=== _thumper_ is now known as thumper === tgm4883` is now known as tgm4883 === android is now known as jussi === alan_g is now known as alan_g|afk [07:27] hello anybody [07:28] is it good to upgrade 10.04 LTS to an 12.04.? [07:40] hey [07:40] twim: supposedly it is possible, I've not done it myself === alan_g|afk is now known as alan_g === MacSlow is now known as MacSlow|lunch === Pendulum_ is now known as Pendulum === MacSlow|lunch is now known as MacSlow === _salem is now known as salem_ [12:25] hey, any idea if bug 806255 is going to be fixed in 12.04 as SRU? [12:25] Launchpad bug 806255 in Compiz Core "Unity/compiz intercepts Super and Alt keypresses from grabbed windows like VMs." [High,Triaged] https://launchpad.net/bugs/806255 === kiwinote_ is now known as kiwinote === rye_ is now known as rye [13:18] Cimi: good morning, could you take a look at https://code.launchpad.net/~jbicha/light-themes/fix-ambiance-in-gnome-panel/+merge/103001 [13:19] jbicha, I think you need to update radiance as well [13:21] Cimi: oh, Radiance didn't seem as bad at least [13:24] jbicha, the match should be the same [13:33] Cimi: thanks, I re-pushed [14:06] Hi. I just upgraded to 12.04 (or rather installed it from scratch). When I tried to reprettify my system, I noticed that there is no ccsm package anymore. Where did it go and what do I do now? [14:15] Baribal, the package name is compizconfig-settings-manager, though you can use the appearance section in control centre now to configure most useful things in unity [14:18] Thanks. I already pilfered through "Appearance", but didn't find any of the things I wanted to tweak, like using the ring switcher. [14:23] Okay, so I did set Alt-Tab and Shift-Alt-Tab as Ring Switchers Previous Window and Next Window. I still get the Unity switcher, though... [14:26] Ah, got it... [14:27] I had to explicitly disable Unitys switcher; telling CCSM to set the key although Unity uses it isn't enough. [14:34] I have five open windows, two of which don't appear in switchers; neither Unitys' nor Compiz'. Why? What can I do? Is that a known bug? [14:38] I'd like to add "Even if I maximize the missing-in-the-switcher windows via the dock, the input focus remains with one of the listed windows" to the last problem. [14:38] didrocks: you there? [14:39] hey tsdgeos [14:39] didrocks: something weird happened with the unity merger [14:39] have a look at https://code.launchpad.net/~aacid/unity-2d/load_24bit_pixbuf_icons/+merge/102471 [14:40] my latest commit was 3 hours ago, gerry approved 32 minutes ago [14:40] but the merger thinks the approval was of an old revision? [14:40] any idea? [14:40] tsdgeos: did gerry reloaded the page? [14:41] tsdgeos: the revision commit is set by launchpad [14:41] i don't know that :D [14:41] ajaxy-thing [14:41] ah [14:41] maybe that was a problem [14:41] you can check it when you "approve" [14:41] it's written "approved rev <…>" [14:41] oki [14:41] I'm just getting the value then and compare to last commit :) [14:41] i'll ask him to reapprove then [14:41] yeah [14:41] just reporting in case you think it might be a glitch on the merger [14:41] and check that the revision number written is the tip of your branch [14:42] greyback: can you reapprove https://code.launchpad.net/~aacid/unity-2d/load_24bit_pixbuf_icons/+merge/102471 then? [14:42] tsdgeos: no, we have one glitch on pre-requisite branches (that is linked to tarmac) [14:42] we'll need to find a proper test case, because everytime I try to reproduce, I can't [14:43] didrocks: I believe I did not reload the page, just pull straight from bzr, test, then approve on the web page [14:43] so the page I was looking at did not mention the last commit [14:43] don't do that, thumper's code allow you to approve whatever revision without checking it's the latest :) [14:43] blame him! [14:43] ;) [14:44] lol ok so :) [14:45] and always reload the page, you always need the latest goodness from the web :p [14:45] will do [15:10] didrocks: still problem merging https://code.launchpad.net/~aacid/unity-2d/load_24bit_pixbuf_icons/+merge/102471 [15:11] greyback: I asked you QA team to ensure that jenkins doesn't give a 404, can you check with them? [15:12] didrocks: ok [15:13] he following tests FAILED: [15:13] 11 - gimageutilstest (Failed) [15:13] greyback: ^ [15:13] tsdgeos: ^^^ passed on my machine [15:13] http://paste.ubuntu.com/942637/ [15:13] meeting, back in 15 or so [15:19] Hi again. The problems that I mentioned earlier vanished once I reactivated the switcher in the Unity Plugin in CCSM. However now I can't use the ring switcher again (though right now that seems to be a result of the Unity plugin using all the key combos that I was using). [15:24] didrocks: brr, that paste looks ugly :D [16:00] With Alt-Tab and Super-Tab hogged by Unity, I need more modifier keys... Where's Meta when you need it? === cyphase_ is now known as cyphase === bdrung_ is now known as bdrung === mick__ is now known as mick0 === hggdh_ is now known as hggdh [19:30] thomi, ping, I think I know why all the ibus test are failing (in jenkins). [19:31] bschaefer: pong - yay! \o/ [19:31] I don't think nux has been re built with those changes I did [19:31] waaaaat!?? [19:31] hmmm, ok [19:31] I can look into that today [19:31] because the only way I can the ibus ap test to fail like that is by removing that line of code [19:31] and getting ime_active stuck on ture [19:32] true* [19:33] cool, I was hoping all those changes would at lease fix most of the ibus ap test, but they are all failing :( [19:34] :( [19:34] lamal666: perhaps you could see if you know why nux isn't being rebuilt? ^^^ [19:34] yeah :(, that is what I have been looking at this morning. I remember jay mention changing the abi number [19:37] thomi, not being built where [19:37] ppa? [19:37] bschaefer, why do they pass on my machine though? [19:38] lamal666, the ibus ap test? [19:38] yes [19:38] they pass locally, fail on jenkins [19:38] lamal666, they all pass on mine too, but I added some changes into nux last friday to fix that [19:38] ah [19:38] so Im not sure if jenkins has rebuilt nux since then [19:39] im hoping it hasn't otherwise Ill have to figure out how it is failing haha [19:41] bschaefer, nux in the ppa is up to date with trunk at rev 613 [19:41] were your changes merged? last commit is 4/19 [19:42] which is yours :P [19:42] hmm let me double check haha [19:42] yeah my rev is 613 [19:43] lamal666, fixing the ime_active bool, which wasn't getting reset...hmm [19:43] that is odd [19:43] yah that's in the staging ppa so AP should be using it [19:43] when i watch the videos the problem seems to be just that IBus doesnt get activated at all [19:43] it just types normal characters into the search box [19:44] yeah, because when it tries to activate it ime_active is true when it should be false :( [19:44] ah [19:44] and it fails an assert [19:44] hmm [19:44] Ill see if i can reproduce it locally then [19:50] lamal666, you said the video shows actual characters being typed? Just english? [19:51] yah [19:51] do you have access to the private jenkins? [19:51] nope :( [19:51] ok let me get you a video [19:51] cool, because usually when i get the assert to fail it wont type engish at all [19:53] I'm back [19:54] bschaefer, hm yah now it's not typing at all [19:54] lamal666: I think the latest videos show nothing being printed - that's because we tightened the tests a bit recently... so they won't type unless im_active is set to true [19:54] ...which, for some reason happens locally but not on jenkins [19:54] is there some way to tell what version of nux is being installed? [19:55] lamal666, hmm that seems like it isn't reseting the im_active bool, which it first selects an engine [19:55] since it has to CreateContext on it, which resets that to false, but it is reporting as true! [19:56] thomi, should be in the output of the buildroot job I think [19:56] * thomi scours the logs [19:57] hmmm, I can't see anything [19:57] oh wait [19:57] the nfsroot job has been failing :( [19:58] does gnome/unity have an equivalent to kde's ctrl+alt+esc (then click a window to kill it) [19:58] ssh: Could not resolve hostname bazaar.launchpad.net: Temporary failure in name resolution [19:58] lamal666: looks like nfsroot job hasn't built due to network issues [20:00] hmmm... even odder - the job on the private instance has been running just fine, but for some reason the results aren't being published [20:00] yah that was happening before too [20:01] private build# = 1151 public build# = 983 [20:12] bschaefer: do you know what version of nux it should be downloading? [20:12] I see it downloading and installing the following: [20:12] libnux-2.0-0 i386 2.10.0+bzr613ubuntu0+333 [20:12] libnux-2.0-common all 2.10.0+bzr613ubuntu0+333 [20:12] that's correct [20:13] that's the PPA version [20:13] thomi my refactor branch is at lp:~alexlauni/unity/unity.ap-switcher-emu [20:13] is there a non-GObject python binding for libunity? [20:13] ahh, the '613' is the bzr revno? [20:13] yah [20:13] bzr613 [20:13] get it :P [20:13] awesome [20:14] thomi, that should be the right number :( [20:14] so the right rev is installed and it is still failing, hmm [20:14] mhr3: is there a non-GObject python binding for libunity? [20:15] mhall119: I don't know for sure, but I doubt it. [20:16] lamal666: OK, I have the branch. You want feedback, or are you having a specific issue with it? [20:17] thomi, its not getting the model [20:17] cant figure out why [20:17] run autopilot.tests.test_switcher.SwitcherWorkspaceTests.test_switcher_can_switch_to_minimised_window [20:17] thomi: is there any way you know of to allow a python/gtk2 app to access the libunity API? [20:17] mhall119: no, you can't use gtk2/gobject2 with gtk3/gobject3 bindings [20:18] the two are incompatible. [20:18] mhr3, ^ [20:18] it's a royal PITA, I know - I've had to deal with the same issue [20:18] mhall119, no [20:18] lamal666: ummm, that test passes for me [20:18] O.0 [20:18] wtf [20:19] :) [20:19] WTF [20:19] computers are awesome! [20:19] mhall119, why? [20:19] mhr3: http://askubuntu.com/questions/123018/how-to-add-quicklists-to-gtk2-application [20:19] thomi, your'e running from my branch? [20:19] you're sure? [20:19] you dont have AP installed or something [20:20] lamal666: ahhh, I had pythonpath set... :( [20:20] yar [20:21] ok, that's better [20:22] mhall119, theoretically it shouldn't be an issue using libunity's gi bindings with gtk2 [20:22] theoretically [20:22] but the reality might be different [20:22] I think packagaing explodes when you do that though [20:22] i know webkit really doesn't like gi [20:23] so maybe there's some interaction going wrong [20:24] oh, i see, now you can't import both old gobject bindings and GObject from gi :( [20:24] that used to work :( [20:27] Hmm. I have a really strange python threading behaviour.. It seems that the main thread is blocking another thread from returning from a urlopen call. That call seems to block until I stop the main thread. [20:29] anyone who seen this before? [20:29] thomi, anything striking you as wrong about that code? [20:29] lamal666: not so far - just adding some debugging.... [20:30] mick0: http://wiki.python.org/moin/GlobalInterpreterLock? [20:30] lamal666: ok, so the model exists, it's the icons attribute that's breaking... [20:31] hm [20:31] lamal666: wait, where does "get_children" come from? [20:32] ..you use it in the model - I think that should be "get_children_by_type" [20:32] hm uh a type likely [20:32] /copy paste error [20:32] get_children_by_type isn't exactly right though [20:32] * thomi tries changing it [20:32] because there are BamfLauncherIcons and DesktopLauncherIcon [20:32] yeah it is, you don't want to get children that aren;'t icons, do you? [20:32] you can use the parent class type [20:32] aye [20:34] ok, when you fix that it gets further, but there's a debugging statement that it barfs on [20:36] ..fixing that makes it get further, but then it ends up cycling through the switcher icons endlessly [20:36] lamal666: shall I push this branch somewhere, or do you want to re-create the fixes I've made? [20:36] yah that's what i was trying to fix- [20:36] that other stuff was in trying to fix the infinite loop [20:36] * lamal666 already fixed [20:37] oh cool [20:37] i mean fixed your fixes :P [20:37] :) [20:37] but not that issue of it cycling endlessly [20:37] oh [20:37] it seems that the tooltip text is always calculator [20:37] * thomi also notices some unused imports and some PEP8 issues in that file :) [20:37] it's not done yet [20:37] sheesh [20:38] :) [20:45] mhall119: Thanks! GIL's a bitch. Swithcing to multiprocesses instead of threading. [20:47] thomi, anything jumping out at you about current_icon not updating? [20:47] are caching introspection somewhere maybe? [20:47] i didnt think so [20:50] mick0: I'm not saying that was the cause of your problem, but it made me think of it [20:51] morning [20:51] thumper, good morning [20:51] morning thumper [20:52] lamal666: no...not yet [20:52] yah i know it was mentioned in potential optimizations [20:52] lamal666: are you sure that the 'active' property is what you want? [20:52] yah [20:53] * lamal666 checks again because you know what maybe not [20:53] pretty sure [20:53] mhall119: It seems to be cause. If it works to switch to multiple processes it sure was GIL. [20:53] if active is set to True on all the icons that might explain the issue :) [20:54] NO. active is not what i want [20:54] grr [20:54] stupid poorly named properties [20:55] heh [20:56] self.icons[self.selection_index] [20:56] much better [20:56] *face palm [20:57] thomi, active means that launchericons's window is focused [20:57] lamal666: ahhh, yes [20:58] lamal666: it's the same icons as the launcher, so that makes some sense at least [20:58] so there's only one where it's true and when i looked at the introspection graph it was the one i had selected [20:59] heh [20:59] oooh, I must remember to fix the introspection graph utility [20:59] * thomi adds a card [21:14] Is any here very familiar with glib? [21:25] j-johan-edwards: you might get more help at #ubuntu-desktop [21:29] thumper: roger [21:51] Ok.. I give up on threading this scope. GIL slaps every solution I've tried. === salem_ is now known as _salem [22:28] Does anyone know how to remove a package that has been built with a recipe and is wrong? [22:29] or do I just have to increase the version number on the new one? [22:29] "INFO unity-scope-spotify_0.1-0~4~precise1.dsc: Version older than that in the archive. 0.1-0~4~precise1 <= 0.1-0~12~precise1" [22:30] the rev 12 is from another branch. [22:40] I increased the version. Done. :) [23:14] mhr3, "Scope net.launchpad.scope. is crashing too often, disabling it" [23:14] 1) That's pretty awesome. [23:14] 2) I feel offended :P [23:20] davidcalle, ?? [23:20] * mhr3 is half asleep [23:20] oh... right :) [23:22] ;) [23:26] * davidcalle still thinks this scope has just the right amount of crashes...