[00:26] * spikeb likes how the dock has intellihide, yet the window manager has stupidplacement [00:32] spikeb: What is "stupidplacement"? [00:33] Omega, so far, compiz has tried to place every single window i've opened under the dock (which then hides) heh [00:40] spikeb: Same problem here, sometimes mine even come under the top panel :< [00:40] haha, nice. === _LibertyZero is now known as LibertyZero [08:02] hey folks [08:11] good morning [08:47] good morning [09:14] loicm, ping [09:33] loicm, GetInputWindowId() does not work on a quicklist [09:34] loicm, I verified it's opened so should have m_input_window set... but still returns NULL [09:35] loicm, not sure what I might be missing, which obviously causes m_input_window not to be set on the BaseWindow of the quicklist-view [09:36] MacSlow: let me check [09:41] didrocks: dbus-daemon stays at a constant 5% cpu usage because indicator-datetime keeps crashing, coming back up, crashing, ad nauseum... known problem? It seems like something needing an urgent fix [09:41] kamstrup: do you have a stack of the crash? [09:41] kamstrup: geoclue has been added recently and I saw some crash with it [09:41] didrocks: i'm not really sure how to get that... [09:42] kamstrup: activate apport is the easiest way to get it: https://wiki.ubuntu.com/Apport#How to enable apport [09:42] kamstrup: ensure to clean /var/crash/ in case you already got something [09:42] kamstrup: I'll even dup it if it's a dup [09:43] kamstrup: I'll let ken knows if you can have something (I'm not upgrading because of nvidia/xorg) [09:44] didrocks: ok, so the bug in i-d is https://bugs.launchpad.net/indicator-datetime/+bug/714763 [09:44] didrocks: and the geoclue-master crash is https://bugs.launchpad.net/ubuntu/+source/geoclue/+bug/715086 [09:45] kamstrup: I'll let ken knows about it, thanks! === smspillaz is now known as smspillaz|swap [09:49] MacSlow: are you sure the input window of the QuickListView is enabled? [09:49] using EnableInputWindow() [09:50] loicm, yes... for testing I called it after Quicklist::Show() which does EnableInputWindow (true, 1) [09:51] MacSlow: did you try with non-override-redirect window? [09:51] MacSlow: removing the second argument or setting it to 0 [09:51] * MacSlow checks [09:54] loicm, no that's not fixing it [09:54] didrocks: eeek, not only do I have dbus-damon at 5%, but unity-panel-service sits at 50% constant churn! [09:55] kamstrup: yeah, this one is known [09:55] kamstrup: tedg was pinged about it, I didn't follow then :) [09:55] kamstrup: kill + respawn works though [09:56] MacSlow: weird, do you have a specific branch to test that? [09:57] or maybe trunk is ok [09:58] loicm, one sec [10:00] MacSlow: btw GetInputWindowId must be called once the InputWindow is enabled [10:01] MacSlow: the code destroys and creates a new X11 window each time EnableInputWindow is called (I think it should be cached somehow once created though) [10:02] MacSlow: you can make sure it's enabled by calling InputWindowEnabled() [10:03] loicm, before calling GetInputWindowId() ?! [10:03] MacSlow: yes, just to test, not for production [10:06] didrocks: sudo rm -rf /usr/lib/indicator-datetime works better [10:06] ;-) [10:07] kamstrup: ahah, you are mean :-) [10:07] didrocks: i don't take no crap ;-) [10:07] loicm, btw lp:~canonical-dx-team/unity/unity.fix-701543 [10:07] heh ;) [10:07] dbarth, ping [10:12] good morning [10:14] MacSlow: I didn't try, but I see you're testing that the InputWindow is enabled in unityshell.cpp when you handle the compiz keyboard shortcut, what's the result of that test? [10:15] loicm, this one fails compWin = self->screen->findWindow (xWin); [10:16] xWin is valid but compWin remains NULL after the call to findWindow() which is on the compiz-side then I'd say [10:17] loicm, I've to poke smspillaz about this too [10:23] MacSlow: maybe you can try with findTopLevelWindow(id, true) [10:27] loicm, tried that... no difference [10:32] loicm, the only difference I see atm between the launcher-window and a quicklist is the way they are created... the launcher-window is kept, since it's always there... whereas the quicklists are created and destroyed as they are needed. [10:33] MacSlow: isn't there always only one quicklist displayed at a time? [10:33] MacSlow: (or none) [10:34] MacSlow: cause if it's the case, maybe you can try to keep that one around, showing/hiding/movig/resizing it when needed? [10:35] loicm, sure only one at a time... but it gets destroyed if not needed / visible [10:35] loicm, to save memory we decided against that iirc [10:38] loicm, never mind... just looked... this was changed... they're now kept around [11:30] didrocks: hey [11:30] didrocks: http://bazaar.launchpad.net/~unity-team/nux/trunk/revision/207 breaks my build! [11:30] loicm: how? I'm just disting additional files! [11:32] didrocks: no it's not just for the dist, it's for the compilation too [11:33] loicm: ok, and those are Win only? shouldn't be compiled? [11:33] didrocks: exactly, it's full of windowseries :) [11:33] loicm: argh, ok, adding to extra dist then, one sec :) [11:33] Win can be a window or windows, I was wondering :) [11:34] hehe [11:35] loicm: rev 208 pushed, sorry for this :) [11:35] didrocks: np, thanks [11:41] apw: pong [11:42] doh, I didn't see you ping earlier [11:42] dbarth, erm, did i ping, i forget [11:42] agateau: there's something still broken with konversation [11:42] agateau: i don't get msg menu alerts anymore [11:42] apw: oops, sorry, i meant API [11:42] but hi anyway ;) [11:42] dbarth: I don't think the new libindicate-qt has reached main yet [11:42] heh phew, thought i was loosing my mind [11:43] nah, you're fine, let me know if you find mine however [11:43] agateau: that would explain why [11:43] agateau: what's blocking it? [11:44] dbarth: nothing afaik, Riddell told me he would do it yesterday [11:44] dbarth: let me check with him [11:44] agateau: thanks, keep me psoted [11:44] posted [11:51] cando_: this time, it's a real bitesize :) === MacSlow is now known as MacSlow|lunch [12:16] didrocks, finally...:):)..on it now... [12:17] * spikeb upgraded his netbook to natty yesterday [12:17] cando_: great! [12:17] didrocks, could you review my previous mp?or it's a DBO's work? [12:17] cando_: this one will be DBO [12:18] OK [12:18] cando_: the bitesize, I know that code quite well, so I can :) [12:18] aside from compiz trying to place windows under things, natty works pretty well right now for me. [12:19] didrocks, https://code.launchpad.net/~cando/unity/fix_692444/+merge/48767 [12:20] cando_: I know about that one, I was anwsering on it [12:20] ah ok:) [12:20] sry [12:20] spikeb: upgrade to today's compiz, it should fix it [12:20] cando_: no worry :) [12:20] cool didrocks === Akshat_ is now known as akshatj [12:29] good afternoon [12:34] has anyone fixed the bug that shows unmounted volumes in launcher? [12:34] because i'm running the bzr version and volumes don't appears [12:34] *appear [12:59] hi all [13:00] hi [13:01] does anyone know how can one put icons in an application indicator? [13:11] didrocks, got it! fixed..:) (the mouse middleclick one..) [13:11] cando_: nice! === artir is now known as afk|artir [13:21] DanRabbit, a tongue-biting moment: http://wstaw.org/m/2011/02/06/plasma-desktopPm5066.jpg [13:22] mpt: OH. MY. GODS. [13:22] njpatel: do you know of any way to run unity compiz in a VM ? [13:23] nerochiaro, kenvandine has it running in virtualbox apparently, but I'm not sure how, sorry [13:23] mpt, my eyes just exploded [13:23] njpatel: i'll ask him then [13:24] kenvandine: when you have time, if you can explain how did you manage to run unity in virtualbox, that would be helpful === afk|artir is now known as artir === oubiwann is now known as oubiwann_ [13:45] nerochiaro, hey [13:45] nerochiaro, yeah... just need to install the guest additions in the VM [13:45] and enable 3d in the settings for the VM [13:45] which i think might be the default [13:50] dbarth: follow up on indicate-qt: it was blocked by plasma-widget-message-indicator and kopete-message-indicator which FTBFS [13:51] dbarth: both fixed by new releases [14:03] does anyone know how can one put icons in an application indicator? [14:05] evfool: what do you mean by "put icons" ? like add launchers? === m_conley_away is now known as m_conley [14:06] vish: no, I would like to add icons before the menu items in a custom application indicator, like the icons in the Sound menu before the media players or like the icons before the statuses in the me menu [14:07] and I haven't found any examples on it [14:08] evfool: iirc there are examples on the wiki, aernt there? [14:08] vish: yep, but only on how to add text items, and I've tried with ImageMenuItems, and they don't really work [14:09] evfool: hmm, no clue then.. maybe ted-g might know.. ;) [14:11] ok, then I'll ask tedg [14:11] thanks vish [14:21] evfool, do you mean just using GtkImageMenuItem's? [14:21] tedg:yep [14:22] i was going to start hacking into unity but i'm having some trouble [14:23] when i log in, i have unmounted volumes in the dock [14:23] when i type "compiz --replace" they disappear === MacSlow|lunch is now known as MacSlow [14:25] gord, Do you know anything about mounted volumes in the launcher? ^ [14:26] what i'm going to fix (try to) is hide the icons until the volume is mounted [14:26] and if the volume is mounted, show it (it doesn't show now) [14:27] ...and, allow it to "keep in launcher" in case a user want to mount it quickly just by pressing the icon [14:29] tedg, njpatel is your man for that [14:29] what i'll try now is just show the icon if i mount the windows partition [14:30] i'm looking at DeviceLauncherSection.cpp now [14:32] i'll open the bugs for what i want to do [14:35] DJKorbit, Cool, that'd be a good idea. [14:35] DJKorbit, Bugs are a good way to track discussions as everyone's not around all the time :) [14:35] DJKorbit, If you post the bug numbers here that'd also be helpful. [14:40] DJKorbit, i think the point is that the partitions show up and you can click to mount them, so its not a bug that they are on the launcher [14:41] tedg: any suggestions on how to show icons of ImageMenuItems in the indicator? [14:43] cando_: any tidbits for the report today? [14:43] yeah, i've fixed #703411 [14:43] proposed an mp [14:43] gord, i can't mount them, so i suppose not being able to mount is a bug too [14:43] but DBO is not here.. [14:44] gord, anyway, i should be able to remove them if i want to, and i can't [14:44] 703411 [14:44] vish: Hey, want to nominate someone to talk about papercuts at udw? [14:44] evfool, http://library.gnome.org/devel/gtk/stable/GtkImageMenuItem.html#gtk-image-menu-item-set-always-show-image [14:44] cando_: oh god, that's been bugging the heck out of me [14:44] lol [14:44] tedg: thanks, I'll look for its python version [14:44] dbarth: word. [14:45] DBO: good morning! [14:45] yeah DBO! [14:45] DBO: https://code.launchpad.net/~cando/unity/fix_703411/+merge/48921 [14:45] morning gents [14:45] DBO: here's your breakfast present! [14:45] i've opened bug 715185 [14:45] Launchpad bug 715185 in unity (Ubuntu) "Non-removable volume doesn't show up in Unity dock after mount." [Undecided,New] https://launchpad.net/bugs/715185 [14:45] DJKorbit, morning [14:45] i'll assign it to me [14:45] wow [14:46] the onslaught begins :) [14:46] DBO and also https://code.launchpad.net/~cando/unity/fix_692444/+merge/48767 [14:46] lol [14:46] DBO: *jump* [14:46] you pinged earlier? :) [14:46] tedg: aha, the person I was looking for [14:46] tedg: who would be ideal for a udw session about papercuts? [14:46] all we need is..DBO! :) [14:46] * tedg *waves hand* this is not the irc nick you're looking for [14:47] nigelb, vish ? [14:47] tedg: well, vish is generally hesistant to do session :D [14:47] tedg: wanna do one yourself? ;) [14:48] tedg: not this cycle.. as we dont pull anything from upstream.. ;) , (/me already getting flamed for that :( ) [14:48] DBO, hi! [14:48] nigelb, Well, I haven't kept up that much with the project.... I'm not sure I could answer all the questions. jcastro, any ideas? [14:48] DBO, i'll begin working on unity now [14:48] desrt, how does one pack a pointer into gvariant? [14:48] nigelb: vish [14:48] it's not possible [14:48] GVariant is a serialisation format. pointers are not generally able to be serialised [14:48] vish: popular opinion nominates you to take a session at UDW [14:48] jcastro: i'v been directing more folks to the bitesize project, instead.. ;) [14:48] :) [14:49] vish: shall I put you down for a session? [14:49] nigelb: but there is not much point in doing a session for a project which cant be active.. [14:50] vish: ok, about bitesize bugs then? [14:50] that would be jcastro :D [14:50] DBO: of course you have the option of storing a pointer into a uint64... [14:50] lol [14:50] desrt, yes but thats a hack [14:50] vish: I wanna jcastro for something else already *twinkle* [14:51] DBO: yup. [14:51] nigelb: put me down for one, pencil DBO with me just like last time, depending on his workload. [14:51] desrt, so the idea here is mostly "you dont"? [14:51] DBO: ya.... [14:51] GVariant was never really intended for shuffling of data around within your program [14:51] jcastro, im with ya [14:51] jcastro: Yessir :) [14:51] but rather to be used as a way to get data between programs or on disk [14:51] vish: thanks :D [14:51] desrt, its quite handy for that though [14:52] chrisccoulson: ping [14:52] DBO: i think the fact that people are finding GVariant to be so damn useful at things it wasn't designed to do is more of a statement about the poor state of GValue :p [14:52] jcastro: monday 28th at 1900 UTC [14:53] DBO: i'd recommend you use it, if it wasn't so difficult :/ [14:53] desrt, you gave us the BFG 9000 when before we only had a pea shooter, what did you expect? [14:53] DBO: you see, not able to abuse GVariant this time (the other trick was still awesome though :p) [14:53] DBO: i expect you to use the pea-shooter, REALLY HARD [14:53] :) [14:54] I am now about to get into easily googled territory, but we can write our own marshallers for gvariant right? [14:54] tedg, how may I help? [14:54] DBO: O_o [14:54] oh, no? okay [14:54] you mean your own functions to serialise it into other formats? [14:54] right [14:54] yes [14:55] you just use the normal API... [14:55] iterate over the structure of the value and do what you want to with it... [14:55] DBO, what _are_ you up to with GVariant? [14:55] hey dereks [14:55] njpatel, evil [14:55] hey desrt [14:55] even [14:55] see GDBus as an example of how that's done [14:55] njpatel: hey :) [14:55] basically passing around the entire launchers set of data [14:56] njpatel, DJKorbit was asking some questions about mounted volumes above. [14:56] * njpatel scrolls up [14:56] tedg, yeah I was helping him with that earlier [14:56] did I miss something before I logged in? [14:56] DBO: at one point we were discussing the possibility of adding a GObject API to allow arbitrary objects to specify ways to encode/decode themselves to/from GVariant [14:56] what gord said [14:56] desrt, I like that, do it [14:57] i think we decided that it was less useful than i originally thought it was [14:57] boo you suck :P [14:57] but the idea wasn't entirely rejected -- just really deprioritised [14:57] if you have some compelling use cases we could probably think about picking it up again [14:57] Make it work for C++ and we'll have tons ;) [14:58] C APIs are perfectly usable from C++ ;) [14:58] but actually, since you mention it, i guess it wouldn't be too hard [14:58] heh, it just looks horrible, unfortunately [14:58] create an interface-only c++ class that you can tie into existing classes with some serialise vfuncs [14:59] problem with c++ is that you can't really get the same sort of access to the type system that GType allows you [14:59] hi m_conley [14:59] ie: our static pseudo-factory pattern doesn't work so good... [14:59] right [15:01] njpatel, btw nux has dnd in trunk now, so if you want it elsewhere, you are good to go [15:01] chrisccoulson: hey - I have a newbie C++ question for ya. So I'm working on my extension that integrates Thunderbird with the Messaging Menu, and that means I'm using the libindicate library...things compile to .o files just fine, but during linking, I'm seeing things like "hidden symbol `[function within libindicate]` isn't defined". Any idea what I might be doing wrong? [15:01] DBO, awesome, and the unity side your working on? [15:01] as we speak [15:01] I have launcher adding .desktop files from menus right now [15:02] m_conley, you're not using the thunderbird build system by any chance are you? [15:02] chrisccoulson: er, I am indeed. I basically used the globalmenu-extension build system as a template. [15:02] chrisccoulson: not a good idea? [15:02] m_conley, did you add the headers to config/system-headers? [15:03] DBO: excellent! ;) [15:03] chrisccoulson: ah, no I didn't (and you mean system_wrappers, right?) [15:03] DBO, excellent [15:03] cd .. [15:03] * desrt hates on his internet connection [15:03] * m_conley sighs, and switches to the right window [15:04] desrt, teach you to live in bizarro australia [15:04] * desrt is stuck on 3G until friday [15:04] Australia? [15:04] DBO: i think you're confused [15:04] m_conley, there should be a system-headers file in config/ [15:04] :) [15:04] desrt, you certainly do live in canada [15:04] chrisccoulson: ah, yes, there it is [15:04] but you share the same bad internet connexion rules :) [15:04] not anymore [15:05] our conservative(of all parties) government stepped in [15:05] chrisccoulson: cool, thanks. :) [15:05] and said to the regulator "fix it, or we'll fix it for you" [15:05] kenvandine: the guest additions don't seem to be able to build in natty. how did you install them ? [15:05] m_conley, you'll probably need to do a make distclean after changing it (and make sure that dist/system_wrappers gets cleaned too) [15:06] with the "we'll fix it for you" shaded by undertones of the same industry minister musing aloud to the media that this particular regulatory body may have just about outlived its usefulness [15:06] chrisccoulson: alright, awesome - I'll give that a shot. [15:08] desrt: waow, quite drastic advice… :) [15:13] can someone confirm for me that 8pm CET is 2pm EST? [15:14] * desrt thinks so, but doesn't want to miss a meeting :) [15:15] desrt, sounds right to me [15:15] cheers [15:16] kamstrup: hey? [15:19] just a question... [15:19] when i mount my windows partition, "mount-added" signal is emitted right? [15:20] the signal is not connected so i suppose connecting it would solve the problem [15:20] sounds logical to me [15:20] not a programmer though [15:21] i'll hack the code a little, just to get familiar with it [15:21] i'm not familiar with all this gnome APIs [15:21] m_conley, i found an issue with the tbird menu this morning. thankfully, nobody noticed it though, because of another packaging bug which meant the extension wasn't loaded ;) [15:21] chrisccoulson: what's the issue? [15:22] m_conley, http://bazaar.launchpad.net/~extension-hackers/globalmenu-extension/trunk/revision/98 [15:22] the compose window wasn't getting a menu if you opened it a second time ;) [15:22] chrisccoulson: nice catch! [15:24] chrisccoulson: ok, so adding those headers to system-headers has changed the error message to something more recognizable: "undefined reference to `[libindicate function]`". Is there a place in this build system where I'm supposed to specify included libraries? I did it manually within the Makefile in the code folder, but there's probably a more correct way...right? [15:24] m_conley, did you add the checks to configure.in too? [15:24] i'm compiling my code now, lets see what happens [15:25] chrisccoulson: er, no [15:25] m_conley, that might be why :) [15:26] chrisccoulson: Did you add checks to configure.in for dbusmenu? [15:26] chrisccoulson: or is there an example I should be looking at? [15:26] m_conley, if you diff configure.in between globalmenu-extension and the one in mozilla-central (from FIREFOX_4_0b11_BUILD2), you'll see what i added to it [15:26] cool, thanks [15:26] desrt: 1 sec. [15:26] easytz is your friend here [15:27] desrt: 1pm EST [15:27] desrt: hey [15:27] 2pm EST, that was [15:28] right. 2pm. :) [15:28] did you guys want to chat on IRC or with voice? [15:28] * desrt has always had really bad luck with voice when the atlantic ocean becomes involved [15:28] desrt: see http://easytz.com/?t=2011-2-8-19-0&d=30&l=1c64-5803-5980-16bf-af8-e3a-3aaa-3eb [15:28] desrt: skype usually works ok'ish if you are not too many on the call [15:28] how can i switch to my freshly compiled unity? [15:29] compiz --replace ? [15:29] skype is bad for me at the moment because my internet connection sucks :/ [15:29] unity [15:29] DJKorbit: ^^ [15:29] desrt: ah, ok [15:29] desrt: tell bregma about it [15:29] * desrt is stuck on 3G until friday [15:30] dbarth: unity-panel-service crashed [15:30] hopefully the installers succeed on that day. this is the 3rd attempt :( [15:30] dbarth, i can't report the bug because i have an old libcups2, i'll update my system [15:31] dbarth, i don't have a window manager now, i'll try running unity from a text console [15:31] desrt: ok, maybe IRC is the safe bett then [15:32] DJKorbit: from a VT you can also use the 'unity' command to restart things propelry [15:33] DJKorbit: or do DISPLAY=:0 metacity --replace [15:33] dbarth, it's working now, but the bugfix didn't work [15:33] DJKorbit: if you're really stuck in a GL issue that prevents unity/compiz from getting out of it [15:34] adding a "mount-added" signal is not enough [15:34] API, rodrigo_: ping? you guys there? [15:34] dbarth, hi [15:34] I made you a ping in the morning, did you see that? [15:34] API: yes, saw it, but late [15:35] konversation fails to either highlight tabs or trigger the menu; that's becoming insane [15:35] hi dbarth [15:35] at least it's staying quite with connman now [15:35] rodrigo_: hi [15:35] dbarth, ok, sorry [15:35] can we do a quick skype call [15:35] fine for me [15:36] ok, a mon, moving to other room [15:38] dbarth, rodrigo_ ready [15:38] API: c.g. right? [15:38] dbarth, c.g ? [15:38] nah, wrong nick [15:39] igalia-corunha [15:46] tedg, looking at bug 714763 , clearly we shouldn't crash with no providers [15:46] Launchpad bug 714763 in Indicator Date and Time "indicator-datetime-service crashed with SIGSEGV in geoclue_master_client_set_requirements_async()" [Medium,In progress] https://launchpad.net/bugs/714763 [15:46] but should we add a hard depends on geoclue and one provider? [15:46] kenvandine, https://code.launchpad.net/~ted/indicator-datetime/bug714763/+merge/48837 [15:46] seems like a problem any other consumers of geoclue will hage [15:46] have [15:46] tedg, i saw that [15:46] kenvandine, Yes, I think it's kinda a bad geoclue thing. :( [15:47] the problem is those recommends don't get installed on upgrade [15:47] Really? That seems broken. [15:47] so i think we should at least change the recommends for geoclue to a depends [15:47] yes really [15:47] strange, i've just connected a usb drive and it doesn't show up in the launcher [15:47] Can we require a "geoclue-provider" and then make every provider "provides geoclue-provider" so they just have to have one? [15:47] but this was working before [15:48] tedg, great idea [15:48] does your unity dock show usb devices? [15:48] DJKorbit, it did for me yesterday [15:48] really strange [15:48] there could be bugs there [15:49] kenvandine, Probably should be "Requires: ubuntu-geoip | geoclue-provider" just to seed the one we want first. [15:49] seb128, what do you think? ^^ [15:49] kenvandine, how can i change to my ubuntu's unity dock instead of the one i'm hacking? [15:49] DJKorbit, what do you mean? [15:49] kenvandine, what? [15:50] indicator-datetime [15:50] kenvandine, ubuntu by default uses unity, i want to change to that unity version instead of the one i'm using now that i installed in /opt/unity [15:50] it really needs at least one provider installed [15:50] DJKorbit, ah... do you have different session files? [15:50] i never run it from local installs... [15:51] hmm, ok [15:51] i don't have different session files [15:51] seb128, so perhaps adding "Requires: ubuntu-geoip | geoclue-provider" to indicator-datetime [15:51] i'm using my regular session [15:51] and then make each provider provide geoclue-provider [15:51] seb128, i am about to upload that other new source package, geoclue-providers [15:52] kenvandine, "Requires"? [15:52] what's that? fancy name for Depends? [15:52] Depends [15:52] copy and paste from tedg :) [15:52] kenvandine, seems fine yes [15:52] * kenvandine does so [15:52] tedg, i'll go ahead and distro patch your fix too [15:54]