[00:01] robert_ancell: hi! did you have a question for me? [00:01] robert_ancell: thanks for fixing seahorse, btw! [00:02] mdeslaur, I just wanted to check which server your key was on, but then I realized the server filter was broken anyway :) [00:02] robert_ancell: ah! hehe [00:02] mdeslaur, I didn't backport it to oneiric - did you want to do that? [00:02] robert_ancell: meh, if someone asks for it, I will [00:03] robert_ancell: you really made my day, I had lost hope that that would be fixed for precise :) [00:04] mdeslaur, np. I think there's still some weird problems there - if I search for robert.ancell@ubuntu.com it does find a key but doesn't show it (but it does find it with my gmail address). But I figure it's better now that it was [00:05] robert_ancell: yes, definitely better :) [01:04] Darxus: We would not enable the wayland backend in an SRU, sorry. [01:05] Okay, thanks. [01:22] chrisccoulson, ping [01:28] ah, I hate paypal/credit cards so much. Why is it so hard to pay with an Australian card and send to a UK address? [01:30] robert_ancell, you international cyber criminal you [01:30] bryceh, I sure am [01:30] Yeah. Clearly up to no good! [01:34] (and why does my bank account still have to be frozen because some person in romania compromised it :( ) === jbicha is now known as Guest73631 [03:17] Guest73631: i love your new nick [03:18] bah, I hate it when that happens === Guest73631 is now known as jbicha_ [03:18] a little better [04:53] Good morning [04:54] cyphermox: I thought gvfs was fixed now wrt. obex [05:19] RAOF: guess I'm not the only one who typos numbers [05:21] smspillaz: Actually, now that I look (and check [05:21] ) again, it seems that directory has the right number. [05:22] *blink* [05:22] *shame* [05:22] never trust me with numbers :( [05:28] good morning [05:30] morning didrocks [05:30] hey smspillaz! [05:30] hope everything is sane today :) [05:31] smspillaz: I hope as well, I just got a crash at startup though :p [05:31] didrocks: I'm building a ppa with my other branches which have yet to be merged so we can get some better testing [05:31] seems to be in unity [05:31] I'll let you know when its ready so we can get it into staging or something [05:31] smspillaz: oh great! [05:31] urgh :( [05:48] ogra_: linaro uploaded nux without using the VCS as well and screwed the testing that happened last night [05:49] * didrocks is loosing again 30 minutes this morning to clean this mess when he should be doing other thing for the release… [05:50] process fail [05:50] I thought everyone should coordinate with you first? [06:13] oupss, didn't see that channel [06:13] pitti: well, especially when it's things we decided to not include in precise, nux and unity opengles [06:13] (compiz and c-p-m was the first steps) [06:14] but ok, it's all sorted now [06:14] didrocks: so reverting isn't a better option then? [06:14] didrocks: it seems to me, if that doesn't work anyway, circumvented the process, breaks the PPA and testing, it'd be better reverted? [06:14] but anyway, you are the master [06:14] pitti: for nux, it's ok, it's just about switching an option at build time on armel only [06:15] pitti: so doesn't really impact us :) [06:15] ogra_: ah, so the patch needed to be updated for c-p-m? Thanks! :) [06:16] hum, there are 2 -0ubuntu10, I'm fixing this :) [06:18] pitti: btw, can you upate to ppa:unity-team/ppa please? [06:18] pitti: just for a final check and no "omg it's broken" at the last minute :) [06:18] (if you have some time, of course!) [06:18] updating [06:19] pitti: various people on the french forum are reporting that they are seeing some softwares being in english right now, is that known? [06:19] (since yesterday evening) [06:20] no, not known to me [06:20] ok, let me try to get more info ;) [06:26] desrt, hey [07:16] pitti: all is fine for you in the ppa? [07:18] didrocks: argh, sorry, trying now [07:19] no worry :) [07:19] about to release (just merging a crash fix I had this morning) [07:20] didrocks: checkbox-unity tool has no icon in teh launcher [07:20] didrocks: first time I open the hud, keys still go into the void (as if it was not focussed) [07:21] otherwise a guest session looks ok [07:21] pitti: grrr, that wasn't from the user testing :/ [07:21] pitti: there is a known bamf issue which is around for a while [07:21] pitti: it's the last piece missing right now, but now, dbo looked at it and can reproduce it [07:21] unfortunatly, no news since [07:21] didrocks, did you know unity has all the gsettings paths set wrong? [07:22] as it appears most of the canonical projects do :( [07:22] robert_ancell: you mean /apps/desktop [07:22] yeah [07:22] robert_ancell: we need to transition that, but for Q [07:22] to com.canonical [07:22] didrocks: will restart my own session in a bit, I'm almost done with jockey [07:22] pitti: ok ;) [07:22] I'm trying to stomp on as many of these as I can, as the more there are the more people copy them [07:22] pitti: the HUD one is worrying me more [07:22] pitti: just tried to restart unity, don't reproduce [07:22] it's similar to what I told the other day about the dash [07:23] keys go into the nautilus typeahead search [07:23] robert_ancell: yeah, but not reasonable to do that without transition [07:23] didrocks: but it's not a regression from what I can see [07:23] pitti: yeah, focusing issue… [07:23] restarting session now, brg [07:23] brb [07:25] didrocks: same problems confirmed in my own session [07:26] didrocks: but the missing icon one doesn't affect the non-PPA, I guess [07:26] pitti: the missing icon is a bamf issue [07:26] pitti: known and under work, should have been done for today [07:26] but well… can't push the devs to communicate [07:27] pitti: we just know it's in libbamf right now [07:27] for the focus issue, I'll ask gord [07:27] pitti: before releasing unity [07:27] I can release the rest of the stack [07:52] good morning everyone [08:02] good morning chrisccoulson [08:07] hi didrocks, how are you? [08:07] chrisccoulson: I'm fine and busy with unity, and you? :) [08:07] didrocks, yeah, pretty good thanks [08:08] hey [08:08] hey chrisccoulson, how are you? [08:08] * RAOF stops frantically pushing the pointer against the screen edge and realises he's in a metacity session, not unity. [08:09] seb128, yeah, good thanks. how are you? [08:09] salut seb128 [08:09] seb128, oh, it seems my compiz freeze issue is a kernel bug ;) [08:09] RAOF: ahah ;) [08:09] chrisccoulson, not fully awake yet, went to bed late and overslept this morning but good I think ;-) [08:09] chrisccoulson, oh, is it? how did you figure it out? [08:09] lut didrocks [08:10] seb128, i didn't. Sarvatt noticed the warnings in my dmesg log though: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/979251/comments/2 [08:10] Launchpad bug 979251 in linux "Compiz hangs every time when undocking or changing display configuration" [Medium,Incomplete] [08:10] tbh, i probably should have looked in there :) [08:10] bad kernel, no cookie! [08:11] I guess I've the same issue, my laptop is frozen even second time I dock,undock it [08:11] I didn't dare docking,undocking too much while the laptop is running for a while, I usually suspend before :p [08:11] oh, mine is normally ok when i dock it. it's when i remove the external screen or turn it off that it dies [08:12] oh, you close the lid when you dock though, don't you? [08:12] hey seb128 [08:12] chrisccoulson, I do yes [08:12] yeah, so i guess it could be the same issue :) [08:12] pitti, hey, wie gehts? [08:13] chrisccoulson, now that you mention it I think I get the issue on lid close after docking [08:13] last time I let the lid open under the desktop just to avoid taking the chance it would freeze :p [08:13] heh [08:13] * chrisccoulson decides not to close his lid to test ;) [08:13] "didrocks@u.c has 400 fixes" [08:14] didrocks, you did reach the 400 mark! [08:14] chrisccoulson, ;-) [08:14] chrisccoulson, now you can probably enjoy playing ping pong with the kernel team bot every time they do an upload :p [08:15] seb128, yeah, that already started just a minute or so after Sarvatt assigned the bug to the kernel [08:15] (I hate reporting kernel bugs, nobody look at them but they auto incomplete it at every single upload for you) [08:15] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/979251/comments/3 [08:15] Launchpad bug 979251 in linux "Compiz hangs every time when undocking or changing display configuration" [Medium,Incomplete] [08:15] ;) [08:15] this is why i hate reporting kernel bugs :) [08:15] hehe [08:15] same here [08:16] seb128: gut, danke! [08:16] seb128: yeah, but I have 80+ pending and they won't count [08:16] seb128: that makes me sad :) [08:16] seb128: you fixed a lot of bugs yesterday, congrats :) [08:16] didrocks, I will cry :p [08:16] seb128: I hope so, it's so unfair! [08:16] hmmmm, i really hope my delivery of beer arrives today [08:17] pitti, thanks ;-) I think I'm mostly done, you can probably catch me today, I plan to do translation updates [08:17] seb128: chrisccoulson: +1, I have the ping pong on one of my bug for over 1 year now [08:17] and the other has no answer as well [08:17] i.e export po files for shared-mime-info, xdg-user-dirs, etc [08:17] same prob reporting xorg bugs? [08:18] bryceh, I didn't report xorg bugs for a while but I don't remember you guys setting the bug to incomplete at every single upload without having an human ever reading your report [08:18] it's pretty frustrating when you describe an obvious bug to have to fight a bot to not have your ticket closed [08:18] pitti: fyi, I asked gord to look at the focus issue [08:19] pitti, sooooo [08:19] * pitti holds his breath [08:20] pitti, gedit lost its french translations with yesterday's langpack update [08:20] I'm an unhappy french user today :p [08:20] seb128: a lot of french users are reporting that [08:20] weird; there was only an update pack [08:20] the -base packages didn't change [08:20] seb128, yeah, agreed. their bugbot derives from mine, but I've toned down my use of it quite a bit. I think they've toned theirs up. admittedly they get a lot of bugs though... [08:20] $ dpkg -S gedit.mo [08:20] I told pitti about it :) [08:20] language-pack-gnome-en: /usr/share/locale-langpack/en@shaw/LC_MESSAGES/gedit.mo [08:20] language-pack-gnome-en: /usr/share/locale-langpack/en_AU/LC_MESSAGES/gedit.mo [08:20] language-pack-gnome-en: /usr/share/locale-langpack/en_CA/LC_MESSAGES/gedit.mo [08:20] language-pack-gnome-en: /usr/share/locale-langpack/en_GB/LC_MESSAGES/gedit.mo [08:21] pitti, I bet you lost the german translations [08:21] indeed I did [08:21] $ dpkg -L language-pack-gnome-de | grep gedit [08:21] $ [08:21] $ dpkg -L language-pack-gnome-en-base|grep gedit [08:21] $ [08:21] WTH [08:21] pitti, it's in language-pack-gnome-en [08:21] pitti, not -base [08:21] pitti: see, you should trust the french forum! :) [08:21] right, but why would it not be in -base? [08:22] pitti, that's the million dollar question! [08:22] ;-) [08:22] so, we'll get a fresh -base export next Tuesday [08:22] people reported terminal, synaptic, gnome-monitor-system [08:22] gnome-system-monitor even :p [08:22] that ought to fix it [08:22] pitti, are we sure gedit will be in? [08:22] terminal is fine here [08:22] didrocks, synaptic is a different issue, it was langpacked this week only, i.e made to use dh_translations [08:23] ah ok ;) [08:23] so it needs a new export I think [08:23] ./sources-base/language-pack-gnome-fr-base/data/fr/LC_MESSAGES/gedit.po [08:23] seb128: yes, as the current export has it [08:23] pitti, ok, good, I hope we will not miss any [08:24] I feel uneasy that this issue is happening now [08:24] if the next export is screwed it's getting late for precise [08:24] seb128, didrocks: sudo apt-get install --reinstall language-pack-gnome-de-base [08:24] it's not the export by itself [08:24] it seems the Replaces: is playing tricks on us again [08:24] trying [08:24] -base originally had it [08:25] then it seems a followup update pack _also_ had it, overwriting the -base one [08:25] and then the next update after that dropped it again [08:25] the last step Should Not Happen(tm), though [08:25] didrocks: I tried it, it works [08:25] wfm as well [08:25] pitti, danke [08:25] you have some speedy computer :) [08:25] it could be that the second-last update was built against a wrong/old -base version or so [08:26] mine is still trying to install :p [08:26] didrocks: 1.055 kB/s *shrug* [08:26] didrocks: oh, that just took some 5 secs here [08:26] pitti: not the download, here, installing a package takes ~40s [08:26] ssd for the win [08:26] I really think my hd is near to death [08:26] * pitti ^5s seb128 [08:26] * pitti proposes foundations-q-usr-on-tmpfs [08:27] ;-) [08:27] ok, gedit is fine as well [08:29] seb128: what do you think of fixing https://bugs.launchpad.net/ubuntu/+source/librsvg/+bug/931802 by doing what Debian did and adding a symlink + some compat magic? [08:29] Launchpad bug 931802 in librsvg "/usr/bin/rsvg disappeared" [High,Confirmed] [08:29] micahg, do we still have anything using rsvg? I though we fixed imagemagick [08:30] micahg, we had an archive rebuild since and no build issue [08:30] seb128: have you looked at all the failures yet? I haven't had a chance [08:30] micahg, I'm unsure we should add a compat layer for something which was deprecated for years and nothing in the archive is using [08:30] seb128: oh, if nothing is using it, that's fine, I was under the impression we still had some failures [08:31] micahg, I will double check, thanks for pointing it, but I think we are fine [08:31] ok, thanks [09:11] didrocks: ah, not using -proposed this time? [09:11] pitti: see #ubuntu-release [09:11] 09:28:28 didrocks | so starting to push the whole unity stack (we will maybe have a bamf release a little bit later) including compiz and [09:11] | c-p-m. There is no ABI break for those, so I can push them directly [09:11] 09:28:40 didrocks | then, for nux/unity/unity-2d I will use -proposed once the release is ready [09:11] didrocks: will the new compiz / c-p-m break current unity? [09:11] pitti: no [09:12] didrocks: ah, thanks [09:33] didrocks, pitti: there is a potential weirdness (or it's my box being weird) with compiz in guest session classic [09:33] didrocks, pitti: I get an "empty" compiz profile here [09:34] like compiz is running but empty of any option, i.e no decorator, no move plugin activate, no resize, no ws switcher [09:34] well it's only when doing a "classic" guest, not a frequent usecase [09:36] seb128: not on your classic normal session? [09:37] didrocks, no, let me try with a new user [09:37] didrocks, maybe apparmor is blocking some profile reading or something for guest [09:37] weird that it works with the unity profile though [09:40] didrocks, same for a new user [09:40] seb128: what your plugin list is telling you? [09:40] I bet it's the default hack that was requested by duflu [09:41] didrocks, what is the gconf key again? [09:41] ok /apps/compiz-1/general/screen0/options/active_plugins [09:42] yeah [09:42] this one and the compizconfig-1/Default … [09:42] seb128: ^ [09:42] Aucune valeur définie pour « /apps/compiz-1/general/screen0/options/active_plugins » [09:42] and the one is the default profile? [09:42] /apps/compizconfig-1/profiles/Default/general/screen0/options: [09:42] active_plugins = [core,composite,opengl,compiztoolbox,decor,vpswitch,snap,mousepoll,resize,place,move,wall,grid,regex,imgpng,session,gnomecompat,animation,fade,workarounds,scale,expo,ezoom] [09:42] weird [09:42] indeed :/ [09:43] so compiz isn't copying for the default profile [09:43] and I was right, the default plugin list is important [09:43] I can't put it it back though, as it's what did the locking [09:43] so need duflu or sam on this one [09:44] didrocks, get going with your uploads, that's not impacting unity [09:44] we can sort that in a second round [09:44] seb128: right, continuing, can you sum up that to him? :) [09:44] yes [09:44] duflu, hey [09:44] thanks seb128 ;) [09:44] * didrocks hugs seb128 [09:44] * seb128 hugs didrocks [09:44] hey seb128 [09:44] duflu, so with the current compiz stack "gnome classic" for new profiles has a weird compiz config [09:45] duflu, compiz runs, but none of the import plugins are activated [09:45] like no resize, no move, no decoration [09:45] import(ant)? [09:45] no workspace switcher [09:45] no alt-tab [09:45] * duflu tries gnome classic [09:45] it's like compiz core only was running and failing to activate any plugin [09:45] gconf has [09:45] /apps/compizconfig-1/profiles/Default/general/screen0/options: [09:45] active_plugins = [core,composite,opengl,compiztoolbox,decor,vpswitch,snap,mousepoll,resize,place,move,wall,grid,regex,imgpng,session,gnomecompat,animation,fade,workarounds,scale,expo,ezoom] [09:46] No value set for `/apps/compiz-1/general/screen0/options/active_plugins' [09:46] [09:46] duflu, didrocks think it's a side effect of the "no default plugin" change done to workaround the locking issue [10:11] pitti, hey, didrocks suggested you might be able to try a fix for the dash/hud input focus thing? lp:~gordallott/unity/horribleworkaroundfordashinputwindow if your willing to build locally [10:11] pitti: can you build it or do you want we try in a ppa first? (but that will take an hour the time for the bot to build it, push to the ppa an so) [10:11] gord: it always takes a while until I can reproduce it again, but I can build it here and then run with it for a while [10:11] seb128, did you have a chance to get the log of the nothing-in-dash issue? [10:12] mhr3, no, it happens only on session start and I'm not quick enough to open a g-t and run a bustle, by the time I do it the session is loaded :p [10:12] gord: oh, the input focus, not the HUD coming up; sure [10:12] yeah [10:12] gord: your hack won't start the lenses on startup? [10:12] seb128, /etc/xdg/autostart ? ;) [10:13] didrocks, nope, i was careful about that [10:13] mhr3, let me try [10:13] ok ;) [10:13] gord: I can reproduce this quite reliably, so sure [10:13] gord: it's nasty, your are nasty ;) [10:13] gord: it also affects the hud, BTW [10:13] what's the bug? focus issue if you type before the hud is being displayed? [10:13] pitti: if you build it, please double check that the lenses aren't start just after login please [10:14] i.e too quick to alt type? [10:14] pitti, yeah, thats also included in it, just didn't want to make the branch name even longer [10:14] seb128, nope not that [10:14] input not going to hud/dash when you open it the first time sometimes but not always [10:14] that's an upstream-ish branch; how can I build and pkg-ify this? [10:14] ok [10:14] gord, didrocks: ^ [10:14] pitti, bzr branch lp:~ubuntu-desktop/unity/ubuntu [10:15] cd ubuntu [10:15] bzr merge lp:~gordallott/unity/horribleworkaroundfordashinputwindow [10:15] bzr bd [10:15] ah, and then merge gor'ds branch [10:15] sure [10:15] cheers [10:15] yw ;-) [10:15] full source upstream and distro in bzr for the win ;-) [10:15] packages? just make install into /usr! [10:16] hum [10:16] pitti: ensure you have the nux -dev package from the ppa [10:16] gord, /usr/local so you can complain that nothing works 3 months later :p [10:16] gord, dx like to do that ;-) [10:17] and they like to poke me about it [10:18] like "I installed the ppa and it's not working, your packages are totally broken" [10:18] "did you have something on /usr/local?" [10:18] worked 100% of the cases from my stats here :) [10:18] didrocks: sorry, which PPA was that again? https://code.launchpad.net/~unity-team/+archive/ppa ? [10:18] pitti: yep [10:22] gord: building [10:23] why does apport say that aptd crashing is a dupe of an ancient bug (639616) and refuses to report it again? [10:23] that bug being closed 2y ago [10:24] apport bug [10:24] ok [10:24] if you could report that and attach the .crash file and pointer to existing bug, that'd help [10:25] ok I will [10:47] pitti: enjoying how unity is long to build? ;) [10:57] didrocks: it finished some minutes ago in fact [10:57] pitti: waow, I'm envious of your hardware then :) [10:59] gord, didrocks: so, that branch seems to fix it, or make it massively better (tried 6 times in a guest session) [10:59] didrocks: I built with -j4 [10:59] gord: confirming that lenses are not started by default [11:00] pitti: if I do that, I can ensure you that my laptop will go to the hoven just after :) [11:00] \o/ [11:00] didrocks: I didn't pay close attention, but was something like 15 mins [11:00] didrocks: how long does it take for you? [11:00] ~45 [11:00] actually, stat'ing the build log tells: [11:01] Access: 2012-04-12 12:21:52.700393898 +0200 [11:01] Modify: 2012-04-12 12:43:08.152455627 +0200 [11:01] 22 mins [11:01] not bad :) [11:01] not to show off or anything, but it takes me about a minute ;) [11:01] can certainly become faster with ccache [11:01] okay maybe to show off a little [11:01] gord: tsssssssss [11:01] gord: pre-built tree, I guess :) [11:01] didrocks, you want a merge request for this then [11:01] ? [11:01] gord: nice work! [11:01] gord: yeah, but please add some /* FIXME */ [11:02] didrocks, yeah thats the plan [11:02] gord: sign off with your blood that you did that, because you had to! :p [11:02] lunch, bbl [11:02] gord: well done, thanks for the timely answer ;) [11:02] pitti: enjoy === MacSlow is now known as MacSlow|lunch [11:11] pitti: thanks for testing btw :) === tkamppeter_ is now known as tkamppeter === mbiebl_ is now known as mbiebl === ogra is now known as Guest6312 === Guest6312 is now known as ogra_ [11:49] pitti, there? === MacSlow|lunch is now known as MacSlow [11:57] installed ubuntu 11.10 desktop and hooked it up to my 32" lcd, it seems the screen is larger than the lcd, and the trashcan is off the screen [11:57] how can I make the desktop fit on the screen [11:58] yaboo, hi, try #ubuntu for user support questions [11:59] seb128 thanks [12:02] seb128: back now [12:03] pitti, so just to check, admin users are either member of "admin" (upgrades) or "sudo" new installs? [12:03] seb128: yes [12:03] seb128: btw, I re-closed the gnome-menus task; the patch is already disabled, so let's keep it that way [12:03] pitti, g-c-c has a similar group handling issues, didrocks is updating the patch to fix it [12:03] jbicha did update it recently but in a buggy way [12:03] argh, another one? I thought we caughht them all [12:04] (did I say that I hate such incompatible changes) [12:04] building the fix with the other g-c-c change :) [12:04] pitti, yes, I wonder how many are still around that we didn't catch [12:06] pitti, does it mean we need to keep the "check for admin" compat patches for ever? [12:06] seb128: unless we do a migration of the old group members, I guess so [12:06] :-( [12:06] I think at some point we should do that [12:06] i. e. add all admin members to the sudo group [12:06] we shouldn't have renamed to start with, but I'm guess I'm late to argue on that :p [12:07] no argument from me :) [12:07] yeah, it seems more reasonable to transition at some point [12:07] I was quite surprised by it, too [12:07] or we will have tons of diff for nothing [12:07] didrocks: locally that's no problem, I just wonder how that works with NIS/krb/smb/whatever other network DBs [12:07] pitti: indeed, it's tricky when you go outside /etc/passwd ;) [12:32] seb128, pitti: I tracked down the bug in gstreamer for bug 966294 [12:32] Launchpad bug 966294 in ubiquity "Ubiquity loops forever from ubiquity_webcam_play" [High,In progress] https://launchpad.net/bugs/966294 [12:33] seb128, pitti: The following change fixes it here: http://paste.ubuntu.com/926326/ [12:33] essentially a revert of upstream f56c6e12255b37d75b1eb949e434fa8e3bb33f51 [12:34] though I don't actually know/understand gstreamer, so before I just upload a new gstreamer with that change reverted, I thought it'd be a good idea to let you have a look at it ;) [12:34] slomo_, hey, do you have an opinion on that ^ [12:34] I basically tracked it down bisecting from oneiric to precise on an affected machine [12:34] you are probably the one knowing gstreamer best here ;-) [12:34] oh right, especially because of: [12:34] commit f56c6e12255b37d75b1eb949e434fa8e3bb33f51 [12:34] Author: Sebastian Dröge [12:35] stgraber: ah, nice detective work! [12:35] ^ that's the commit I'm reverting (well, part of it) [12:35] seb128: let me check [12:35] slomo_, thanks ;-) [12:35] pitti: rather boring detective work, but yeah ;) took me almost 2 days to track it down to that and get a working version [12:35] did you open an upstream bug about the issue btw? [12:36] where's the useful information in the bugreport? :) don't want to read all the comments that also contain unrelated stuff [12:36] slomo_, comment #31? [12:36] and how could this commit cause any "looping"? [12:38] hm, do you have a testcase or something i could look at? or gstreamer debug logs? [12:38] slomo_: sure, starting cheese on any affected machine will have it freeze using 100% of CPU [12:39] slomo_: I can give you a GST_DEBUG=5 log of cheese starting up if you want [12:40] stgraber: please do :) cheese works fine for me (what's an affected machine? special v4l2 driver?) [12:40] slomo_: we haven't identified this clearly yet, but a large range of netbooks at least [12:40] they are using uvcvideo with v4l2 [12:40] but so are some working machines [12:40] so that's more specific than just that [12:41] oh, that's why the installer hasn't been working for me except in virtualbox; I had to use the alternate CD [12:41] yes [12:41] stgraber: i expect this to be a problem in camerabin(1) btw [12:42] slomo_: I did too but both camerabin and camerabin2 are affected and the issue also happens when I run oneiric's packages with just libgstreamer on the new version [12:43] so AFAICT it's not a bug that was recently introduced in camerabin that's breaking it, but it may be an existing bug that became visible with the new gstreamer [12:43] stgraber: oh with camerabin2 too? i'd be interested in a debug log for that too (and please file an upstream bug about this with the debug logs) [12:44] slomo_: yeah, ubiquity uses camerabin2 and that's where we found the problem initialy (and why the bug is marked as critical) [12:47] slomo_: I'm getting you the debug logs with cheese and ubiquity with/without my hack now. I'll attach the details on the LP bug and will get someone with an account to file one on the gstreamer bug tracker (I have other omg-critical bugs to take care of now ;)) [12:47] slomo_: just wondering, do you expect the world to explode with that revert? we have final freeze today and having a working installer would be kind of nice [12:52] stgraber: what's ubiquity again? [12:53] pitti: struct group *admin; admin = getgrnam ("admin"); printf ("group id: %i\n", admin->gr_gid); gives 27 to me [12:53] slomo_, the ubuntu installer [12:53] and 27 is the sudo group [12:53] the admin one is supposed to be 120 [12:53] didrocks: admin is not a static group, it's a dynamic one [12:53] stgraber: no, reverting that commit should be "safe" (for some definition of safe...). it could only cause suboptimal caps to be negotiated in the pipeline (thus give worse quality than possible or hurt performance), but it's always better than what you have now [12:53] didrocks: but yes, that looks seriously wrong [12:54] seb128: why is there camera support in the installer? :) [12:54] pitti: yeah, because when you cycle on the users group (old user) [12:54] you find 120, but not 27 [12:54] jbicha_: can you look at this please? https://bugs.launchpad.net/ubuntu/+source/gnome-shell/+bug/973845 [12:54] Launchpad bug 973845 in gnome-shell "network manager keeps trying to connect to random wireless connections when already connected to wired connection" [Undecided,New] [12:54] pitti: so no match [12:54] slomo_, because it takes a picture of you for your account [12:54] seb128: ah, nice idea :) [12:54] ;-) [12:55] didrocks: I still don't understand [12:55] $ python -c 'import grp; print grp.getgrnam("admin").gr_gid' [12:55] 118 [12:55] didrocks: how can this possibly say 27 for you? [12:56] pitti: just printing the value from g-c-c [12:56] /etc/group agrees as well [12:56] didrocks: your admin group really has ID 27? [12:56] that's supposed to be sudo [12:57] pitti: no, sudo is 27 [12:57] admin in 120 [12:57] that what is told by /etc/group [12:57] didrocks | pitti: struct group *admin; admin = getgrnam ("admin"); printf ("group id: %i\n", admin->gr_gid); gives 27 to me [12:57] and that's what I get when I cycle on my group list [12:57] pitti: yeah, that's what is wrong [12:57] it should gives 120 [12:57] if this code prints 27, then something is seriously screwed [12:58] stgraber: so... if you get my debug logs and an upstream bugreport i can try to look at it in the next days [12:58] didrocks, pitti: does calling getgrpnam twice works? [12:58] unless it's not the actual code [12:58] pitti, that code call it twice [12:58] checking for admin and sudo [12:58] i.e old and new group [12:58] hum, let me see the traces for the first call [12:58] seb128: the result is static [12:58] pitti, that's gnome-control-center [12:58] the proxy patch [12:58] weird, it's returning a strut, so should be fine [12:59] seb128: i. e. you can't call it twice for admin and sudo, and expect the former to still be intact [12:59] but let's see [12:59] pitti, seb128: Can one of you forward bug 966294 upstream for slomo_? I attached the debug logs now [12:59] Launchpad bug 966294 in gstreamer0.10 "gstreamer hangs when accessing webcam (on specific hardware)" [High,In progress] https://launchpad.net/bugs/966294 [12:59] didrocks: the return value is static memory and gets changed by the next call [12:59] ok [12:59] stgraber, pitti: I can do [12:59] so that's where the issues lay [12:59] didrocks: ok, *phew*, so that was not the entire code :) [12:59] heh ;) [12:59] pitti: yeah, sorry tried to stripped down, but too much [13:00] pitti, do you have code we can copy for the "check if user is admin"? [13:00] I think I'll have to call for each :) [13:00] seb128: it's easy to deal with, just a little longer [13:00] seb128: accountsservice should have it [13:03] done ;) [13:03] didrocks, \o/ [13:03] * pitti hugs didrocks [13:04] I wonder if my coming upload will increase CD space use [13:04] pitti might track me down if it does [13:04] compare the .debs against what's in the archve? [13:04] will do [13:04] well, they will likely be bigger, unless you build with pkgbinarymangler [13:05] pitti, I planned to do shared-mime-info xdg-user-dir etc builds with rosetta translation exports since those are not langpacked [13:05] I hope that will not add too much [13:05] should be trivial [13:08] PSA, I'm disabling gnome-keyring tests [13:11] * didrocks hugs pitti back, just tried in both guest session and new user and it works where we want it to work ;) [13:13] mterry, ok [13:13] slomo_, https://bugzilla.gnome.org/show_bug.cgi?id=673991 [13:13] Gnome bug 673991 in don't know "gstreamer hangs when accessing webcam (on specific hardware)" [Normal,Unconfirmed] [13:14] Humm, is it possible to tell compiz to update window content while resize is performed? I'd like to get rid of the annoying resize frame. :/ [13:15] seb128: thanks [13:15] slomo_, thank you for looking at it, I hope the bug has what you need, I tried to summarize what was useful from the launchpad log [13:15] BigWhale: ccsm :D [13:16] desrt, I'd like to do that just for my application ... :> [13:16] seb128: yes, the debug logs and knowing that the state change never returns should be enough information to start debugging :) [13:16] I guess there has to be some flag for that [13:16] BigWhale: i think you're out of luck [13:17] * BigWhale bursts into tears. [13:18] BigWhale, that seems like a bad idea to have one application behaving differently from everything else wm wise [13:19] but it makes your app look special ! [13:19] seb128, I know. But. I have a very good use-case: https://bugs.launchpad.net/kazam/+bug/979814 Something I wanted to implement from the beginning [13:19] Launchpad bug 979814 in kazam "record region should update dimensions on the fly" [Undecided,New] [13:20] BigWhale, you can probably calculate the pointer move? [13:21] without being the wm [13:21] I can do that alright. Not a problem, but I have to display the coordinates somewhere. :) [13:21] I'd prefer to have them in the same window. [13:22] I'll think of something. [13:32] kenvandine, how do I tell gwibber to stop spamming my .xsession-errors? [13:33] File "/usr/lib/python2.7/dist-packages/gwibber/microblog/storage.py", line 71, in maintenance [13:33] self.db.execute("VACUUM") [13:33] sqlite3.OperationalError: unable to open database file [13:33] kenvandine, it keeps doing that [13:33] seb128, ugh! [13:34] seb128, file a bug :) [13:34] i wonder why you don't have a db [13:34] there is a trick there, I know it :p [13:37] kenvandine, https://bugs.launchpad.net/ubuntu/+source/gwibber/+bug/979869 [13:37] Launchpad bug 979869 in gwibber "unable to open database file" [Undecided,New] [13:37] thx [13:37] kenvandine, yw [13:38] seb128, and do you have ~/.config/gwibber/gwibber.db [13:38] $ ls .config/gwibber/ -l [13:38] total 6928 [13:38] -rw-r--r-- 1 seb128 seb128 7079936 janv. 20 11:18 gwibber.sqlite [13:38] -rw-r--r-- 1 seb128 seb128 0 janv. 20 11:33 gwibber.sqlite-journal [13:41] so it is in some corrupt state [13:41] seb128, to stop the spamming, just move those out of the way [13:41] but i might want them to investigate later [13:41] so please don't delete them [13:42] kenvandine, ok, I just mv the dir [13:42] thx [13:42] kenvandine, thanks [13:53] pitti, hi [13:53] hello tkamppeter [13:54] pitti, is there any chance to get a fix for bug 968956? I get around two reports of duplicates for it per day. [13:54] Launchpad bug 968956 in apparmor "apparmor_parser: Unable to replace "/usr/lib/cups/backend/cups-pdf". Profile doesn't conform to protocol" [High,Confirmed] https://launchpad.net/bugs/968956 [13:56] pitti, tkamppeter: there is a pending apparmor upload that should fix that [13:57] jdstrand, OK, when will it appear? [13:57] tkamppeter: will be uploaded soon [13:57] tkamppeter: ie, today, before final freeze [13:57] jdstrand, OK. Thanks. [13:58] wtf @ bug 978509 [13:58] Launchpad bug 978509 in firefox "I dunno. I really don't. I'm a newbie. And think mister bill gates isn't perfect either. " [Undecided,New] https://launchpad.net/bugs/978509 [13:58] strangest description ever [14:00] lol [14:00] that's a new winner on the "how bad can bugs be" scale [14:03] * kenvandine heads out for about an hour [14:06] hate automake [14:06] Makefile.am:15: `pkgdatadir' is not a legitimate directory for `PYTHON' [14:07] pitti: I filed bug 979753 and added 'need-amd64-retrace' but still no retrace run [14:07] Launchpad bug 979753 in aptdaemon "aptd crashed" [Undecided,New] https://launchpad.net/bugs/979753 [14:07] tjaalton, the amd64 retracer has a > 60 bugs backlog (or had when I looked after lunch) and seems very slow [14:07] like it was taking 15 min by bug [14:08] urgh? [14:08] the load on the box was around 9 [14:08] it should be a minute or less [14:08] ah, ok [14:08] so I/O congestion [14:08] like even ssh takes 10s to give you a prompt [14:09] seb128: ah, ok [14:12] jani 20256 5.1 93.0 6490212 5687252 pts/3 Sl+ Apr11 70:55 ./6.out [14:12] that might explain it :p [14:12] I pinged jani [14:16] mvo: why does apturil recommend libgtk2-perl ? [14:17] pitti, debconf [14:17] ah, so similar to what synaptic had? [14:17] yes [14:17] I thought we had gotten rid of the gtk-perl stuff by now already [14:17] we still didn't get ride of that [14:17] we need to land cdebconf for that I guess [14:18] that stuff costs us 1.2 MB [14:23] slomo_, do you have any opinion on how risky it is for us to revert http://cgit.freedesktop.org/gstreamer/gstreamer/commit/?h=0.10&id=e6d2da7cf85c6fdd6b1c458d8a2c526246060051 or the part of it pastebined earlier? [14:28] pitti: yes, for debconfs gtk frontend [14:28] ok [14:29] who doesn't hate automake and can tell me how to tell it to be happy with "nobase_pkgdata_PYTHON" [14:30] i hate automake [14:30] but maybe i can help :) [14:31] 'sup? [14:31] desrt, Makefile.am:15: `pkgdatadir' is not a legitimate directory for `PYTHON' [14:31] ah [14:31] is what I get when running automake [14:31] using "nobase_pkgpython_PYTHON" works but I don't want to move those files from /usr/share to /usr/lib just before the hard freeze [14:32] that's a new restriction introduced by automake 1.11.2 or 1.11.3 (i can't remember) [14:32] yes... [14:32] there's an easy workaround [14:32] someotherdir = $(pkgdatadir) [14:32] nobase_someother_PYTHON = ... [14:33] desrt, thanks [14:33] * desrt balances karma [14:35] pitti: *grumble* I was taking care to clean that up again. [14:35] Sweetshark: but it keeps creeping back in? [14:36] pitti: btw, we will very likely need to include this in 3.6/p+1 because of https://launchpad.net/~ubuntutemplates [14:36] Sweetshark: yeah, in 12.10 we'll have bigger media [14:36] pitti: nah. I cleaned up how that was implemented. I could have spared myself that work. [14:38] pitti: btw, you are mean ;) [14:38] err [14:38] I know! [14:38] pitti: your little wanted patch it making our life hard (some weird crashes on sandybridges) [14:38] :) [14:38] didrocks: how did I hurt you this time? [14:39] didrocks: oh, the focus issue? [14:39] yeah, you are making compiz crazy in some case! ;-) [14:39] didrocks: well, I can live with it :) [14:39] didrocks: ITYM "slightly more crazy" [14:39] pitti: gord is trying to decipher a little bit more, but I think I'll revert it yeah [14:39] heh ;) [14:40] pfiou [14:40] I though for a minute didrocks was going to finger point my unity bug :p [14:40] seb128: it was between pitti's one and your :) [14:40] pitti: http://anonscm.debian.org/gitweb/?p=pkg-openoffice/libreoffice.git;a=commitdiff;h=dc456e24eac6f8f7ea212dba9252d63a38313eb9 [14:40] seb128: I accused your first of course :p [14:40] ;-) [14:41] seb128: danke sehr! [14:42] Sweetshark, ^ [14:42] seb128: reverting that commit should be "safe" (for some definition of safe...). it could only cause suboptimal caps to be negotiated in the pipeline (thus give worse quality than possible or hurt performance), but it's always better than what you have now [14:42] $ dpkg -L oneconf|xargs grep gtk [14:43] didrocks: ^ why does oneconf depend on python-gtk2? [14:43] slomo_, would you revert the full commit or just the part in the pastebin? [14:43] pitti: urgh, surely a leftover from the software-center plugin time [14:43] pitti: want me to remove it? [14:43] didrocks: let me check that it's the only rdepends [14:43] didrocks: just spotted on http://people.canonical.com/~ubuntu-archive/germinate-output/ubuntu.precise/desktop.depends [14:44] * pitti loves that file, kudos cjohnston [14:44] err, cjwatson [14:44] didrocks: ah, nevermind [14:44] not the only one? ;) [14:44] if nothing else, system-config-printer needs it [14:44] ok, I'll still drop it for next upload [14:44] seb128: what part in pastebin? [14:45] :-( [14:45] didrocks: yes, mostly that; but staging it in bzr can't hurt anyway, for seeing what's left [14:45] slomo_, http://paste.ubuntu.com/926326/ [14:45] cjohnston: I love status.u.c., too! [14:46] seb128: only the pastebin part will break stuff [14:46] seb128: revert the complete commit [14:46] slomo_, ok, thanks, good that I asked ;-) [14:47] pitti: ok, it seems that gord found a way to keep you happy. kudos to him. Testing under progress :) [14:47] pitti: heh... I actually haven't done anything with that for a while... :-( Summit has become pretty much a fulltime job [14:47] gord: more cookies for you! [14:49] chrisccoulson: sorry about that, it really is a kernel bug but i guess i should put it on x-x-v-intel so it gets attention anyway.. we'd ask you to test newer kernels first too though but not in the bug bot fight to the death style :) [14:49] pitti: how do you see that system-config-printer needs it on the previous page btw? [14:50] (it's not in the desktop.depends view?) [14:50] didrocks: I don't; I trieed to purge python-gtk2 in a live system [14:50] ah ok ;) [14:50] didrocks: you can see it in the rdepends/ tree, though; but that's harder to read [14:50] didrocks: http://people.canonical.com/~ubuntu-archive/germinate-output/ubuntu.precise/rdepends/pygtk/python-gtk2 [14:51] pitti: ah ok, the depends view will just show like the output of aptitude why <…> just one cause, not all of them [14:51] didrocks: yes, there is just one reason [14:51] * pitti off for a bit [14:52] pitti, should cups depend on appmenu-indicator? bug 979877 comment #3. [14:52] Launchpad bug 979877 in cups "package cups (not installed) failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,Incomplete] https://launchpad.net/bugs/979877 [15:38] slomo_: heya. Full revert is: http://paste.ubuntu.com/926588/ [15:38] slomo_: does that look sane? I had to revert by hand as there were too many changes in that part of the code for a simple reverse diff to work [15:38] (the change above builds fine and runs fine in my limited test environment) [16:05] pitti, ok, you win for precise ;-) [16:06] * desrt marks it down [16:11] seb128: still two weeks to go :) [16:11] tkamppeter: most certainly not, let me look [16:12] pitti, well, hard freeze and you scored like 10 bugs today [16:12] 4 on udisks, 4 on jockey and I stopped counting :p [16:12] seb128: you are 15 ahead on me on http://reports.qa.ubuntu.com/reports/bug-fixing/precise-fixes-report.html [16:12] I was trying to catch up a bit :) [16:12] but somehow this and the desktop-team charts are quite a bit different [16:13] pitti, ahah [16:13] indeed [16:13] * seb128 hugs pitti [16:13] ok, you win desktop, I win precise :p [16:13] well didrocks win both ;-) [16:13] * pitti hugs seb128 [16:13] :p [16:13] and I win people screwing the prod at the last minute ;) [16:13] not fun game! [16:13] as imprecise as this page is, it at least shows a nice increase, and it was a fun race :) [16:13] indeed [16:14] tkamppeter: no, that must be some other weird effect; cups does not use indicators at all [16:15] pitti, btw I uploaded shared-mime-info, based on previous cycle that should be a 25k increase in the non stripped part (the .xmls) [16:15] pitti, so that should be fine [16:16] pitti, thanks [16:19] time for some exercice, bbl [16:29] enjoy seb128! [16:38] good night everyone! === m_conley_away is now known as m_conley [17:35] so i am running apt-get dist-upgrade on my dell duo right now, and it wants to remove unity. [17:36] dobey: tell it 'no', wait a bit, try again [17:37] eh [17:45] dobey, do you use precise-proposed? [17:47] seb128: i don't know if it's turned on, but probably it is [17:48] dobey, you shouldn't proposed pre-release is used as a staging for builds [17:48] dobey, which is why you get this issue, just use precise, binaries will be pocket copied when ready [17:49] seb128: i suppose update-manager -d doesn't disable proposed repo then? it just migrates it to the new one? [17:49] dobey, I guess [17:50] shouldn't it disable proposed? [17:50] which makes sense for post release upgrades [17:50] well it makes sense if you don't use -d perhaps [17:50] well, using proposed as a staging is new and only during unstable cycles [17:50] "normal" users will not run into that issue [17:50] true [17:50] what you get there is just that what happened before we used proposed as a staging :p [17:50] we just moved the issue from the release to the proposed pocket [17:51] i'm going to start calling people who don't run the development version "abnormal users" [17:51] :) [17:51] ;-) [17:51] dobey, in any case infinity is looking at fixing those issues for good next cycle [17:51] ok [17:51] i.e patching apt and launchpad to deal with builds not in sync between archs [17:52] though, i don't think this is the first time apt-get dist-upgraded has wanted to remvoe unity for people. i recall some people having the issue before the proposed repo was opened up, no? [17:52] dobey, as said " we just moved the issue from the release to the proposed pocket" [17:52] dobey, it would happen to anyone before, now it happens only to those who have proposed enabled [17:53] right, ok [17:53] dobey, if you don't have proposed the binaries will be copied in sync between all archs to precise which avoids the issue [17:54] on that, dinner time [17:54] bbl [17:54] seb128: bon appetit [17:54] dobey, merci! ;-) [19:10] * didrocks waves good evening [19:19] ello [19:49] hellooo [19:49] anybody speack romanian? === m_conley is now known as m_conley_away [22:11] * desrt wonders why /usr/lib/lightdm/lightdm is in his $PATH [23:11] robert_ancell, morning, busy ? query wrt launchpad.net/bugs/949782 [23:13] ritz, can you make that into a merge request? [23:14] robert_ancell, sorry, new to this. is it this -https://wiki.ubuntu.com/UbuntuDevelopment/Merging ? [23:17] ritz, np, do a bzr push lp:~ritz/unity-greeter/bug-949782, then request a merge from https://code.launchpad.net/~ritz/unity-greeter/bug-949782/+register-merge [23:17] robert_ancell: hey. question [23:17] ritz, since this is an upstream project you just request to merge right into trunk [23:17] robert_ancell: why is /usr/lib/lightdm/lightdm in my path? [23:18] desrt, to run a special version of gdmflexiserver. Yes, it's a hack [23:18] I think we'll be able to drop that soon [23:18] robert_ancell: think it's possible to fix that up before release? [23:18] no [23:18] ouch [23:18] desrt, why? It's a required hack [23:19] couldn't you have done a diversion or something? [23:19] desrt, but what if you're running gdm? [23:19] good point, i guess [23:19] * desrt suspects gdm isn't working so great on ubuntu these days, though [23:19] I think it works fine [23:20] though I think the indicators dropped some support for it [23:20] hmmm, how does one merge request ? [23:20] why not have a different environment variable for GDMFLEXISERVER=lightdmflexiserver or such [23:20] and modify the site of use... [23:20] desrt, gdm does not do this, afaik [23:21] robert_ancell: I fixed the use of gdmflexiserver in gnome-screensaver...was there anything else calling it directly? [23:21] ritz, that link should have a form to submit it, otherwise go to https://code.launchpad.net/~ritz and find the branch [23:21] mdeslaur, I think gnome-shell will need to call it for switching too [23:21] robert_ancell: ah, yes, that is possible [23:22] robert_ancell, so, this https://code.launchpad.net/~khadgaray/+junk/unity-greeter.lp949782 ? [23:22] it was one of those cases where we couldn't know exactly how many things used it, so the safe option was to just provide backwards compatibility [23:22] ritz, yes, that works. But note you don't need to use +junk if you're branching a known project [23:22] sigh @ https://bugzilla.mozilla.org/show_bug.cgi?id=744847 :( [23:22] Mozilla bug 744847 in XPCOM "crash in nsAppShell::ProcessNextNativeEvent @ g_main_context_. with Linux kernel 3.0.0-18 on Ubuntu 11.10" [Critical,New: ] [23:23] desrt, any ideas for how to track down an issue like that? [23:23] desrt, any particular complaints apart from the yuckiness of it? [23:23] robert_ancell: well... i had PATH hardcoded in my .bash_profile [23:23] and it broke user-switching [23:23] desrt, well, that's just wrong to start with... [23:24] i consider it to be no worse than what you've done :) [23:24] robert_ancell: was it intentional that lightdm only tries to spawn X twice now instead of 6 times before giving up? [23:24] Sarvatt, no [23:24] desrt, your env is set via PAM, so your potentially overriding anything that PAM might have needed to do [23:25] anyway... mostly -10 points for style [23:25] that is my main complaint [23:25] desrt, agreed [23:25] robert_ancell: http://paste.ubuntu.com/927207/ seems to imply that it stopped attempting more than 2 tries before giving up sometime just after oneiric release [23:26] Sarvatt, that shows 6 X logs [23:26] yeah the dates though [23:26] i've hit failsafe many times since then [23:27] Sarvatt, note that it does recycle display numbers [23:27] though I wonder if it's run out of VTs or something [23:27] desrt, btw, have you seen the dconfalypse of everyone sticking stuff in /apps recently? I've been trying to file bugs to stop everyone copying the people doing it wrong [23:28] ugh [23:28] like who? [23:28] we really need an easy way to migrate broken configs - gsettings doesn't give enough info [23:28] desrt, and why is it still not documented anywhere what format the path should be? [23:28] * desrt ponders doing 'unfriendly' things to people who use /apps/ [23:28] robert_ancell, will this be a debian patch, or merged into main code base ? [23:29] ritz, main code base [23:29] hmm doesn't seem to be the case here at the moment, it does 0 and 1 then hits Xorg.failsafe.log, we're hitting some problems where nvidia isn't ready when X is started, and it used to work when it tried 6 times but now it doesn't [23:29] robert_ancell: from a practical standpoint we're talking about gsettings here, right? [23:29] actually i'm wrong, it's only doing Xorg.0.log then going straight to Xorg.failsafe.log [23:30] desrt, accerciser, beatbox, brasero, cheese, epiphany, gnome-nettool, gnumeric, indicator-session, notify-osd, onboasrd, seahorse, shotwell, telepathy-logger [23:30] christ... [23:30] desrt, yes [23:30] * desrt is thinking of a warning in glib-compile-schemas [23:30] "use of /apps /system and /desktop is deprecated" [23:31] desrt, and new projects like beatbox / geary are just copying what older projects are doing. So I've patched those ones and they don't need to worry about migration but things like cheese are a problem [23:31] and basically it seems all of unity has done it wrong [23:31] also in /desktop is some gnome crypto stuff, vino, gstreamer [23:31] migration is quite easy [23:32] and /system contains org.gnome.system.dns_sd [23:32] dconf dump /app/whatever/ | dconf load /org/proper/path/ && dconf reset -f /app/whatever/ [23:32] desrt, yeah, do you think we should encourage the migration to be done in the dconf layer? [23:32] i think i don't want a migration API [23:33] i want people to shell out to that ^^ [23:33] desrt, and if they weren't using dconf as a backend? [23:33] robert_ancell: that's hypothetical enough that i don't really care [23:34] desrt, ok, so I've files some bugs, please help squash them too [23:34] I'll do a PSA announcement [23:34] robert_ancell: i'm going to add the warning to the schema compiler [23:34] that'll make one hell of a lot of noise [23:34] i bet those ubuntu punks will vendorpatch it out of existence [23:34] desrt, also the gconf migration tool does the wrong thing last time I checked [23:34] oh? [23:35] robert_ancell, do I need to fill in the reviewer ? [23:35] desrt, It copies the gconf patch names which is probably the main cause of the problem [23:35] * desrt recently caught the migration tool doing the wrong thing and fixed it [23:35] path names [23:35] robert_ancell: oh... the schema migration tool [23:35] interesting [23:35] ritz, just leave it blank and it will assign it to the unity-greeter team [23:36] robert_ancell: btw: unity-greeter's config story is a bit of a nightmare [23:36] desrt, could you recommend the migration on https://bugzilla.gnome.org/show_bug.cgi?id=673965 [23:36] Gnome bug 673965 in general "Settings in wrong path" [Normal,Unconfirmed] [23:36] desrt, ? [23:36] you have a keyfile in /etc that's ignored entirely [23:37] desrt, yes, it was replaced with gsettings [23:37] meanwhile you have a dconf database in the lightdm user's homedir with no obvious way of editing it [23:37] desrt, you control it with overrides [23:37] robert_ancell: what if i'm a sysadmin? [23:37] desrt, is this the most official statement on what gsettings paths should be http://mail.gnome.org/archives/desktop-devel-list/2011-February/msg00064.html? [23:37] desrt, then you can set overrides right? [23:38] robert_ancell: sysadmins should not place schema overrides [23:38] robert_ancell: yes. that email is correct. [23:38] desrt, also ask mterry, he changed it [23:39] desrt, the main issue with the u-g config was derivatives couldn't easily override the config if it's all in a keyfile [23:39] robert_ancell: derivatives are fine [23:39] robert_ancell: they're just like distros, really -- overrides are meant for them [23:40] imho, sysadmins should not be mucking around in /usr [23:40] desrt, well, then you need a system dconf database... [23:40] ya. that's sort of what i'd recommend, in fact [23:40] empty by default [23:41] vendors can continue to set their preferred defaults in the schemas... [23:41] one day probably you will want some kind of automated tool for controlling these things [23:41] ...and one day i will finally add dconf-service system database support [23:42] \o/ [23:42] i want to spend this entire cycle working on dconf and gsettings improvements [23:42] wanna wager if i'm gonna get a chance to? :) [23:42] I'd like that too [23:43] desrt, btw, any change we can make dconf-editor into gsettings-editor anytime soon? Note that UDS is coming up and that is the traditional time to improve the code [23:44] desrt, depends on how much crack is going into 12.10, and being post LTS I suspect a good amount :) [23:44] that's one the list of stuff to fix :) [23:44] robert_ancell: did the wayland experiments get anywhere this cycle? [23:44] anyone step up? [23:44] desrt, nope [23:44] desrt, I'd like to focus on that for 12.10 [23:45] wanna wager on that one, then? :) [23:46] desrt, though I suspect the answer is going be the nvidia drivers screwing things up. I need to ring RAOF today and pick his brains over that [23:46] desrt, I think I have a good chance :) Putting together a proposal document for it now [23:52] * desrt has made the evil change [23:53] you're right -- the unity cats are just going to love this [23:54] warning: Schema 'com.canonical.indicator.session' has path '/apps/indicator-session/'. Paths starting with '/apps/', '/desktop/' or '/system/' are deprecated. [23:55] desrt, mwuhaha! [23:55] * desrt is gonna get shit for this [23:55] desrt, btw, didrocks says he's all over this. So they're going to try and migrate for 12.10 [23:56] good [23:56] desrt, the worrying part is seeing elementary and yorba just copying the same mistakes. And I'm sure there's the 3rd party apps we have no idea about doing it too [23:57] perhaps we should patch dconf-editor to have a warning on these keys - "this key is in using a deprecated name" [23:57] I used /apps for my app [23:58] mdeslaur, how did you decide that path to use? [23:58] pushed [23:58] may god have mercy on my soul [23:58] robert_ancell: I think I read a python gsettings tutorial somewhere that had that, and when I looked in dconf-editor, a bunch of other stuff was using it [23:59] mdeslaur, cool, that confirms what I suspected - I think patching dconf-editor will help there then [23:59] desrt, can you update http://developer.gnome.org/gio/stable/GSettings.html - it makes no mention of what format the path should be [23:59] robert_ancell: what are your plans for dconf-editor? [23:59] robert_ancell: sure [23:59] desrt, ignore it and hope it goes away? [23:59] "I think patching dconf-editor will help there" you said [23:59] desrt, nah, I'd like to convert it to the gsettings api, and then hope other people write patches instead of just complaining