[01:48] has anybody else run into a compiz issue with exiting from the expo view when unity isn't loaded? i can't seem to find a bug on it, but it's like compiz isn't repainting the desktop fully after exiting [01:49] http://web.mit.edu/broder/Public/compiz-expo-bug.png - look at the edges of both the screen and the terminal window [01:52] only seems to happen when unityshell isn't loaded, and i haven't triggered it with any other effect yet === ScottSanbar is now known as SanbarComputing === SanbarComputing is now known as ScottSanbar === ScottSanbar is now known as SanbarComputing [03:06] cjwatson, ping [03:16] is anyone reproducing bug 863119? [03:16] Launchpad bug 863119 in lightdm "~/xsession-errors is owned by root, incorrect consolekit session breaking DRI" [High,Fix committed] https://launchpad.net/bugs/863119 [03:16] (the CK/DRI breaking part) [03:17] jasoncwarner_: ping [03:17] jasoncwarner_: the stacking bugs that you get. do you have mumble open when they happen ? [03:17] * smspillaz may have fixed that one [03:21] ^^^ anyone else who still gets stacking bugs [03:21] broder: yeah, I am not sure what is causing that ATM [03:21] smspillaz: I had a bug earlier today where my launcher was behind all my other windows but after a restart I don't see it [03:22] jbicha: running mumble ? [03:23] smspillaz: nope, I don't have mumble, it was immediately after logging in after logging out of gnome shell [03:23] ok [03:24] jbicha: reproducible ? [03:27] smspillaz: looks like it [03:29] my launcher even manages to hide behind the desktop which is pretty weird [03:29] ok [03:29] it's inconsistent though, I don't know what triggers it & what clears it up [03:35] robert_ancell: I believe didrocks is reproducing that; he's a couple of hours away, though :) [03:40] argh. this tomboy logout thing is really annoying. does anybody have any clue of how mono events/dbus-sharp event dispatching works? [03:42] broder: Laney was looking into it yesterday. I wonder if he documented anything in backscroll? [03:42] oh, ok. i'll check [03:42] Bah, insufficient backscroll. [03:43] This seems like something I could help with, though. [03:43] i don't see anything [03:44] i managed to trace it to the point that DBus# doesn't think there's anybody queued up to handle QueryEndSession [03:44] but i don't know enough about mono's managed event signal-alikes to really follow it much further [03:44] there is a DBUS_VERBOSE environment variable that will confirm that [03:44] Oh, so dbus-sharp is recieving the dbus message and is just not dispatching it? That shouldn't be hard to trace. [03:44] that was my conclusion [03:45] does mono...have a functional debugger? i couldn't find any documentation on one - that was going to be my next step [03:46] Monodevelop's debugger is pretty reasonable. [03:46] can it do cross-library stuff? [03:46] And with a couple of gdb macros you can get a pretty reasonable gdb session. [03:46] jbicha: ok, fingers crossed it's fixed [03:47] * smspillaz has to go [03:48] smspillaz: thank you [03:58] broder: Ok, I see what you're seeing. [04:08] Good morning [04:11] smspillaz: not sur eif it is mumble or not...I didn't notice when I was having the issues. However, I haven't had any (knock on wood) issues this weekend with stacking. [04:11] morning pitti :) [04:11] good long weekend? [04:11] Morning pitti :) [04:12] jasoncwarner_, RAOF: hey, how are you? [04:12] yeah, we were in Dresden again, and met family and friends again [04:13] Bopping along to The Dandy Warhols ? 'Cause everydaa aa aay should be a holiday ? [04:13] pitti: nice [04:14] chromium builds stopped a month ago [04:14] jbicha: yeah, fta left, we'll get them back online later this month, stable will be updated this week [04:16] micahg: thanks [04:18] argh, another compiz crash [04:18] after only 20 minutes of running.. [04:25] ah, bug 863303 [04:25] Launchpad bug 863303 in unity "compiz assert failure: *** glibc detected *** compiz: free(): corrupted unsorted chunks: 0x0000000004a1d390 ***" [High,Fix committed] https://launchpad.net/bugs/863303 [04:27] broder: Aha, think I found it. [04:29] src/Connection.cs adds a match on Signal.Sender, which is ':1.194', but the handler is matching on .Sender = org.BansheeProject.CollectionIndexer. [04:29] Furthermore, this is fixed in trunk by not adding those matches. [04:30] (Although I guess it might want to be fixed by resolving the sender to the peer's bus name, but whatever) [04:30] trunk of dbus-sharp? [04:31] Yes. [04:31] But in two easily SRUable commits, which I'm cherry picking right now. [04:31] :) [04:31] let me know if there's anything i can do to help [04:31] What's the LP bug for this? [04:32] The fix is either a 0-day SRU or a rapid upload of dbus-sharp; I'd like to know which :) [04:32] bug #816950 [04:32] Launchpad bug 816950 in tomboy "Oneiric shutdown : Tomboy not responding" [High,Triaged] https://launchpad.net/bugs/816950 [04:34] Ok, it's got release-manager tracking happening. Time for a rapid upload. [04:34] :-D [04:37] are we going to make a brand new PPA for the Oneiric GNOME 3.2 stuff? [04:37] robert_ancell: hey, how are you? [04:38] robert_ancell: as the apparmor stuff was a bit urgent, I uploaded the patch on Friday already, I hope you don't mind? [04:38] no, of course not [04:38] robert_ancell: if you have any questions about the upstream merge proposal, I'm at your service :) [04:39] it should work fine for non-AppArmor systems as well, it just won't be protected then === Zdra is now known as xclaesse [05:10] bryceh: still awake? [05:12] bryceh: I reject the bash-completion upload; a debian-changes-1:1.3-1ubuntu5 patch crept in which partially reverts one of the patches; can you please fix/reupload? [05:12] hey pitti , I saw this on ubuntu one a week ago. wondering if you know why this might be happening for someone - http://twitter.com/#!/humphreybc/status/121088043485642752/photo/1 [05:13] jasoncwarner_: hm, clicking on the link just brings me back to the same twitter text page [05:14] seems you need to be logged in to look at it or so [05:19] pitti: oh...one sec...it basically says that someone can't ubuntu-bug unity b/c unity is a not a genuine ubuntu package [05:19] jasoncwarner_: presumably he got it from the PPA then [05:19] or he's just a little bit out of date [05:19] pitti: oh, interesting...hadn't realized that. [05:20] so if I have desktop-ppa, I can't submit bugs for those packages in the PPA? [05:25] pitti, ok [05:27] jasoncwarner_: right [05:29] pitti: does tracker qualify for the GNOME standing freeze exception? [05:30] wtf, debuild -S adds that debdiff in there [05:30] jbicha: not quite [05:30] er, debian-changes [05:30] bryceh: are you building from UDD? there's often trouble with pre-applied patches [05:31] yeah the packaging has some sort of magick in it [05:31] going to try starting over from scratch [05:35] bryceh: do you need your upload? [05:35] bryceh: you can dpkg-source it, quilt pop -a, drop the patch, and build again [05:36] don't you love it when your bios dies [05:42] pitti, ok reuploaded [05:42] bryceh: thanks [05:42] pitti, I don't need it myself; just was on patch pilot today and figured bash completion was a nice low-risk thing to work on [05:43] plus it's got tons of simple patches filed against it :-) [05:44] Morning === tkamppeter__ is now known as tkamppeter === rickspencer3__ is now known as rickspencer3 [06:07] good morning [06:07] Hey didrocks! [06:07] hey didrocks [06:07] There's a line forming to pester you with testing requests :) [06:07] hey RAOF, rickspencer3 [06:07] morning didrocks :) [06:08] RAOF: heh, you worked on the gnome-desktop3 check gl cras? :p [06:08] hey jasoncwarner_ [06:08] didrocks: And robert_ancell would like you to test lightdm, I believe :) [06:08] hey didrocks I logged a bug, thought you might like something to triage this morning: [06:08] bug #865913 [06:08] Launchpad bug 865913 in unity "Launcher Edge Activation Does not Work with Auto Run Apps" [Undecided,New] https://launchpad.net/bugs/865913 [06:08] I file bugs because I love [06:09] rickspencer3: sweet, I was just about filing it! [06:10] didrocks, yes, can you reproduce that bug? [06:11] robert_ancell: well, I get it quite randomly, yesterday, didn't get it before the 8th/9th reboot [06:11] didrocks, but you think it was caused by ~/.xsessions-errors having the wrong permissions? [06:12] robert_ancell: hum, I'm not sure, I didn't check that [06:13] bonjour didrocks, ca va? [06:13] smspillaz: your patch is quite invasive, I prefer that we postpone that as a SRU [06:14] pitti: guten morgen. I'm ok, thanks! And you, How was your week-end? [06:15] smspillaz: and you c-p-m patches containing still some debug printf… [06:22] does a g-s-d update like yesterday's require rebooting to go into effect & fix the bugs [06:28] robert_ancell: so, do I have something to test? [06:30] didrocks: pretty nice, thanks! were in Dresden again [06:31] pitti: enjoyed some nice weather there? [06:32] pitti: John is already warned that we can't without changing the translations retitle to "Ubuntu Destkop" for bug #865150 my pick would be just "Desktop" and remove the logo [06:32] Launchpad bug 865150 in ayatana-design "UIFe: Remove Ubuntu logo again from desktop title" [Critical,Fix committed] https://launchpad.net/bugs/865150 [06:33] pitti: not like if I didn't warn them about the confusion and the bfb when the first request was down, but I was turned down… [06:35] mvo: hey, how was you long week-end? [06:37] didrocks: You could test out gnome-desktop for me if you wanted to :) [06:37] RAOF: yes, I can try to do some reboot, but as yesterday, as it's pretty random, I can't say 100% "it's fixed" :) [06:38] didrocks: Oh, that won't fix it. Well, it will make it so that it doesn't segfault and will print something to check that I'm not mad in .xsession-errors. [06:39] Yeah, just keep an eye out when you're testing lightdm; it's not a critical bug. Basically the only side-effect will be crash files in /var/crash. [06:39] RAOF: so, I bet compiz will crash again then :) [06:39] if .xsession-errors is sick [06:40] Fedora's advanced partitioner is amazingly complex (reading Planet GNOME) [06:41] Yes. [06:41] Unnecessarily so, it seems. [06:42] didrocks: very nice and relaxing! good morning to you as well [06:42] yeah, can be more useful for a graphical server install, but for a client… [06:43] mvo: great! FYI, I reassigned you 3 piston-miniclient crashes that seems to be due to recent proxy changes [06:43] mvo: jibel has such a setup [06:43] didrocks: thanks, and *weeeh* [06:43] sorry ;) [06:44] didrocks: no, just the opposite, thanks a lot for this! [06:44] mvo: the ascii plage is spreading to pitti as well bug #865394 [06:44] bot? https://bugs.launchpad.net/ubuntu/+source/apport/+bug/865394 [06:44] didrocks: Error: bug 865394 not found [06:44] didrocks: did jibel setup work before and the proxy change broke it? [06:45] mvo: I don't know at all, he can easily trigger it with just trying to sync oneconf-service [06:45] mvo: I just got all those 3 bugs this week-end [06:46] didrocks: ta [06:46] yw :) [06:46] mvo, didrocks are you talking about bug 863717 ? [06:46] Launchpad bug 863717 in piston-mini-client "oneconf-service crashed with HTTPError in __negotiatehttp(): (403, 'Forbidden')" [High,Confirmed] https://launchpad.net/bugs/863717 [06:46] jibel: exactly [06:47] cjwatson, ping [06:47] mvo, I get it everytime I update the package cache with apt-get update for example. [06:47] jibel: good morning! [06:48] mvo, and good morning :) [06:48] jibel: thats good as it means it easy to reproduce :) what is your "echo $http_proxy" output ? anything "unusual" about your setup? do you use squid-deb-proxy? [06:49] didrocks: ah, I think they should be removed though in later patches. I'll refresh them [06:50] smspillaz: do you think that bug #865913 can be a stacking issue? [06:50] Launchpad bug 865913 in unity "Launcher Edge Activation Does not Work with AutoMaxized Apps" [High,Triaged] https://launchpad.net/bugs/865913 [06:50] didrocks: nope [06:50] it's the hide machine then? [06:50] didrocks: probably. It's best to check with other edge triggering actions [06:50] and last time we try to poke the debug, it crashes, just saw a workaround that lamalex committed :/ [06:50] smspillaz: yep, I need to add one though [06:51] mvo, http_proxy = http://nnn.nnn.nnn.nnn:nnnn/ (with digits instead of Ns of course) [06:52] mvo, it is a squid proxy without auth and that transparently redirects deb requests to apt-cacher-ng. [06:52] didrocks: it seems to work fine here fwiw, but I'll need to update my other system and check [06:53] smspillaz: multiple people get it randomly [06:53] like I said, I will update my other system and check. I don't know my way around the launcher hide machine that well though [06:54] * smspillaz is better at fixing things like ... stacking [06:54] jibel: thanks, let me look at this for a minute too see [06:54] mvo, in the proxy logs there are a couple of 403 "TCP_MISS/403 448 GET http://archive.ubuntu.com/ubuntu/dists/oneiric/universe/i18n/Translation-en_US - FIRST_UP_PARENT/localhost text/html" [06:54] I don't know if it is related [06:55] pitti, I've got patches for lightdm to fix bug 862427 and bug 863119 and I'm waiting for cjwatson to confirm a patch for bug 864618. Should I upload the first two or wait for all three? [06:55] Launchpad bug 862427 in lightdm "lightdm-gtk-greeter does not set translation domain" [High,Fix committed] https://launchpad.net/bugs/862427 [06:55] Launchpad bug 863119 in lightdm "~/xsession-errors is owned by root, incorrect consolekit session breaking DRI" [High,Fix committed] https://launchpad.net/bugs/863119 [06:55] Launchpad bug 864618 in lightdm "UTF-8 locale no longer set" [Critical,In progress] https://launchpad.net/bugs/864618 [06:56] didrocks: I'll ask jason about it when he wakes up. Though, if you are able to reproduce it, can you post your xwininfo of the very left edge of the screen ? [06:57] smspillaz: I guess it's the compiz arc, jono already pinged Jason about it a week ago [06:57] that's where we discovered that the debug machine is broken [06:59] jibel: that dosn't look releated, does something show up in your apt-cache logs? I wonder if it maybe redirected you accidently? [07:01] mvo, nothing of interest. [07:07] * robert_ancell goes off for diner, but will be back to check up on bugs later tonight [07:09] robert_ancell: enjoy [07:11] didrocks: what do you mean by "the compiz arc" ? [07:13] smspillaz: IIRC, xwininfo qualifies this small window as "arc" [07:14] jibel: what does ' echo -e "GET http://reviews.ubuntu.com/ HTTP/1.0\n\n" | nc nnn.nnn.nnn.nnn NNN ' print? any useful error message or the html page of reviews? [07:17] didrocks: I suspect that #84280 and #865175 are not releated to the proxy, just "odd" network setups or hickups [07:17] mvo, HTTP/1.0 301 Moved Permanently http://paste.ubuntu.com/702107/ [07:17] mvo: ok, I didn't assign it clearly because of that. only jibel's one as it's known as being clearly a proxy issue :) [07:18] jibel: ohh, if you change http to https in the echo, what does it say? [07:18] didrocks: yeah, thats definitely one :) [07:27] mvo, I'm not sure how to do an ssl request with nc ? [07:27] oh, that is a good question [07:27] I guess I should actually look up the real url instead, hold on a minute or two [07:28] use openssl s_client [07:28] aha, thannks maxb! [07:28] openssl s_client -connect host:port [07:29] maxb, and to make this command pass through the proxy ? [07:31] Morning Desktoppers! [07:31] Sweetshark: morning... [07:32] er, proxy? [07:32] hey Sweetshark [07:33] robert_ancell: if it will take a bit longer, two uploads are fine; it's a small package [07:34] didrocks: yeah, I get these all the time; python 3 porting should fix these for good [07:34] pitti: I'm wondering why it happened recently though, what changed? Barry isn't sure [07:34] pitti: it makes crashing a lot of project for a month approx. mvo is impacted as well [07:36] perhaps we got more translations since then? [07:36] hum, perhaps, weird for projects that didn't change at all like sotware-properties-gtk though [07:37] (it was already 100% translated in French last cycle) [07:42] jibel: oh, sorry - I was amusing myself with IRC on my phone, on the bus to work, and didn't have full scrollback [07:42] nautilus segfaults if nautilus-open-terminal is installed? do I need to open a bug to request removing it from the Oneiric archives or is bug 865297 sufficient? [07:42] Launchpad bug 865297 in nautilus-open-terminal "nautilus crashed with SIGSEGV in gconf_client_get()" [Medium,New] https://launchpad.net/bugs/865297 === rickspencer3_ is now known as rickspencer3 [07:44] that's the latest release of n-o-t & there haven't been any git commits since release 7 months ago [07:44] maxb, np, thanks [07:45] jibel: Are you able to explicitly connect to the apt-cacher-ng directly? [07:45] morning [07:45] jibel: I know whats goind on and work on it now [07:46] mvo, ok great. finally here is the output of the request over https http://paste.ubuntu.com/702117/ [07:46] I guess n-o-t is just trying to ready gconf settings that don't exist any more [07:47] hey mvo [07:47] mvo, I have a gnome-desktop branch with a fix for https://bugzilla.gnome.org/show_bug.cgi?id=660482 [07:47] jibel: it appears the current httplib2 just supports CONNECT based proxying which TBH is rather silly for our use-case [07:48] Gnome bug 660482 in power "dpms interferes with g-s-d screen blanking counter" [Normal,Assigned] [07:48] mvo, can you test it? [07:48] rodrigo_: cool, yes, any debs already ;) [07:49] rodrigo_: have you seen comment #7 ? [07:49] mvo, no, no debs, you'll have to build it, as well as lp:~ubuntu-desktop/gnome-settings-daemon/ubuntu [07:49] mvo, no, looking [07:50] mvo, hence the g-s-d branch, it has a call to gnome_rr_screen_set_dpms_mode on startup [07:50] rodrigo_: aha, cool [07:51] mvo, ok, the branches are lp:~rodrigo-moya/ubuntu/oneiric/gnome-desktop3/fix_863038 and lp:~ubuntu-desktop/gnome-settings-daemon/ubuntu [07:54] seb128, hello [07:54] robert_ancell, hey! [07:54] robert_ancell, how are you? ;-) [07:54] robert_ancell, thanks for staying up, I guess you wanted to catch up on things? ;-) [07:55] seb128, trying to find a cjwatson. There's 3 bugs I want to get in before rc; bug 862427 and bug 863119 and bug 864618 [07:55] Launchpad bug 862427 in lightdm "lightdm-gtk-greeter does not set translation domain" [High,Fix committed] https://launchpad.net/bugs/862427 [07:55] Launchpad bug 863119 in lightdm "~/xsession-errors is owned by root, incorrect consolekit session breaking DRI" [High,Fix committed] https://launchpad.net/bugs/863119 [07:55] Launchpad bug 864618 in lightdm "UTF-8 locale no longer set" [Critical,In progress] https://launchpad.net/bugs/864618 [07:56] robert_ancell, those are basically the one I listed to RAOF yesterday, so seems we have the same list, good ;-) [07:56] robert_ancell, do you plan to upload 1.0.1 or to cherrypick? [07:56] seb128, have you heard of anything else critical that we should deal with? I expect to do at least one SRU, and provide a PPA for oneiric for those that have edge functionality that may not work perfect [07:56] seb128: bonjour [07:57] seb128, I've pushed the first two as patches into the ubuntu branch. I'm wary about taking all of 1.0.1 just in case there's any regressions [07:57] robert_ancell, let me look at the bugs assigneed to you, I assigned what I though was important [07:57] robert_ancell, the 1.0 commits seems fine to me but backporting works for me [07:57] pitti, hey, how are you? had a nice w.e? [07:57] seb128, yeah, I noticed you assigned me a LOT of bugs. I've unassigned the low priority ones [07:58] robert_ancell, sorry I didn't know how much subscribing would reach you, I used basically as a "robert, it would be nice it you could read this one and decide if it's important or not" ;-) [07:58] seb128, heh, you are sneaky like that [07:59] heh :p [07:59] robert_ancell, it's a service I'm providing you, stop complaining ;-) [07:59] salut seb128 [07:59] robert_ancell, bug #851612 is somewhat some we should fix at least in a sru if it still happens [07:59] Launchpad bug 851612 in lightdm "Logging out from a FUS session does not reliably return to VT7" [High,Triaged] https://launchpad.net/bugs/851612 [07:59] seb128, if we do take 1.0.1, it has the change to make the GTK greeter select the previous user on login - do you think that would need a FFE? [08:00] robert_ancell, otherwise your list seems to be those I had as well [08:00] seb128, yeah, there's that and the 100% CPU one, but I've no idea what's causing them [08:00] Good morning seb128 :) [08:00] lut didrocks [08:00] hey RAOF [08:00] robert_ancell, hum, good question [08:00] pitti, ^ seb128, if we do take 1.0.1, it has the change to make the GTK greeter select the previous user on login - do you think that would need a FFE? [08:00] there's also the multi-monitor stuff, which is not great, but the best we can do. We're really leaning on X to "do the right thing" here [08:01] robert_ancell, X should do the right thing! [08:01] I think the Xubuntu guys want that change [08:01] robert_ancell, if not you need to pay extra beers to RAOF until it does ;-) [08:01] robert_ancell, yeah, I would argue it's an important bug fix and not a feature [08:01] it sucks usability wise to always have the wrong user selected on boot [08:02] robert_ancell, let me look at the actual code diff [08:02] seb128: I need to leave a bit earlier today; I don't expect that we need to discuss much on the meeting, but would you be able to chair? [08:03] pitti, can do [08:03] seb128: merci [08:03] pitti, http://bazaar.launchpad.net/~lightdm-team/lightdm/1.0/revision/1220 <- now or sru? need a ffe? [08:03] pitti, yw [08:04] seb128, do you happen to know what hours cjwatson does? [08:04] robert_ancell, ok, so you fixed the xsession-errors owned by root [08:04] robert_ancell, it's 9am for him, he should be soon around I think [08:04] seb128, yeah, I don't see why it would case the CK issue though [08:04] robert_ancell, right, I was going to ask "what about ck" [08:04] which is the issue we care about ;-) [08:05] seb128, I've never seen it, didrocks said he saw it, but only after 9+ reboots [08:05] so it's likely that the ck issue is not fixed right? [08:05] seb128, I can't see a connection between the two [08:05] robert_ancell, jibel says he see it every 5 boots or something [08:05] robert_ancell: btw, did you ever talk to the security team about the three remaining items on https://blueprints.launchpad.net/ubuntu/+spec/desktop-o-lightdm ? [08:05] robert_ancell: well, as told, I'm not sure it's xsession-errors owned by root, RAOF think that the crash at startup I can get is because of it [08:05] rodrigo_: its installeld now, give me some minutes to test it (will require 10min of inactivity ) [08:05] robert_ancell: and we have more than 30+ similar bugs [08:06] jibel, didrocks, can you get a lightdm.log after the issue? I'd expect the log to show what CK said [08:06] robert_ancell: I'll try to reproduce today, but later [08:06] didrocks, np [08:07] robert_ancell, let me know the infos you need and I will try to get those [08:07] robert_ancell, I got the xsession-errors permission issue in a guest session yesterday but ck was right and 3d working [08:07] robert_ancell, so I'm pretty sure they are different issues [08:07] robert_ancell, trying now. do you need any other info ? [08:07] pitti, hello? [08:08] seb128, lightdm.log logs all the CK calls, and the responses from the CK daemon. So I'm hoping either lightdm incorrectly didn't call it, or CK returned an error [08:08] seb128: looking [08:08] pitti, thanks [08:08] jibel, the logs should be all I need [08:08] pitti, I asked micahg, and he put me onto someone else and then I forgot sorry [08:08] seb128, robert_ancell: remember last user> oh, I thought we already got that fix? it certainly works for me now [08:08] I think they're happy however [08:09] before that fix it always selected the wrong user, now it's "martin" by default [08:09] pitti, it's fixed for the unity-greeter [08:09] ooh wait, "gtk greeter" [08:09] pitti, that fix is for the gtk one [08:09] robert_ancell, seb128: fine for me to upload it now [08:10] robert_ancell, you should roll a 1.0.1 today still if you can and upload [08:11] seb128, ok, I'll do that [08:11] I just want to wait for cjwatson. [08:11] robert_ancell: could you merge the apparmor fix for that? [08:13] pitti, sure [08:13] mvo, ok, no hurry :) [08:14] robert_ancell, did you need cjwatson for the locale bug? pitti can maybe reply as well [08:14] seb128: I'll poke bug 840292 [08:14] Launchpad bug 840292 in nautilus-share "uses synaptic to install samba which is not on the CD" [High,Confirmed] https://launchpad.net/bugs/840292 [08:14] Actually pitti, you might know the issue cjwatson is getting - he now has LANG=en_GB set by lightdm. The lightdm logic is a) get a LANG from pam b) set LANG based on the user configured language (dmrc/accountsservice) c) set LANG and others in ~/.profile [08:14] pitti, thanks, I was going to ask mvo for review today of the aptdaemon code there [08:14] LANG is set by scanning the output of 'locale -a' and matching the first one. [08:15] I think he must have both en_GB and en_GB.utf8 in 'locale -a' which is why lightdm is picking en_GB (I don't in mine). So the proposed fix is to scan the list twice, once for lang.codeset and then for lang if it doesn't match. Does that make sense? [08:19] robert_ancell: I also have a few en_XX in locale -a [08:19] but these don't look valid [08:19] e. g. [08:19] en_IN [08:19] en_IN.utf8 [08:19] (I don't have en_GB, though) [08:19] robert_ancell: but parsing "locale -a" for setting $LANG doesn't make sense [08:20] robert_ancell: what is that supposed to do? [08:20] pitti, well I have a language, and I need to make a valid LANG value. But I have no idea what codesets are available [08:20] seb128: I'll sponsor the apturl one, as proposed by hyperair; it's a lot simpler [08:20] pitti: thanks. [08:20] pitti, ok, thanks [08:21] robert_ancell: ah, you parse it for the first line that matches the language, not just for the first line? [08:21] yes [08:21] still, shouldn't /etc/profile / /etc/default/locale already set the default $LANG, if we don't have one? [08:22] pitti, yes, but I have to override it if dmrc defines one [08:22] jasoncwarner_, bug #865990 [08:22] Launchpad bug 865990 in nux "compiz crashed with SIGABRT in raise()" [Undecided,New] https://launchpad.net/bugs/865990 [08:22] and then we override that in ~/.profile with the full locale settings [08:22] robert_ancell: so, en_GB would be the ISO-8859-1 variant [08:22] robert_ancell: they are not actually supposed to be built, but it seems they are for some reason [08:22] robert_ancell: would it be possible to scan /usr/share/i18n/SUPPORTED instead of `locale -a`, and only grab UTF-8 ones? [08:23] pitti, I'd like to just slap .utf8 on the end, but some codesets are still used right? [08:23] robert_ancell: e. g. en_IN is actually UTF-8 [08:23] robert_ancell: it's not that easy [08:23] robert_ancell: we only really support UTF-8 ones; I meant that you can't just append .utf8 [08:24] robert_ancell: /usr/share/i18n/SUPPORTED is the definition of valid locale names, and `locale -a` shows the ones that are built on the local machine [08:24] so I think you'd need both [08:24] well, actually I think you need neither :) [08:24] if the user doesn't have a locale set in ~/.profile, it should just use /etc/default/lcoale [08:24] pitti, it's only there because derivative don't have a method of setting ~/.profile [08:25] robert_ancell: ah, this happens for (b) only? [08:25] and I hope we can remove it in the future [08:25] i. e. from .dmrc/accountsservice? [08:25] yes [08:25] so, it's upgraded users that see it [08:25] and xubuntu etc [08:25] and GNOME shell [08:26] didrocks & smspillaz https://bugs.launchpad.net/ubuntu/+source/nux/+bug/865990 [08:26] Launchpad bug 865990 in nux "compiz crashed with SIGABRT in raise()" [Undecided,New] [08:26] robert_ancell: so, locale -a -v has the codeset, or you use /u/s/i18n/SUPPORTED to limit to UTF-8 codesets [08:27] robert_ancell: but perhaps your proposal would do for now as well, and be a lot simpler [08:27] robert_ancell: i. e. grep all '^ll_.*', take the first .utf8, and if it doesn't exist, take the first [08:27] pitti, yeah, it's really just a sticking plaster anyway [08:27] robert_ancell: at least that will cause a lot fewer error cases :) [08:28] pitti, so the only case where it might break is if you have two or more codesets, but I'm hoping that's pretty irrelevant [08:28] jasoncwarner_: was supposed to be fixed though… [08:28] jibel, hey, any luck getting the lightdm,ck,compiz not starting issue today? [08:28] robert_ancell: the only case I know there is sr_RS vs. sr_RS@latin [08:28] jasoncwarner_: adding to the target, asking for dx to look at [08:29] didrocks, btw I assigned you one bug and subscribe you to another one yesterday (iirc) [08:29] didrocks, the assigned is that the lens need a gettext domain in their .ini (to test if that's enough) [08:29] seb128, I'm currently trying to reproduce it [08:30] didrocks, the subscribed is dobey's launcher migration code, he put it in g-s-d to migrate ubuntuone-control-panel.desktop to ubuntu-installer.desktop ... dunno if you prefer to do that in the unity migration code? [08:30] didrocks, he did it in g-s-d [08:30] seb128: you subscribe me to this one, isn't it: bug #865593 [08:30] Launchpad bug 865593 in ubuntuone-client-gnome "Ubuntu One launcher disappears on upgrade to Oneiric" [Critical,In progress] https://launchpad.net/bugs/865593 [08:30] didrocks, right, just in case you would prefer do that migration in the unity script [08:30] didrocks, i.e I wanted your opinion on it [08:30] seb128: no, I don't think we want to support all those .desktop file change [08:31] ok, great [08:31] so feel free to unsubscribe [08:31] ;-) [08:31] seb128: btw, changing the gsettings key won't work [08:31] but well :) [08:31] didrocks, why not? can you comment saying so on the bug? [08:31] not sure dobey tested it… [08:32] didrocks, no, he couldn't boot his oneiric yesterday [08:32]