[00:02] bryceh: Create the /var/lock directory in your pbuilder chroot; base-files removed it, and pbuilder/sbuilder don't know to create it. === mclasen is now known as mclasen_afk [00:05] mm, the bug that keeps on giving === m_conley is now known as m_conley_away === chaotic_ is now known as chaotic [05:02] Good morning [05:04] dobey: replied as well [05:15] Good morning pitti! [05:15] hey RAOF, how are you? [05:16] Alright. [05:16] Managed to do a little running around lunchtime when it wasn't raining (when I started). That was nice :) [05:17] Hows about you? [05:23] still a bit 'oneiric', went to bed rather late [05:23] it seems the Brits finally sent us their bad weather! [05:23] Hah! [05:24] pitti, can you kick unity-greeter out of the new queue? [05:24] ooh [05:24] robert_ancell: sure [05:25] Also, Would you be a dear and create a git repository for colord and shared-color-profiles on alioth in collab-maint? :{) [05:25] you can install it and play around with /usr/lib/lightdm/unity-greeter --test-mode [05:25] or be risky like me and actually use it as your greeter :) [05:25] robert_ancell: Do you have multiple monitors? [05:26] RAOF, no, and no idea how it will behave... [05:26] My guess is fine — the first time :) [05:26] RAOF: can DDs actually do that, or does it require an admin? [05:26] Because that's what the current greeter does :) [05:26] pitti: Any DD can; you've got write access there. [05:26] "create git repo" still sounds like having to tear apart a chicken [05:27] I suggest re-using the chicken-tearing apparatus found in /git/pkg-cli-apps/create_repo :) [05:27] handy [05:27] ok, doing Robert's NEW, and then getting to that [05:28] Also, I'd just like to put it on record that I hate copyright, and it should go and die somewhere where I don't have to worry about it. [05:28] duly noted [05:28] I don't actually mind it much as long as it comes with a free license [05:29] Perhaps I should rephrase: I hate that *other people* don't care as much about copyright as we do, so I run into things like shared-color-profiles. [05:29] robert_ancell: it could do with some copyright/license comment headers in the vala files [05:29] pitti, oh, can do [05:30] robert_ancell: i. e. where did you take the "Copyright: 2011 Canonical Ltd" in debian/copyright from? [05:30] robert_ancell: or did you write it yourself? [05:30] pitti, well, I filled it out manually :) [05:31] robert_ancell: can you please fix Format-Specification: to http://dep.debian.net/deps/dep5/ ? [05:31] pitti, I have copyright headers pushed to master, do you want a new release? [05:32] robert_ancell: that's fine; won't be the last upload in oneiric, I guess [05:32] "New upstream release" is a bit confusing, but [05:33] robert_ancell: lightdm is only a recommends, not a depends? [05:33] pitti, the library is a depends, the daemon not technically [05:34] robert_ancell: so, with the copyright headers and fixed format-specification: I'm happy (neither requires an upload) [05:36] pitti, I've pushed copyright changes to the packaging branch, please check [05:37] robert_ancell: I think it's "Upstream-Name:" now [05:38] and Maintainer: -> Upstream-Contact: [05:38] robert_ancell: and you didn't push the copyright headers yet? [05:38] RAOF: do you mean /git/pkg-cli-apps/setup-repository ? [05:39] whoops updated copyright (dropping those fields as they're optional). Headers are pushed to lp:unity-greeter [05:39] ah, trunk branch [05:39] pitti: Yes, I do. You won't be able to use it as-is, because pkg-cli-apps uses a rather particular setup and that script will set the repository up to publish things to #debian-cli, but the core commands in that are all you need. [05:39] robert_ancell: thanks, looks good [05:40] The Yamma profiles are Copyright by those well known contributors, :( [05:42] RAOF: I created colord.git [05:43] http://anonscm.debian.org/gitweb/?p=collab-maint/colord.git;a=summary doesn't exist yet, I guess it takes a while to catch up? [05:43] pitti: Rockin. [05:43] I think it does, yes. [05:43] Now, to get me access to that… [05:43] * RAOF wonders if it would really be *that* much effort to get DD. [05:44] RAOF: DM wouldn't be, anwyay [05:47] Hm. That looks positively friendly. [06:35] Just an FYI, the new gnome-control-center depends on the "glxinfo" command to print details on the GPU in "System Settings" [06:35] Without the mesa-utils package containing glxinfo, it returns "Unknown" [06:37] Does it really spawn glxinfo and parse its output, rather than doing a simple GL call? Ugh. Anyway, have you filed a bug? :) [06:38] bahahahhaaha [06:38] wow that reminds me of [06:38] ... beryl [06:39] Heh. Kernel SRUs are big, and cause lots of bugmail. [06:43] I'm making a mental note to note dist upgrade ever [06:43] *not [06:44] I was just about to upload the compiz release yesterday and then boom, my system hangs halfway through a kernel update and now its trashed [06:44] (and by yesterday, I mean 4:30 am, when I was just about to go to bed :/) [06:46] smspillaz: hm, could that still be bug 807306 ? [06:46] Launchpad bug 807306 in udev "[oneiric] Keyboard & mouse not working in X - incomplete migration to /run" [High,Fix released] https://launchpad.net/bugs/807306 [06:47] pitti: no, it was because I'm using nouveau and did something the driver didn't like [06:47] pitti: basically, nouveau gets cranky if you run out of video memory [06:48] pitti: I also get 807306 too, but I have an external mouse, so its not that bad [06:48] that should be fixed now, though [06:48] (by trashed, I mean it fails to mount / even though fsck says its fine) [06:49] *My* laptop started correctly today :) [06:49] pitti: basically, its the case of, I had to force power off the machine during a dist-upgrade [06:49] that always breaks things :) [06:55] RAOF: hm, are we turning on nouveau 3d by default now ? [06:56] smspillaz: Yes. [06:57] nice [06:58] is there just a whitelist or ... ? [07:03] It's just turned on. [07:03] ah ok, I didn't know it worked in most places [07:03] That's part of what we're trying to discover ;) [07:03] But it works fine on all of *my* hardware, so ship it! [07:04] :) [07:09] on a gts 250 with nouveau3d the system hard freeze after few mins!! [07:09] so it is hard to install also from livecd [07:09] :) [07:10] * smspillaz dist-upgrades and hopes for the best [07:10] from my chroot [07:16] wohooo! I have X back again! [07:16] isn't it amazing that you can leave your system in complete ruins and somehow recover from that? [07:22] lucazade: See, that's the sort of feedback we're after. Well, not exactly, but it's important to know! [07:24] RAOF: I have this issue from a couple of Ubuntu releases when using experimental-dri [07:27] Is there a bug on launchpad about this? [07:27] yes, let me search [07:28] 'Cause that's where we'll be looking when it comes time to keep 3D enabled or disable it. [07:28] RAOF: maybe it would be good to have a whitelist [07:28] smspillaz: That would be another option. [07:28] yeah [07:28] I know that the 8600GT does pretty well under average circumstances [07:28] I just do insane things to my gpu [07:29] RAOF: bug 696427 ... gts250 1gb, it is quite recent as card [07:29] Launchpad bug 696427 in mesa "Unity visual glitches and freeze (libgl1-dri-mesa-experimental)" [Wishlist,Confirmed] https://launchpad.net/bugs/696427 [07:29] lucazade: Ta. [07:30] RAOF: I was going to ask you actually, is there a sane way to enable pcsnv at the moment ? [07:30] smspillaz: Looking to do some CUDA? [07:30] (No) [07:30] mmm, power management actualy [07:30] Ah! [07:31] so that, you know, my house doesn't catch fire for a reason that isn't my inability to cook one day [07:32] I had a look into it a week ago but it doesn't like there's a non-crack way of doing it [07:32] IIRC pscnv doesn't actually play with the rest of the 3D stuff, and the kernel module does things like not use TTM. If you really want power management you could pull the nouveau kernel patches from the mailing list and build them? :) [07:32] yeah, I suppose I could do that [07:32] I have to run my own kernel most of the time anyways for wayland stuff soo ... [07:34] AFAIK the power management code works fine, they're just being a bit paranoid about reverse-engineering stuff that can, you know, /fry your card/ [07:35] eh, my card is already frying things [07:35] I'm lucky I haven't had to bake it yet though [07:39] Oh, you've got one of those bakeable laptops too, have you? Lucky! :) [07:40] yeah, the most notorious one [07:40] the XPS M1530 [07:40] same laptop as sabdfl iirc [07:40] Hi, can someone just rebuild this build https://launchpad.net/ubuntu/+source/unity-greeter/0.0.1-0ubuntu1/+build/2623446 ? The build-dep is now in the archive [07:40] smspillaz: What's bad about that laptop? [07:40] smspillaz: baking seems to work well for didrocks, though [07:40] pitti: indeed, he's done it twice now :) [07:40] cdbs: it's already building [07:41] * cdbs refreshes [07:41] ah okay [07:41] robert_ancell did that probably :) [07:42] cdbs: http://www.engadget.com/2008/07/10/all-nvidia-8400m-8600m-chips-faulty/ [07:43] the 8600M is basically an overlocked 8400M [07:43] only a handful of laptops ship with it [07:43] that being the 2008 MacBook Pro and Dell XPS M1530 [07:43] (which is basically just a 2008 MacBook Pro hardware wise) [07:43] smspillaz: the news is from 2008 [07:44] I know :) [07:44] smspillaz: haven't the chips been replaced since then? [07:44] this laptop is from 2008 [07:44] well, yes [07:44] yeah, but still, through a product advisory or recall [07:44] in most laptops you'll find a fermi or something now [07:44] cdbs: nope, they never recalled it [07:44] all dell did was extend the warranty for "graphics related issues" and that's only in the US too [07:45] Seriously, all tech companies think 99% of their target audience and consumer base lives in the US [07:45] well, it's only practical to do it in the US [07:46] people stateside tend to have the capacity to make the most noise [07:46] if you keep them happy you'll usually keep the world happy [07:53] morning [07:55] good morning everyone [07:56] Howdie and good morning chrisccoulson. [07:56] hey RAOF, how are you? [07:57] I'm an awesome force of nature. [07:57] Well, perhaps not. [07:57] But pretty good, anyway :) [07:57] heh [07:58] Hows about your fine self? [08:02] hey chrisccoulson [08:02] hi pitti, how are you? [08:02] pretty well, thanks! [08:03] wasting my time with packaging python3-cairo and then python3-gobject [08:04] pitti: python3-cairo ? Does it use GI ? [08:04] pitti - oh, i thought the python bindings were all disappearing? [08:05] abhinav-: no, it's a static binding; apparently cairo isn't very suitable for GI [08:05] Yeah :-| [08:06] might also be a performance issue [08:06] py3cairo is written in cpython [08:06] for throwing buttons around the GI overhead doesn't matter, but for vector drawing it might, but I don't know [08:08] Well not having GI bindings for cairo seems a drawback, since Gtk3 removed some of the API just because cairo provides those functionalities. I think Pycairo was lagging behind the C API of cairo [08:08] why drawback? [08:08] it doesn't matter much if you do "import cairo" or "from gi.repository import Cairo"? [08:09] for the "lagging behind", certainly, yes [08:09] Yes, so while the current C API of cairo is ready to be augmented with Gtk3, I am not sure same can be done with the Python bindings [08:10] I am referring to my experience when I was trying to play around with Gtk3 to add screenshot feature to apport [08:10] Althoug it was more than a month ago [08:13] I wouldn't expect the GI overhead to be a huge issue for cairo; you should generally be doing a small number of calls to produce a large image. [08:14] anyway, py3cairo is there, and GI bindings are not [08:14] so I'm packaging these for now to unblock python3 progress [08:23] hey [08:24] bonjour seb128 [08:24] hey pitti [08:24] how are you? [08:25] good morning, gents [08:25] * pitti hugs seb128 and desrt [08:25] what are the news today? ;-) I saw an unity greeter upload, did anybody try it? [08:25] hey desrt [08:25] seb128: pretty well, thanks! just got my py3cairo package working [08:25] * seb128 hugs pitti [08:25] pitti, \o/ [08:25] now off to packging python3-gobject [08:25] seb128: just built/building, needs binNEW still [08:25] (greeter) [08:25] ok [08:28] hey seb128, how are you? [08:29] hey chrisccoulson, I'm fine thanks, how are you? [08:29] seb128, good thanks [08:29] chrisccoulson, great work on g-s-d, you won beers for the next UDS ;-) [08:29] argh, LP down [08:29] heh :-) [08:30] seb128, i wanted to fix the crash in exit() too, but i'm not sure of the best way to do that yet :/ [08:30] do you understand what is wrong and ponder how to fix it? or do you need to figure what is wrong? [08:30] seb128, yeah, i know what is wrong there [08:31] great [08:31] 1 of the plugins links against gconf, which creats an exit handler [08:31] but the plugins get unloaded before exit [08:31] tell rodrigo_ so he doesn't spend time trying to debug it as well [08:31] so it crashes ;) [08:31] perhaps desrt has some pointers on what to do there [08:31] he will tell you "don't use gconf" I guess ;-) [08:32] lol [08:32] chrisccoulson, yes, that seems to be the cause for http://launchpad.net/bugs/788710 [08:32] Ubuntu bug 788710 in gnome-settings-daemon "gnome-settings-daemon crashed with SIGSEGV in exit()" [High,In progress] [08:34] chrisccoulson, you fixed bug #805908 right? [08:34] Launchpad bug 805908 in gnome-settings-daemon "multimedia keys not working" [High,Triaged] https://launchpad.net/bugs/805908 [08:34] seb128, yeah, that ones fixed [08:34] great [08:34] sorry, i should have checked for existing bugs ;) [08:34] bah, launchpad is ro [08:36] yeah, if you turn off the keybindings plugin, g-s-d doesn't crash on exit, as that's the only one still using gconf [08:36] that and the gsettings to gconf gateway plugin [08:38] seb128, oh, that one isn't activated by default here [08:38] hum, indeed, it's not there either [08:42] chrisccoulson, how was g-s-d handling the gconf handler in 2.32 when it was still using gconf? [08:45] seb128, i think that it just used to exit uncleanly when it lost the connection to the x server (so it wouldn't have unloaded plugins) [08:45] those issues on logout are annoying [08:46] they create lot of apport noise for no real issues [08:47] seb128, i guess that this would have broken it: http://git.gnome.org/browse/gnome-settings-daemon/commit/gnome-settings-daemon/?id=c3505969ad2bb61863965e49f04000e7773f3231 [08:48] well, I will not complain especially if that means gnome-session respawn it when needed ;-) [08:52] ups [08:53] don't ctrl-R in xchat [08:53] what happens? [08:53] blank bug day mail? [08:53] Laney, evo bug #809379 [08:53] Launchpad bug 809379 in evolution "Evolution is sending a blank body email" [High,Triaged] https://launchpad.net/bugs/809379 [08:53] fun [08:53] Laney, you can ctrl-u and read the source ;-) [08:54] it's a display issue [08:54] seb128: not sure I can, using mutt [08:54] oh, you mean the one from pedro [08:54] he resent it [08:54] bad moderators then :P [08:55] oops ;) [08:55] lol [08:55] oh indeed [08:55] Laney, the bugsquad list got the second email, not the announce one [08:55] chrisccoulson, tried the ctrl-R? ;-) [08:56] seb128, yeah ;) [08:57] RAOF, hey, is there known xorg breakage on intel with dual screen? [08:57] on current oneiric [08:57] seb128: I don't know of any; what are you seeing? [08:58] oh, i hope my dual screen setup isn't going to break ;) [08:58] i haven't restarted in several days [08:58] RAOF, it's not me but ronoc is having issues since he upgraded [08:58] ronoc, ^ could you describe how it's broken for you? [08:59] WFM, modulo compiz *still* being a pile of fail at hotplug. [08:59] RAOF, like bug #808677 [08:59] Launchpad bug 808677 in compiz "Failed to activate external monitor (dual monitor mode)" [High,New] https://launchpad.net/bugs/808677 [08:59] RAOF, is that a compiz issue? [09:00] smspillaz: While you're hear to shout at, when is compiz going to start drawing on *all* of my desktop after I plug a second monitor in? [09:00] or bug #808685 [09:00] Launchpad bug 808685 in compiz "Corrupted display after switching to external monitor (1 active monitor)" [High,New] https://launchpad.net/bugs/808685 [09:01] RAOF, well I guess those are rather compiz issues than xorg ones [09:01] seb128: Right. That's actually a combination of compiz *still* being totally broken, the crazy “Is your monitor OK” dialog getting continually bigger, and a genuine Xserver bug when the dialog reaches a critical size and crashes the server. [09:02] great combo [09:02] oh, i get the corrupted display when attaching an external monitor, but it does that in natty too [09:03] i always have to restart compiz after connecting it, and then it works [09:03] Yeah. Compiz 0.9 has hated XRandR since forever. [09:03] yeah, it's pretty broken [09:03] :-( [09:03] works in mutter though ;) [09:03] bug #795458 got fixed in the natty sru though [09:03] Launchpad bug 795458 in unity "Display garbled upon connecting external displays" [High,Fix committed] https://launchpad.net/bugs/795458 [09:03] chrisccoulson: Corrupt, or just black and any compiz drawing (like the window switcher) doesn't get cleared? [09:04] ronoc, wb [09:04] RAOF, i get compiz drawing correctly in a very small rectangle in one corner of one screen, and the rest of it is black as you describe (with any drawing not getting cleared) [09:05] ie, i can open an indicator menu, and it's contents get left on the screen after i close it [09:05] Yeah. That's compiz failing to be a compositing manager in those areas. [09:06] if i could figure out how compiz worked, then i would just go ahead and fix it ;) [09:06] it's just code, how hard can it be?! ;-) [09:09] i tried reading the compiz code once ;) [09:09] what's the bug ? [09:10] chrisccoulson: RAOF if that's happening you either hve output detection disabled or XineramaGetScreenInfo is returning some very odd values [09:10] smspillaz: Where you enable a second monitor and compiz doesn't notice that the desktop is bigger and so doesn't actually composite on that. [09:10] RAOF: are you running unity ? [09:10] Yes. [09:11] RAOF: I'm pretty sure this was a unity bug where the clip region gets reset incorrectly [09:11] This is also possible. Unity knows about the second screen; it draws the panel, dash, etc just fine. [09:11] RAOF: unity does some rather ... insane things [09:11] * RAOF doesn't really care. He *does* care that plugging in a second monitor doesn't work. [09:12] yep :) [09:12] although, I thought we fixed that [09:12] njpatel: ^ ? [09:12] RAOF: in other news, we should be able to handle the window size > mts case now [09:12] Yay! [09:12] RAOF: the plugin to do that was broken in natty but I fixed it [09:12] smspillaz, yes? [09:13] so we can probably enable it [09:13] njpatel: that bug where you plug a second monitor in and it's black, was that a) a unity bug where the clip region was reset incorrectly b) was it fixed ? [09:13] since I've heard a multitude of things about it [09:13] smspillaz, a) no idea, was on yours and jay's plate b) definitely not fixed, just happened to me five times [09:13] ok [09:14] - Display garbled upon restoring original resolution (lp: #795454) [09:14] - Display garbled upon connecting external displays (lp: #795458) [09:14] were fixed in a sru [09:14] Yeah, that's a different bug. [09:14] seb128, different bugs [09:14] right [09:14] RAOF: try seeing if you get it without unity on [09:14] so the other ones didn't get fixed [09:14] smspillaz: I'll do so later. [09:15] sandybridge is very, er, delicate at the moment [09:15] RAOF: :) [09:15] xv kills my xserver [09:15] njpatel: try nouveau [09:15] smspillaz, on intel? [09:15] move a window with metacity, oops, your X server died [09:15] ah [09:15] nice [09:15] njpatel: yes, try it on intel :) [09:15] smspillaz, okay, I'll dedicate today to making it work on intel ;) [09:15] njpatel: though it won't really have any effect [09:16] also, my laptop doesn't suspend all the time [09:16] njpatel: I did a dist-upgrade yesterday [09:16] NEVER AGAIN [09:16] it hanged halfway through [09:16] I just did it this morning, it went fine, I have one more thing working that yesterday [09:16] so I thought "well crap, force reset" [09:16] nope trashed my system [09:16] so I should be happy I guess [09:16] didnt' [09:16] event boot [09:17] had to do some interesting recovery with init=/bin/bash [09:17] smspillaz: Well, without unity it doesn't suffer that bug. If only because compiz crashes when I enable the second monitor and then respawns. [09:17] heh [09:18] Oh! Second time lucky. [09:18] if there is one good thing about running arch linux for a few months before I switched from fedora to ubuntu it was that you get to learn how to do these things because every single time you do a pacman -Syu world you have to reconfigure your entire system from scratch again [09:18] Second monitor enabled *and* compiz didn't crash *and* it's actually drawing on *both* heads. [09:19] RAOF: so I just hotplugged my monitor like 10 times and I didn't get the bug [09:19] and that's without unity [09:19] so unity bog it seems [09:20] oh, monitor hotplug is awesome [09:20] seb128: well, not conclusively [09:20] it could just be on certain hardware that compiz gets that [09:20] well RAOF gets the issue with unity and it seems to work without unity [09:20] seb128: I think I need to do more testing [09:20] like, checking if I get the issue with unity :) [09:21] since I don't usually hotplug this monitor [09:21] I should do some testing as well, my external monitor use is usually "boot with the laptop docked" [09:21] which works fine [09:21] smspillaz, if you know what it is, any chance of a quick mail to jay cc'ing jason and I so he knows what to fix/where to start? [09:21] yeah [09:21] njpatel: sure [09:21] smspillaz, thanks! [09:21] np [09:22] I wonder if there's a sane way I can work around this 32 character long key limitation in gsettings [09:23] that doesn't involve the use of --allow-any-key in glib-compile-schemas [09:23] smspillaz, did you check with desrt? [09:23] smspillaz, why do you need over 32 chars? [09:23] desrt: ^ :) [09:23] seb128: some of the option names in the plugin metadata in compiz are over 32 chars [09:24] maximize_window_horizontally_key [09:24] for example [09:24] O_o [09:24] I'm more curious to know why that limit exists [09:24] max-win-horiz-key!! [09:25] desrt: yeah, I'm not too keen on rewriting half the plugins so they'll work with one settings backend :) [09:25] i think it was a case of "32 ought to be enough for anybody" [09:25] dconf limits names to 65535 characters [09:25] so there's no technical limitation [09:26] i wonder if it was a windows registry consideration. i'm trying to dig through the history. [09:26] smspillaz, I get that on 3 different intel cards, arrandale, pineview and i915, but only with Unity. monitor hotplugging works fine on unity-2d [09:26] since if I change maximize_window_horizontally_key to max-win-horiz-key then I'd need to change bcop to use - instead of _ for delimiting [09:26] and then also change anthing like optionGetMaximizeWindowHorizontallyKey to optionGetMaxWinHorizKey [09:27] in the source [09:27] desrt: yeah, that wouldn't surprise me at all [09:27] GSettings: merge the schema compiler [09:27] doh. [09:27] ahahahah [09:27] i kept the branch around though [09:27] let me check there :) [09:28] clean up schema compiler [09:28] [09:28] enforce key name restrictions [09:28] duh. [09:28] not much better :) [09:28] desrt: well, maybe a better question to ask: is allow-any-name going away any time soon ? [09:29] you may not use it [09:29] because if that's the case I could just g_setenv ("GSETTINGS_BACKEND", "dconf") and be done with it [09:29] desrt: :( [09:29] smspillaz: the problem is that when you install schemas it's fine [09:29] but when someone else does it they're going to run the schema compiler without that switch [09:29] then you're in trouble [09:29] ah yeah [09:30] I mean, the brutal workaround here [09:30] hold on hold on [09:30] is to install all the schemas to $PREFIX/share/compiz/glib-2.0/schemas and then add $PREFIX/share/compiz to $XDG_DATA_DIRS [09:30] i'm doing a bit of research that may lead to a substantial lifting of the restriction [09:30]