[01:37] jono_, pong [07:45] morning all [07:50] Why do I see random changes in po/unity.pot on *every* merge request? [08:01] hey everbody! [08:03] hey Mr MacSlow [08:03] hey kamstrup [08:12] good morning [08:12] ’mornin’! [08:14] definitly morning [08:15] 3:15am here [09:03] gord: morning dude - I posted some comments on your places tiles you might wanna have a look at === _LibertyZero is now known as LibertyZero [09:33] smspillaz: hey, what file the spewer plugin does generate on SIGSEV ? [09:37] didrocks: /tmp/compiz_internal_state${PID} [09:37] didrocks: though really, I'm not sure of the point of that, considering that it spews to stderr as well and that goes straight to .xsession-errors [09:37] smspillaz: hum, I don't have such a file [09:37] didrocks: weird [09:37] didrocks: maybe it crashed while spewing ;-) [09:38] smspillaz: hum, let me try again… [09:38] bbl [09:38] didrocks: btw, I noticed that there sees to be a truncation of the beginning of the file, so I'll need to look into that [09:39] smspillaz: ok, let's see if I have something generated at least first [09:41] smspillaz: doesn't work [09:41] seb128: do you have a minute to confirm? [09:42] confirm what? [09:43] seb128: can you enable the Debug Spewer plugin in ccsm, and send a sigsev to compiz. You should then have a file in /tmp/compiz* [09:43] on current natty? [09:43] yes, please [09:43] with last week compiz, you should get the spewer plugin [09:46] no [09:47] that doesn't do anything, got a crash when ticking the option in ccsm then I sent it a -11 [09:48] seb128: you mean, you sent the -11 after restarting compiz, isn't it? [09:48] (after the crash ticking it in ccsm) [09:48] yes [09:48] seb128: thanks for the test :) [09:48] got it to crash when opening ccsm now [09:48] still nothing dumped [09:48] you're welcome [09:48] smspillaz: was it really working for you? how hard did you test it? ^^ [09:49] I can't add it to the unity hook then and can't enable it by default :/ === MacSlow is now known as MacSlow|lunch [11:03] smspillaz: ^^ [11:08] didrocks: it works fine for me [11:08] didrocks: maybe the packaged version is broken [11:08] didrocks: I'll look into it when I get some time [11:08] (currently working on the menu bug) [11:09] didrocks: actually, I noticed something else today - are you getting an error which says "couldn't load because opengl or composite isn't loaded?" [11:10] for debugspewer [11:10] smspillaz: how the packaged version can be broken? I just took your distro patch [11:10] *shrug* maybe my patch was broken [11:10] it's possible [11:10] smspillaz: where? when there is a crash or when it's started? [11:10] didrocks: when compiz starts [11:10] smspillaz: let me look [11:10] smspillaz: no, nothing [11:11] didrocks: weird [11:11] didrocks: is apport catching the compiz crashes ? [11:11] smspillaz: yeah, there is the generated file in /var/crash [11:12] weird [11:13] didrocks: I'll look into it when I get a chance [11:13] didrocks: does the apport hook give us .xsession errors ? [11:13] smspillaz: yes, it does, as well as a lot of Xorg info, gdm and such… [11:14] didrocks: ah ok, in that case we can probably nuke the file writing stuff in debugspewer then [11:14] since it all ends up in .xsession_errors [11:14] (fprintf (stderr, "blahblah")) [11:15] smspillaz: no, having this file in addition is better, don't spam .xsession-errors please [11:15] smspillaz: also, as the session crash, I think a lot of people will log again [11:15] didrocks: ahh right! [11:15] so, .xsession-errors will be erased [11:15] didrocks: well we have .xsession-errors.old [11:15] smspillaz: and if they start more than twice? (apport can report the bug later…) [11:16] smspillaz: I match the pid in the hook for being able to report the right spewer file [11:16] well… once it will be there :) [11:16] didrocks: but I buy your other point more - the debugspewer output is like 40,000 lines long, wouldn't want unity stuff lost in there [11:16] exactly! [11:16] didrocks: :) [11:16] didrocks: I'll look into it after I fix the stacking/menu bug today [11:16] didrocks: it's not really a top priority [11:17] didrocks: speaking of this stacking bug - do you know of any circumstances that make it easy to reproduce? [11:17] smspillaz: no, it's not, it's just annoying that it didn't work after the work effort you put in it [11:17] smspillaz: well, nothing changed since the past month… it's still "random", most of the time, present at startup [11:17] didrocks: the way I'm doing it now is dragging an icon rapidly from one side of the screen to the other until the icon is invisible, but it only happens rarely [11:18] didrocks: but it's annoying since I get used to dropping the icon and I forget to not drop it when the icon is in fact invisible (wrongly stacked ;-)) [11:18] :) [11:19] didrocks: as for the debug stuff, I suppose that sucks that it doesn't work right now. The good thing is that kvalo and others have confirmed it to work with the keybinding mode anyways, so maybe we can make an apport hook to watch /tmp for a dump and do apport-bug automatically [11:20] since most of the bugs aren't really crashers but rather when something weird happens WM wise and the user presses that button so that we can get some info of what's going on [11:21] didrocks: I just rememberd - I may have commented out the signal handlers in the patch I gave you [11:22] njpatel: do you know if unity-panel-service is creating and destroying the menu windows every time the user hovers them or is it just creating them once and then mapping/unmapping them ? [11:23] since it really should do the latter [11:23] smspillaz: I think both are needed [11:23] didrocks: yeah, agreed - although I suppose we can hook up the keybinding mode for now if I can't get enough time to figure out what's going on till the next upload in (in 3 h) [11:23] smspillaz: ok, if you can just test that later, ensure it works with apport, but again, it's not a priority, just nice to have that once we will turn that by default [11:23] didrocks: indeed [11:23] smspillaz, i'm pretty sure it's the latter (we get the mennus from the indicator, the service isn't creating them itself) [11:24] smspillaz: well, it's more complicated to trigger apport on file watch, it's more on crash [11:24] njpatel: hrm - I wonder why it's so laggy then ? [11:24] njpatel: oh I know why [11:25] njpatel: it's because we're doing the actual mapping setup and such in the same thread as the compositor which needs to show them [11:25] smspillaz, no, service is it's own thread [11:25] njpatel: hrm ? [11:25] njpatel: so the service is the one doing the mapping and unmapping then ? [11:25] smspillaz, it could be laggy becuase we need to request the showing of the menus of D-Bus, but still, it's not laggy here [11:26] njpatel: it's *extremely* laggy on this i945 [11:26] njpatel: it's like compiz is getting the getting the window and it hasn't even finshed rendering yet [11:27] smspillaz, hmm, can you file a bug please and assign it to me [11:27] njpatel: sure [11:27] smspillaz, yes, even here I get a white flash, it's somethign to do with the timing [11:27] njpatel: like I said earlier, it certainly feels like there's some unecessary overhead going on - since a simple XMapWindow and XUnmapWindow is not that expensive [11:27] unless you do XSync or something right afterwards [11:28] didrocks: btw, on what signal are you doing the auto-maximization of windows ? [11:28] smspillaz, probably ,there is some horrid code in the service that I need to fix (it needs to track pointer motion right now). I was planning to take a look after A2, so let's revisit it then [11:28] njpatel: urgh, yeah that's probably it :) [11:31] didrocks: it might be worht adding a signal to UnityWindow::UnityWindow since that happens on MapRequest so the window isn't visible at that point [11:31] that way you don't see the window get mapped unmaximized and then maximize [11:32] didrocks: even better would be to wrap CompWindow::place, but that won't work because unity loads after place :/ [11:36] smspillaz: it's not done on Reparent [11:37] didrocks: what is it done on ? [11:37] CompWindowNotifyReparent ? [11:39] didrocks: oh actually [11:39] didrocks: you should be able to wrap CompWindow::place [11:39] didrocks: that way you can maximize the window there [11:39] and it should appear maximized before it appears [11:40] hm, intersting, the place plugin sends a maximization request ClientMessage to itself for maximization [11:40] I wonder why that works [11:41] smspillaz: actually, I see it directly maximized here [11:41] didrocks: that's because you have a fast machine :) [11:41] smspillaz: TBH, I think there are others things to focus on now :) [11:41] that can wait as a bug fi [11:41] fix* [11:42] didrocks: yeah I know :) just things that I'm noticing while using a slow laptop [11:42] and I'm already really busy with other part of the stack [11:42] and you too :) [11:42] didrocks: yeah :) [11:42] didrocks: btw, we definitely need to tweak some of the wall plugin's settings for lower end laptops, doing things like painting textures twice and re-rasterizing them in different places is hugely expensive I've noticed [11:43] or at least have a fallback mode [11:45] smspillaz: ok, we'll discuss that once the most annoying bugs are gone :) [11:45] didrocks: indeed === MacSlow|lunch is now known as MacSlow === oubiwann is now known as oubiwann_ === oubiwann_ is now known as oubiwann [13:11] Kaleo, so unity-2d in natty installs fine and properly fires up all bits, but klicking on anything in the dash doesn nothing, nor does the searchbar give any results [13:13] ogra: OF COURSE [13:13] ogra: sorry for the caps [13:13] ogra: silly keyboard [13:13] missing places ? [13:13] ogra: the place daemons you removed the deps on [13:13] ogra: yep [13:13] ah, k [13:13] ogra: but it's coming back according to kamstrup [13:13] i was suspecting that [13:13] :) [13:13] :) [13:13] so sad.. [13:14] well, i'm happy it installs wi5hout any issues [13:14] Kaleo, ogra: when todays release hits the archives the place daemons will be let loose once again [13:15] ah, k [13:15] should the API still work the same ? [13:20] ogra: there are changes that I need to track [13:20] kamstrup: great [13:20] kamstrup: any ETA on that? [13:20] Kaleo: ask didrocks [13:21] didrocks: :) [13:21] depends on when kamstrup will make a place release :) [13:21] didrocks: i think i already did, lemme check [13:21] kamstrup: what? without telling me! [13:22] that's mean :p [13:23] hehe [13:23] didrocks, just so that we can chase you for not uploading :P [13:23] slacker !!! [13:23] lol [13:23] *g* [13:23] didrocks: yes, u-p-f 0.5.34 and u-p-a 0.2.28 should be good to fly [13:24] didrocks: I didn't shout it from the rooftops because it didn't make much sense to package them before after todays release [13:24] kamstrup: nice! [13:24] or with todays release what ever suits you [13:25] kamstrup: depends, I still can prepare the update, I just want to ensure it works first :) [13:25] sweet [13:26] didrocks: it'll "work" with unity because they don't hook up the places yet [13:26] lol [13:26] that's one definition of work [13:26] didrocks: unity-2d will require some updates to reflect the dee-0.5 api if Kaleo isn't on top of that already [13:26] kamstrup: I will take care of that [13:26] as I outlined in the email [13:26] kamstrup: yep [13:26] Kaleo: sweet [13:26] kamstrup: ok, so it's a wasted upload, you mean? :) [13:27] didrocks: it depends - if Kaleo pushed a new unity-2d then that should be able to rock the house [13:27] Kaleo: just tell me once you will have updated it, I can push them in a ppa, but I want to avoid broken upload (broken as "no way to test before") [13:28] ogra: btw, did you rename the session? [13:28] didrocks: as you wish but it cannot be more broken than before.. [13:28] didrocks, not yet [13:28] ogra: can you do it, so that we can have the fallback working as expected? [13:28] didrocks, first i need to clean upt the copyright files and i'm still waiting for feedback from NCommander on one issue [13:29] ogra: oh, it's not in natty yet? [13:29] once i'm through with that one I#ll take care of the session renaming [13:29] ogra: ok, no worry [13:29] it is in natty but all branches point to upicek still [13:29] urgh :/ [13:29] it installs and runs fine btw [13:29] i just tested+ [13:29] apart from places :P [13:30] didrocks, we didnt have public branches ready, they were prepared *while* we were rushing the packages in ... thats what you get from early leakage :P [13:31] team to transition :) [13:31] its all fine now, we only need places back, the session rename and then the merge of unity-2d-default-settings into unity-2d [13:31] then we should be golden === bregma is now known as bregma|afk [13:33] having it installable and running properly was the first target [13:33] right [13:37] ogra didrocks what do we want to rename the session to? [13:37] Kaleo, ubuntu-2d [13:37] Kaleo, i'll care for that in natty [13:37] no need to do it in maverick [13:37] Kaleo: 2d-ubuntu rather [13:37] ogra: what is it now? [13:37] ah [13:38] oh [13:38] that's odd [13:38] (since the whole session handling changed between the two) [13:38] Kaleo: I'm talking about the session in /usr/share/gnome-session/sessions/ [13:38] not the one in /usr/share/xsessions/ [13:38] didrocks: right [13:38] Kaleo, its just the two files [13:38] or even one [13:38] i didnt look closely yet [13:39] you have of course to update the file in /usr/share/xsessions/ to gnome-session --session=2d-ubuntu of course [13:39] and the description in the session/desktop file i suppose [13:39] indeed [13:39] yeah, look at the other to have some examples [13:39] i'll get to that next ... as soon as i have spoken to NCommander and fixed the branch locations in all packages [13:39] * didrocks is so happy to be able to defined starting component without a big hack in gconf :) [13:40] apart from the merge i plan to be done with everything this week [13:40] (i also have to file MIRs and get them through, would be good to have unity-2d in the alpha2 images) [13:41] * ogra eagerly wants to get rid of the efl launcher [13:41] (talking about armel only indeed) [13:46] didrocks, looking at bug 705413 ... shouldnt we have a unity-common package for that ? [13:46] Launchpad bug 705413 in unity-2d "[packaging] include applications.png and files.png from unity in maverick packages" [High,Confirmed] https://launchpad.net/bugs/705413 [13:46] (in natty indeed) [13:48] there is already an unity-common, they will have the applications.png and files.png (or maybe it will be in the unity-asset-pool package) once the places will be back and they will ship it [13:48] k [13:48] i just need something that doesnt pull in all of the 3D stuff [13:48] ogra: but what's in /usr/share/unity/3 should be in common, indeed [13:48] dont care how its called [13:49] ogra: well, unity-common is already there for that :) [13:49] Kaleo: do you use anything in /usr/share/unity/3 ? [13:49] while we ship both in the arm images i know that there will be people that build custom images only shipping 2d [13:50] didrocks: I don't know that folder [13:50] didrocks: let me have a look [13:51] Kaleo: in any case, I'll put that in -common, makes more sense [13:51] didrocks: please do [13:52] Kaleo: I think you will use the dbus unity-panel-service as well on day? [13:52] didrocks: we don't use files from /usr/share/unity/3 *yet* [13:52] didrocks: if you put them in common then we can depend on common [13:53] didrocks: I don't knwo what that service is nor does [13:53] Kaleo: this is the service acting as a proxy between the indicators and you [13:53] didrocks: is it new? [13:54] agateau: do you know that ^ [13:54] Kaleo: for natty, yeah [13:54] ok [13:54] * agateau reads backlog [13:54] any doc anywhere? === oubiwann is now known as oubiwann_ [13:55] well, ask the unity-team :) [13:56] didrocks: how does it expose the indicators? [13:56] ogra: I fucked up https://bugs.launchpad.net/ubuntu/+source/unity-2d/+bug/705182 [13:56] didrocks: using libdee? [13:57] Kaleo, hmm ? how ? [13:57] i thought thats handled by the metacity patch [13:57] ogra: I set it to fix released [13:57] ogra: yeah but look at the version [13:58] ogra: it says maverick [13:58] odd [13:58] well, set it back to New or something [13:58] I cannot [13:58] :( [13:58] agateau: it's using dbus AFAIK. It's not linked against dee [13:58] i can [13:58] done [13:58] :) [13:58] thx [14:00] njpatel: ping, is there some doc about unity-panel-service somewhere? do you think this is something we should use in unity-2d? [14:02] agateau, i think we have to to make it work in natty [14:03] ogra: ok, but i'd like to know how to [14:04] indeed [14:04] though i see all indicators in the panel on my natty install [14:04] seems they still all work [14:06] we should try to share as much code as possible, though, and indicators implementation in unity-2d is not exactly beautiful, so that's another reason to use this service [14:07] indeed [14:13] agateau, no documentation at the moment. I think you could use it, but there are some hacks in there as I haven't had a chance to work on it...probably after A2 [14:13] njpatel: does it handle all indicators? including the appmenu one? [14:13] agateau, yep [14:14] agateau, it has a dbus interface if you want to poke at it with d-feet [14:14] agateau, com.canonical.Unity.Panel.Service [14:14] njpatel: ok, will have a look [15:17] MacSlow, hello, is it by design that critical notifications show up on the unlock screen? [15:27] dbarth: got any new fresh bitesizers for me? [15:45] njpatel: normal release today? [15:45] jcastro, yeah, with the first cut of the dash home screen (no search, though :( ) [15:46] http://people.canonical.com/~njpatel/gordlovesme.png [15:46] very nice [15:47] htorque, no that's a bug [15:47] MacSlow, should i file a report then? [15:48] htorque, if it's not filed already... yes please === oubiwann_ is now known as oubiwann [16:25] jcastro / chriscoulson: hey - I'm Mike Conley, that new MoMo hire, working on Thunderbird + Unity integration === james is now known as Guest26472 [17:04] m_conley: what's a MoMo? [17:05] thumper: hey - that's a short-form for Mozilla Messaging. :) [17:05] m_conley: so you work for Mozilla> [17:05] ? [17:05] thumper: Mozilla Messaging, with Mozilla as parent. So yeah. :) [17:05] cool [17:06] I know a few Mozilla guys [17:07] thumper: it's a pretty awesome situation. :) [17:07] bbl - lunch === bregma|afk is now known as bregma [17:40] howdy m_conley [17:40] <-- just got back from lunch [17:43] jcastro: which TZ are you in? [17:43] I am EST [18:03] back [18:04] jcastro: hey! just wanted to introduce myself, I guess. [18:04] awesome, welcome! [18:05] chrisccoulson: ^^ [18:06] hi m_conley! [18:06] how are you? [18:07] chrisccoulson: I'm good! Very excited to get started on this. I'm just sorting through all of the material out there on this project - in particular, looking at your globalmenu-extension [18:07] as someone who's a bit new to the Mozilla framework, XPCOM, etc, it's quite a bit to digest [18:07] but I think I'm getting there. :) [18:07] m_conley, awesome :) [18:08] i'm pretty new to this too, so the globalmenu work has been a steep learning curve ;) [18:08] but interesting nonetheless :) [18:08] but, if you have any questions, feel free to ask [18:08] chrisccoulson: awesome, thanks a bunch! [18:09] excellent [18:10] chrisccoulson: actually, I do have a question for you [18:11] chrisccoulson: I'm attempting to compile the globalmenu-extension, and I'm hitting a small dependency problem: ./configure is looking for dbusmenu-gtk-0.4 ...but this package is, as far as I can tell, not listed in the Ubuntu packages. [18:12] m_conley, which ubuntu version are you using? [18:12] I'm on Lucid [18:12] ah, that will be why :) [18:12] this is uuber bleeding edge stuff? [18:12] yeah, this currently only works on the current development version (it depends on some functionality we recently added to dbusmenu) [18:13] i think it will compile if you drop the "-0.4" from the pkg-config check, but the menus won't work ;) [18:14] if you don't want to run bleeding edge, it might be worth setting up natty in virtualbox [18:14] chrisccoulson: oh. well that's no good. :/ Hm... yeah, a VM might be the way to go here. [18:14] cool, thanks. [18:14] i think a few people have done that recently, and managed to get unity running successfully in it too [18:14] kenvandine, did you have unity working in virtualbox? [18:14] chrisccoulson, yup [18:14] using it now [18:15] awesome sauce. Alright, that's what I'mma do. Thanks! [18:15] kenvandine, and it's pretty easy to get going? [18:15] just need to install the guest additions [18:15] m_conley, so, it should be ok in virtualbox :) [18:15] thanks kenvandine [18:15] so it will fallback to classic gnome until you install those [18:15] you mount the iso, /usr/share/virtualbox/VBoxGuestAdditions.iso, in the VM [18:16] and run the install script in it [18:16] cool [18:16] thanks. :D [18:16] m_conley, i'll provide some packages for the extension later [18:16] gotcha. alright, i'll let you know if I hit any walls. thanks again! [18:17] i got tbird 3.1 working last week, but i have a crash that only occurs with lightning installed. i'm not sure if you saw my mail last week to the thunderbird-unity group? [18:17] chrisccoulson: I did see that, yes [18:17] it would be good to figure that out at some point :) [18:17] ok, noted. :) [18:20] m_conley, do you have a launchpad account? [18:21] no, not yet. I suppose I should get that going? [18:21] m_conley, yeah, it could be useful :) [18:21] I just got my head wrapped around Mercurial, and now you've thrown Bazaar at me. :p [18:24] heh :) [18:24] bzr is pretty easy, thankfully [18:25] chrisccoulson: alright, cool [19:51] Hai y'all :) [19:52] Was there an ayatana UX meeting around 14:00CET today? [20:02] Anyone? [20:19] chrisccoulson: hey - so I've been trying to get the i386 Natty daily onto a VirtualBox for the last hour or so. After install, it refuses to boot (lots of system messages, numbers, scary OS stuff). Is there a version of Natty I can get thats known to work nicely with VirtualBox? Or, the other way around, is there a version of VirtualBox known to work nicely with Natty? (I've got VirtualBox 4). [20:19] or should I be trying this with VMWare Player, or does it really matter? [20:20] Vbox 4 should be ok. you could try the alpha 1 ISO, but there would be a lot of updates to do afterwards [20:20] hopefully one of them will work :) [20:20] coolbeans [20:20] i don't think vmware player has 3d support does it? [20:20] unsure [20:23] do you know how far it boots? do you get to a login screen, or does it fail before then? [20:23] i might try the latest daily in a bit [20:23] natty is quite volatile right now ;) [20:26] well, I can do a full install, and then I reboot, and then I get a bunch of that scary business. [20:26] no login screen [20:26] i'm trying the alpha [20:26] i'll let you know. :) [20:29] vmware I dont think so [20:29] m_conley, which video card? [20:30] coz_: no video card on the machine I'm running the VM on. On board. [20:31] ah ok [20:34] Having gone through the Ubuntu installer a few times now, I have to say: lovin' the UI for it. Very smooth, folks. [20:52] chrisccoulson: hey - I hate to bug you with support questions, but my Alpha 1 install of Natty is also hitting the fan. Install hangs after displaying the following error: http://imgur.com/K4AJ4 . This occurs whether or not I choose to install updates while installing Natty. Any suggestions on how to proceed? [20:53] easier to see: http://i.imgur.com/K4AJ4.png [20:55] m_conley, i haven't used the installer for a while? does it give an option to not update packages on install? [20:55] it does [20:55] and whichever I choose, same result. [20:57] hmmm, strange. it might be worth popping on #ubuntu-devel and asking in there [20:57] chrisccoulson: cool, thanks [20:57] cjwatson might have a clue about what to do :) [20:57] maybe using the alternate installer? [20:58] m_conley, oh, just as i pop in there, i see they're talking about bug 701954 [20:58] Launchpad bug 701954 in console-setup (Ubuntu Natty) "initramfs hook fails with ".: 17: Can't open /etc/default/keyboard" when /etc/default/keyboard doesn't exists for some reason" [High,Fix committed] https://launchpad.net/bugs/701954 [20:58] might be your issue ;) [21:05] chrisccoulson: cool, thanks for the heads up. :) [21:27] ah, i figured out why my menu icons aren't working now :) === m_conley is now known as m_conley_away [22:21] chrisccoulson: What was it? [22:21] oh, i'm trying to figure out how to decide which icons to show in the menubar from firefox [22:22] Alright :) [22:22] and i was checking the visibility attribute of some anonymous content which never gets created, because the frames for the menus are never created [22:22] (because the menu is in the unity panel instead) [22:22]