=== Optichip` is now known as Optichip [05:02] Good morning [05:06] Good Morning. [06:07] I wonder if anyone has a free time slot to upload a g-s-d patch for me, bug 950160. [06:07] Launchpad bug 950160 in gnome-settings-daemon "Unity blocks other programs from binding globally to Super+* or Alt+* (* = any key)" [High,In progress] https://launchpad.net/bugs/950160 [06:12] * micahg leaves that for a desktopper === zyga-afk is now known as zyga === alan_g is now known as alan_g|afk === tkamppeter__ is now known as tkamppeter [08:00] hey [08:01] Good morning! [08:03] RAOF, hey, how are you? [08:03] Good! Yourself? [08:04] RAOF, good, I think, I'm not fully awake yet but I will be able to tell after coffee ;-) [08:05] Got the bottom of the SRU queue done, and am now meandering up the queue into some easier ones :) [08:06] RAOF, can I point gwibber? ;-) [08:07] RAOF, it's fixing harder what the current SRU half fixed [08:07] Heh [08:08] RAOF, thanks for the SRU work btw, and sorry for regularly ping you guys about the queue... [08:11] No problem; your pings for high-value SRUs are useful :) [08:12] RAOF, yeah, I still need to corner a SRU team member wanting to ack the libreoffice one ;-) [08:12] Looking at it now. [08:12] brave man! [08:12] It's... urgh. [08:13] RAOF, yeah, I think there is no way any of us can review that, we basically need to trust upstream and the testing Sweetshark got out of the ppa users [08:13] Yeah. === fenris is now known as Guest13007 [08:18] more people now, I wonder if anyone has a free time slot to upload a g-s-d patch for me, bug 950160. [08:18] Launchpad bug 950160 in gnome-settings-daemon "Unity blocks other programs from binding globally to Super+* or Alt+* (* = any key)" [High,In progress] https://launchpad.net/bugs/950160 [08:22] Anyone from the USC team here? [08:23] I would like to know if the toolbar used in the USC is custom made and if so, is there a lib/api available? [08:25] mvo will probably know, right? :) [08:26] try #software-center maybe? [08:27] but I guess it's a custom hack made in python in s-c and not a shared lib [08:29] seb128, software-center isn't on the list of ubuntu channels... so I didn't even bother to try it :> [08:29] kengyu, hey, did you subscribe ubuntu-sponsors to it? we will try to make sure it's reviewed this week but there is another SRU already uploaded so ideally we need to get that one validated before doing another upload [08:30] we can probably get the fix in quantal to start though [08:31] RAOF, if you are not done yet with SRUs, can you review the g-s-d upload in the precise queue? I would like to get that one rolling out to make space for another set of fixes to upload next week (I don't want to batch too much in one upload) [08:32] RAOF, I will be done with SRU requests for the day (maybe the week) then ;-) [08:32] seb128, yes, I did subscribe ubuntu-sponsors. thanks for the help. :-) [08:32] kengyu, yw, thanks for the work! [08:32] seb128, :-) [08:32] :) [08:47] seb128: Gah vala! I don't suppose it's possible to do the gwibber thing *without* valac generating ~7K lines of noise in the diff? [08:48] RAOF, hum [08:49] It makes reviewing the diff unnecessarily frustrating. [08:50] RAOF, that diff looks a bit weird to me [08:50] RAOF, feel free to skip it, I will sort it out with ken today [08:50] Ta. [08:57] There you go; have a gnome-settings-daemon SRU. [09:04] RAOF, thanks ;-) [09:11] not sure what changed recently (GTK?), but right now the theme seems pretty messed up to me; is that a local problem here, or just what quantal is right now? [09:13] If you mean square edges on everything, I think that's just quantal right now. [09:13] pitti: it's a known problem with gtk+3.0 3.5(.4) [09:13] I think the new light-themes yesterday was supposed to make it better, but not perfect. Or something. [09:14] buttons are still without rounded corners and menu entries disappear under a white box when hovering [09:14] * pitti will just be patient then, seems it's not a local problem [09:14] thanks [09:18] pitti, better to be patient, that's going to be broken for a while [09:18] btw, you should be able to sync glib2.0 from exp now [09:19] pitti, basically gtk 3.5 is continuing their css,theming work, dropping stuff which were available but not public but that unico is using [09:19] pitti, that breaks unico and Cimi, #ps have no time to allocated to fix it [09:19] Laney: \o/ [09:20] pitti, so it was either "we block on them for ages, or we unblock as we can and break a bit the theme while doing that" [09:20] we opted for the second option [09:21] The Cloud™ was quite nice for testing that [09:21] * Laney will be using it more [09:21] Laney, what is that? [09:21] Laney, btw did you figure your lightdm issue? [09:21] I just mean that I used cloud instances to test build and test run it [09:21] no [09:22] somehow when I restarted I could startx though, but it just exited almost immediately [09:22] Laney, you get a full desktop session in the cloud to test? or does that work? [09:22] just X forwarding [09:22] ah, I see [09:22] seb128: seems fine; as long as it's not too broken to actually use it [09:23] Laney, startx just returning, that's what I though it would be ... do you have the .xession-errors or log? [09:23] xsession-errors is empty [09:23] pitti, with the light-themes update from yesterday it should be good enough to be usable [09:23] let me get the Xorg log. It showed X starting up and then exiting but no obvious errors [09:23] Laney, even after restart? when you tried yesterday startx didn't work at all [09:23] yeah [09:24] * Laney boots it [09:28] seb128: hm, I have that version already [09:28] anyway, menus are not that crucial; there's the HUD, after all! :-) [09:28] seb128: http://paste.debian.net/175234 [09:28] it's probably just a test run to drive people towards the HUD, isn't it? [09:28] pitti, did you restart your session since you got it? [09:28] lol [09:28] pitti, shush, that's a secret :p [09:29] 2012-06-19 07:19:10 status installed light-themes 0.1.10-0ubuntu1 [09:29] indeed, I did't! [09:29] * pitti STFU, sorry [09:30] pitti, lol, I'm not saying that a restart will fix all the issues but maybe start by trying that ;-) [09:30] it should be better [09:30] Laney, what videocard do you have? [09:30] seb128: at least this thing feels a little bit like a dev release agagin! [09:30] oh, er, erm [09:30] hehe [09:30] It's a Macbook Pro 7.1, whatever they have [09:30] all this stability, meh [09:30] some nvidia thing [09:31] I switched to nouveau /after/ I got this problem, was using nvidia before [09:31] and anyway the login manager works … [09:31] Laney, ok, I was going to say that nouveau was in the upgrade run [09:32] Laney, do you have a .xsession? [09:32] "source ~/.Xdbus" [09:33] pitti: You could always use orca to read the menus to you. :p [09:33] Laney, hum, can you try to edit .xinitrc and put "exec gnome-session" in it? then try again the startx [09:34] doing [09:34] yep [09:34] that worked [09:42] Laney, that worked == you get a session? [09:42] seb128: that's right [09:42] Laney, but lightdm is still buggy? [09:43] i'm just dist-upgrading then will restart and see [09:43] ok [09:43] Laney, well at least it shows that the issue is not xorg [09:43] I guess the xinitrc won't make a difference for lightdm? [09:43] Laney, startx was just returning because it had nothing to run [09:43] i.e. I don't need to remove it [09:43] yeah [09:43] Laney, you don't no [09:43] I didn't know that's the file that startx executes to start a session [09:44] Laney, I know it uses it, that's how I use to change my desktop back in the years, putting an exec wmaker in there ;-) [09:45] hah, yeah — I must have known that too at some point :P [09:45] ancient unix knowledge bitrotting away [09:45] Laney, but what is weird is that I though that it would run a session anyway even without that file, here startx starts a gnome session [09:45] maybe there's a default somewhere [09:46] hum, no mlankhorst on this channel? [09:46] RAOF, still around? do you know what is the default session for "startx" if there is one? [09:48] * Laney reboots [09:52] ok, so same problem with lightdm [09:52] it affects all users, even the guest session [09:52] that's going to be fun to debug :-( [09:52] Laney, can you mv /var/log/lightdm/* away, restart lightdm, try to log in and copy the logs somewhere? [09:53] I would expect /etc/alternatives/x-session-manager [09:54] there is an xinitrc in /etc/X11/xinit [09:54] it comes down to run-parts on /etc/X11/Xsession.d [09:55] Laney, lightdm runs /etc/lightdm/Xsession [09:55] doesn't exist here [09:55] it lists the ones in /usr/share/xsessions though [09:56] Laney, ?! [09:56] hey [09:56] ls /etc/lightdm gives lightdm.conf unity-greeter.conf users.conf [09:56] Laney, ok, maybe that's a conffile leftover on my box [09:56] what 's going on? [09:56] mlankhorst, hey [09:57] mlankhorst, do you know what session "startx" is supposed to run in absense of a .xinitrc? [09:57] xterm probably [09:58] ok, weird [09:58] it runs nothing for Laney, startx just bails out [09:58] it seems to run gnome-session for me [09:59] Laney, ah, it's /usr/sbin/lightdm-session nowadays [10:00] http://people.canonical.com/~laney/lightdm/ [10:01] Laney, [+6.84s] DEBUG: Session 3690 running command /usr/sbin/lightdm-session gnome-session-xmonad [10:01] Laney, what is that xmonad? [10:01] another window manger i was using [10:01] Laney, does that command work or is available? [10:01] it's the session i chose to run, but also happens with unity [10:02] yes [10:04] Laney: what's exactly going on? [10:05] I get thrown back to the login manager after logging in, with any user including the guest session [10:05] startx does the same with no .xinitrc, but with one it functions as expected [10:06] tkamppeter: hey Till, how are you? [10:06] Laney, can you hack /usr/sbin/lightdm-session to echo "something" >> /tmp/lightdm.debug and see if that happens when you try to log in? [10:06] tkamppeter: I'm currently looking into Debian bug 640939 [10:06] Debian bug 640939 in cups "hplip-cups: asks for root password when cups updates PPDs" [Serious,Open] http://bugs.debian.org/640939 [10:07] tkamppeter: I already fixed the other RC bug in bzr, want to do an upload today (urgent to fix armel FTBFS) [10:07] tkamppeter: woudl you know a way for lpadmin to never ask for a password? [10:07] s/a/root/ [10:07] seb128: ok [10:07] it already does echo stuff though; do you know where that goes? [10:09] right then, that's some information [10:09] it doesn't get past the loop over $xsessiondir [10:11] Laney, we had issues in the past where a parsing error in any of those files were creating login fail issues [10:11] Laney, so my bet is that one of the file in that dir is buggy [10:11] it gets right to the last one [10:11] /etc/X11/Xsession.d/99x11-common_start [10:12] Laney, can you echo $STARTUP from it? [10:12] or echo $... > /tmp/log [10:13] not sure where stdout is directed [10:13] Laney: how much disk space do you have? [10:13] adding some more echo statements [10:13] and anything in ~/.xsession-errors ? [10:13] Laney, yeah, is your disk or partition full? ;-) [10:13] mlankhorst: 70G free [10:14] or ~/.xsession-errors-:* [10:14] nothing relevant, a perl warning from msva [10:18] Laney, did you try to pring $STARTUP? [10:18] Laney, it seems 99x11-common_start just runs that command [10:18] yes, it gave me a lead [10:19] ok, got it [10:19] its a bug in monkeysphere [10:20] strange though, that hasn't been updated in ages [10:21] Laney, we should really try to improve things to make such errors easier to spot and maybe not bail out when they happen [10:22] seb128: that is unfortunately the single point of failure :( [10:23] Laney, once you get to the bottom of it can you do a summary of what was wrong and how it make lightdm bail out? [10:25] tkamppeter: I played around a little, and there doesn't seem to be a simple way; I'll postpone this and upload current bzr for now [10:26] seb128: Yeah, but I don't see how to get around it with this design. The scripts can construct a command that the display manager executes (in $STARTUP), and this package was inserting something buggy in there [10:29] Laney, ok, very weird that it was working before though... [10:29] perhaps it was perl or some random library that monkeysphere uses [10:29] Laney, in any case good that we figure it out [10:29] figured [10:29] yeah, was a fun debugging trip [10:29] thanks for the help [10:29] yw! [10:29] mlankhorst, thanks as well ;-) [10:30] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=677565 [10:30] that [10:30] Debian bug 677565 in msva-perl "msva-perl: Insecure dependency in socket while running with -T switch at /usr/lib/perl/5.14/IO/Socket.pm line 80" [Grave,Open] [10:31] as you can see it's a new bug :-) [10:32] Laney, is monkeysphere changing $STARTUP in some way? or the fact that it errors out breaks further scripts? [10:32] yes it inserts itself into $STARTUP [10:33] Laney, ok, not a lot we can do against that then [10:33] I suppose that it could try to execute a fallback if $STARTUP fails [10:33] the whole $STARTUP thing concept seems flaky though [10:33] yeah, it should [10:37] mlankhorst: do you know if that would be possible? [10:37] theoretically.. [10:37] assuming you use bash [10:37] why does it need to assume that? [10:38] can't it exec some fallback session? [10:38] because exec will ordinarily cause the whole Xsession to fail if it doesn't succeed [10:39] it returns the error code of exec [10:39] yeah, I mean try to exec something else if the normal one fails [10:39] in this case it was because something polluted $STARTUP, so if it fell back to a clean one it would have worked [10:40] I mean bash will usually do this: exec foo becomes exec foo || exit $? [10:40] but I forgot which 'set' option changed it [10:40] hm [10:40] so make it nonfatal in 99_whatever and have one more after it which is fatal [10:41] but the file is sourced, not executed, … [10:41] it's execfail [10:41] If set, a non-interactive shell will not exit if it cannot execute the file specified as an argument to the exec-builtin command. An interactive shell does not exit if exec fails. [10:43] before the final exec, however that doesn't guarantee it will work.. probably best to handle in lightdm instead [10:45] and then spawn a session that DOESN'T use /etc/Xsession at all [10:46] pitti, what are you talking about? [10:46] tkamppeter: cups.postinst's trigger calls lpadmin [10:47] tkamppeter: and in some configurations that asks for a password [10:47] which it must not [10:48] I don't suppose we could throw /etc/Xsession away entirely as a generally bad idea? :) [10:48] RAOF: No :( [10:49] pitti, about Debian bug 640939, most probably the user has a configuration with a client.conf, pointing all CUPS clients (lpadmin, lpstat, system-config-printer) away from the local CUPS daemon. The remote CUPS daemon asks for the password then. [10:49] RAOF, I don't think it would take a lot to convince robert_ancell, he was close of doing that with lightdm :p [10:49] Debian bug 640939 in cups "hplip-cups: asks for root password when cups updates PPDs" [Serious,Open] http://bugs.debian.org/640939 [10:49] tkamppeter: I thought we already solved this case with the -h /var/run/cups.sock thing [10:50] pitti, there are already measures applied against this, CUPS tools called with the "-h" option especially. What has to be done is to check through all maintainer scripts of CUPS whether nowhere the "-h" was forgotten. [10:51] tkamppeter: it's the two lpadmin calls in ppd_updater() [10:51] tkamppeter: I guess the reporters have some weird configuration [10:52] pitti, these two have "-h" AFAIR. Perhaps the reporters have done away with /var/run/cups.sock. But should lpadmin then not simply error out? [10:53] tkamppeter: it does for me [10:53] $ sudo lpadmin -h /var/run/cups/cups.sock.invalid -p testraw -P /usr/share/ppd/cupsfilters/textonly.ppd [10:53] lpadmin: Unable to connect to server: Datei oder Verzeichnis nicht gefunden [10:56] pitti, or can it be that the reporter has restricted root's rights ending up that root is not allowed to administer local CUPS (is this possible?). [10:56] tkamppeter: I haven't debugged this any further really; it's certainly possible with some weird cups configuration [10:57] tkamppeter: perhaps with an auth type different than "basic" [11:03] pitti, perhaps you ask the reporter for his cupsd.conf. and client.conf (both /etc/cups and ~/.cups/client.conf). === MacSlow is now known as MacSlow|lunch [11:07] tkamppeter: done === MacSlow|lunch is now known as MacSlow [13:07] good morning! [13:08] cyphermox, hey, how are you? [13:08] seb128: hey, not bad [13:08] and you? [13:09] cyphermox, I'm good thanks ;-) [13:14] hi guys [13:15] what's up with the unthemed indicator menus in unity-2d in Q? [13:15] hey Chipaca [13:15] i don't know what they look like in unity-2d, but there is currently some bumpiness with gtk 3.5 and light-themes [13:16] ah, ok [13:16] in unity 3d a couple of them have the wrong bgcolor [13:16] perhaps unthemed, but it is known [13:16] Chipaca, at least they still work :) [13:16] it's like it's 2002 all over again [13:16] hehe [13:17] getting gtk 3.5.x in has been a bit of a pita, mostly due to themeing changes [13:17] is anybody outside of my team running the u1 test suite on Q? [13:17] not me [13:17] because syncdaemon in Q isn't connecting, and we don't have a jenkins slave up for Q yet [13:17] (which is why it's in the state it is) [13:17] :) [13:18] Chipaca: not connecting as an SSL error? I reported a bug about that yesterday [13:18] cyphermox: yup [13:18] well, it's not a ssl error, although it is ssl that complains [13:18] cyphermox: bug #? [13:19] Chipaca, fwiw it was working a week or so ago [13:19] i just noticed last night that it wasn't working [13:19] kenvandine: *should* platform be running our QA suite? [13:19] but i have files synced to my Q laptop that are a week old [13:19] Chipaca, i don't think so... but i think we should care that your tests are passing before they land in the distro [13:19] we've had tests break by the gtk upgrade for example [13:20] interesting... [13:20] or was that scrolling related? [13:20] what kinds of tests do you have that need gtk? [13:20] Chipaca: https://bugs.launchpad.net/ubuntu/+source/ubuntuone-client/+bug/1014654 is what I have reported [13:20] Ubuntu bug 1014654 in ubuntuone-client "ubuntuone-syncdaemon fails to sync (fails to verify SSL certificate)" [Undecided,In progress] [13:20] kenvandine: test that test that a certain label has a certain color [13:20] interesting.. [13:21] cyphermox: thanks [13:22] we aren't really prepared to run the tests for everything that is in the distro [13:22] kind of tricky [13:22] something like gtk we try to beat the crap out of it manually [13:22] since it could affect so many things [13:22] and we found quite a few blockers with gtk 3.5 which delayed landing it [13:22] several weeks at least [13:23] Chipaca, what is that test for? the nautlius plugin? [13:23] kenvandine: ralsina knows more [13:24] Chipaca, have you seen the scrollbar issue in Q with the control panel? [13:24] kenvandine: (because i sure don't :) ) [13:24] hehe [13:24] i can't think of anything else that uses gtk [13:24] oh, maybe the music store [13:25] kenvandine: it was a test in ubuntu-sso-client [13:25] Chipaca, bug 1007421 [13:25] Launchpad bug 1007421 in ayatana-scrollbar "ubuntuone-control-panel-qt has no scrollbars on quantal" [Medium,New] https://launchpad.net/bugs/1007421 [13:25] ralsina, ^^ [13:25] not sure it is something u1 should/can fix [13:25] kenvandine: yes, I am aware of that one, I don't see what I can do about it though :-) [13:25] :-( I meant [13:25] probably more for overlay-scrollbar [13:26] indeed [13:26] damn, still un-assigned [13:26] * kenvandine assigns that [13:26] kenvandine: the scrollbars were made into a gtk module, but the app the bug is against is qt? [13:26] yeah, it is a weird gray area afaict [13:26] why bother making the scrollbars a gtk module? it's not like it's going to not be loaded... [13:27] it means we know long patch the crap out of gtk [13:27] it is much more maintainable [13:27] ralsina: that way it can be slow [13:27] * Chipaca takes off his evil hat [13:27] the weird thing is why that affects the qt app that way [13:27] kenvandine: probably the qt theme that loads styles from gtk is not making the scrollbars module load [13:28] kenvandine: or something evil like that [13:28] ralsina, doubt that... if so then it would just have regular scrollbars [13:28] hmm [13:28] and it isn't like it is just not drawing the thumb [13:28] the indicator for the scrollbar isn't there either [13:29] kenvandine: could be getting wrong size info and keeping it 0-width [13:29] uh, hi [13:29] hi dobey! interject! ;-) [13:30] Chipaca, ralsina, kenvandine: we do want to start running testsuits from selected rdepends before i.e gtk uploads [13:30] ralsina: OTOH, the control panel displays nothing [13:31] seb128: s/i.e/e.g/ :) [13:31] Chipaca, ralsina, kenvandine: that's not done yet but on the roadmap for this cycle [13:31] seb128: awesome [13:31] Chipaca, ups, e.g indeed, thanks ;-) [13:31] Chipaca, kenvandine: I am setting up a Q VM right now, if I can make everything work on it I can try to see what's wrong with the scrollbar info [13:32] seb128: excellent. When that happens, we're using jenkins for QA ourselves, so you should find it simple to port into your stuff [13:32] Chipaca, great [13:32] seb128: well i'm trying to get our test suites running during builds in q, but unfortunately there are some things that prevent it happening for all our stuff :-/ [13:32] ralsina, that would be cool [13:33] kenvandine: no promises but I'll try :-) [13:33] appreciated [13:33] i have no idea when cimi will look at it [13:33] and he will likely blame QT [13:33] etc [13:33] so would be nice to cover the bases there [13:33] and of course, the python3 thing doesn't make it any easier [13:33] haha, sure [13:37] ralsina: confirmed the list of folders in the control panel is empty in Q [13:37] Chipaca: sigh [13:37] Chipaca: installing Q it is [13:41] Chipaca, my list of folders isn't empty [13:41] and i can scroll with my mouse wheel to see them all :) [13:41] kenvandine: maybe i need to disable proposed? [13:41] in Q? [13:41] yes [13:42] Chipaca, you just confused me... quantal-proposed? [13:42] kenvandine, proposed is used as a staging in q [13:42] i doubt that would make a difference [13:42] yeah [13:42] but I doubt we have anything in there [13:42] but doubt there is much in there [13:42] kenvandine, ;-) [13:43] i get a bunch of (python:13774): Gtk-CRITICAL **: IA__gtk_widget_style_get: assertion `GTK_IS_WIDGET (widget)' failed [13:44] and the debug log does show the control panel *getting* the folder list [13:44] but it's not displayed [13:44] Chipaca, isn't that dialog using qt? [13:45] seb128: yes [13:45] Chipaca, why do you get gtk style errors? [13:45] seb128: yes it is, but it loads gtk to get info about how to draw the widgets [13:45] seb128: i presume because of the qt stylorator === zyga is now known as zyga-afk [13:45] probably the qt theme stuff coming from gtk [13:45] kenvandine, do you have that qt glue installed? is that coming by default? [13:46] should be by default [13:46] I would not be surprised if gtk 3.5 broke that code... [13:46] i have a ton of updates available [13:46] upgrading now [13:46] only a day behind though [13:46] gtk 3.5 broke my heart. [13:46] dobey: is this like asteroids, where you now have two, smaller hearts? [13:47] i get all that noise and my folders are listed [13:47] probably [13:47] my heart shrank 3 times that day [13:47] dobey: i'm told qml is quite nice [13:48] i'm liking fox [13:48] go go go [13:48] :-D [13:50] man, fox-toolkit.org still uses frames even [13:51] oh... cross platform... [13:52] which of course means it'll never be more than mediocre... [13:53] yeah, stuck with the likes of gtk+ and qt [13:53] no comment :) [13:54] there is something to be said for focusing your energy on creating a toolkit that is awesome on one platform [13:55] i know right. borland forever. [14:18] hi kenvandine, here's a question for you -> http://askubuntu.com/questions/152882/how-do-i-post-to-gwibber-from-python :) [14:21] i think i can answer that :) [14:21] dpm, that SRU is still waiting for approval :( [14:22] s-fox gave a bad example just 6 hours ago... sigh... i gave him an example that uses GIR just yesterday [14:24] heh [14:25] kenvandine: well, technically s-fox did that on the 15th, and someone else posted it 6 hours ago, giving credit [14:25] albeit the example script won't work for other reasons (main loop is never actually started) [14:27] kenvandine, is the upload not yet in -proposed? [14:28] ah [14:28] dpm, no... it is waiting for approval [14:28]