=== ion_ is now known as ion === ubuntu is now known as spikeb [06:59] morning [07:01] that. [10:03] hi, one question about using compiz settings manager on natty [10:03] Im trying to install it [10:03] but I get this error [10:03] The following packages have unmet dependencies: [10:03] python-compizconfig : Depends: python (< 2.7) but 2.7.1-0ubuntu3 is to be installed [10:04] that means that I also need to compile and install compiz and his related packages? [10:04] (as explained here https://wiki.ubuntu.com/Unity/InstallationGuideFromSource) [12:07] Ok im working on the unity --replace bug ill be done in a few minutes if someone could tell me how to enable the plugin [12:08] I wont be able to test the patch because this computer would take a month to compile the code [12:10] fagan, hey, you didn't get your new one yet? [12:10] akshatj: waiting on it to be delivered to dublin to go get it [12:10] It should be today or tomorrow [12:11] Its a pain but it cant be helped [13:58] klattimer: awesome, when did you start doing indicator code reviews? [13:58] dbarth just the other day [13:58] fagan: which bug is that you're working on? [13:58] klattimer: that's pretty awesome [13:59] jcastro: the one for --replace [14:00] * jcastro looks for it [14:00] it sounds bitesizeable [14:01] yeah it is [14:01] can you tag it so please? [14:02] it already is I remember [14:02] ah oh a "visual" bug http://dl.dropbox.com/u/132551/screenshots/Screenshot-CompizConfig%20Settings%20Manager.png [14:02] hmm, I don't see it on the list [14:02] it was there before :-/ [14:04] there is this https://bugs.launchpad.net/ubuntu/+source/unity/+bug/685520 [14:04] Ubuntu bug 685520 in unity (Ubuntu) "When attempting compiz --replace in terminal" [Undecided,Incomplete] [14:04] but i remember there was a bug already open [14:04] oh thats compiz --replace [14:05] I believe it should be just "compiz" which performs the --replace etc now ...yes? [14:06] I just do unity --replace when it's all messed up [14:06] well unity --replace shouldnt work [14:06] I just do "unity" now I dont t hink the --replace i necessary [14:06] sorry, I mean unity --reset [14:07] oh unity works by itself the bug about --replace must have been marked as fixed [14:07] it would be nice if it worked for --replace too [14:07] file a wishlist [14:08] just for equality with gnome-shell and gnome-panel [14:10] jcastro: well I could do that if someone could point me to where in the code is the main method [14:10] * fagan doesnt know his way around unity [14:10] fagan: file the bug and bitesize it, and in there ask that [14:10] and I'll have someone answer asap [14:11] kk [14:11] getting it on the bitesize list puts it on my radar basically [14:11] fagan: if you run into bitesizeable bugs ping me or didrocks, and we can tag them [14:11] I want a nice healthy list for people to be able to pick and choose [14:12] well ill tag them as bitesize myself and give it a go to fix it if im not busy [14:12] its just im not too familiar with the layout of unity's code its a bit hard to get around [14:15] reported Bug #692569 [14:15] Launchpad bug 692569 in unity (Ubuntu) "unity --replace would be nice" [Undecided,New] https://launchpad.net/bugs/692569 [14:17] jcastro: I marked it as assigned to me and set it as a bitesize bug [14:17] rock and roll! [14:19] Its kinda funny how apport collected all of that info and uploaded it to launchpad and its not at all needed [14:20] I wish you could go ubuntu-bug --no-info or something [14:20] fagan: Then we wouldn't get info when we need it :) [14:21] Amaranth: I mean for things like a miss spelled word or like the bug I just did you dont need any of that info at all [14:21] and it takes up un needed bandwidth too [14:22] fagan: Sure but it would be abused by people who don't want to wait for all of that to upload [14:22] Amaranth: true but we could just not document it anywhere and have it as a nice querk for developers in the know [14:24] * fagan has insane ideas sometimes :) [14:32] sometimes? [14:32] DBO: howdy [14:32] jcastro: yes "sometimes", a bit of insanity is always needed to get things done [14:32] heh [14:43] is there any way to test out the dash yet [15:19] fagan: nope [15:23] jcastro: :( === JanC_ is now known as JanC [15:56] DBO, is merging window decorations into the panel fully enabled yet? [16:31] lamalex, no really [16:31] not [16:31] yeah ok [16:31] I didn't think s [16:31] we're getting a lot of bug wrt the decorations merging into the panel but I didn't really think they were fully implemented [16:31] jcastro, re: long time ago. Hi :) [16:32] hello [16:32] DBO: I told you you'd be explaining it over and over now [16:32] not really [16:32] I sent an email to the ayatana-dev mailing list, but I can't actually see the mail on the list... [16:32] you are probably doing it more than me [16:32] evaluate: did you get a moderation message? [16:33] also, are you subscribed? [16:33] jcastro, didn't get any message at all... [16:33] jcastro, yes, I am subscribed, but at the time I sent the mail, the email address I sent it from wasn't in my account on launchpad [16:33] ah [16:33] that's probably why [16:34] I've been told by the people on #launchpad that it shouldn't be a problem though, and that it should only be accepted by someone on the list... [16:34] hrm, weird [16:35] jcastro, do you have access to the moderation queue? [16:36] if it isn't in there, I could send it again... [16:36] DBO: I've been kind of pawning it off as "jason will come along shortly and explain" and other sorts of promises you can't cash [16:36] evaluate: yeah I guess try it again [16:36] jcastro, ok [16:36] jcastro, been doing the same, but with your name :) [16:38] ok, it worked now :-) [16:50] DBO: lamalex: since dbarth and njpatel aren't here if you have anything you can tag bitesize by EOD today that would be swell. [16:50] DBO, jcastro: I've been telling people njpatel broke something and ran away :P [16:50] jcastro, sure [16:50] Amaranth, thats perfect, because I think hes blaming you [16:50] jcastro, you got it bud [16:50] DBO: The circle of life [16:50] lamalex, when is your last day? [16:51] was I fired? [16:51] DBO: Let's say it's due to excessive reparenting and blame smspillaz too [16:51] lamalex, of the year [16:51] lamalex, before vacation [16:51] erm, I need to do that right now :\ [16:52] what? [16:52] my ca was broken last time I tried [16:52] said I had 364 days of vacation.. [16:52] Sounds like it's time for a year hiking around the world [16:53] :) [16:56] lamalex, hi, do you know if you will have a time slot today to read the mail I sent you? [16:56] is just to coordinate myself and know where to start [16:56] API, oh hey, yes I read it earlier and it got forgotten about. I'm doing some bug triaging but I will get to it asap [16:57] lamalex, ok, thanks just to know [16:57] yup [16:57] if just to know if I can start to review one part or the other [16:57] there are a lot to see here ;) [16:57] basically short part is "there is nothing" [16:57] gtk_init was loading the module [16:59] hmm, well, as I told you on our little chats on #a11y [16:59] as with GNOME Shell [16:59] I bet that it would be required to load it by hand [16:59] and in fact [16:59] try to avoid the loading from gtk_init [17:00] but anyway, I will wait for the "long part" meanwhile [17:11] anyone up for a couple of reviews? :) [17:13] of what? [17:14] a bug fix it seems [17:14] https://bugs.launchpad.net/bugs/691812 [17:14] Ubuntu bug 691812 in unity (Ubuntu) "Window border doesn't get restored" [Undecided,Confirmed] [17:14] tareth: right? [17:14] hi DBO [17:14] that's one of them! [17:14] hi dbarth [17:14] I also have https://code.launchpad.net/~mrasmus/unity/unity.fix-691765/+merge/44146 [17:14] tareth: are you already in contributor-canonical-agreemnt? [17:15] well now [17:15] that first bug is just an awesome [17:15] I sent the agreement to the agreement address that was listed, and I was also told to e-mail you with it (which I did) [17:16] so I'm not sure if anything went wrong there :/ [17:16] tareth: ah ok [17:16] dbarth, is it possible to get me on the CC list since I seem to be doing a lot of these reviews? [17:16] tareth: well, there's a manual process at some point, but we can already review the merge prop. meanwhile [17:18] DBO: the best is to ask people to cc you when they send the agreement [17:18] i do [17:18] DBO: i've checked tareth's submission and it's fine [17:19] though usually they are resending it to me is all [17:19] okay thanks [17:19] I am going to review and if all is good, merge [17:19] tareth: thanks Matthew, i received your email [17:19] right [17:19] whew. I was worried for a moment :) [17:19] tareth, rocking :-) [17:22] How does touch interaction work with the new appmenu? [17:24] touch? [17:24] we dont do TOUCH [17:25] DBO: Don't touch me [17:25] Amaranth, just wait till dallas [17:26] I think I'm in a different hotel [17:26] wtf [17:26] Amaranth: hey, got any bitesize compiz bugs? [17:26] I'd like to add a few to the unity report. [17:27] * Amaranth hides [17:27] Amaranth: just tag em up! [17:27] I can look, I don't think any compiz bug can be considered bitesize [17:27] yeah [17:27] I got a list of kind of beginner tasks from Sam [17:28] but none are bugs per se [17:31] tareth, congrats on two new approved branches [17:32] Thanks :D [17:35] DBO: are there news about the compiz issue from last week that was creating problems at session startup? [17:36] dbarth, no [17:36] no sam, no news [17:36] I thought we got a workaround for that [17:36] It's a glib issue but the workaround causes issues with menu stacking so the workaround was disabled then for some reason enabled again [17:36] the workaround seems to cause issues [17:37] i dont understand how its a glib issue [17:37] *shrug* [17:37] That's what I heard [17:37] I need to understand the problem better [17:37] tareth, branches are merged, you are once again hero of monday [17:38] - workaround in delaying the plugin init as the gconf backend hangs when [17:38] trying to contact/launch the gconf daemon. This only happens when you [17:38] don't change your current profile, at session start. [17:38] Whatever that means [17:38] dbarth, I might need to check out for a bit on account of skull crushing headache, I will make up the hours if I do so. just a heads up [17:38] Amaranth, I [17:38] will look at it [17:38] see where the hang is [17:39] btw, if you thought compiz plugins were complicatingly interconnected now wait until I add egl support [17:39] what, why? [17:40] the opengl plugin is going to have a function similar to registerBindPixmap that returns an instance of a child of GLScreenInterface depending on what backend you have and GLScreen will proxy calls to it [17:40] so you are making a subclass setup [17:41] big deal? [17:41] it could be the migraine talking here [17:41] DBO: It's going to be a little trickier due to not being able to link to libGL.so and libGLESv2.so at the same time [17:41] but it seems to me that if we exclude those plugins who are making direct opengl calls [17:41] there should be 0 API change [17:41] Right, outside of the opengl, egl, and glx plugins you shouldn't notice [17:42] But I hope you never have to dig in one of those :) [17:42] fantastic, you are a scholar and a gentleman [17:42] btw do we have egl on the desktop? [17:42] Yep [17:42] at least on all mesa-based drivers [17:43] Although you have to do something extra to get it setup correctly for drivers not using gallium [17:43] I think nvidia has support too, not sure about fglrx [17:47] okay [17:57] DBO: ok, no worries, it's no use trying to bang your head on the gconf issue if you already have a headache === DanRabbit|afk is now known as DanRabbit [18:45] lamalex, wanna do be a review favor? [18:47] DBO, sure [18:47] DBO, link? [18:47] making it [18:48] I really need a music collection titled "So you grew up in the 90's" [18:48] then I can ignore all new music [18:48] https://code.launchpad.net/~canonical-dx-team/unity/unity.spec-conformance/+merge/44259 [18:56] lamalex, ^^ [18:56] DBO, reading [18:59] DBO, are these things speced to be customizable? [18:59] nope [19:00] but they are specced multiple ways [19:01] boost? [19:01] I thought boost was disallowed [19:02] lamalex, I didn't use boost [19:02] DBO, I don't understand why we're implementing these as options [19:03] this seems like stuff that should all be in animations spec, and if it's spec'd differently then there's a bug in the spec [19:11] lamalex, because they want to do user testing with it [19:11] lamalex, and because its simply for me to maintain the option rather than switch out the code [19:11] lamalex, also neil gave me the +1 to add options we can rip out later [19:20] DBO, ok [19:20] just checking [19:20] my job as QA :) [19:25] DBO, did something change in unity/bamf recently? A lot of my launchers aren't matching and are creating new icons when I click them [19:25] lamalex, the boost usage you see is from compiz [19:25] lamalex, nothing in bamf has changed recently... [19:26] what isn't matching? [19:26] chrome, banshee, thunderbird, probably others [19:27] Ok, I think I'll probably need to break opengl plugin API/ABI [19:27] the fudge... [19:27] why API? [19:27] Nothing in the opengl plugin will be able to reference any opengl functions or types [19:28] I dont htink it di [19:28] It does [19:28] wow, just assume that was typed right [19:28] what does it? [19:28] types at least in the public API [19:29] GLuint, GLenum, etc [19:29] plus I'm going to have to change how the Private* stuff works [19:29] okay so very easy fixes [19:29] Well, maybe [19:30] Some of the API will likely have to go away [19:30] Or at least be done differently [19:30] The current API assumes we can set global state on things still [19:31] Amaranth, okay, well I would audit the code and see how much they are really needed [19:32] DBO: Right now I'm just concerned with making an API that both backends will be able to actually implement [19:32] DBO, final thing, we're sure we want to merge this into trunk and not keep it as a seperate branch and build packages for design team to use in user testing? [19:32] After that the things that don't fit will get that kind of checking [19:32] Or I could ignore sam and just fork the opengl plugin into an opengles plugin [19:32] because you know OMG ubuntu will write a story OMG UNITY OPTIONS HERE'S HOW ASK FOR MOAR [19:33] I could mark all the options [debug] [19:33] so in CCSM they say debug next to them [19:33] I wish I could talk to sam... [19:33] dont we all [19:33] to be honest [19:33] I dont care what OMG Ubuntu does [19:34] DBO, just seems like a good chance to leverage our infrastructure with ppas and whatnot, could honestly be a good workflow in general. Anyway +1 [19:35] maybe I'll bring it up on the list [19:35] I think I'm just going to fork the plugin for now and figure out a proper API with sam in january, I need to get something done [19:36] lamalex, I am double checking with neil [19:36] It could honestly be a pretty good workflow to have a -proposed ppa or something [19:36] where design team and interested users can test maybe features [19:36] before we commit them to trunk [19:36] Then omgubuntu posts that and everyone switches to it :) [19:37] and we say tough shit, we told you not all of these features would make it into release [19:39] There are advantages and disadvantages to the idea though [19:39] fewer eyes on new features landing for bug reports [19:39] but also fewer bug reports for issues we know about [19:40] lamalex, I am going to likely migrate the options to gsettings [19:41] away from compotions? [19:41] actually nope [19:42] sticking it under an "experimental" heading [19:46] in compiz [19:47] I think it would actually make sense to have the unityshell plugin directly use gsettings to store its setting with the rest of the unity stuff [19:51] DBO, I +1'd that merge otherwise, the code looks fine [19:51] lamalex, thank you for the review [19:51] you're welcome my friend [19:51] I have more work to do, and thank you for asking the right questions :) [19:52] tell dbarth how great I am next time you chat ;) [19:52] so i wonder if anyone can help out an old guy here... when I log in to (natty) with either unity- or classic sessions I just get stuck in a blank screen with a mouse pointer... [19:54] kamstrup: is the cursor spinning? [19:54] Amaranth: nope [19:55] normal idle [19:55] and 'ps aux' indicates that all processes for the session is running [19:55] except the window manager [19:56] and starting either compiz or metacity doesn't cut it [19:56] well, it sounds like you didn't get a panel in the classic environment, which is bug #1 [19:56] Launchpad bug 1 in Ubuntu "Microsoft has a majority market share" [Critical,In progress] https://launchpad.net/bugs/1 [19:56] and you don't have the unityshell plugin enabled in compiz, which is bug #2 [19:56] Error: Could not parse data returned by Launchpad: 2 (https://launchpad.net/bugs/2) [19:56] eep, silly ubot5 [19:57] so starting a window manager isn't going to get you anything visibly different as you have no shell [19:57] and no open windows [19:57] nautilus should still paint the desktop [19:57] Yeah, dunno about that [20:01] Actually, come to think of it; I did have unity working fine for a few brief minutes today. Then I pulled in the latest updates... [20:01] it was after that [20:04] ah, got it. compiz dies silently when launched [20:07] nice [20:07] lamalex, can you do me a favor (as our QA guy) [20:07] see if we can find any C++ static analyzer [20:07] I cant find one [20:08] DBO, http://sourceforge.net/projects/cppcheck/ [20:08] come on dude all you have to do is google [20:08] a good one [20:08] does that one suck? [20:08] yes [20:08] tried it [20:08] Ah great. Apparently the unity session thinks it's completely fine that compiz exits with code 1 during startup [20:08] kamstrup, gdb + break on _exit [20:08] there's a hudson plugin for cppcheck [20:08] so we could automate [20:09] in that case I'll find a way to make it do what I want [20:09] I want to audit render code for complex pathways (5+ possible branches per method) [20:09] DBO, it's also packaged and the kernel uses it [20:09] it still sucks [20:10] I wanted something like resharper for visual studio [20:10] it's the only one listed in wikipedias list of static anyalzers for cpp [20:10] DBO: no need... vv [20:10] Program exited with code 01. [20:10] (gdb) No value set for `/apps/compiz-1/general/allscreens/options/active_plugins' [20:10] DBO, there are a bunch of commercial ones [20:10] if we can get funding we can probably buy something nicer [20:12] but I guess that's a dbarth issue? [20:13] + is it ok to use non-free tools [20:13] DBO, did you ask in #cppcheck for doing complex pathwyas? [20:14] yes will do [20:14] DBO: so it looks like I have no dconf schema for compiz... who's responsible for installing that? [20:17] i would guess compiz [20:18] compiz uses gconf [20:20] kamstrup: and the message there is because it's /apps/compiz-1/general/screen0/options/active_plugins [20:20] screen0, not allscreens [20:20] So the 0.8->0.9 conversion that either didrocks and smspillaz wrote is doing something funny, I guess [20:55] DBO: breaking on _exit only gives me a stack trace inside libc. I'm trying to locate which piece of code triggers this, but my greps to up with bupkis... Any tips on how to nail this? [20:55] can I see the stack trace? [20:56] DBO: http://paste.ubuntu.com/546097/ [20:57] wtf [20:57] DBO: just my words [20:57] how is it start... [20:58] compiz --replace ccp [20:59] I meant to say [20:59] how is it in start? [21:00] The only reason compiz used to exit is if you didn't have composite or opengl support but now that just affects plugin loading [21:00] I can't think of a reason compiz would exit without giving a message first anyway [21:01] oh, is there more than one thread? [21:02] I think there are loads of threads [21:02] well, no [21:02] There is going to be 1 or 2 [21:03] I see 3 possible locations of EXIT_FAILURE [21:03] I think there is just 1 then [21:03] two are in the bailer plugin [21:03] The other is in my "move checks to compiz" patch [21:04] cando: bug #683466 is fix released right? I can't reproduce it [21:04] Launchpad bug 683466 in unity (Ubuntu) "There are 2 Quit menu options in Quicklist" [Low,Triaged] https://launchpad.net/bugs/683466 [21:04] In either case it should have printed something but would still only exit there if you didn't have proper support for compiz [21:04] jcastro, yup...:) [21:05] DBO has already pushed it to master... [21:05] The only time my patch would cause exit 1 is printf ("Detected GNOME failsafe session.\n"); [21:05] cando: it has to be in the last upload though, the one on friday? [21:07] jcastro, dbo pushed it 14th december.. [21:07] and if you were stuck in my move checks to compiz stuff it would have at least launched an xterm [21:07] yeah he likely forgot to update the status, I fixed it [21:07] it should never have actually exited [21:08] ok...thanks..:) [21:08] kamstrup: ^ [21:08] Amaranth: ok... my hunt goes on then :-) [21:09] kamstrup: Perhaps glib is pulling the rug out from under us for some reason [21:09] although afaik the only time it does that is when g_malloc fails [21:11] According to the code what you're seeing shouldn't be possible :/ [21:17] Amaranth, DBO: ! it was the "bench" plugin... disabling that and now it runs... [21:17] kamstrup, [21:17] wha? [21:17] :) [21:17] Amaranth: wha?! indeed :-) [21:17] kamstrup: The only thing I can think of is the bailer deciding your system isn't fast enough due to results from the bench plugin [21:18] But I can't see why it wouldn't have printed something [21:18] Amaranth: I tried with and without the bailer, and it ran with both [21:18] Amaranth: I started manually bisecting this with 'compiz --replace plugin1 plugin2 ...' [21:18] Amaranth: it worked until I hit "bench" [21:19] so I just noticed bench doesn't seem to be in compiz or compiz-plugins-main [21:20] so it was dying because that plugin doesn't exist [21:20] kamstrup: does `compiz --replace fancyfakething` exit too? [21:22] Amaranth: that works ok... somehow the fact that the bench plugins was ticked in ccsm caused this [21:22] lemme just try a full session [21:23] Amaranth: are you in natty? [21:23] jcastro: yep [21:24] Amaranth: can you snag me a screenshot of the launcher drag feature? [21:24] or do you know how? [21:25] jcastro, what do you want? [21:25] just a ss of a launcher being drug? [21:25] dragged [21:25]