[01:28] darkxst, can you reproduce bug 1278467? [01:28] Launchpad bug 1278467 in gnome-settings-daemon (Ubuntu) "[xsettings]: gnome-settings-daemon crashed with SIGSEGV in notify_have_shell()" [Medium,Confirmed] https://launchpad.net/bugs/1278467 [01:34] robert_ancell, I haven't seen that crash in quite some time [01:34] also it should have been fixed already by http://bazaar.launchpad.net/~ubuntu-desktop/gnome-settings-daemon/ubuntu/revision/448 [01:34] (if its the same bug) [01:35] darkxst, it's the number 3 crasher in trusty right now [01:35] there's a typo in debian/patches/90_set_gmenus_xsettings.patch which is probably causing it. [01:35] But I can't reproduce [01:36] darkxst, #3 on https://errors.ubuntu.com/?release=Ubuntu%2014.04&period=month [01:45] apparrently I can reproduce it [01:45] do you have a patch? [01:45] darkxst, can you try lp:~ubuntu-desktop/gnome-settings-daemon/ubuntu? [01:48] robert_ancell, ok, will do [01:52] robert_ancell, that code should not even get run on a gdm/gnome-shell session [02:00] oh it does actually, but setting have_unity = TRUE there is just wrong [02:07] TheMuso, what time is the desktop meeting? [02:11] robert_ancell, that seems to fix the crash, but can you also fix [02:11] 104 [02:11] 104 [02:11] + manager->priv->have_unity = TRUE; [02:11] before you upload [02:11] darkxst, I'll let you fix that one [02:11] darkxst, I've already uploaded [02:11] I was optimistic :) [02:12] so dobey, when you guys going to stop hogging the top spots on errors.ubuntu.com? ;) [02:17] robert_ancell, lp:~darkxst/gnome-settings-daemon/have-unity [02:18] robert_ancell: I have no idea what meeting you are referring to. So far as I know, there hasn't been a desktop team meeting for a while. I know there are system settings meetings on Friday EU time, but thats all. [02:19] TheMuso, there is one but just not a good time for us https://wiki.ubuntu.com/DesktopTeam/Meeting [02:21] And note the lack of any 2014 meeting notes... [02:22] Sed meetings seased in June last year. [02:24] TheMuso, they're on IRC, no-one updates the Wiki anymore (the last one http://irclogs.ubuntu.com/2014/04/01/%23ubuntu-desktop.html) [02:24] h ok. [02:28] qengho: you should ask to get testimonials prior to the next membership boards meeting [02:28] Ah, right. [02:28] to add to your wiki page, and/or people to come cheer for you during the meeting [02:28] darkxst, I don't think that's needed - g_bus_watch_name will call one of the callbacks so in this case you will immediately get a name_vanished in GNOME [02:28] https://developer.gnome.org/gio/unstable/gio-Watching-Bus-Names.html#g-bus-watch-name [02:29] cyphermox: How many testimonials do you recommend? [02:30] cyphermox: and, how did you notice my page? [02:30] qengho: no number in particular, one or two, the more the better of course [02:30] darkxst, yeah, and that value is only used after one of the callbacks so it should be fine [02:30] I'm on the 1200 UTC board and I was wondering if tomorrow was a meeting [02:31] I've been feeling so-so for a while so waking so early up is hard [02:35] robert_ancell, in which case that line should not even be there! [02:35] darkxst, yeah, I guess so! [02:40] All: Most of you are out-of-phase from my normal work-day, but if you know me, chad, please make a note on my wiki page about my suitability for Ubuntu membership. https://wiki.ubuntu.com/ChadMiller [03:01] beuno, dobey, jcastro, slangasek: Hi, diverse selection of colleagues. I'm applying for Ubuntu membership. If you don't think that is a terrible idea, please consider adding a note to say so on my wiki page. https://wiki.ubuntu.com/ChadMiller [03:02] Oh, smuxi irc client, you so crazy. [03:02] Wrong channel. [03:06] qengho, https://wiki.ubuntu.com/DeveloperMembershipBoard [03:08] darkxst: you're pointing me to the desktop delegationm? [03:13] qengho, DMB handles memberships for devs and they have a template for collecting endorsements [05:26] Good morning [06:30] pitti, hi, the chinese language pack is not updated to latest. Our QA team is doing the translation tests these days. Would you please help to rebuild this package? [06:31] maclin: latest is https://launchpad.net/ubuntu/+source/language-pack-zh-hans/1:14.04+20140403 [06:32] maclin: today we'll get a new automatic update [06:32] right, it was built last night: https://translations.launchpad.net/ubuntu/trusty/+language-packs [06:32] maclin: for some reason Launchpad didn't produce an update last Friday [06:32] pitti, yes, we have updated new translations these days. [06:34] these changes are not include in the latest ISO [06:46] pitti, could you help to confirm the problem of language pack? [06:46] maclin_: which problem? [06:46] maclin_: probably not, I can't read Chinese; i. e. if some translation is wrong I won't be able to spot taht [06:46] maclin_: if you mean the rebuild, that's going to happen automatically in a few hours [06:47] (see above) [06:52] pitti, I lose the connection just now. you said there is a problem of building of package last Friday. Normally the pckage will be built automatically everyday, however the last version is 20140403. [06:53] maclin_: yes, but LP did build an update yesterday, so today we'll get new langpacks [06:59] pitti, I get it, thanks a lot:) === maclin_ is now known as maclin [07:44] morning! [08:04] what is the probability of this fix in precise - https://bugs.launchpad.net/ubuntu/+source/glib2.0/+bug/1214352 ? [08:04] Launchpad bug 1214352 in LibreOffice Productivity Suite "GUINT32/64_SWAP_LE_BE macros do not enclose val argument in parentheses" [Critical,Confirmed] [08:06] guten morgen [08:10] bregma, I'm sorry if I'm starting to annoy you but I just found another bug related to the launcher (no gestures). bug 1304882 [08:10] Launchpad bug 1304882 in Unity "[Regression] Launcher doesn't reappear fully after hiding from DND" [Undecided,New] https://launchpad.net/bugs/1304882 === beidl_ is now known as beidl [08:32] tjaalton: did you see the wayland change? Might be something that's needed in Debian too [08:33] don't think so [08:33] they never had the old pkg names iirc [08:33] on a shipped release anyway [08:33] I thought they did the same libwayland0 -> client/server split [08:33] https://packages.debian.org/source/stable/wayland [08:34] hum, ok [08:35] I don't really get what situation made this necessary, so I can't say if it is a problem there too [09:36] man, I like epiphany [09:49] + g_spawn_command_line_async("killall -q gnome-screensaver", nullptr); [09:50] This makes me feel weird [09:52] Laney: ugh [09:52] why? [09:52] unity wants to own org.gnome.Screensaver [09:52] I guess gnome-screensaver was getting there first [09:53] and gnome-screensaver doesn't allow replacement? [09:53] patching it to do allow that might be the better route [09:54] Don't know [09:54] how do you do that? [09:54] killall is racy [09:54] pass G_BUS_NAME_OWNER_FLAGS_ALLOW_REPLACEMENT do g_bus_own_name() [09:54] dbus-activation or upstart could be faster than your own [09:54] and G_BUS_NAME_OWNER_FLAGS_REPLACE in unity [09:55] right, that sounds like the more proper way to do it [09:55] yes, it does [09:55] I didn't know about this facility [09:55] Trevinho: ^^^^^^ [09:55] pitti: I love how you came up with an actual problem of this line instead of just being disgusted by it (like I was) ;) [09:56] haha [09:56] If we're fixing the line, don't use killall [09:56] it does something very very different on solaris [09:56] also pass -u :-) [09:56] well, eww for killall and using shell, too, of course :) [09:57] 'pkill', '-u', geteuid(), 'gnome-screensaver' [09:57] (with appropriate stringificatino and syntax of course) [09:57] Laney: argh, dbus-glib [09:58] larsu: somehow I'm not surprised [09:58] Laney: so pass DBUS_NAME_FLAG_ALLOW_REPLACEMENT to dbus_bus_request_name(). Want a patch? [09:59] ALLOW_REPLACEMENT on a security-based component isn't overly... secure [09:59] *shrug*, you can also just kill it [09:59] mhr3: if you're on the bus it's too late [09:59] ALLOW_REPLACEMENT is merly avoiding a race condition, it's not making the whole thing any less secure [09:59] you can even call "SetActive" on it [09:59] also, what pitti says [10:00] larsu: it's okay, I'm already there [10:00] coolio [10:03] I guess it should handle NameLost too,e h [10:09] wow, it adds a match for _every_ nameownerchanged [10:09] *sigh* [10:10] and then doesn't do anything with it ... [10:10] (not even when its own name changeS) [10:10] * larsu should stop reading that code === tkamppeter_ is now known as tkamppeter [10:24] Laney: larsu: borderless metacity theme is unusable under a normal desktop (e.g. edubuntu) so now i'm thinking what can I do about ubiquity. It has darkcolor/normal white/darkcolor layout and an ugly 1px white hairlines on the left and right around the dark bits under metacity. [10:24] last time around trying to switch to compiz, proved unreliable as it was crashing a lot. [10:25] i wonder if i can make ubiquity window borderless - e.g. with Window hints? set_border_width? i'll experiment with it later. [11:06] xnox: you can remove decrations entirely, but that will remove the titlebar as well [11:07] I don't think there's another way to control the border width [11:10] larsu: that wouldn't be that bad. We don't support mimimising/maximizing and for most part "the x" button doesn't do anything, and we have a dedicated "quit" button anyway. [11:10] larsu: how would i do that? [11:10] (to try out, clearly not for trusty) === MacSlow is now known as MacSlow|lunch [11:10] xnox: gtk_window_set_dectorated [11:16] no seb128 today? [11:17] travelling [11:17] k [11:32] ^^ no seb today! [11:33] Laney yeah, larsu yeah I thought about use the replacement thing... but gs does not use gdbus [11:34] pitti: yeah, I've done that with killall -u ... I just didn't push the change yet as it needs a rebuild of the silo [11:34] Trevinho: trying it now, hopefully works [11:35] didrocks, pitti: since seb is AWOL, would one of you sponsor a final LibreOffice upload for me today then? minimal change: to heal the worst of bug 1300283 with one upstream commit reverted? [11:35] Launchpad bug 1300283 in libreoffice (Ubuntu) "LibreOffice does not start in a KDE 4 session" [Critical,In progress] https://launchpad.net/bugs/1300283 [11:35] Trevinho: that's a dbus feature which should work with any client library === Sweetsha1k is now known as Sweetshark [11:35] and yes... I would have loved to use upstart to stop gs,. but that would have made things more complex [11:37] larsu: yeah, I know, it's just that i wasnt' unsure how quick to do it was there [11:38] larsu: as... if I override the name, then the client gets it back automatically as soon as it's available? [11:39] however, ihmo the best solution would just be to use upstart to start the service, instead that using a dbus rule [11:39] but not sure how quick that can be [11:45] Sweetshark: here is seb128 for you :) [11:45] I wonder where he was slacking :p [11:45] 13:35:12 Sweetsha1k | didrocks, pitti: since seb is AWOL, would one of you sponsor a final LibreOffice upload for me today then? [11:45] | minimal change: to heal the worst of bug 1300283 with one upstream commit reverted? [11:45] Launchpad bug 1300283 in libreoffice (Ubuntu) "LibreOffice does not start in a KDE 4 session" [Critical,In progress] https://launchpad.net/bugs/1300283 [11:46] Trevinho, as long as upstart activation and dbus activation are two separate things, it's never going to work properly [11:46] fortunately we're moving to systemd :) [11:47] mhr3: isn't this working https://lists.ubuntu.com/archives/ubuntu-devel/2010-December/032295.html [11:47] mhr3: at least, that's what I meant... [11:47] mhr3: i read about that some time ago, but never tested [11:48] nor I can find any *.service using UpstartJob [11:48] if that works then indeed that's the way to go [11:48] good afternoon desktopers [11:48] Trevinho, mhr3: what are you trying to do? [11:48] hi seb128 [11:49] seb128: well.... see if I can clanup the killall thing for gs (that I hate, but it's the faster, and works) [11:50] seb128: so if it would be a way to start the g-s service using upstart on name activation, then it would be just trivial to stop it in a cleaner way [11:51] I'm looking at the replacement, don't bother duplicating [11:51] Laney, thanks [11:51] hey seb128 [11:53] didrocks: ha! [11:53] Trevinho, so that feature is available for over 3 years and there isn't a single service using it? [11:54] mhr3: it looks like that :o [11:54] that's just... sad [11:54] Trevinho, mhr3: dbus activation through upstart? [11:55] yea [11:55] Laney: looking at the replacemenet of gs name? [11:55] yes [11:55] why did we have the awful hud hacks then [11:56] guess i'm not the only one who didn't know about it [11:57] I read about that loong time ago, always in my mind, but now not looking references for that makes me worry [11:57] seb128: Im right now preparing an update for just the libreoffice doesnt start on KDE issue with just one upstream revert (the other fixes seem too risky), will you sponsor today? [11:57] Trevinho, mhr3: the dbus patched needed for the upstart-dbus-activation got dropped since that email (or said different upstart doesn't support upstart activation) [11:57] Laney: ok, thanks... what made me worried about it how to recover the name and functionalities when the other owner goes away... If that is somewhat possible... Also at that point I would have preferred not to keep a not-needed instance of gs running [11:58] :°( [11:58] Trevinho, why is g-s running at all? [11:59] seb128: I didn't debug, but it always run on startup for some service/app/indicator requiring it... Anyway also if it would not run, we still need to kill it in case that you switch off one option that doesn't need gs anymore [11:59] seb128: also, have you tried the ppa? And, can you trigger a rebuild or only bregma can? [12:01] as I've one line fix for 1291088 [12:02] btw in recent updates I've noticed another thing... When clicking on the username in the indicator-session, logind doesn't emit the Lock signal anymore... So tty7 (say) stays unlocked [12:04] Trevinho, since when? with that ppa only? [12:04] Trevinho, no I didn't test the ppa, it failed to build on the a11y schemas issue yesterday and I was travelling/offline this morning [12:07] seb128: ah ok... [12:07] seb128: no, since before [12:07] wfm [12:07] the sessions got locked here [12:08] Sweetshark: sure! do you have a .dsc and source.changes to sign and upload somewhere? (debdiff would be nice too, but I can generate that myself) [12:08] seb128: mh, it does here by calling the dbus service manually... let me check [12:08] Trevinho: I made it exit if the name gets replaced [12:09] and it's got D-Bus activation so if unity goes away then the next call will start g-s [12:14] Laney: ah nice [12:14] pitti: (re)generating them right now, Ill ping you with a link, when they are all done. Thanks! [12:18] Trevinho: btw how do you build unity packages? [12:18] I'm finding that the dbus part of the testsuite fails when I use sbuild [12:18] Laney: I quite never build packages for unity... and so when I do that I just use the manual way :/ [12:19] Laney: but dbus should work there... I mean, we create a fake context [12:19] ok, well the branch I'll propose (if it works) contains a change in debian/rules to let you skip the testsuite [12:19] I was getting connection refused errors === MacSlow|lunch is now known as MacSlow [12:20] Laney: mh, let's see [12:20] it's not urgent to fix as it obviously works on the buildds [12:20] just was wondering [12:29] Trevinho: how do I test the onboard/a11y fallback? [12:29] hold on, I installed the wrong g-s [12:29] Laney: just enable OSK or screen reader from ucc [12:31] turning on OSK didn't do anything visible [12:31] I can't make it come up [12:32] however it did make g-s start and the osk worked in its lock screen [12:36] pitti: http://people.canonical.com/~bjoern/trusty/4.2.3/libreoffice_4.2.3~rc3-0ubuntu2_source.changes <- here is the new package, and here is the debdiff: http://people.canonical.com/~bjoern/trusty/4.2.3/libreoffice_4.2.3~rc3-0ubuntu2.diff [12:39] Laney, could you update me what's going on with the lockscreen stuff? Do you have an additional patch for g-s to upload to my silo and are going to make an MP for Unity? [12:39] Sweetshark: do we need -l10n as well? [12:40] bregma: correct [12:40] killall gnome-screensaver is bad and unnecessary [12:41] pitti: I generated, signed and uploaded one just in case, but we certainly dont need one for the fix (which will only touch libreoffice-kde). So no, dont sponsor a new libreoffice-l10n. [12:42] Laney, agreed, do you know when it'll be ready? [12:42] Sweetshark: ack; libo uploaded [12:42] bregma: imminently [12:42] Trevinho: can you check lp:~laney/unity/g-s-replace and merge it into your branch if you like it? [12:44] Laney: yes [12:44] Laney: do I need also a gs irght? [12:45] larsu: http://paste.ubuntu.com/7226032/ for g-s [12:45] Laney, Trevinho: what vcs do you use? I commited yesterday's silo upload to lp:~ubuntu-desktop/gnome-screensaver/ubuntu [12:45] I'll bzr it in a sec [12:45] Laney: ok, yhsnkd [12:45] thanks* [12:45] I didn't do anything fancy like checking the owner [12:46] pitti: We only ran with the -l10n debian revision diffing from the libreoffice debian reversion once in early trusty and there havent been any issues then with dependencies, so I assume that to be safe. I prepared the -l10n package only for the worst case scenario that there is an overspecific dependency relation somewhere still. But no, I dont want it to be uploaded. [12:46] because it's exiting if it doesn't get the name that situation shouldn't happen [12:46] pitti: thanks a lo for the upload. [12:47] Laney: yeah, change is fine (it's the same that I did, and then stashed :)) [12:47] nice [12:47] seb128: you mentioned a touch bug to jasoncwarner? [12:47] well, if someone can review that g-s debdiff and see what they think then we can do some bit pushing [12:48] also can't reproduce the chromium bug, probably hw specific :/ [12:49] mlankhorst, the latitude inspiron we got for testing gets lot of input error in the xorg logs when using the touch screen [12:50] what errors exactly? [12:50] mlankhorst, the chromium issue got fixed with tjaalton's mesa uploaded yesterday [12:50] ah k [12:50] Laney: where it is? [12:51] marking dupliate then [12:51] mlankhorst, I don't have my touch laptop here for the xorg error, let me try to get that [12:51] Trevinho: pastebinned it up ^^^ there [12:51] Laney: ah sorry [12:52] Laney, you should commit to the packaging vcs directly? [12:52] just did that [12:52] Laney: exit(0) can't be replaced by quitting the loop? [12:55] seb128, would that be the [ 25508.732] (EE) BUG: ../../dix/touch.c:644 in TouchConvertToPointerEvent() xorg error? [12:56] bregma, could be, I tried that some weeks ago ... do you get stack of thoses in the log? [12:56] mlankhorst, ^ [12:56] bregma, can you pastebin/share your xorg log maybe? [12:56] Trevinho: Maybe could gtk_main_quit() instead [12:58] Laney: yeah it's probably better [12:58] Laney: I don't think you need to listen to both nameownerchanged _and_ namelost [12:58] Just thinking if it can happen before gtk_init [12:58] larsu: wasn't clear to me which one you are supposed to get [12:59] Laney: ah, well main_loop_quit might still be fine, isn't it? [12:59] it's the gtk mainloop [13:00] Laney: namelost should be enough and results in much less dbus traffic (if you also remove the watch) [13:00] not sure if this patch should fix existing bugs, though [13:00] but getting woken up for every name change is a bit excessive... [13:00] otherwise it's fine. Thanks! [13:00] yeah... [13:00] let me check if NameLost works [13:01] you might ned to add a mtch rule for that (not sure though) [13:01] seb128, mlankhorst, Xorg.0.log with MT errors: http://paste.ubuntu.com/7226096/ [13:02] bregma, thanks [13:02] the on-screen symptom is the pointer gets entangled with MT gestures [13:02] mlankhorst, ^ [13:02] Laney: is that patch based on lp:~3v1n0/gnome-screensaver/not-autostart-in-unity? as it fails to merge there [13:02] now if I could figure out how to get my external monitor configured correctly, I'd be happier [13:02] damned lying EDIDs [13:02] Trevinho: lockscreen-disable-with-osk [13:02] Laney: ah, ok it's based on the pkg that is in the silo [13:03] Laney: no, I meant the gs one [13:03] ah yes, that [13:05] non-emulating touch event :s [13:05] testing with NameLost only [13:06] Laney: how is it going? [13:06] 2 mins [13:08] Trevinho: works, also with gtk_main_quit() too [13:08] Laney: cool [13:12] Trevinho: pushed, try lp:~ubuntu-desktop/gnome-screensaver/ubuntu [13:12] Laney: thanks [13:16] lunch now, will fix up when I get back if necessary [13:16] otherwise looking forward to silage [13:20] bregma: do you have a reproducible testcase? [13:20] or just a sequence that lets you trip things reliably [13:20] Laney: it works like a charm, cool! [13:22] xnox, is it known there is no indicator-power on ubiquity-install-mode? [13:23] seb128: from my POV, Laney's patch is fine to go into the silo [13:23] Laney: fyi i've merged the unity branch into mine as well [13:23] Trevinho, wfm, I don't have access to the silo for uploads [13:24] so you need a lander to do it [13:24] I all. someone could take a look at https://bugs.launchpad.net/ubuntu/+source/gtk+3.0/+bug/1305016 ? [13:24] Launchpad bug 1305016 in gtk+3.0 (Ubuntu) "Windows have only borders if __GTK_FRAME_EXTENTS is not supported by WM" [Undecided,Confirmed] [13:24] ah, ok... bregma ^ [13:26] Trevinho, he can't upload either [13:26] maybe Laney can [13:29] seb128: /usr/share/unity/indicators/com.canonical.indicator.power doesn't declare a "ubiquity" profile and ubiquity doesn't exec it, so yeah there is no indicator-power in ubiquity. [13:29] seb128: if it's desired/required/wanted then we need to update the indicator and add a line to exec it to ubiquity-dm [13:29] xnox, so both need to be changed? [13:30] seb128: yeap. Do we need/want power indicator? is it e.g. shown on the login/lock screens? [13:30] it does [13:30] I'm installing trusty on a new machine which is unplugged and I've no clue what's the battery charge level [13:31] would be useful to have that info [13:33] xnox, also unity-control-center is mostly empty because XDG_CURRENT_DESKTOP is not set (so you can't add e.g keymaps) [13:34] xnox, I'm going to change u-c-c to always display all its icons though [13:34] mlankhorst, I could give you evemu device and recordings files [13:34] need to check with Laney if that's fine for release though [13:34] bregma: might help, let me upload x1.15.1rc1 real quick, perhaps it helps some [13:35] uploaded to x-staging [13:38] mlankhorst, evemu device description http://paste.ubuntu.com/7226247/ and event recording http://paste.ubuntu.com/7226249/ [13:39] those were recorded at the same time as those logged errors [13:45] seb128: yeah, i lost track in updates and changes to the unity-* jobs events. [14:03] bregma: can you test xserver from https://launchpad.net/~canonical-x/+archive/x-staging/+packages ? [14:04] seb128: I'm applying for chromium upload rights via normal means. https://wiki.ubuntu.com/ChadMiller/DeveloperApplication I don't understand the desktop-team Ubuntu-membership delegation or which I should choose. [14:04] qengho, just apply for a ppu for chromium [14:05] seb128: Okay. I think ^that's it. [14:05] looks about right yes [14:06] Good. Thanks. [14:10] hey [14:10] you want me to upload gnome-screensaver to the silo? [14:12] Laney, ouais [14:13] hold tight [14:14] Laney: works nicely, thanks for fixing my bad code :) [14:14] but I went with the faster thing as I had to improve the things in unity code itself, first [14:15] Trevinho: np - actually I didn't know enough about replacing names on dbus to appreciate this solution [14:15] so thanks go to larsu for telling me [14:15] I was thinking about adding a method to make g-s quit itself ... [14:16] * larsu can't count the number of people anymore that he taught aobut this feature [14:16] *about [14:16] yeah, I knew about it, but I didn't think it was that straight forward to do in gs code [14:16] λn. suc(n) [14:17] Trevinho: that's a fair assumption given that it still uses dbus-glib (it turned out to be fairly easy though) [14:17] Trevinho, is everything ready for one last rebuild in the screensaver silo? [14:17] larsu: yeah, when I've read that gs-dbus code the first time, I just thought that it wasn't the case :) [14:18] bregma: we need to trigger a rebuild of unity [14:18] bregma: and at this point add also https://code.launchpad.net/~brandontschaefer/unity/lockscreen-caps-lock-detector [14:20] bregma, hey, for info I tried a unity8 session on a desktop and u8 crashed. I filed bug 1304959 [14:20] Launchpad bug 1304959 in unity8 (Ubuntu) "unity8 crashed with SIGABRT in qt_message_fatal()" [Medium,New] https://launchpad.net/bugs/1304959 [14:20] bregma, is the a minimal system requirement to run unity8? [14:21] s/the/there/ [14:25] ok it is silo'd [14:26] seb128: what are you planning on doing with u-c-c? [14:26] Laney, listing all the panels in its main UI, without respecting OnlyShowIn [14:26] still to make it ignore OnlyShowIn? [14:26] I think it makes sense [14:27] ok, great [14:27] I ran into the issue where the ubiquity install mode has indicator-keyboard opening u-c-c with 2 icons listed [14:27] not including the keyboard settings [14:27] (no XDG_CURRENT_DESKTOP env there) [14:27] it means my fixes (that I forgot to push) to set XDG_CURRENT_DESKTOP from the indicators won't be needed [14:28] great [14:29] seb128: hang on a minute, inside ubiquity install mode there is no XDG_CURRENT_DESKTOP and no session upstart running. [14:29] I guess g-c-c could do that too [14:29] seb128: if you want me to fake one, when on Ubuntu Desktop product (and/or when we launch unity-settings-daemon) i'm happy to set it to anything that is needed. [14:30] seb128: ubiquity-dm (install mode) is not launched by normal display manager, thus environment is weird. [14:30] xnox, I'm just going to make unity-control-center not rely on an env [14:30] seb128: sounds the best way. [14:53] xnox, I just did an install of a recent daily and the screen reader was on in the user session after first boot, is there any log than can tell me if that's something ubiquity did? [14:59] seb128: pastebin /var/log/installer/syslog [14:59] it might have useful things there. [15:04] xnox, http://paste.ubuntu.com/7226601/ [15:05] xnox, Apr 9 14:29:51 ubuntu /plugininstall.py: running /usr/lib/ubiquity/target-config/30accessibility [15:05] xnox, is that normal/wanted? [15:05] hum [15:05] Apr 9 14:30:00 ubuntu ubiquity: Setting capabilities for gnome-keyring-daemon using Linux Capabilities failed. [15:05] (that's not the same issue but spotted that in the log) [15:06] what a11y settings did you end with in dconf-editor? [15:07] I don't know how to diff that compared to a normal install should be [15:07] seb128: hmm.... i ponder if the unity lock screen "shortcut" clashes with the screen reader shortcut.... [15:07] xnox, org.gnome.desktop.a11y.applications screen-reader-enabled true [15:08] aha, so did you ever triggered screen-reader whilst installing? [15:08] I doubt it [15:08] (cause if you did we propagate that...) [15:08] right. [15:08] doesn't it turn itself on automatically if you wait for $time? [15:08] i'll do test install here to check. [15:08] the installed was silent/I didn't touch the keyboard [15:09] Laney, screen reader? [15:09] Laney, why would it? [15:09] * seb128 confused [15:14] So that people who can't see can still install [15:14] But it's not true, you have to press ctrl-s I think [15:22] Hi seb128! [15:22] FYI unity packages are in ppa for testing [15:22] silo-6 [15:24] https://launchpad.net/~ci-train-ppa-service/+archive/landing-006/+packages === olli__ is now known as olli [15:27] GunnarHj, hey [15:27] seb128: Since you are working with l-s... When you launch it, a new icon shows up in the Launcher. The colors are nice (Swedish flag), but still. [15:28] GunnarHj, right, I was just looking at that [15:28] seb128: Ok, good. [15:28] GunnarHj, but maybe you can help me [15:29]