=== TheMuso` is now known as TheMuso [04:00] morning desktop [04:11] Good morning [04:16] o/ pitti === maclin_ is now known as maclin === work_alkisg is now known as alkisg === alkisg is now known as work_alkisg === jibel_ is now known as jibel [06:51] Hello, world! [06:51] hey mlankhorst [06:52] morning === work_alkisg is now known as alkisg [08:04] hey hey hey [08:04] morning Laney! [08:04] good monring Laney, didrocks and seb128 [08:05] Guten Morgen mvo :) [08:05] hello didrocks et mvo, how's it going? [08:05] I'm good thanks! yourself? [08:05] not too bad, thanks! [08:06] good morning desktopers [08:06] hey hey hey Laney didrocks mvo [08:06] Laney, how was your long W.E? [08:06] i'm good [08:07] people daring taking bank holidays :-) [08:07] it was fun thank you! [08:07] + an extra day even ;-) [08:07] we got a canal boat and sailed around a bit [08:08] working the locks and stuff, I recommend it! [08:09] nice weather? [08:09] i'm looking a bit pink ... [08:09] waow :) [08:09] * didrocks won't make any comment on all people living "in the north" like seb128 or you ;) [08:09] :P [08:09] (and getting easy sunburn :p) [08:14] it's surprising that it can happen even if it doesn't feel that sunny [08:17] yeah, clouds are a double trap, you don't feel that warm and don't think you can burn [08:18] wind as well [08:18] especially when you are on the water [08:22] hum, I can't assign the thinkvantage button as a shortcut anymore on trusty :/ [08:23] * didrocks tried to change the shortcut script path it pointed at and it's broken now [08:23] assign to alt + u works though, it's clearly related to thinkvantage button [08:25] ok, at least, same behavior in a guest [08:26] [org/gnome/settings-daemon/plugins/media-keys/custom-keybindings/custom0] [08:26] binding='Launch1' [08:26] command='/home/didrocks/foo' [08:26] name='creds' [08:28] Sweetsha1k: hey Bjoern [08:29] Sweetsha1k: we finally fixed the autopkgtest timeouts, now LibO fails "properly" [08:29] looks like a missing @builddeps@ or build-essential or similar dependency, depending on what the test needs to to [08:45] seb128: can you promote appdata-tools please? https://bugs.launchpad.net/ubuntu/+source/appdata-tools/+bug/1315019 [08:47] Laney, is anything (build-)depending on it yet? [08:47] cheese [08:48] ok, so "yes" [08:48] doing the promoting then [08:48] thanks! [08:48] yw! [09:04] xnox: any thoughts on https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=675008 yet? === alan_g is now known as alan_g|afk === vrruiz_ is now known as rvr === alan_g|afk is now known as alan_g [09:38] hum, seems that the thinkvantage button issue would be due to compiz [09:38] popey: FYI ^ [09:38] Trevinho: do you know about why a newly assigned shortcut won't work under unity? [09:38] ☹ [09:39] popey: assigning it to compiz directly works as a workaround [09:40] how? [09:40] popey: what's your reaction if you hear about ccsm? :p [09:40] Laney: in ubuntu we source /etc/profile.d/*.sh from /etc/profile, similar to what fedora does. [09:41] Laney: however, fedora also sources /etc/profile.d/*.sh from /etc/bashrc [09:41] didrocks: meh, i have it installed, use it to enable various random things - what's the spell to cast? [09:41] popey: in the "general" area, you have "commands" [09:42] yeh that is the cause of the whole gnome-terminal vte.sh sourcing issue! [09:42] popey: assign a command line [09:42] go to the next tab [09:42] and assign to the corresponding number the keyboard shortcut [09:42] seems that the compiz <-> u-c-c integration is clearly broken [09:43] bug 1132700 [09:43] and mainly why we are still stuck with gnome-terminal 3.6 [09:43] yes, that's what i'm asking about [09:43] Laney: i really see no harm in not sourcing /etc/profile.d/* since i'm yet to see snippets dropped there that "may not be ever executed more than once" [09:44] Laney: i'm making a patch now, and will send it to bts & doko and will nmu that. [09:45] Laney: i guess you do want it in debian, and not just ubuntu? [09:46] didrocks: woohoo! thanks! [09:46] popey: yw! ;) [09:46] xnox: cool, good luck with that ;) [09:46] I don't see any reason this would be ubuntu specific, indeed [09:46] didrocks, u-c-c custom commands wfm [09:46] seb128: it works here as well, but not for the thinkvantage button, if I reassign [09:46] seb128: popey confirms it as well [09:47] I don't know what "thinkvantge" is [09:47] I got the issue this morning when trying to change the script path [09:47] seb128: big blue button on thinkpad laptop [09:47] seb128: blue button on a thinkpad ☻ [09:47] oh ok [09:47] * seb128 has Dell [09:47] well, normal keybindings work [09:47] I just assigned ctrl-alt-e to "gedit" [09:47] and it runs it [09:47] yeah, not sure how or why that one broken [09:47] and started to be broken when I reassign the script [09:48] the thinkpad special keys are handled slightly different [09:48] the name between GNOME and compiz aren't the same [09:48] like, it's Launch1 under GNOME [09:48] and XF86Launch1 for compiz [09:49] popey: after a dconf dump /org/compiz/integrated/ [09:50] on the laptop you never reassigned [09:50] did you have anything setup? [09:53] GNNNN. I would tell you if my trusty desktop hadn't just booted up in low-graphics mode [09:54] popey: he felt you wanted to hurt it! :) [09:54] ☹ [09:54] this shouldn't happen ☹ === dbarth_ is now known as dbarth [10:28] Anyone want to try gnome-terminal from lp:~ubuntu-desktop/gnome-terminal/ubuntu-current ? I re-added a patch we used to have to use the transparency that the theme sets for the background. [10:39] larsu: do you think rishi would be interested in doing something like that in his transparency patch btw? It adds a property which makes g-t read the transparency from the theme instead of the profile's settings [10:40] Laney: oh, speaking of which the light theme needs activating for gnome terminal (ambience?!) [10:40] Laney: can you pester me at the sprint again? have to do some real work ;p [10:40] activating? [10:40] hey, this is real work >:| [10:40] Laney: so far i've subverted all of his argumentation =) [10:40] good work [10:41] Laney: don't know. Let's ask him when he's around [10:41] okay [10:42] this is for g-t 3.12? [10:42] his stuff would be [10:42] k [10:42] that patch I just (re)added was an old one we already had [10:42] our theme already sets the value [10:42] we just dumped the patch for some reason [10:43] Laney: yeah, so light-themes has "dark" gnome-terminal (radiance?!) but the "light" one is commented out (ambience?!) [10:45] xnox: Looks like they're the same if you ignore the commented out stuff [10:47] Laney: are they?! hm. i thought there was something special about it. let me test it out. [10:48] i thought tabs were suppose to get nicer matching colors, or some such. [10:49] It just sets the default transparency & background & text colour [10:50] meh, ok. [10:50] I think the rest of it is standard gtk stuff [10:50] didrocks: only has alt for show-hud, nothing else, on desktop [10:53] pitti: I just remembered our conversation on Friday --- do you think http://irclogs.ubuntu.com/2014/05/02/#ubuntu-devel.html#t16:32 could explain why it doesn't pass? [10:54] Laney, that patch won't work with the new(er) gnome-terminal, they remove the api's [10:54] darkxst: that's what I just asked Lars about [10:55] there's a patch (or intention to write one) to keep transparency [10:55] pitti: (xnox:) going to skip dbus-test-runner for now anyway [10:56] Laney: as in force-bad-test? and thus gvfs will migrate?! [10:57] it's a possibility [10:57] Laney: it looks like it's bustle related, since "needs-build" does a build without bustle, whilst adt-test does _with_ bustle installed (or at least that's the intention...) and something somewhere doesn't like that. [10:59] I think the idea is to not recompile in the actual test, just to run the testsuite [10:59] it's pretty weird [10:59] Laney, xnox: BTW, gvfs' test failure is real (due to new libgphoto2), on my list after the bigger fires were putout [10:59] oh [10:59] I think that britney has forgotten about that one [11:00] pitti: do you want to fix it before gvfs goes in or later? [11:01] Laney: if it holds up lib transitions, feel free to override [11:01] Laney: it's not a simple fix, and as I'm also patch pilot today and also have to do something for the OEM team, I don't expect to get to it today [11:02] It doesn't hold them up, but only because (adt-)britney has forgotten about it [11:03] Laney: that's one of the "urgent" thinks I'm currently busy with, BTW -- reviewing jibel's fix for that very britney bug :) [11:03] thinks -> things [11:03] great, I've been hoping for that one :) [11:13] uploaded g-t to the desktop ppa, btw [11:22] Laney: \o/ britney still has dementia gvfs is in =) [11:23] nice [11:23] plist hooooo [11:24] apart from haskell, there is only mpich transition left and bits&bobs of boost transition [11:24] yeah ;-) plist is now done. [11:24] Laney: do you know how to get a shell inside the test environment with "adt-run" & qemu? [11:25] nope, I've not used that yet [11:25] it was easy with the old run-adt-test [11:25] yeah. [11:25] pitti: how to get shell inside test environment with "adt-run" & qemu runner? "--shell" doesn't seem to do that for me. [11:26] xnox: it doesn't directly throw you into a shell, right; it prints out various commands (in descending "niceness") how to log in [11:26] xnox: i. e. usually with ssh, but if that's broken you can also use minicom or even nc [11:26] xnox: -s is more useful, btw (only shell on failed tests) [11:27] xnox: or do you just want a virgin environment without any autopkgtset stuff run before? [11:27] pitti: http://youtu.be/2xW5xK3UeXs =) [11:27] (that's about minicom & nc =)))) ) [11:28] pitti: i want with autopkgtest just finished running (or failing) [11:28] # For a native build we would also want to append /usr/lib/pkgconfig [11:28] # at this point; but this is a cross-building script, so don't [11:28] grr! [11:29] xnox: yes, then --shell should do exactly that; what does it say after the test? [11:29] pitti: well let me try again. [11:31] xnox: that's how it's supposed to look like: http://paste.ubuntu.com/7409845/ [11:31] xnox: I think it's currently broken if you use --output-dir; do you? [11:32] i am rebuilding my cloud image. only had -proposed enabled and that's it i think. [11:32] xnox: yes, with --output-dir I get some "IOError: [Errno 2] No such file or directory: '/proc/15433/fd/pipe:[967428]'" gobbledigok [11:33] (on my list, but medium prio) [11:33] xnox: ah, you are not using the adt-buildvm-ubuntu-cloud ones? [11:34] i am. === MacSlow is now known as MacSlow|lunch [11:46] pitti: no i don't get that, i believe this is because package build fails and it declares needs-building. [11:46] (fails due to test-suite error) [11:54] xnox: aaah [11:54] xnox: indeed, --shell doesn't work on package build failures [11:54] xnox: that's a good point, mind filing a bug? (LP or Debian, doesn't matteR) [11:54] pitti: removed that, changed tests, trying now with -B .// option [11:56] Bug #1317055 [11:57] dpkg-checkbuilddeps: Unmet build dependencies: build-essential:native [11:57] pitti: what do i need to upgrade to get that ^ [11:57] * xnox does a full-distupgrade [11:57] lunch time. [11:57] xnox: not released yet, I'm afraid, still in git [11:58] xnox: git clone http://anonscm.debian.org/gitweb/?p=autopkgtest/autopkgtest.git [11:58] ok... [11:58] =( [11:58] xnox: and /where/ever/run-from-checkout instead of adt-run [12:04] * ogra_ sighs about the new FF UI ... [12:05] i dont mind the new look, but why oh why did they have to make the close buttons for tabs so super tiny [12:10] ogra_: hehe, the term FF is a bit overloaded, I was pondering for a moment what ui FeatureFreeze has and then when you got a new FullFrame camera ;) [12:10] lol [12:14] Laney: uploaded into utopic, patch+NMU intent sent to debian. [12:41] xnox: haha, did he give in? [12:42] noon [12:43] Laney: no, complained once he read utopic-changes. [12:43] :D === MacSlow|lunch is now known as MacSlow [12:44] Laney: i pointed that i'm going to nmu this into debian as well as per patch on the debian bug report. [12:44] Laney: and talking to him, gave me more rebutal points to include in the patch email i sent to BTS =) [12:44] not sure that'll sooth the flames ... [12:44] ogra_: you actually use the close button on tabs? I only ever middle-click there === alkisg is now known as work_alkisg [12:44] ogra_: (or ctrl+w) [12:44] 'd' (vimperator ftw) [12:44] Laney: my plan is to duck & hide in malta and bribe doko with beer [12:45] pitti, no middle click on my touchpad ... [12:45] I don't even see close buttons [12:45] pitti: you must not be wearing ubuntu glasses [12:45] you dont have a small X in the active tab ? [12:45] xnox: anyway, thanks, that'll unblock gnome-terminal once we get the patch to bring transparency back [12:46] larsu: ↑ fyi [12:46] ah, cool [12:46] xnox: no, probably too many tabs :) [12:46] you could pop round to doko's house and massage his shoulders or something [12:46] :) [12:47] * Laney brb [12:47] Laney: i don't think that would fly at all. [12:47] someone would fly [12:52] i nominate infinity === gatox is now known as gatox_lunch [14:13] mlankhorst, looks like you have a fix! nice [14:13] tseliot, nice [14:14] :) [14:17] yeah hopefully, pending review :P [14:27] what bug was that? [14:27] the "can't dnd the uitk apps from the touchscreen"? [14:27] or the segfault one? [14:27] seb128: the former, but I can't reproduce the latter either [14:29] I had some errors in valgrind [14:30] and I can reproduce the segfault [14:30] but not under valgrind [14:32] seb128: can you try xserver-xorg-core from the same ppa, please? [14:32] what ppa is that? [14:33] seb128: https://launchpad.net/~canonical-x/+archive/x-staging [14:33] seb128: the new qt and xserver-xorg-core are what you should test [14:33] as per https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1307701 [14:34] they might fix the second bug you mentioned too [14:36] tseliot, thanks [14:38] om26er, hey, did you have slots to look at the u-s-s CI issues? [14:41] seb128, sorry I didn't look at that yesterday. Looking at that now and it seems its failing on the very start of the test. [14:41] the phase where it actually expects a page to be opened [14:42] seb128, which MR is it related to ? [14:47] om26er, it happens on all the mps for a month or so [14:47] well the settings ones [14:47] om26er, https://code.launchpad.net/~diegosarmentero/ubuntu-system-settings/ignore-updates-autopilot/+merge/218481 is a recent one [14:47] or https://code.launchpad.net/ubuntu-system-settings/+activereviews if you prefer the list [14:48] seb128, it passes on mako, only otto is failing. I'll run them on my desktop first and see if it happens or not and then contact francis [14:48] om26er, thanks [14:49] om26er, I've no issue locally either, so I'm unsure how to debug === gatox_lunch is now known as gatox === m_conley` is now known as m_conley [15:18] tseliot, mlankhorst, pmcgowan: the xserver/qt in that ppa works great for me, I can dnd the apps without issue and I didn't hit the xserver segfault yet (though I couldn't easily trigger it before, so it might be that I just got lucky so far) [15:18] seb128, awesome [15:19] seb128: goodie, I've written an xorg-integration-tests testcase too [15:19] :) [15:36] om26er,^^ can you try xserver-xorg-core from x-staging ppa and verify it fixes 1307701 for you too and that X stops crashing [15:36] it's qtbase that's important :P [15:36] om26er, or what mlankhorst said :) [15:37] the xorg-server parts probably only matters if you run in valgrind [15:38] enable the ppa and upgrade [15:38] those are the only trusty updates in there [15:58] I'll sru it tomorrow after people tested it some more, needs a small fix to prevent the bug from occuring on tablets. :P [16:10] Laney, do you some free slot to discuss the unity8-desktop image/where we start to bootstrap that? [16:12] seb128: yeah, now? [16:12] Laney, well, just IRC chatting on the channel ... now if it works for you, yes ;-) [16:12] do you have opinion on where we start? [16:13] what do you think the idea is? [16:13] is it to provide all of the applications and everything? [16:14] the goal ? or how I would start ? [16:14] mir or x? [16:14] the goal [16:14] the goal seems to be "unity8 on Mir + new core apps + some more desktop apps" [16:14] well I assume on Mir [16:14] like unity8-desktop-session-mir [16:14] Rick said it's fine to have one app on screen at time/fullscreen [16:15] bregma, there? your input might be useful in that discussion [16:15] so it's -desktop-session-mir and the apps from touch [16:16] I don't know how we'd do click apps ... [16:16] let's start by not doing click [16:16] seb128, what? Are you planning to do an ISO or something? [16:16] bregma, yes [16:16] bregma, Rick wants us to do a new desktop iso with unity8 [16:17] that's what I thought [16:17] bregma, the idea is to not disrupt too much the current desktop, so rather than being too agressive swapping pieces, we build a new iso and ramp that up [16:17] it feels like a more sensible approach to me [16:17] right, need a new metapackage to replace ubuntu-desktop [16:17] we can still do swapping but there's less pressure to get stuff out there [16:17] because it already is [16:21] bregma, basically we want to replace unity7 by unity8-desktop-session-mir right? [16:21] well I guess lot of stuff to drop as well [16:22] we don't need any of the applications that won't work yet [16:22] right [16:22] so do we start from an empty seed [16:22] or do we "derive" from some existing one? [16:23] seb128, I'd start by copying the ubuntu-desktop package, stripping anything X-related (gnome, unity) but leaving the rest, and adding unity8-desktop-session-mir, then seed that [16:23] mlankhorst, seb128 it's been hard but I'm able to reproduce 1307701 with the package from the PPA [16:23] I get Xlib: sequence lost in reply type 0x23 on the console [16:23] I'll try to find a reliable way to reproduce it [16:24] jibel, thanks, what did you do to reproduce it? [16:24] jibel, is that the same stacktracE? [16:24] seb128, no crash for the moment, just DnD is broken [16:24] jibel, oh, ok [16:25] seb128, I've 2 gallery-app, 1 gnome-terminal and 1 gedit. I'll reboot and try again [16:25] bregma, do you know what's the status of XMir? [16:25] jibel, ok [16:25] seb128, as far as I know it is not longer being developed and not really supported [16:25] ok [16:25] so not really an option if we want to get ubiquity working [16:26] the right solution is either to get the GTK Mir backend working then, or to port ubiquity to qt5 [16:26] bregma: someone from mir team was looking into getting ubiquity to run under Mir was that at all invested in? [16:26] bregma: can qt4 work under Mir? [16:26] bregma: else i'll need to look into porting ubiquity to qt5 urgently. [16:26] yeah [16:27] bregma: was that answer to 1st, 2nd, or 3rd statements? =) [16:27] bregma: or none ? =) [16:27] or just include an X server to run ubiquity on a separate session maybe? [16:27] bregma: that's what ubiquity already does by default. [16:28] then there should not be a rush to get it working on Mir [16:28] bregma: so it pre-empts lightdm, and it starts it's own X server & ubiquity-dm "display manager". [16:28] well, that image is not going to have X [16:28] bregma: if one quits that, lightdm does auto-login into whatever default is. [16:29] unless we include it just for ubiquity, as a transition [16:29] stage 1; include X, stage 2: port uniquity [16:29] seb128: seeding X just for ubiquity sounds good. I can actually go ahead and add an explicit dependency in ubiquity just in case. [16:29] well, how would that work? [16:29] like you boot the live session [16:29] end up with Mir/unity8 [16:30] how do you run ubiquity/what happens? [16:30] seb128: oh, you will not be able to start ubiquity from live session. [16:30] seb128: you can boot into installer or you can boot into live session -> the way the boot menu is currently presented on UEFI that is the case. [16:30] ok [16:30] so you have both options [16:30] test/install [16:30] but the test/liveCD can't lead to an install [16:30] that wfm [16:31] the goal is to have GTK/Mir ready this cycle anyway [16:31] will it install X then? [16:31] Laney: yes, i will add an X dependency on ubiquity-frontend-gtk. [16:31] that sounds weird [16:31] we'll be able to seed whatever we want [16:31] Laney: ok. [16:31] I was more meaning that it's not that desirable to install it in this case [16:32] but I guess it doesn't matter that much [16:32] Laney: well if i port kubuntu's qt4 frontend to qt5 that may work, but will pull in python3-qt5. [16:32] Laney: unless i use something like pyotherside. [16:32] that image is not going to be built from main [16:33] seb128: oh, cool! [16:33] so doesn't bother about component mismatch [16:33] well, at least not in a first time [16:33] we don't want to block it on promoting all the core apps [16:34] doesn't lightdm require X11 anyway? Or is there a Unity8 Greeter for lightdm I'm missing? [16:34] we don't need a greeter to autologin [16:34] but good point [16:34] we might as well keep xorg on the iso for now [16:34] I think we'll have to [16:35] I'd just rather avoid pulling it by accident if that's possible [16:35] not the end of the world for now though, for that reason and probably others [16:37] I think a good first goal is to get something you can boot and install even if it doesn't have much content beyond unity8 [16:39] and the core apps, there's no point in installing a system without sudoku [16:39] bregma: hahahahahahahaha [16:39] or, um, the terminal app [16:39] I meant to say [16:39] what's the one that is a grid of stallmans [16:40] bregma: way less funny! :p [16:40] we should open that by default [16:40] Laney: +1 [16:40] otherwise, we can't do Linux Action Shows! [16:40] but actually click packages present issues [16:40] :) [16:40] in Mir world even? [16:40] almost certainly not for the first images [16:41] for a more desktop style iso I think so [16:41] I didn't look into how they are preinstalled for touch [16:41] but I'm anticipating issues [16:41] livecd-rootfs has a hackish hook for that [16:41] ogra_: ^ [16:41] ah :) [16:41] hahaha [16:41] reassuring [16:41] faster :p [16:41] as I understand, the problem is not click packages per se but the installer back end is incompatible between apt and click [16:41] Laney: they are preinstalled into livefs and that's about it. [16:42] Laney: users can "unregister" or "upgrade" them. [16:42] so the concept of installing the image probably breaks that [16:42] Laney: that is -> make newer version my version, or make it not available to me. [16:42] the hook reads http://people.canonical.com/~ubuntu-archive/click_packages/click_list and just installs them [16:42] Laney: nah, it's all preserved. [16:42] Laney: as long as /usr/ is copied across you'd have everything afresh & registered on the installed imaged. [16:43] Laney: is this image going to use system-image updates? [16:43] is it ... [16:43] Laney: in that case ubiquity installer is pointless. [16:43] no [16:43] do you guys at least properly start with a system-image install ? [16:43] no [16:43] lame [16:43] ! [16:43] not lame [16:43] the idea is to get something which works [16:43] super lame ... [16:44] ogra_: that is patent encombared you mean "ogg" don't you? [16:44] we are going to need to sort out system-image, and how to keep apt working [16:44] clicks as well [16:44] also the question is before the tb [16:44] xnox, lol [16:44] it is pretty much out of scope [16:44]