[01:07] Anyone have experience with the Launcher.cpp source code to help with a quesion? [01:07] and signals [01:11] DBO can you help for a little bit? [01:11] ^^ [01:19] jcastro are you around? I want to know how you want this bug fixed. [01:30] heh [01:31] wish i could help. [01:34] as do I [01:42] anyone know how to get the unity launcher to scroll in the c++ code? what do I call? [02:01] rbnswartz, the edge scrolling is going to be one I do [02:02] sorry :/ [02:02] okay thanks are you going to be around tomorrow morning? [02:02] yep [02:03] 9-5 (sometimes later), M-F, EST [02:03] okay I'll probably have more questions then thanks. bye [02:03] cheers === akshatj|bed is now known as akshatj === oubiwann-holiday is now known as oubiwann === oubiwann is now known as oubiwann-holiday [07:10] morning [07:37] hey kvalo === evilvish is now known as \vish === \vish is now known as vish [08:26] d [08:35] * boulabiar watching Mark conf. in Tunisia http://ustre.am/rrdQ [09:24] anyone on natty that can help me out with a quick compilation check? [09:24] kamstrup, ! [09:24] http://paste.ubuntu.com/543961/ [09:24] njpatel: ^^ [09:24] kamstrup, checking now [09:25] kamstrup, http://paste.ubuntu.com/543963/ [09:26] njpatel: W.T.F [09:26] njpatel: and that's on natty you tell me? [09:26] kamstrup, he didn't use the command in your example [09:26] kamstrup, yep [09:26] kamstrup, you need the object before the lib [09:26] otherwise it breaks [09:27] oh, woops [09:27] seb128: ? [09:27] kamstrup, gcc .c -o bin -l... [09:27] not gcc -l .c -o bin [09:27] the --as-needed in natty is stricter about that [09:27] the object needs to be before the libs [09:28] hmm, it doesn't compile [09:28] you need to move "$(pkg-config --libs --cflags glib-2.0)" at the end of the line [09:28] otherwise it will not find the symbols [09:28] seb128: ok, but I don't think it matters for this particular case... debugging a va_list issue? [09:28] http://paste.ubuntu.com/543967/ [09:29] seb128, right, but I think kamstrup's issue is different [09:29] well he didn't say what his issue is [09:29] he just asked a test build [09:29] ok :-) [09:29] sorry to be noisy, I will go back to my corner [09:29] kamstrup, also, for compiling tests and examples on natty, you need to add $(DEPS_LIBS) to LDADD in both those makefiles, otherwise the linking doesn't work [09:29] seb128, no, no, we like it when your not in your corner [09:29] ;-) [09:29] seb128, it means launchpad is quicker [09:29] ;) [09:30] lol [09:30] seb128: it's ok - I learned something anyway - just not the thing I was asking about :-D [09:30] kamstrup, so, yeah, it compiles and runs fine. I spent some time looking into this last night too, so I could do the review, but no luck :/ [09:30] seb128: so the deal is that it seems that gcc's varargs implementation has changed in some way between M and N [09:34] njpatel: can you try this one then... this time with fixed compilation instructions ;-P http://paste.ubuntu.com/543969/ [09:36] kamstrup, fails http://paste.ubuntu.com/543970/ [09:41] njpatel: Woohoo. Or I mean "Woohoo" - so know we can reproduce it in a simple way [09:41] yep :) === agateau_ is now known as agateau [11:00] njpatel, hey, any idea what the compiler would accept as the default value for a gsettings entry of type "(as)" [11:01] needs the default node to compile [11:01] can't find any docs on this, just scanty tuts [11:02] ronoc, hold up, unity uses (as) [11:02] ah /me checks [11:03] ronoc, http://bazaar.launchpad.net/~unity-team/unity/trunk/annotate/head:/com.canonical.Unity.gschema.xml [11:03] thx njpatel [11:11] njpatel: ok, one last stab - this should work I think... http://paste.ubuntu.com/544011/ [11:13] njpatel: if this works for you then I have a solution that'll work on natty as well [11:14] okay, hol dup [11:14] kamstrup, it works! [11:20] njpatel: holy moly! [11:30] njpatel: pushed changes to the dee-gdbus branch, so it should build (and even work!) on Natty ;-) [11:34] kamstrup, awesome dude, I was just getting to the point that I needed to start using dee for the place backend stuff :) [11:34] kamstrup, will take a look as soon as I'm done with my current stuff [11:35] njpatel: sure thing [12:22] kamstrup, approved :) can you bzr pull before merging, though, as I made a small build fix to the branch [12:22] for natty [12:35] njpatel: thanks a bunch! [13:02] kamstrup, ping, I forgot something [13:02] kamstrup, time to bump the dee major version? [13:03] kamstrup, also, the daemons need updating too? [13:03] njpatel: yes, it'll be a dee-0.5 series [13:04] kamstrup, i meant the lib versioning and gir etc 2.0? [13:04] kamstrup, as it's breaking api/abi/kittens [13:09] * hyperair is beginning to see a correlation between memory leaks in his system and anything ayatana comes up with. [13:09] * hyperair sighs [13:10] dbusmenu is one hell of a leaky library [13:10] you know you've gone wrong when you've allocated 9M of memory for strdup just to set some text value somewhere [13:18] njpatel: hmmm... the biggest reason for that would be to allow parallel installations I think - since unity was the only consumer [13:18] njpatel: so any new consumers just have to check version >= 0.5 [13:18] that should, i guess [13:19] kamstrup, are you going to update the place daemons and release everyhting in one go? [13:19] njpatel: if there had been other consumers then I'd agree [13:19] njpatel: I was thinking that yes [13:24] kamstrup, okay, so time frame? :) [13:24] njpatel: today [13:24] kamstrup, as i can't really use dee until it's in natty [13:24] kamstrup, oh, wow [13:24] :) [13:25] kamstrup, is too cool for school [13:25] damn comma [13:27] njpatel: you may address Mikkel "über trapper, and code ninja" Kamstrup [13:27] I believe that would be the appropriate etiquette [13:28] heh [13:28] I think I'll have some lunch and then do some more places stuff [13:28] njpatel: bonne appetite dude [13:28] kamstrup, let me know if/when you update a place, so I have something to test against === njpatel is now known as njpatel_ [13:32] ronoc: it seems that a lot of memory leaks come from dbus_g_proxy_end_call allocating data and expecting libdbusmenu to free it, which it doesn't. [13:52] hyperair, okay good to know, I'll talk to ted when he comes from holidays (after xmas) [13:52] ronoc: sure [13:52] ronoc: but what i find strange is that nobody else with spectacular uptime have seen memory leaks before. =\ [13:52] really strange. [13:55] hyperair, yeah I haven't noticied, although alot of time i have my indicator-panel down cause i'm in dev mode [13:55] ronoc: ah, i see. [13:55] ronoc: vish seems to have pretty long uptimes though, and it doesn't seem like he has memory leaking issues either =\ [13:55] weird [13:56] yes, very weird. [13:56] hyperair, going to push ahead and get this new sound menu registration work this afternoon, one less dependency for banshee :) [13:57] ronoc: new sound menu registration? [13:57] O_o [13:57] hyperair, yes strictly using mpris and gsettings, [13:58] gsettings for removing yourself from the menu [13:59] ah, nice [13:59] but does gsettings have C# bindings? [14:02] ronoc: oh lol, all the GError's haven't been freed either. [14:06] hyperair, which file are you looking at ? [14:06] hyperair, http://live.gnome.org/GnomeGoals/GSettingsMigration [14:06] ronoc: client.c in libdbusmenu, and listener.c in libindicate. [14:07] hyperair, it appears it is on their todo list [14:08] hyperair, considering banshee is the default application this is not so important immediately [14:08] hyperair, users just won't be able to remove banshee from the menu until those bindings get sorted [14:10] ronoc: will it be a blocker for natty? [14:11] hyperair, i'll find out [14:11] ronoc: do you have commit access to libdbusmenu? [14:11] hyperair, yes [14:11] ronoc: alright, please take a look at this patch... [14:11] http://paste.debian.net/102478/ [14:11] ronoc: ^^ [14:13] ronoc: and http://paste.debian.net/102479/ for libindicate [14:14] hyperair, looks good, why don't you file a bug a place the patches on each bug report. This really is ted's domain, he will come back to you in the new year [14:14] ronoc: right. [14:16] ronoc: you didn't touch indicator-applet, right? so i should upload the patch for that as well? [14:17] hyperair, you should file a bug and patch for any observations/fixes you happen upon, that way all the karma goes to you ! [14:18] ronoc: heh i'm not really bothered with launchpad karma. i get plenty from automated PPA uploads ;-) [14:18] hyperair, fair enough [14:21] morning [14:22] hey ronoc [14:22] lamalex, hey [14:22] HOW WAS GODSPEED [14:22] EXCELLENT [14:22] good two hour set [14:22] no encores or nonsense like that [14:22] very "lift your fists ..." [14:23] morning metal fans! [14:23] a bit rusty to start with [14:23] hey jcastro [14:23] but they soon warmed up [14:23] no merchandise I'm afraid [14:23] lamalex, ^ [14:23] it was packed so was straight in and out [14:23] nbd [14:24] ronoc, that was their first show in 10 years I think [14:24] I'm pretty sure London was the start [14:24] they played atp at the weekend on the south coast (england), they were curating, played all three nights apparently [14:24] * ronoc finds a link [14:25] brb [14:27] yeah yeah that's right [14:30] back [14:35] ronoc: https://bugs.launchpad.net/libindicate/+bug/690668 [14:35] Ubuntu bug 690668 in libindicate "libindicate, libdbusmenu, indicator-messages-service memleaks" [Undecided,New] [14:39] hyperair, cool thx [14:53] njpatel_: is the top bar multimonitor not-finished-yet bitesizeable? [14:53] wow, what a sentence [14:54] jcastro, yep, though I'd like to add some notes on the implementation [14:54] jcastro, got the bug and I'll brain dump in there now? [14:54] njpatel_: so I have an idea based on what jason did on a bug that was a little bit over the bite size [14:54] he just braindumped in the bug [14:54] with like a "here's how you should do it" [14:56] Bug #675862 [14:56] Launchpad bug 675862 in unity (Ubuntu) "Top panel isn't multimonitor aware" [High,Triaged] https://launchpad.net/bugs/675862 [14:56] jcastro, dude, is that not what I just said? :) [14:56] jcastro: I was typing that too [14:57] great minds think alike ... etc. [14:57] jcastro, ah, okay, sorry, I looked away and it seemed like yours came in a bit later [14:59] urgh dual monitor global menubar issues [14:59] what do you mean? [14:59] seb128: since didrocks is not here I'll give you the heads up: I just rolled libzeitgeist-0.3.0 and dee-0.5.0. Both are abi+api breaks [15:00] seb128: the place daemons will follow any minute, and njpatel_ will handle the unity side [15:00] so by tomorrow we should be rolling [15:02] kamstrup, so should those be uploaded today? [15:02] or will they break unity until it's updated? [15:02] seb128: probably not [15:02] seb128: I think it'll be better with a batch tomorrow if njpatel_ also wraps up by then [15:02] tell me you changed the soname [15:03] * kamstrup whistles innocently [15:03] I'm not uploading an abi break without soname change [15:03] jcastro, done [15:03] kamstrup, do place daemons compile against latest libunity? [15:04] seb128: ok, np. i'll roll with updated sonames tomorrow then [15:04] kamstrup, thanks [15:04] njpatel_: hehe, they compiled against the specially handcrafted libunity here on my hard disk... is that not good enough for you? [15:04] kamstrup, is there any reason you can't do it today? [15:04] njpatel_: this is the last multimonitor bug afact that is a regression from the old things, after this you'll be parity with the other stuff [15:04] seb128: i might get around to it [15:04] kamstrup, lol, no, sorry :) [15:04] kamstrup, should be just a configure numbers change [15:04] seb128: my top prio right now is to give njpatel something to play with [15:05] kamstrup, we'll kinda need them to compile against libunity 1.0 :) === njpatel_ is now known as njpatel [15:05] kamstrup, sorry to be annoying about that, but if you don't change the soname we will break unity [15:05] jcastro, yep [15:05] if we do it means we can have a safe upgrade path [15:05] njpatel: there's a second bug for what to do about a menu on the second monitor iirc [15:05] current unity depends on the current soname [15:05] seb128: yeah, but the configure script is a bit annoyin in this regard because it computes a bunch of stuff from the soname [15:05] jcastro, I think we can do better with the launcher, making it stay on the primary monitor too, but not right now [15:05] jcastro, yep, should maybe defer that commenter on this bug to that one, so he can register his idea there [15:06] njpatel: ok so I can just say in this bug, fix the one panel, if you care about the menus and stuff on the 2nd monitor see the other bug [15:06] kamstrup, if you change the soname I can land the new dee in natty today [15:06] which should make testing easier [15:06] njpatel: man, it's like we're reading each other's thoughts. [15:06] seb128: so changing the soname also breaks a lot of other stuff unless I hack the configure script to not be "too clever" :-) [15:06] jcastro, showing the menu bar on multiple displays is already in the spec. What's the bug number? [15:06] kamstrup, ok, sorry about that [15:06] kamstrup, did you figure anything out with dee? [15:06] jcastro, heh :) [15:06] kamstrup, I just know how upgrades work [15:07] kamstrup, if we make the new dee Breaks unity << people will upgrade dee in their dist-upgrade [15:07] which will remove unity [15:07] kenvandine: there is a 0.5.0 up - but beware that it breaks abi without updating the soname [15:07] that's always a fail [15:07] you can't break with abi breaks if you don't update the soname [15:08] njpatel: considering the above discussion I have with seb128 it might make sense to call in dee-2.0 after all... [15:08] mpt: Bug #683084 [15:08] Launchpad bug 683084 in Ayatana Design "Global menu doesn't work well with more than one screen" [Undecided,In progress] https://launchpad.net/bugs/683084 [15:08] that also means you'll like me better right njpatel? ;-) [15:08] thanks [15:09] njpatel: so u-p-f trunk together with dee-0.5.0 and libzeitgeist-0.3.0 should work... at least somewhat... :-) [15:09] hi ppl [15:15] Hello eveyone onboard [15:17] I am an undergrad student, a FOSS enthusiast, who wants to start his contribution in open source world with Unity [15:18] kamstrup, I'll love you more, yes :) [15:19] hey theone [15:19] theone, awesome! check out http://goo.gl/tiheb [15:20] and http://unity.ubuntu.com/getinvolved/#coding [15:21] Thnx for the links, I have gone through them and they are very helpful. for a start, I ve downloaded a daily build of ubuntu 11.04 and tried it on Virtualbox but it starts with message --> "3d acceleration not supported by hardware. Switch to classic mode". Any ideas why? 3d support is enabled in VBox [15:22] virtualbox doesn't support as much 3d as unity needs [15:22] we require something called frame buffer objects, which virtualbox doesn't support [15:22] you need to run unity on real hardware [15:23] lamalex: Not even VmWare will work/. [15:23] jcastro, that's probably something we should put somewhere [15:23] theone, afaik only real hardware supports fbos [15:23] the steps say to use a usb key [15:24] I will add a note though [15:24] lamalex, ok then I will go for the USB method [15:24] lamalex: poing people to this page first: https://wiki.ubuntu.com/Unity/Bitesize [15:25] theone: ^^^ [15:25] so much fragmentation [15:25] theone: when you're ready ping me and we'll get you started! [15:25] lamalex: well, this page is for the smaller bugs, the getinvolved page is for all of unity. [15:26] jcastro, sure. I will have to install it on my laptop first [15:28] one more thing, in the next year I will be applying for GSoC too. So. will Unity be there in participating projects? [15:28] it will make my schedule easier [15:28] hmm, that's a good idea [15:28] gsoc rules [15:29] money to hack all summer [15:29] <3 [15:29] njpatel: ok, u-p-a trunk also updated to new dee, but it's untested so far [15:29] it's like .. having a job [15:29] kamstrup, niice [15:29] kamstrup, will try and compile it later today [15:30] * apw wonders if the unity folk hang out here [15:31] * apw has just updated and right click in the unity bar is making the whole of the unity bar go dissappear [15:32] seems to be a crash [15:32] do you get a .crash from apport? [15:33] seb128, (if you are talking to me) i don't think its died, just invisible ... when i hit left again on it it comes back [15:33] also the menus arn't working any more for me [15:33] what unity and libnux version do you use? [15:34] seb128, wahtever is in the archive as of the top of the hour [15:34] natty proper? no ppa? [15:35] seb128, as far as i know [15:35] seb128, let me check there is nothing spammer on here and get back to you [15:35] dpkg -l | grep libnux [15:35] dpkg -l | grep unity [15:36] ii libnux-0.9-0 0.9.10-0ubuntu2 Visual rendering toolkit for real-time applications - shared lib [15:36] ii unity 3.2.6-0ubuntu2 Interface for Ubuntu Desktop Edition [15:36] [15:36] seb128, both look to be stock, and i have no PPAs on this box [15:37] ok, weird [15:37] those didn't change since last week [15:37] so weird that things break today [15:37] could be an xorg or driver issue... [15:38] seb128, the menu stuff has changed since last time i updated cause they are now more pretty, more sculpted to look at [15:39] so you didn't update for some days before that? [15:39] seb128, looks to be menu related to me, while the bobble menu for 'pin' is meant to be visible the whole bar and menu itself are not [15:39] right click again and the bar is back [15:40] njpatel, ^ any clue? [15:40] seb128, and the menu is 'there' as i can click 'new window' on it [15:40] seb128, and clicking on the top menu items for the application gives me the change in the top bar indicating there is a menu open [15:40] but no menu at all, though again i can select from it [15:41] needless to say, its pretty useless [15:41] Any of you speak spanish? [15:41] we didn't get bugs similar to this one I think [15:41] weird [15:42] apw: I haven't upgraded to the new menu yet, let me see if I can reproduce [15:42] apw, can you take a screenshot of the issue? [15:43] what new menus? [15:43] njpatel, can i take a screenshot in unity mode? [15:43] jcastro, ^? [15:43] apw, gnome-screenshot -i? [15:43] seb128: I just saw an upload scroll by for the indicator appmenu [15:43] oh my god new tron action figures look so cool [15:43] jcastro, seems apw is rather talking about the launcher [15:44] kamstrup, [15:44] /usr/bin/ld: find-events.o: undefined reference to symbol 'g_type_init' [15:44] /usr/bin/ld: note: 'g_type_init' is defined in DSO /usr/lib/libgobject-2.0.so.0 so try adding it to the linker command line [15:44] /usr/lib/libgobject-2.0.so.0: could not read symbols: Invalid operation [15:44] apw, if you want to take a screenshot with a delay, to give you time to open the menu etc, then open a terminal and type in gnome-screenshot --delay 5 (where 5 = seconds) [15:44] seb128: sorry, I should have said "hey I will upgrade today too and try to reproduce" sorry [15:44] kamstrup, libzeigeist fails to build [15:45] kamstrup, that's the same issue that was in dee, just need to add the $(DEBS_LIBS) to the LDADD lines [15:45] seb128: ok, it's probably the same issue as your patch for dee [15:45] right, will [15:45] do [15:45] would be nice if you guys could make sure your code works on current ubuntu ;-) [15:45] njpatel, ok managed to get some screenshots yes [15:46] seb128, tell me unity gets tested with up to date natty before upload [15:46] apw, it does [15:46] apw, could you upload them to imgur.com or something? [15:46] seb128, in a clean install? [15:46] given it never seems to work on anything i install it on [15:46] apw, I don't think so, on a normal natty box [15:46] didrocks's or mine usually [15:47] njpatel, ok i have screen shots [15:47] though I think didrocks test on several machines [15:47] though taking a screen shot in one case makes the image change slightly [15:48] njpatel, http://people.canonical.com/~apw/misc/unity/ [15:48] seb128, ^^ ... first one is with the right click 'pin' menu open, the second is with the 'file' menu open [15:50] seb128, and my logout button is gone ... which app is broken for that [15:50] seb128, oh an window stacking is broken, top bar menus are behing my terminals [15:54] apw, looks like clipping/damage issues [15:55] apw, is this with latest natty or latest natty + daily PPA? (sorry if seb has already asked) [15:55] njpatel, thats whats in the archive [15:56] apw, video card? [15:56] njpatel, basic atom machine [15:56] so i915 [15:56] kamstrup, njpatel: libzeitgeist breaks abi without soname change [15:57] zeitgeist_log_install_monitor_finish has been dropped but was part of the public api [15:57] njpatel, ok i logged out and in a couple of times and got the same behaviour, this last time i see nothing unusual [15:57] zeitgeist-1.0.gi: [15:57] kamstrup, ^ that seems buggy as well since the symbol has been dropped [15:58] apw, hmm, that's not good [15:58] njpatel, ok logged out and in and its back [15:58] apw, can you file a bug with those screenshots please + some info about your system [15:59] apw, and this started with your last update, or it's always an issue with Unity? [15:59] njpatel, ok the trick to 'fixing' it seems to be to right click on the desktop [15:59] that materialises the normal background menu and things are resolved after that one action [15:59] both menus are restored to normal behaviour [15:59] njpatel, this is all new behaviour as of top of the hour [16:00] apw, what did you update before the issue started? [16:00] seb128, yes, this is a direct result of updating to what was in the archive at 15:00 [16:00] can you copy your dpkg.log online? [16:00] or at least what got upgraded in that run? [16:01] apw, and before that, when was your last update? We havent' update unity since last friday [16:01] i had what was in the archive before the doko mass python update and had held off until now [16:01] when was said python update? [16:02] * lamalex is still trying to figure out how to get his panel back [16:02] lamalex, what happened? [16:02] it's empty [16:02] lamalex, empty or gone? [16:02] killall unity-panel-service; /usr/lib/unity/unity-panel-service & [16:02] lamalex, ^ [16:02] does it come back with windows-E [16:03] seb128, dpkg.log at the same place as the images [16:03] apw, this screams of a driver issue, btw [16:03] njpatel, I think it's my dbus [16:03] lamalex, is it not moving? :) [16:04] nope, there it is. re-running the panel service wasn't working before [16:04] njpatel, fun [16:04] apw, hmm, you did have a unity update too [16:04] apw, oh, but you had intel update as well [16:04] hmm [16:04] njpatel, yes indeed, the one which gives you the alt-LETTER to get your menus up [16:04] right [16:05] though its utterly useless as none of the shortcuts work on it, and paste is even worse as it no longer works [16:05] apw, you mean on the fallback menu? [16:05] njpatel, no the new 'alt-F' menus ... the application menus [16:06] njpatel, they move focus out of the window, so an alt-E, P, return (cause alt-E p no longer just does it) does nothing [16:06] apw, oh, man, that's a *huge* bug [16:07] so much for the test plan [16:07] I'm not sure how well that worked before, tbh [16:07] it seems like an appmenu bug [16:16] njpatel, the fact that right clicking the background and closing the menu there seems to fix things ... does that not indicate away from a driver issue ? [16:17] apw, no, because by doing that your causing a screen paint, so if there was an issue with the paint/fbo etc, your basically flushing it [16:17] apw, but it still could be a damage issue, as compiz was also updated [16:17] apw, the only thing that you could do is downgrade compiz and see if it still happens [16:17] if that fails, downgrage unity too [16:18] see if we can figure out where the problem is introduced [16:18] hopefully it is in unity as it'll be easier to fix than driver [16:18] njpatel: so how do we generate the Config namespace with cmake? [16:19] it's proving pretty tricky to get all the old tests running here on my end [16:19] kamstrup, oh, fun, you need to make a config.vapi.cmake or something and do what I do with config.h.cmake in trunk [16:20] kamstrup, the libunity tests? I had most of them working, but I did disable a few. Are you using that branch? [16:25] seb128, all good now thx [16:26] njpatel: yep, I'm merging that branch into my own - it's easier that way [16:26] yep [16:26]