski: welcome back! ;) [07:46] sil2100: hey! [07:46] sil2100, that's what I'd like to verify [07:47] sil2100, it just started fine [07:48] jibel: indeed, so far so good [07:48] didrocks: hello :) sdk packaging changes of adding devscripts to have licensecheck seems legit, can I publish? [07:50] * sil2100 ACKed the change himself to FTBFS [07:50] ;) [07:50] hey Mirv! welcome back [07:50] Mirv: can you check with seb128? I'm on the IoM this week sprint, so not really available for packaging reviews :) [07:51] Mirv: I trust that between you and sil2100, you'll get through it! :) [07:52] seb128: morning! [07:52] sil2100, hey [07:52] seb128: can we have your approval on http://10.97.0.1:8080/view/cu2d/view/Head/view/SDK/job/cu2d-sdk-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_ubuntu-ui-toolkit_0.1.46+13.10.20130729-0ubuntu1.diff ? [07:52] didrocks: thank you :) ok, have a good IoM [07:53] thanks! [07:53] sil2100, Mirv: seems fine to me [07:54] Mirv: green light! [07:54] Mirv: publish ;) [07:57] seb128, sil2100: thanks [07:58] sil2100: and thanks for taking care of my stacks in my absence :) [07:59] sil2100: any idea of those ATI unity test failures? (platform) [08:00] hm, not sure [08:00] Mirv: maybe let's poke upstream maybe? [08:01] sil2100: the other states D-Bus problem [08:01] sil2100: it's not platform upstream but normal unity autopilot tests that are failing [08:02] Mirv: what I would do - I would poke unity guys (or even fill in a bug about those and inform them) and publish anyway - or re-run the tests to see it it passes the second time [08:02] morning! [08:02] Laney, hey, how are you? [08:02] didrocks: sorry to bother you, but maybe on the sprint you can poke someone now - I need to be part of ~cordova-ubuntu, or one of my sub-teams has to [08:03] didrocks: since I can't re-deploy webapps to fix the stack ;/ [08:03] hey seb128, not bad thanks [08:03] sil2100: rerunning, a good way to find out also if it kills the ATI machine again.. [08:03] part of my bike got stolen over the weekend so I had to carry it two miles home, but other than that ... [08:04] Laney, :-( [08:04] Laney: hi! Part of your bike? [08:04] Laney, which part? [08:04] intel succeeded fine [08:04] yeah [08:04] the front skewer [08:04] just the skewer? that sounds strange! [08:04] it was either just to be annoying or in the hope that i'd abandon it and they could come back and take the whole thing when it was more quiet [08:04] those are my theories anyway [08:05] Laney: ugh, I'm sorry for you; that suck! [08:05] s [08:05] but instead I carried it back and the nice man in the shop gave me a new one for free the next day [08:05] pitti, salut, ca va bien ? [08:07] seb128: salut Sebastien ! oui, je vais mieux, il ne fait plus chaud [08:07] seb128: et toi ? [08:07] pitti, c'est la même chose pour moi, ça fait du bien un peu de frais [08:08] mlankhorst, hi [08:09] Laney, I would normally say zip ties can fix anything, but probably not in your case ;) [08:09] sil2100: there is nobody from this team AFAIK here :/ [08:09] Mirv: it passed now \o/ [08:09] darkxst: heh [08:09] didrocks: :( Too bad then, I hate when they create new teams without assigning them to others [08:09] I have some "pinheadz" on my other bike after having a similar incident there [08:09] will get some for this one now too [08:12] sil2100: wow, nice! [08:14] Laney, so stealing skewers is regular there? here they just take the whole bike ;( [08:15] and they thieves don't care about it being busy. I had one bike stolen during peak hour on one of the busiest streets in Melbourne [08:16] :/ [08:16] I had two good locks on it so maybe it would have required more time/effort [08:17] disable bike, go away to fetch more tools, come back in middle of night to take it [08:17] seems like lot of efforts to steal a bike [08:17] seb128, angle grinders cut through even the best locks like butter! [08:18] haha [08:18] and I guess nobody would do anything if you look confident enough [08:26] seb128: so, once ibus 1.5.3 moves out of -proposed, it will be hosted officially on the image, right? [08:26] s/hosted/installed [08:26] sil2100, correct [08:26] sil2100, why? [08:27] seb128: ah, just asking, since unity stack failed because it was requiring libibus-1.0-5, which is in proposed still - wanted to know for sure if it will be installed by default [08:28] Since there was some discussion I guess [08:28] sil2100, oh right ... CI doesn't build with proposed right? [08:28] sil2100, we are going to need an upload of unity build with ibus 1.5 to move it out of proposed though [08:30] seb128: it's building with proposed, but it fails if a package is not mentioned in packages: and not installed by default - which currently is the case with the new ibus, although for the time of transition I'll fix it up [08:30] sil2100, ok, great [08:30] seb128: so we'll have a new unity released with the new ibus [08:30] sil2100, excellent [08:30] today? [08:30] seb128: I'll fix the config now and let's see how the tests roll [08:30] great [08:34] seb128: https://code.launchpad.net/~sil2100/cupstream2distro-config/temporary_unity_new_ibus/+merge/177333 <- can you take a quick look and approve? I'll redeploy and re-run unity stack once all other stacks are green, since unity is hogging the test machines [08:35] sil2100, ok, approved [08:36] tkamppeter: did you recheck 56578 with the xserver I uploaded + a newer version of onboard? [08:38] Eek! [08:47] mlankhorst, no, how do I get the new xserver and the new onboard? [08:47] mlankhorst, can I do this with saucy? [08:47] yeah needs saucy.. xserver in canonical-x/x-staging, onboard bzr branch lp:onboard iirc [08:50] mlankhorst, xserver-common and xserver-xorg-core are 2:1.14.99.1-0ubuntu0.0~ppa3 and xserver-xorg-video-intel is 2:2.21.12-0ubuntu1~ppa1. Is this correct? Then all what I need is the onboard from BZR? [08:53] probably [08:53] intel version doesn't matter much, it was uploaded for other reasons :) [09:02] mlankhorst, it is unbuildable. I get [09:02] dpkg-source: error: unrepresentable changes to source [09:02] dpkg-buildpackage: error: dpkg-source -b onboard gave error exit status 2 [09:02] cd .. [09:03] ignore, or build with debuild -b [09:03] it's probably trying to include the .bzr directory if you build a source file [09:06] mlankhorst, even after "rm -rf .bzr" I can only build with "debuild -b", but at least I can test onboard now. [09:14] seb128: can I ask you for a packaging ACK before publishing? http://10.97.0.1:8080/view/cu2d/view/Head/view/Phone/job/cu2d-phone-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_address-book-app_0.2+13.10.20130729-0ubuntu1.diff [09:14] seb128: (it has the qtdeclarative5-ubuntu-contacts0.1 package in it) [09:16] sil2100, +1 [09:17] mlankhorst, onboard works principally, but right-click emulation leads to a segfault. [09:18] mlankhorst, onboard also does not take the black theme of the desktop. [09:19] seb128: thanks! Another (last) quick one: http://10.97.0.1:8080/view/cu2d/view/Head/view/Apps/job/cu2d-apps-head-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_webbrowser-app_0.21+13.10.20130729-0ubuntu1.diff [09:20] tkamppeter: yeah but I only care about stuck buttons atm [09:20] sil2100, +1 [09:22] seb128: thanks! [09:25] Laney: yikes, here they try to steal your bike by adding another lock to the bike, wait until you unlock your own then steal the bike when you go leave it unattended :/ [09:25] mlankhorst, I get no stuck button, seems to work no (typing this with onboard on touch screen) [09:25] mlankhorst: haha, that's quite obvious [09:27] Laney: but unfortunately, it works.. [09:27] why would you leave it? [09:27] without your lock on [09:27] seems to be a common reaction when it happens when people are not aware of that trick [09:28] mlankhorst, so I can do tablet mode with the only problem of a missing right button, missing screen rotation, and xbmc not working. [09:28] tkamppeter: the only useful data point to me is... do you still get stuck buttons like that? [09:29] and does xorg stay stable [09:31] mlankhorst, stuck button and X stability seems to be perfect, rest are apps which are not your scope. [09:32] ok good, so I guess the stuck button was partially caused by onboard grabbing the mouse too then.. [09:33] or w/e, maybe there is still a bug in xserver but at least there's a workaround for it.. === vrruiz_ is now known as rvr === Sweetsha1k is now known as Sweetshark [10:24] mlankhorst, I have updated https://bugs.freedesktop.org/show_bug.cgi?id=56578 now. [10:25] Freedesktop bug 56578 in Server/Input/Core "race condition with active/passive grabs when opening menus with touch" [Normal,Assigned] [10:25] thomi: hi! You in IoM too? [10:25] sil2100: yup [10:26] thomi: awesome! Quick question - would you mind if I proposed a new version of python-evdev to ubuntu? [10:26] sil2100: sure, but please make sure you don't break autopilot - what's the change needed for? [10:27] thomi: the upstream developer released a new version and simply asked me to sync it with Ubuntu, so no urgent fixes [10:28] sil2100: cool - in the past they;ve broken API, so we need to double check that [10:28] it's nice that they've realised we pushed it into distro though :) [10:28] thomi: right, I'll poke you about testing later, at least for some guidelines how to check if AP is still working [10:29] sil2100: that's easy: bzr branch lp:autopilot; cd autopilot; autopilot run autopilot === jdstrand_ is now known as jdstrand [10:29] Ah, thought that jsut some specific tests use evdev ;) [10:30] And/or only on certain configurations [10:30] sil2100: yeah, you could probably just run the input tests [10:30] nah, we run on all cofigs [10:30] as long as you have write acces sto /dev/uinput [10:30] sil2100: feel free to ping me and I can run the tests, if you like [10:31] thomi: thanks :) === sil2100_ is now known as sil2100 [10:56] thomi: hi! Can you give me power over launchpad.net/python-evdev ? [10:57] thomi: it would be nice to be able to edit the data, could you make me the maintainer? === MacSlow is now known as MacSlow|lunch [11:11] seb128: hmm [11:12] sil2100, I don't like that hmmm :p [11:12] seb128: I have a question related to releasing new upstream versions of packages where we're not upstream, but which I'm a maintainer of [11:12] ah [11:12] seb128: when releasing a new upstream version then, in the changelog, should I list all the changes that have been made inbetween versions? [11:12] sure, ask [11:12] seb128: or just write 'New upstream version'? [11:12] your call [11:12] there is not written rules [11:13] I tend to list things that close a bug [11:13] I like to put the interesting things from NEWS [11:13] and interesting changes [11:13] Ok ;) Thanks guys [11:13] * Laney tries sbuild --arch armhf u-s-s on his desktop [11:17] Laney, u-s-s is a small build, building on the n7 takes less than a minute [11:17] * seb128 usually does build there [11:17] it's just a bit annoying having to reinstall the build-deps when putting a new image on it [11:17] yeah I don't have those atm [11:17] thought this might be fast but it's not really [11:18] Laney, let me know if cross building work [11:18] not cross building, qemu [11:18] is working though [11:18] ah [11:20] I think it's slower :P [11:22] oh, segfaulted in the testsuite [11:30] ah good, that makes it more responsive [11:30] * Laney breathes [11:30] brilliant indicators are finally crashing ubiquity panel =) [11:30] * Laney locks Qt in the cupboard under the stairs as a punishment [11:32] I think i will disable ubiquity panel as a workaround for bug 1161058 [11:32] Launchpad bug 1161058 in ubiquity (Ubuntu) "panel crashed with SIGSEGV in indicator_object_get_entries()" [Medium,Confirmed] https://launchpad.net/bugs/1161058 [11:42] Right, larsu if you need any help testing / porting ubiquity's panel to new indicator stack feel free to ping me =) [11:45] xnox, did you read the friday discussion on the topic? [11:46] seb128: hm, I don't think I did. Was it here? [11:47] xnox, yes, short summary: larsu is working on it, he might have it done today (if IRC crazyness stops pulling him off work) [11:47] seb128: there is a mega plan in the back of my head to drop ubiquity-panel and drop ubiquity-dm, and instead use upstart user session to start ubiquity on login, with dash hidden / not-loaded and with a few thing pre-emptied. [11:47] seb128: ok. cool. === MacSlow|lunch is now known as MacSlow [11:56] sil2100, how is unity looking? did you retry it yet? [11:56] seb128: I did, let me look at the status [11:58] seb128: it finished, but intel has a few too many failures [11:58] sil2100, real bugs/issues? or system flackyness? [11:59] Seems like system flackyness, since ATI was fine [11:59] so another retry? [12:08] Yes, it's running already [12:08] thomi: hi! I installed the new evdev and tested the test_input_stack tests and they passed, but if you have a moment, could you also test on your machine? I pushed the package to my PPA: [12:08] thomi: ppa:sil2100/testing [12:19] good morning freedom lovers [12:19] . o O ( hmm... that didn't work... ) [12:19] PANTS OFF! [12:24] desrt, howdy! [12:24] desrt, you are trying for GUADEC? ;-) [12:24] * desrt invokes jdub [12:24] seb128: i made a present for you over the weekend [12:25] desrt, I noticed, thanks for working on that ;-) [12:25] when is next tarball due? ;-) [12:25] well, you need to get it merged first I guess :p [12:25] i think this won't be done until after guadec [12:25] i've been generating a whole lot of work lately that he needs to review [12:25] ok, I might just distro patch it if I need it [12:25] whether in gio, file monitors, desktop files... [12:26] warning: i may change the name [12:26] ok [12:26] g_file_measure_disk_usage() sounds nicer than _query_disk_usage() i think [12:26] well, I've enough stuff to do that I can delay that for a few weeks [12:26] I should probably just wait for you to land it properly [12:29] sil2100: sure [12:29] seb128: once I get this new package I am maintaining tested, how should I proceed to get the new version into the Ubuntu queue? Who should I poke and with what? [12:29] thomi: thank you! [12:30] sil2100, either do a merge proposal against the ubuntu: vcs or open a bug with a link to the ppa, or with the package files attached to the bug, and subscribe ubuntu-sponsors [12:31] seb128: thanks :) [12:31] yw [12:49] thomi: and then as well my previous question - could you assign me as the maintainer of launchpad.net/python-evdev? [12:50] sure [12:50] sil2100: done === dednick is now known as dednick|afk [12:57] sil2100: I don't get a new python-evdev with your PPA added [12:58] hmmm [12:58] thomi: hm, let me see! [12:59] sil2100: it doesn't look like it's in your PPA yet (not published anyway) - I'll try again later [12:59] thomi: ah, right, amd64 still pending ;/ Sorry about that! [12:59] thomi: and thanks! [12:59] :) [12:59] no worries [13:09] Laney, did you hit weird gsettings missing schemas bugs happening randomly with system-settings recently? [13:10] GLib-GIO-ERROR **: Settings schema 'com.ubuntu.touch.system' does not contain a key named 'r' [13:10] just hit that (but it's not the first time, it seems to try to read random keys) === dednick|afk is now known as dednick [13:16] seb128: no [13:16] no idea how to trigger it? [13:17] it happens on start every few runs here [13:17] with random key names, looks like corruption [13:17] could be an issue with gsettings-qt [13:18] that one in particular comes from orientation-lock [13:18] I can't see how that could be a bug in system-settings [13:18] have you seen it on the device or desktop or both? [13:19] desktop [13:19] but I don't test on the device that much [13:19] I mostly do check a few specific thing, otherwise run it when I update the image just to have a look [13:19] same [13:22] aloha [13:24] wondering who is the developer driving scopes? question on their functioning. each time I lauch one in the dash it's instintive for me to double click but that just makes it disapear then I go and click it again and then have to hit launch. Is there a way to change this so that if I find it in the dash and double click it, it opens? [13:27] czajkowski, you want to talk to mhr3_ [13:28] seb128: cheers [13:28] robru: hey Robert, how are you? [13:28] maybe I'm just too used to double clicking [13:28] robru: would you mind having a look at friend's failing autopkgtest? it's blocking other packages like the new pygobject from entering saucy === dednick is now known as dednick|lunch [13:58] thomi: if anything, the amd64 binary has been built and published! [14:11] Laney, larsu: [14:11] GLib-GIO-ERROR **: Settings schema 'org.gnome.desktop.background' does not contain a key named 'picture-u\xa8-d\xb0\xb1' [14:11] there is something weird with gsettings-qt :/ [14:19] seb128: aren't you glad that resulted in an abort with a helpful error message instead of strange difficult-to-trace-down behaviour? ;) [14:20] desrt, don't get me started on the abort thing... ;-) [14:20] * seb128 is pondering storing the ringtone in a ini style file instead of gsettings just to make sure the phone app on the phone is robust [14:21] when did you first start seeing this behaviour? [14:21] since we started using gsettings [14:22] interesting [14:22] locale related? [14:22] dunno, why would it? [14:23] it's a corruption [14:23] seb128: there's a much easier way to ensure the app is robust: install the schema [14:23] it works 9 times on 10 run [14:23] :) [14:23] desrt, robust means = if your disk is corrupted, the schemas missing, etc, you still have a chance the app keep working [14:23] it might not [14:24] if disk corruption results in part of the application going missing then it's true that the application may not run properly [14:24] but the chance is higher when it doesn't abort just because e.g the configured ringtone schemas is missing [14:24] * desrt raises a glass to not having old arguments [14:24] ;-) [14:24] I told you, don't get me started on that :p [14:24] fair enough! [14:25] we have to agree on disagreeing there [14:25] it also annoys me that I've to make system settings depends on the phone app, which is not available in the archive yet, to be able to start it [14:25] I would prefer have a if(key not there) hide that ui [14:26] this API exists now [14:26] and it's fast and convenient [14:26] g_settings_schema_source_lookup (g_settings_schema_source_get_default (), "your.schema.name") [14:26] yeah, I just need to get larsu to provide a qt api in gsettings-qt [14:27] since I'm calling from qml [14:27] i think he's already doing that, no? [14:27] with his dont-abort-like-desrt patch [14:27] it doesn't abort on missing key [14:27] but it does on missing schemas [14:27] ah [14:28] * desrt wonders how it could ever have aborted on missing keys === dednick|lunch is now known as dednick [14:30] desrt: g_settings_get_range() with a missing key aborts iirc [14:30] seb128: hm, when do you get this message? [14:30] larsu: but don't you have to enumerate the keys in the schema in the first place in order to spin up the property map? [14:31] ie: how is it even possible that you'd get a request coming in for a key that's not in the map? [14:31] larsu, that's happening in system-settings on start, once every 10 run ... seems like some sort of corruption :/ [14:31] larsu, the code doesn't do anything special, it just use a GSettings and read one key [14:31] desrt: there's an api settings.schema.choice(nameofkey) [14:31] desrt: pass the wrong nameofkey, boom. [14:31] ah [14:32] ya... you should protect against that [14:32] hey larsu, wie gehts? [14:32] this isn't C, after all :) [14:32] larsu, I'm hitting it in different panels [14:32] pitti: hey... i wanted to ask you about your jetlag system [14:32] desrt: ya, that's the dont-abort-like-desrt patch [14:32] pitti: gut danke! How was your weekend? [14:32] larsu: brainmelting :) [14:33] seb128: weird. I'll have a look. [14:33] larsu: we've been to Dresden, but didn't get that much done really aside from BBQ and idling at the beach :) [14:33] pitti, when do you start you holidays? I though you said they were conflicting with GUADEC, must be soon? [14:33] seb128: yes, next week [14:34] pitti: ya, same here :) [14:35] seb128: soo, can we mir lp-solve quicky? it seems to have been in main at least twice (see bug 305790) and has been on upstream version 5.5.0 since at least lucid: https://launchpad.net/ubuntu/+source/lp-solve [14:35] Launchpad bug 305790 in xom (Ubuntu Jaunty) "MIR - move to main for openoffice.org 3 build-depends" [High,Fix released] https://launchpad.net/bugs/305790 [14:36] seb128: trouble is: I am using libreoffices own copy of lp-solve and its causing all kinds of stupid dependency trouble ... [14:36] desrt: how do you recommend fixing bug 1195481? [14:36] Launchpad bug 1195481 in gnome-control-center (Ubuntu) "[power]: gnome-control-center crashes if indicator-power is not installed in Unity" [Medium,New] https://launchpad.net/bugs/1195481 [14:36] Sweetshark, check with doko or mterry maybe? but it has been in main it should be fine [14:37] jbicha: lots of good ways to deal with that [14:37] jbicha: could move the custom indicator to the indicator-power package like datetime does it [14:38] seb128: \o/ Tests passed \o/ [14:38] seb128: can I publish unity? [14:38] jbicha: could move the key to g-c-c and add a dep from the indicator to g-c-c [14:38] sil2100, excellent, please do! [14:38] jbicha: could add a dep from g-c-c to the indicator [14:38] seb128: we'll have it in proposed and ready for the transition [14:38] jbicha: could move the schema to some other package like gs-d-s and make both depend on that [14:38] or you could do the conditional thing i mentioned above [14:39] desrt: yeah the last one is the one I'm looking for [14:39] that's the least-good one :p [14:39] but ya... might make sense in this case [14:39] i kinda like the datetime approach, fwiw [14:40] desrt: because you're hoping gnome-control-center gets forked? [14:40] not because of that [14:40] you need to think of schemas like you think of libraries.... [14:40] alex-abreu: ping! [14:40] alex-abreu: hi! [14:41] so either a) they're API or b) they're private [14:41] sil2100, hi! [14:41] alex-abreu: could you add me to ~cordova-ubuntu ? [14:41] having all users of the schema in the same package means that you get option (b) which is all-around nicer [14:41] sil2100, sure [14:41] doko, mterry: see above -- do we need to do all the redtape for getting lp-solve 5.5 into main for the third time? [14:41] plus... why have custom code installed (even if disabled) to deal with an indicator that's not installed? [14:41] alex-abreu: since I need to fix the cu2d webapps stack and need to be part of the team to do that ;) [14:42] alex-abreu: thanks! [14:42] sil2100, done .. welcome in the team ! :) [14:42] ;p [14:42] Sweetshark, looking [14:43] mterry: note its still supported in main in lucid anyway with that upstream version. [14:43] Sweetshark, yeah. that's fine. [14:43] approved :) [14:44] * Sweetshark gives mterry a hug. [14:45] desrt: I attached a MP to the bug which is one fix (having Unity depend on indicator-power) but bschaefer wasn't so sure that was right [14:46] jbicha: move the panel is your best bet, imho [14:46] keeping patches out of the g-c-c package and shipping override panels alongside indicators is the best way [14:47] desrt: that's a lot more work though and it will mean that the Ubuntu versions will likely get outdated (but at least the UI would be more stable from release to release) [14:49] also there's bug 1197662 [14:49] Launchpad bug 1197647 in indicator-datetime (Ubuntu) "duplicate for #1197662 Date & Time applet in control center does not start" [High,Fix released] https://launchpad.net/bugs/1197647 [14:50] oops I mean bug 1187981 [14:50] Launchpad bug 1187981 in gnome-control-center "symbol conflicts in libtimezonemap1" [High,New] https://launchpad.net/bugs/1187981 [14:52] fun! [14:52] maybe we should fork g-c-c... [14:53] (why didn't we do this already?!?) [14:54] because it's work and nobody did it (yet), and because have 2 different g-c-c on top of 1 g-s-d is impractical [14:54] so it requires some thinking about g-s-d, or also forking that ... but then it starts being tricky [14:54] because we can *almost* make things work without forking [14:54] it might be easier to just 'fork' a few panels [15:14] * qengho had to flip to a VT, run "startx -- :4 -ac; sleep 5; DISPLAY=:4 xterm" to bootstrap a GUI this morning. Ouch. [15:20] jbicha, seems like only plasma-widget-kimpanel is missing for the ibus transition ... did you talk to the kubuntu guys about it/do you plan an upload? [15:21] seb128, can we upload a new accountsservice? [15:22] attente, hey, sure, what change do you need? [15:23] (I don't see a new upstream version nor a pending sponsoring request) [15:23] seb128, it's for the input sources [15:23] it's already proposed upstream [15:23] do you have the bug url? [15:23] but it's still pending [15:23] ok [15:23] https://bugs.freedesktop.org/show_bug.cgi?id=63086 [15:23] Freedesktop bug 63086 in general "Add KeyboardLayouts property" [Enhancement,New] [15:23] did you talk to Stef about it? [15:23] yes [15:24] i should follow up again though [15:24] right [15:24] I would prefer if he could ack the name/format before we backport it [15:25] so we don't build on a variable type/format that might change [15:26] seb128, ok, sounds good [15:26] attente, let me know when you hear back from Stef, then we can upload [15:27] seb128, sure [15:27] attente, btw ibus 1.5 is in saucy-proposed, you can ping alesage about daily landing I think (that was blocked on ibus to be available iirc?) [15:29] seb128, ok I'll get that job going [15:29] alesage, thanks ;-) [15:29] attente ^^ [15:51] pitti: https://bugs.launchpad.net/ubuntu/+source/upower/+bug/1203655 ... is that the bug that causes my thinkpad to not go to sleep when i close the lid? [15:51] Ubuntu bug 1203655 in upower (Ubuntu) "Hangs in pthread_join in libusb_exit" [Undecided,Confirmed] [15:52] larsu: ^ [15:52] let me know if you want me to extract something fro debugging or so [15:52] asac: I think it's related (I have the same issue), but I haven't had time to verify it yet [15:52] asac: I don't know; could be [15:52] asac: my laptop suspends just fine, and larsu couldn't reproduce it yet, I don't have any other data [15:53] * larsu had the unsuspended-laptop-in-backpack-problem because of that [15:53] I suspended/resumed it a lot over the weekend [15:53] pitti: I didn't have it in the last couple of das either. I'll keep an eye on it, though [15:54] i can suspend through menu [15:54] not when closing the lid [15:55] it just goes in heat mode [15:55] asac: does it suspend when you Ctrl+Alt+F1 ? [16:15] Laney, larsu, kenvandine: hum, are the gsettings key value supposed to update on key change? [16:16] I'm doing [16:16] GSettings { [16:16] id: soundSettings [16:16] schema.id: "com.ubuntu.touch.sound" [16:16] ... [16:16] SilentModeWarning { visible: soundSettings.silentMode } [16:16] but the visibility doesn't toggle when the key change [16:17] if I do use a prop and do "onChanged: { if (key == "silentMode") silentMode = value" it works though [16:17] seb128: they are supposed to and there are tests that they do... [16:17] hum [16:17] I wonder what I'm doing wrong there [16:17] me too :) [16:17] it looks totally fine to me [16:18] any warnings when loading the script [16:18] ? [16:18] no [16:18] how are you changing the visibility, with gsettings set? [16:18] oops, silentMode of course [16:19] yes [16:19] and it works if I do "onChanged: { if (key == "silentMode") silentMode = value [16:19] in my GSettings {} [16:20] ah, right [16:20] that doesn't work [16:20] and you're sure the visible thing works? [16:24] larsu, http://paste.ubuntu.com/5925776/ [16:24] $ gsettings set org.gnome.desktop.background show-desktop-icons false [16:24] $ gsettings set org.gnome.desktop.background show-desktop-icons true [16:24] does it change for you dynamically? [16:26] larsu: the previous version doesn't reflect changes for me, that one does: http://paste.ubuntu.com/5925784/ [16:26] kenvandine, hey, "that doesn't work" ... you confirm what I'm seeing? [16:27] seb128: same problem here. Definitely a bug [16:27] larsu, ok, let me file one [16:27] seb128: let me finish ubiquity first please, I've had too many distractions today already :) [16:27] seb128: thanks! [16:28] larsu: yes, please, that =) [16:29] larsu, remember it doesn't automatically bind the property [16:30] at least that is what i seem to recall [16:30] kenvandine: well, it should in that direction [16:30] kenvandine: the problem was that it didn't in the other (checking a checkbox didn't update the key) [16:30] oh right [16:30] larsu, https://bugs.launchpad.net/ubuntu/+source/gsettings-qt/+bug/1206181 [16:30] Ubuntu bug 1206181 in gsettings-qt (Ubuntu) "the properties don't get updated when the key value changes" [Undecided,New] [16:30] sorry for the confusion... :) [16:31] seb128: thanks [16:31] that did work :) [16:31] kenvandine: no worries [16:31] * kenvandine goes back to ofono [16:31] :) [16:31] larsu, sorry for adding to your day IRC ping-o-load (welcome to most of our days :p) [16:31] * seb128 goes for some exercice [16:32] seb128, btw i am able to get those serviceNumbers from ofono-qt just fine [16:32] seb128: enjoy! [16:32] just fighting a little exposing it to qml [16:32] kenvandine, nice! [16:32] should have it ready today [16:32] seb128: usually it doesn't bother me, but otday was just crazy [16:34] larsu, yeah, I know how it is when pings don't stop and you can focus 15 min on code ;-) Anyway going for some exercice, maybe you manage to get an hour of quiet hacking :p [17:34] xnox: https://code.launchpad.net/~larsu/ubiquity/fix-indicators/+merge/177438 [17:52] seb128: thanks for bringing this to my attention, it was a regression that I introduced with the c++ split. Fixed and MRed, would love a review ;) [17:56] larsu, great, thanks for fixing it! [18:08] pitti, hi, yes, I can have a look. where are the failures? === sam113101 is now known as sam113101_afk === sam113101_afk is now known as sam113101 [19:27] hi , could someone look at https://bugs.launchpad.net/indicator-session/+bug/1206231 [19:27] ritz: Error: ubuntu bug 1206231 not found [19:27] wrong channel [19:28] hmmm, ubuntu-security ^^^ [19:29] hrm, I can't load it either. === racarr is now known as racarr|lunch [19:30] sarnold adding you [19:30] not a very high prio [19:30] but annoying [19:32] sarnold mark this as public ? === ritz is now known as ritz|dead === VD is now known as Guest37498 [19:46] seb128: hey. been following the conversation? [19:47] i think we should drop per-user layouts at the login screen. they don't make sense. [19:47] desrt, hey, hum, no ... where was that? #gnome-hackers? [19:47] ya [19:47] ah. you're not there. [19:47] (I'm not on there anymore most of the time) [19:47] i think we should just have one global system default [19:48] the idea that the keyboard layout changes after you type the username but before the password is _really_ weird [19:48] (and only slightly less weird if you pick the user by mouse) [19:48] we don't type the username [19:48] we don't have a username entry [19:48] we click on it [19:48] i know. please read again :) [19:48] and it makes sense to type the password in the keyboard layout you use [19:49] right... but a computer has a keyboard attached to it, right? [19:49] was that discussion mostly between qwerty users? [19:49] one particular kind of keyboard [19:49] like... in your case, we have a setting on a harddisk that's physically bolted into a machine that has an azerty keyboard on top [19:49] I'm typing azerty on qwerty keyboard (ok, not the majority but still) [19:49] so why not always use azerty at the login screen? [19:50] what about university kind of setups? [19:50] or libraries [19:50] or internet cafés [19:50] maybe... [19:52] well, the feature is no expensive to support and is handy in some cases [19:52] I guess we can keep it as Ubuntu specific if upstream is not interested [19:52] i'm trying to decide if upstream should be interested [19:52] or if ubuntu should drop the feature :p [19:52] the cost is lost, do they just object on principle? even if it's useless for 90% of users, it's not cluttering UI or creating work... [19:52] do you have your system keyboard layout configured to azerty? [19:53] is *low* [19:53] yes [19:53] but let's say I do a guest account for dholbach because I stay at his place for a while and he's going to use my laptop [19:53] he doesn't know azerty [19:54] he's probably going to use qwerty [19:54] well, I don't get what's the cost/issue to just have one list of values by user there [19:54] it's not the most useful feature in the world [19:54] but the cost is low... [19:54] well [19:54] my issue with it is that we now have the list of keyboard layouts used by the user in two places [19:54] accountsservice and gsettings [19:54] and we have to keep them in sync [19:55] also: it's not clear how we should select the default [19:55] why does gsettings need the keyboard layout if accountsservice can do it? [19:55] probably we don't want the user changing the layout in the session with a hotkey to change what is used at the login screen next time [19:55] jbicha: accountsservice doesn't do it yet... at least not upstream [19:56] desrt, does changing in the session change what you get in the session at next login? [19:56] right but when it does... [19:56] desrt, for me it's simple, we have a ordered list configured in gnome-control-center [19:56] the first item is the default [19:56] seb128: i don't know. should it? [19:56] on the login screen and in the session [19:56] no [19:57] these are the questions i am currently trying to think through :) [19:57] i guess it's a backwards compatibility issue? do we remove the schema if accountsservice handles it, or keep it and sync.. [19:57] I think we should respect the configuration and always apply defaults on fresh boot/login [19:57] that's not how the session side currently works [19:57] it remembers the one from last time [19:57] the cycling you do in a session is in session state [19:57] whic is arguably bogus... [19:58] well [19:58] depends [19:58] do you use "keymap by win" option [19:58] lol [19:58] most people don't use the same input method/keymap in their im client and text editor [19:58] you code in C/english [19:58] you chat in your language [19:58] "lol" but welcome to keymap fun ;-) [19:59] default state should really be whatever you defined the default, at every login [19:59] then users change the config according to the context [19:59] i think i agree === racarr|lunch is now known as racarr [20:58] ritz_: public makes sense to me === thumper--afk is now known as thumper [21:07] does somebody want to maintain a browser? [21:07] lol [21:07] oh no, I'm not falling for that [21:07] :) [21:09] IT'S A TRAP! [21:09] i'm just going to pop in to random IRC channels this week and ask the same question [21:24] chrisccoulson, which one are you giving away? ;-) [21:25] seb128, i only have one to give away ;) [21:25] chrisccoulson, I saw you uploading chromium recently! :p [21:25] seb128, oh, i just sponsor stuff for qengho