[00:03] RAOF, robert_ancell, TheMuso: meeting time? [00:04] yep [00:04] I believe that it's 9am, yes :) [00:04] Jason's away to the airport and asked that I chair the meeting today [00:04] alrighty, robert_ancell want to kick us off? [00:05] how's lightdm coming? :-) [00:05] coming along good. Seems to be working well on most peoples boxes [00:05] Sure. [00:05] I believe it just got on the seed and displaced gdm for Oneiric so preparing for floods of reports! [00:06] sweetness [00:06] Is the plymouth integration madness hooked up? [00:07] I've implemeted it, and it's working until someone tells me otherwise :) [00:07] Heh, ok. Looks like it's time to reboot a box or two :) [00:08] robert_ancell, saw the discussion about keyboard layout stuff, sounds like you have a sane approach worked out, to the limits of my own keyboard layout knowhow [00:09] people seem not shy about filing bug reports when keyboard stuff isn't right ;-) [00:09] The next main feature in 0.4.0 will be guest logins and some stuff under the hood to make it more stable (particularly unittests) [00:09] yeah, keyboard/language is a bit of a black art! [00:10] ok, moving on... TheMuso how's your week been going? [00:10] Not too bad, got all of at-spi2 into main as of the next dailies, it should be used by default. [00:10] Still doing Qt a11y testing, will be pushing patches for Qt to didrocks by EOW. [00:11] And still got to get the qt at-spi bridge uploaded and into main. [00:11] Also started working on ubiquity a11y work. [00:11] Thats the extent of my week pretty much. [00:11] cool. pardon my ignorance, what is at-spi? [00:11] (just idly curious) [00:12] ah assistive technology service provider interface [00:13] TheMuso, ok great [00:13] Yeah, its the core API/ipc for getting a11y stuff from apps, and getting it to a screen reader/magnifier etc. [00:13] RAOF, up to your eyeballs in mesa merging I should guess? [00:13] Yup. [00:13] how's that going? [00:14] Got some nice savings in the gallium drivers by factoring out another shared library. [00:14] They're now about a third as large (so ~3MB → ~1MB) [00:14] nice [00:15] had much chance to look at bug reports on mesa? anything interesting there? [00:15] The upload is pending a llvm-2.9 MIR, which is pending a build of openjdk-6b18 on armel to check for test-suite regressions. [00:16] Nothing particularly interesting has landed in my Mesa bugmail box (yet ☺). [00:17] RAOF, would you mind jotting down about our current package merge status in https://wiki.ubuntu.com/DesktopTeam/Meeting/2011-06-07 when you get a moment? I neglected to do that last week [00:17] speaking of mesa, I traded a couple emails with keithp on the topic of post-release mesa updates [00:18] bryceh: Certainly. I'll do it before rebooting to check that plymouth integration isn't broken :) [00:18] I outlined what we're facing as SRU requirements in getting mesa patches (let alone point releases) accepted, and he seemed receptive and said they were consistent with the requirements he wants to impose on mesa 7.11 point releases [00:19] so even if we can't do full point releases, *maybe* we'll have more actionable mesa patches this time through [00:20] he did suggest that there might be a regression test for mesa we could look at hooking up... I'll have to follow up on that more [00:20] I think keithp has a pretty good understanding of what's required in a stable branch. [00:20] agreed [00:20] bryceh: Other than piglit? Interesting. [00:20] sounded like it [00:21] It'll run into the problem of needing to be run on actual hardware though. [00:22] yep [00:22] other than this, I've been toolsmithing and sponsoring various bugfix patches. Lots of bug reporters needing debugging advice too. [00:22] I also took the role of Ubuntu stakeholder for Launchpad, so have been conversing with folks about ideas on how we could improve Launchpad to Ubuntu's benefit. Fun stuff. :-) [00:23] Yay! I can't think of a better LP liason than you. [00:23] ok, did we miss anyone? If not, AOB? [00:23] Nope [00:24] excellent, ok guys have a good day. :-) [00:24] Oooh, mesa talking about a 6-monthly cadence for releases? Yay! [00:26] Hm. That would see 7.11 scheduled for release between A2 and A3. I could live with that! [01:52] Is it just me, or does compiz currently have rendering issues in oneiric with intel? [01:52] i.e nothing gets cleared. [01:54] Works fine¹ here. [01:54] You're not running into the background-setting problem? [01:55] (Where, unless you explicitly set a new background g-s-d doesn't draw anything on the root window so whatever was last drawn is left there?) [01:55] Oh thats what it is. [01:55] Thats terrible. [01:57] ahhh. [01:57] Thats better. [01:57] Heh. [01:57] Yeah. I thought didrocks was fixing that? [01:58] But, it is a dev cycle after all. [02:00] RAOF, where is a good description of the X protocol (i.e. the binary messaging) [02:01] robert_ancell: The various x11-proto-* docs. They all have a description of the protocol messages and replies. [02:01] RAOF, on the web? Or packaged? [02:03] robert_ancell: They're all packaged - in the same package as the development headers. [02:03] (For that part of the protocol) [02:03] RAOF, ok, cheers [02:04] So, /usr/share/doc/x11proto-core-dev/x11protocol.txt.gz for the core protocol, which is huge, or /usr/share/doc/x11proto-dri2-dev/dri2proto.txt.gz for the dri2 protocol, which is svelte, sensible, and easy to read :) [02:52] So, the Liferea upstream has no problem with integrating indicator support upstream provided I maintain it there [02:59] So, you shouldn't begin sentences with so. ;-) [03:04] heh [03:10] * RAOF goes “hm, now that it's stopped raining it might be time to go and get stuff for lunch”… and it starts raining again. [03:10] Yeah that does suck. [03:11] Now that NSW has had a butload of rain, it seems its Tassy's turn now. [03:11] s/now/this time/ [05:28] Good morning [05:38] Good morning pitti. [05:40] hey TheMuso, how are you? [05:40] pitti: Not too bad thanks. Yourself? [05:40] still a bit tired, but well otherwise [06:31] * bryceh waves to pitti [06:32] hey bryceh, how are you? [06:32] great; closed a ton of bugs today [06:33] oh, always good! [06:33] pitti, speaking of which, I have an SRU for a really critical bug that would benefit from being fast-tracked [06:33] oh, sure [06:33] bug 774978 [06:33] Launchpad bug 774978 in xserver-xorg-input-synaptics "xserver crashes in RecordAReply when XRecord is enabled in syndaemon" [High,Fix released] https://launchpad.net/bugs/774978 [06:34] pitti, fix is uploaded to natty-proposed but I haven't filled in the paperwork and stuff. I'll do that now. [06:34] adding natty task, processing [06:34] heh, weird fix :) [06:36] that indeed calls for some further explanation about regression potential of building without libxtst [06:36] bryceh: processed [06:38] pitti: damn. Had I known you were going to upload casper, I would have asked you to wait. [06:38] No matter, I just had to make another upload. [06:38] it's a small package, the buildds will survive two uploads :) [06:39] Yeah. [06:48] pitti, thanks! [06:49] pitti, yeah I know... I wish we had the root cause figured out, but I suspect an earlier fix to record functionality that Chase did for a utouch bug exposed the codepath [06:58] pitti, aha not chase's fault; debian introduced the xtst/record dependency, and it was just plain buggy. Chase's fix was a partial fix that papered over a worse problem but (apparently) left this crash bug. [06:58] * bryceh will update bug report with findings [06:58] ah, thanks [07:04] morning bryceh and pitti [07:05] hey glatzor, wie gehts?/ [07:05] pitti, I am fine. Making my first steps in gobject introspection land :) [07:05] ooh, what are you porting? [07:05] glatzor: you know https://live.gnome.org/PyGObject/IntrospectionPorting ? [07:06] heya glatzor [07:06] * bryceh needs to learn PyGObject at some point [07:07] pitti, I am writing an aptdaemon client library in C to even get introspection python bindings :) [07:09] glatzor: ah, so you are learning about https://live.gnome.org/GObjectIntrospection/Annotations? :-) [07:11] right [07:32] pitti, ooooh the story of this bug is getting really interesting! [07:54] good morning everyone === chrisccoulson_ is now known as chrisccoulson [07:55] good morning [07:55] hey didrocks [07:57] mvo: guten Morgen [07:57] hey didrocks and pitti [07:57] hey chrisccoulson, how are you? [07:57] guten morgen pitti [07:57] hey chrisccoulson! [07:58] pitti - good thanks, how are you? [07:58] I'm great, thanks! [07:59] hey pitti and didrocks and chrisccoulson! good morning [07:59] hi mvo :) [07:59] hey mvo [08:00] mvo: update-notifier is in a dire situation now, as all the notify_notification_new_with_status_icon() calls don't work any more [08:00] mvo: i. e. in natty they stopped working (no status icons), and in oneiric it doesn't even build any more due to the new libnotify [08:00] mvo: but don't we just open update-manager/apport/etc. anyway? can we just rip out all the notification stuff? [08:02] pitti: right, I kept it for compatibility with e.g. xfce [08:02] pitti: I have a look at it today [08:02] chrisccoulson: ok to remove xulrunner-1.9.2-gnome-support xulrunner-2.0-gnome-support from oneiric? [08:02] (oneconf as well is called by update-notifier) [08:02] chrisccoulson: no rdepends, depends on NBS libraries [08:02] morning mvo [08:03] pitti: I need to look at u-n anyway for the apport hooks (re your mail from today) [08:03] glatzor: hey glatzor! [08:03] mvo: I fixed the constructor call (not committed yet), but I wondered whether you just want me to remove the other notification stuff [08:03] pitti - which libraries do they depend on? note that those packages literally only exist to pull in the dependencies. the actual dependencies are in the in the main packages [08:03] chrisccoulson: ah, ignore me, it's not a separate source [08:04] chrisccoulson: libnotify1 [08:04] ah, ok [08:04] i guess they need a rebuild to pick up the new version [08:04] but, i'm not sure if there's much point, seeing as we're going to kill them soon anyway ;) [08:04] chrisccoulson: the xulrunner packages depend on libreadline5-dev [08:04] (and i made a start on that last night) [08:04] chrisccoulson: should be libreadline-dev now, but why -dev in the first place? [08:05] pitti - yeah, that's bogus. i already dropped that from thunderbird and firefox [08:05] ah, tahnks [08:05] mvo: do we still need src/reboot.c? we already have that in the session menu and the logout dialog [08:06] pitti: right, again for compat with other environments like xfce [08:06] mvo: hm, so it seems this would all need to be ported to use an appindicator? [08:06] ... which we then would need to suppress in unity/gnome [08:07] pitti: if we rip stuff out I would prefer to just rip all the auto open out and essentially leave u-n as it is and replace with a new tool that is massively simpler [08:07] pitti: i.e. one for xfce/ldxe etc and one for unity/indicators/auto-open [08:08] mvo: ok, so the simpler fix for u-n would just be to display notifications, without attaching to status icons [08:08] want me to do that for now, so that the package builds and works again? [08:08] with notify-osd they aren't "attached" anyway [08:08] pitti: I actaully wanted to work on this for natty but then there was the of upstart session plans and I kind of hoped to use that [08:08] yeah, that'd make it much easier [08:08] pitti: so notify-osd break this api? compared to notification-daemon? or is it removed in notification-daemon too? [08:08] so perhaps just _new_with_status_icon -> _new() would do it for now [08:09] mvo: no, thet new libnotify 0.7 doesn't have it any more [08:09] gnome-shell doesn't provide this any more either [08:09] it's an API change [08:09] pitti: ok, thats fine then, please just remove it then [08:09] mvo: ok, will do [08:09] pitti: thanks! [08:09] thanks for the heads-up [08:09] yw, once thats in I will add the apport check addition and upload [08:20] cdbs: your autojoin comment made me smile (: [08:20] Tm_T: why? [08:21] perhaps I'm so used to have you around that it's a bit weird you're not (: [08:21] mvo: hm, any idea how I actually get notification bubbles? haven't seen one from u-n in ages, it just pops up windows [08:22] (it builds fine now, but testing wouldn't hurt, I guess) [08:25] mvo: oh, /apps/update-notifier/show_apport_crashes ? [08:26] hm, that's already true [08:27] presumably it never gets around to actually showing it, as gtk_status_icon_get_visible(ta->tray_icon) never becomes true [08:28] pitti: could you try setting /apps/update-notifier/auto_launch to false? [08:29] mvo: trying that, also setting the interval to 0 [08:32] mvo: look mom! notifications! [08:33] mvo: tested for crashes and updates now, I'm happy [08:33] mvo: pushed [08:33] mvo: do you want me to upload, or do you want to work on it further anyway? [08:34] * didrocks likes the "tested for crashes … I'm happy" (puttin things out of context can be so funny) [08:34] pitti: I will add a additional checking before calling apport, but feel free to upload dosn't hurt if it will be uploaded twice [08:34] * pitti mumbles something about "gsettings migration" [08:34] didrocks: that's simple: no crashes -> no work for us -> we'll get fired [08:35] pitti: arghhhhh, right! let's ensure we have crashes :-) [08:53] morning [09:06] hey [09:06] hi seb128 [09:06] hi seb128, how are you? [09:06] hey rodrigo__, how are you? [09:06] chrisccoulson, hey, I'm fine thanks, you? [09:06] hi chrisccoulson [09:07] seb128, yeah, good thanks. started cleaning the python-gtkmozembed rdepends last night :) [09:07] \o/ [09:07] hi rodrigo__, how are you? where is rodrigo_ and rodrigo? ;) [09:08] rodrigo is always taken by someone in brazil [09:08] by the end of the day, will you have become rodrigo________________? [09:08] :) [09:08] not sure where rodrigo_ is :) [09:08] yeah, with all the disconnections I get recently :) [09:09] what we should do is all have an arbitrary number of underscores in our names, and then remove autocomplete from IRC clients ;) [09:09] that might reduce the number of pings people get [09:13] right, good idea [09:13] so, you patch xchat and I do empathy to remove the autocompletion? :) [09:14] heh :) [09:14] you need to break the text selection to prevent people to do copies ;-) [09:14] right [09:14] seb128, oh, i'm sure we could do that too [09:14] and we could even generate a random number of underscores automatically [09:14] imagine the chaos! [09:15] easier you could just disable highlighting ;-) [09:15] that's too simple ;) [09:18] hey rodrigo__, seb128 [09:18] lut didrocks [09:18] didrocks, how are you? [09:19] I'm fine, thanks, and you? [09:19] didrocks, users complain that they can't install unity today due to you ;-) [09:19] didrocks, I'm fine thanks! [09:19] (nightly build of Qt unfortunatly didn't fix it with -O1) [09:19] oh, binNEw [09:19] didrocks, (seems you need to new compiz binaries) [09:19] yeah [09:19] seb128: can you have a look at them? as I don't really want to new them myself :) [09:19] ok [09:20] compiz*-default [09:20] bonjour seb128 [09:20] pitti, hey! how are you? [09:20] seb128: pretty well, thanks! [09:20] gdm is making progress, but still a bit buggy [09:20] half of the time the greeter doesn't draw properly [09:20] that's with and without our patches [09:20] it's runtime buggy? [09:20] ok [09:21] you can navigate the window, but the user list isn't visible sometimes [09:21] pitti, did you figure if we still need the plymouth integration patches and why those are not upstream? [09:21] we still need them [09:21] no idea why this "plymouth integration branch" was never merged :/ [09:22] http://pkgs.fedoraproject.org/gitweb/?p=gdm.git;a=blob;f=plymouth.patch;hb=HEAD [09:22] fedora still ships it as a patch as well [09:22] ok, I will ask on #gdm today [09:25] didrocks, seems like kde.xml got dropped from the compiz-plugins binary but not added to -default [09:25] seb128: it's beeing added to compiz-kde [09:26] * move /usr/share/compiz/kde.xml wrongly set in compiz-plugins to compiz-kde [09:26] in the changelog [09:27] didrocks, oh right (I didn't look at the changelog, just at the debs) [09:27] no worry :) [09:28] didrocks, ok, compiz newed, looking at c-p-m [09:28] seb128: thanks [09:32] pitti: I send two branches your way for apport, please let me know what you think [09:33] didrocks, ok, I didn't check but compiz-plugins-main-dev could need a Replaces on compiz-fusion-plugins-main as well for lucid upgrades [09:33] didrocks, otherwise looks good [09:33] didrocks, i.e I didn't check if they have common files but I assume they do [09:33] seb128: hum, I think files moved, but I'll have a look [09:34] didrocks, ok [09:34] (was compiz-fusion dir= [09:34] didrocks, anyway not an issue for oneiric so newing those [09:34] seb128: thanks :) [09:34] didrocks, well things like ./usr/lib/pkgconfig/compiz-animation.pc [09:34] or what that fusion namespaced as well? [09:34] should be easy to check anyway [09:34] let me quickly check, I think it was in 0.8 [09:35] mvo: ah, will do [09:35] seb128: you're right, I'm adding the change to the vcs [09:35] didrocks, thanks! [09:35] gah, today's live images failed to build due to compiz-plugins-main [09:36] thanks for pointing it :) [09:36] didrocks, yw [09:36] compiz is now the land of C/R/P :-) [09:36] * pitti binNEWs [09:36] doh [09:36] "FAILED: compiz-plugins-main (Queue item already accepted)" [09:36] sorry seb128 [09:37] pitti, hum, did you just hijack what I was discussing with didrocks for 5 minutes? ;-) [09:37] pitti, no worry, I just finished checking, out of the lucid replaces they were ok [09:39] pushed in vcs, will wrap that next upload [09:42] seb128: sorry, I just got the "failed to build" email from cdimage, and acted on it [09:42] bad mid-air collision [09:43] pitti, no worry ;-) [09:43] * seb128 hugs pitti [09:43] * pitti hugs back seb128 [09:45] Are there any themes yet for gtk3-engines-unico? [09:47] lucazade is working on one and has it on his launchpad ppa or vcs [09:49] lucidfox: Yes, its Ambiance being developed by lucazade at lp:~lucazade/+junk/ambiance-gtk3 [09:49] Ooh, nice [09:53] mvo: I did https://code.launchpad.net/~mvo/apport/apport-checkreports-addition/+merge/63819 now [09:55] Okay, it works. Doesn't look very Ambiance-y at the moment, but I presume this will change by the final release :) [09:56] pitti: awsome, thanks! === Guest99970 is now known as gord [10:01] hmm, did anyone else see hangs in oneiric when gnome-user-guide is installed ? [10:02] i get "hung task" timeout messages from the kernel here for dpkg ... [10:02] only for that one package and reproducable [10:04] mvo: done as well; want me to upload that to oneiric now? [10:04] ogra_, I didn't but I don't know if that one got upgraded or installed since I use an upgrade from natty [10:04] it could be still the natty version [10:05] well, on armel it definitely gets updated during a move from natty to oneiric [10:05] so i would expect that on armel too [10:05] what bothers me is that it seems to hang in the configure step ... usually i would expect the postinst to misbehave here [10:06] you are sure it's not just being slow? [10:06] or that you have io issues? [10:06] but that would likely point to a different binary in dmesg and also i cant seem to find a postins script for the package [10:06] pitti: please! [10:06] seb128, dpkg stops right after the dmesg message shows up [10:07] i left it sitting there over night [10:07] so i doubt its slowness [10:07] and i just heard the same issue from another armel user [10:08] seems armel specific then [10:08] but what does that package do that others dont ? especially in the configure step [10:09] i cant find a postinst, but there must be some command running that other packages dont use [10:09] nothing, it doesn't even has a postinst [10:09] yes [10:09] well maybe not [10:09] is there a trigger like scrollkeeper or so = [10:09] it's maybe the content of the deb which hits a dpkg bug [10:09] ? [10:09] dunno [10:10] no, scrollkeeper has been deprecated [10:10] and triggers are run after the configure anyway [10:11] it seems likely the deb content hits a dpkg bug [10:11] hmm, k [10:11] i wonder why only this package hits it though [10:11] you should better take that discussion to #ubuntu-devel, out of mvo who might be able to help you I think it's out of the desktop domain [10:12] installing or upgrading g-u-d was always slow in the past though [10:12] like 5-10min just for that package [10:12] it has quite some content [10:12] yep [10:12] i checked the .list file ... [10:12] nearly 7000 files [10:21] One of my friend bought Braid from Software Center, but sound is not working. Where should I report this? [10:26] hey saamm, we don't currently have a official bugreport channel, its probably best to contact the braid publisher for now and report the problem there [10:31] uhm isnt there a way that we file a bug in launchpad and it is reported in upstream as Ubuntu developers may have better contacts with app providers [10:32] braid has no support forum, page or anything...just an email :( [10:33] pitti, mvo: is there any reason update-notifier still need a systray at all nowadays? [10:34] seb128: compatbility with xfce and friends [10:34] mvo, how compatible? what does it use under unity which has no systray which is not available in xfce? [10:34] I though it was a service spawning commands only nowadays [10:34] like it does open update-manager or apport [10:35] seb128: in xfce they don't use it in auto-open mode (last I checked) [10:35] seb128: the longer term goal of me is to get rid of u-n by default and replace it with some simpler version that just supports the auto-open stuff [10:35] seb128: actually having the tray icons makes it all more complicated [10:36] well, why not just dropping the icon then? [10:36] xfce could use it in auto-open mode [10:36] saamm: I think we need a support channel for issues like this, but there is none currently, I will enquire about this. [10:36] seb128: but why force it on them if they don't like auto-open mode? [10:37] seb128: its low maintainence (for the most part) [10:37] mvo, why lower the ubuntu quality just to accomodate other desktops? [10:37] pitti - will you let me know when you're ready to start building language packs for natty? [10:38] mvo, I've seen several bugs with syslog errors about the systray icon being rejected from unity, some users get clearly confused under unity by trying to use it [10:38] mvo, so either it should we an indicator or dropped imho [10:38] mvo, ok thanks...I try to contact them through email... [10:41] pitti, is it worth downloading an oneiric daily iso to use in kvm or will it give no desktop between unity-3d not working and the unity-2d,qt,gcc issues? [10:50] seb128: dpeends on what you want to test [10:51] with -vga vmware you can get to a desktop without the unity 2d drash [10:51] dash [10:51] you can ctrl+alt+t for a terminal and run things there [10:54] chrisccoulson: yes, I will; still waiting for the export to happen [10:56] seb128: I tried installing gnome-session-fallback in a live system and then choosing gnome session in gdm, that works fine, too [10:56] the applications menu is empty, though -> missing dependency? [10:56] pitti, weird, gnome-menus should be installed [10:57] pitti, I will try that, I don't need much, I want to check the rdepends for the deprecated lib on a clean system [10:57] that's fine in a terminal [10:57] I've some non default things installed on mine and I might have uninstalled some others [10:57] g-menus is isntalled, yes [10:57] seb128: I have yesterday's CD running, want me to try some stuff? [10:57] not sure why the menu doesn't work then [10:57] like purging, and teling you what's left? [10:58] pitti, no, I just want to poke around, do rdepends list and dump them on the etherpad [10:58] pitti, I will just rsync the iso over lunch and do that later but thanks for asking! [10:59] hm, I can purge libgail-gnome-module and libpanel-applet2-0; only a recommends? [10:59] libpanel-applet2-0 is the old gtk2 bonobo lib [10:59] ah, this image still has gnome-mag and libatspi1.0-0 [10:59] the current version is 4 [11:00] whatever is using libpanel-applet2-0 is broken [11:00] so yeah, as we figured out, tomboy and gbrainy are the only two culprits [11:03] I want to start building gtk2 and gconf rdepends lists and "what ship schemas" as well [11:04] I might use a pbuilder could be easier to script those [11:05] http://cdimage.ubuntu.com/daily/20110608/ [11:06] wohoo! [11:06] all but i386 are within limit [11:06] dropping lsb-* and build-essential really helped [11:07] pitti: Congrats! [11:07] pitti: But, that's the Alternate CD, isn't it? [11:07] we removed 12.6 MB, added 0.3, and changed -4.1 MB [11:07] cdbs: yes [11:07] desktop failed to build today [11:08] due to the compiz split in NEW [11:08] the lives are still huge [11:08] but it should be in the archive now, I'll retry a daily build [11:09] seb128: gbrainy can certainly be unseeded; I don't quite remember, did we say yesterday that we could build totem with a --no-gnome something configure option? [11:10] * pitti starts a new live build [11:10] seb128: ^ you might want to wait for that [11:12] ugh, a lot of icons have disappeared after some cherry-picking update of gnome packages [11:13] rodrigo__: --verbose ? [11:13] rodrigo__: I recently split gnome-icon-theme [11:13] oh [11:13] so, should I install gnome-icon-theme-full? [11:13] so that it mostly uses humanity, and laternatively depends on g-i-t-full [11:14] ok [11:14] rodrigo__: I guess that would help, but it shouldn't break for a normal desktop in the first place [11:14] so I'm intersted what you did [11:14]