[06:01] <hikiko> Hi
[06:36] <pitti> Good morning
[06:37] <pitti> Laney: offline updates> is that about installing at shutdown? that's a really bad idea IMHO
[06:38] <happyaron> morning hikiko, pitti
[06:41] <hikiko> hi happyaron pitti happy new chinese ywar
[06:41] <hikiko> year*
[06:41] <happyaron> ty, :P
[06:42] <pitti> hey happyaron, hello hikiko!
[07:20] <happyaron> seb128: hey seb
[07:20] <seb128> good morning desktopers
[07:20] <seb128> hey happyaron
[07:21] <happyaron> fcitx's gtk3 support requires gtk2 one to be present, but qt4 and qt5 are seperate
[07:21] <happyaron> so you want to drop qt4 from image?
[07:25] <seb128> yes
[07:26] <seb128> why does gtk3 requires gtk2 (just curious, we are not going to drop gtk2 this cycle since firefox/libreoffice still use it)
[07:36] <happyaron> because gtk2/gtk3 share the same environment variable, when the variable is set but gtk2 support isn't present, the user has no way to input anything
[07:37] <happyaron> I'll propose an MP for dropping qt4 then
[07:46] <pitti> didrocks: bonjour ! ça va ?
[07:46] <pitti> didrocks: just FYI: we can now run autopkgtests triggered by github PRs (https://plus.google.com/u/0/+MartinPitti/posts/fjGYfENHo15)
[07:47] <didrocks> hey pitti, ça va
[07:47] <didrocks> pitti: oh nice!
[07:47] <pitti> didrocks: in case this is useful for u-make
[07:47] <didrocks> pitti: the results point to the artefacts directly?
[07:47]  * didrocks looks at the post
[07:47] <pitti> didrocks: took me some time to implement the plumbing etc., but now it should be simple to enable it for more projects
[07:47] <didrocks> yeah, sounds good! :)
[07:48] <pitti> didrocks: I send the link to log.gz for now, statuses_url only accepts one result URL
[07:48] <didrocks> pitti: we should do that on Ubuntu Make after mwc
[07:50] <pitti> didrocks: yeah, I'll watch systemd PRs for the time beingn, and there's still some small cleanup to be done
[07:51] <didrocks> great! thanks for the head's up?
[07:51] <didrocks> s/\?/!/
[09:02] <Laney> word up ma gangstaz
[09:03] <happyaron> hey Laney
[09:03] <willcooke> morning all
[09:03] <willcooke> tgi Friday
[09:03] <happyaron> hey willcooke, :)
[09:05] <seb128> hey Laney willcooke
[09:05] <seb128> how is u.k today? still holding to the UE?
[09:05] <Laney> pitti: https://wiki.freedesktop.org/www/Software/systemd/SystemUpdates/
[09:05] <Laney> hey seb128
[09:05]  * happyaron supper time
[09:05] <Laney> oh yeah the news was all about this
[09:05] <Laney> hey happyaron!
[09:06] <Laney> the next fun part is that the conservatives get to shit themselves over it for a few months
[09:11] <willcooke> spent a few hours just playing with 16.04 on my Inspiron last night.  I know it's a fresh vanilla install, but man it felt fast.
[09:11] <willcooke> Found some bugs too
[09:11] <willcooke> :)
[09:12] <seb128> reported them?
[09:12] <seb128> oh, and which ones? ;-)
[09:12] <willcooke> not yet, want to check them with you first
[09:13] <willcooke> it's only a few
[09:13] <willcooke> and on of them is certainly opinion
[09:13] <willcooke> *one
[09:13] <willcooke> :)
[09:14] <willcooke> The first one is the Bluetooth "add a device" windows and the "Set up a mobile broadband connection"
[09:14] <willcooke> there is no padding it seems.
[09:14] <willcooke> Is that "just" styling?
[09:15] <seb128> unsure, needs investigation
[09:15] <seb128> but it's a gtk thing
[09:15] <seb128> it does the same in gtk3-demo -> assistant
[09:15] <willcooke> ah, oki
[09:15] <willcooke> I'll log it then
[09:15] <willcooke> which project is best?
[09:15] <seb128> gtk+3.0
[09:15] <willcooke> thx
[09:15] <willcooke> next:
[09:15] <seb128> yw
[09:16] <willcooke> In my opinion, when you use the mouse wheel over the playback position indicator in Rb, it goes the wrong way
[09:16] <willcooke> mouse wheel down goes back, up goes forward
[09:17] <willcooke> it seems the opposite in 14.04
[09:18] <seb128> up for forward seems to make sense to me?
[09:18] <willcooke> I think of it as a long document.  Up to get to the top.
[09:18] <willcooke> ain't no thing
[09:18] <willcooke> just opinion
[09:19] <willcooke> Next:
[09:19] <willcooke> LibreOffice menus.  The ugly
[09:19] <willcooke> I'll speak to Bjoern about that when he's on
[09:19] <seb128> what's wrong with those?
[09:20] <willcooke> MASSIVE right arrows (to indicate sub menus) , white border around the menu
[09:21] <seb128> are you menu integrated to unity?
[09:21] <willcooke> ah, no, that's a good point
[09:21] <seb128> you likely have libreoffice-gtk3 installed
[09:21] <seb128> try without it
[09:21] <seb128> we discussed it the other day, it got pulled in by error, going to be remove in the next upload
[09:22] <willcooke> yeah, I remember now
[09:22] <willcooke> thanks
[09:22] <seb128> yw
[09:22] <willcooke> and finally :) ...
[09:22] <willcooke> chrisccoulson, can we drop the default U1 bookmark from Firefox?
[09:22] <seb128> haha
[09:23] <willcooke> Ubuntu and Free Software -> Ubuntu One - The personal cloud that brings your digital lif..............  dead
[09:27] <seb128> willcooke, I think the gtk/padding issue is due to https://git.gnome.org/browse/gtk%2B/commit/?id=cfd85de
[09:28] <seb128> mclasen wrote on https://bugzilla.gnome.org/show_bug.cgi?id=750631
[09:28] <seb128> "Unfortunately, I'm not convinced this is compatible enough - it will visually break quite a few assistants out that that assume the padding."
[09:28] <seb128> I wonder why he commited anyway
[09:28] <seb128> sometimes their logic fails me :-/
[09:29] <pitti> Laney: ah, that's applying them on boot, not on shutdown; that's better IMHO
[09:29] <seb128> pitti, -1
[09:29] <pitti> Laney: btw, I didn't yet find out what's up with teh trusty/ppc64el tests (the "libc6 not found blabla")
[09:30] <seb128> pitti, I like better my laptop to be stucked in the evening when I stop working that having it to block me in the morning when I want to start working
[09:30] <pitti> on shutdown is just about the worst possible time
[09:30] <pitti> consider: when do people shut down their computer?
[09:30] <seb128> like I can walk away and let it shut down 10 minutes later
[09:30] <pitti> either not at all (suspend) -> then you don't get updates
[09:30] <seb128> me? every evening after work
[09:30] <pitti> or, when they actaully want to *leave*
[09:30] <pitti> the last thing you want to do then is to twiddle your thumbs for 15 minutes
[09:31] <pitti> yes, me too
[09:31] <pitti> and when I have to leave, I want to leave *now*, not wait for an indefinite time
[09:31] <seb128> well, I'm fine walking away knowing the computer is going to turn down after 10 minutes
[09:31] <pitti> and if your computer is stuck, people will just switch it off
[09:31] <seb128> there is a chance that you are not taking the computer with you
[09:31] <seb128> especially on a desktop
[09:31] <seb128> but when you turn on the computer there is all chances that it's because you want to use it
[09:32] <seb128> and not sit watching updates being applied
[09:32] <tjaalton> Trevinho: ping
[09:32] <pitti> seb128: well, that's cf "no good time"
[09:32] <seb128> right
[09:32] <pitti> TBH, I still prefer the in-place updates most
[09:32] <seb128> in session updates ftw :p
[09:32] <pitti> it hardly ever breaks even for the devel series
[09:32] <pitti> and a lot of things change there
[09:32] <seb128> right
[09:33] <pitti> for a stable release we don't reorganize files or drop packages or what not
[09:33] <Laney> you update system libraries and want them to be used
[09:33] <Laney> that's exactly where the benefit of offline updates lies
[09:33] <pitti> so holding up the rather theoretic case of "in-place might disrupt your session" against "annoy everyone who wants to leave when they shut down", is a bad trade IMHO
[09:33] <seb128> seems people are working the reverse way now though
[09:33] <seb128> they are teaching the kernel to reload bits without reboot
[09:33] <pitti> Laney: different use case
[09:33] <Laney> what?
[09:34] <pitti> Laney: servers and even many laptops are rarely booted
[09:34] <seb128> ideally you want to never have to reboot
[09:34] <pitti> so for those, applying them on boot or shutdown doesn't work anyway
[09:34] <Laney> the option in gnome-software is "restart and upgrade"
[09:35] <pitti> Laney: that actually sounds like a good solution for interactive upgrades
[09:35] <pitti> i. e. if you trigger them through a knob, if ubuntu then tells you "you need to reboot to apply this security update", then the user has control over that
[09:35] <pitti> works less well on servers, of course -- there, doing it in-line is still the least bad way IMHO
[09:35] <Laney> no
[09:35] <pitti> (inline and cron'ed)
[09:36] <Laney> I just got an email fro IS telling me to restart to upgrade glibc
[09:36] <pitti> yeah, me too (for prodstack)
[09:36] <Laney> it's the safest way to upgrade libraries
[09:36] <Laney> doesn't particularly matter at which exact point they are installed
[09:36] <Laney> the restart is the important bit
[09:37] <Laney> I should go do that actually O:)
[09:38] <Laney> good to see that landscape is working
[09:39] <seb128> :-)
[09:42] <willcooke> seb128, should I still open a LP bug?
[09:43] <seb128> willcooke, about libreoffice no, it's being handled, the gtk issue yes please, the rb one ... as you wish but I'm unsure it's a bug, would need to see what other platforms are doing
[09:43] <willcooke> @ padding ^
[09:43] <meetingology> willcooke: Error: "padding" is not a valid command.
[09:43] <seb128> yes
[09:44] <willcooke> grrrr
[09:44] <willcooke> lp timeouts
[09:44] <Laney> hmm, that didn't quite come back up cleanly
[09:45] <seb128> :-/
[09:45] <seb128> willcooke, same here
[09:45] <Laney> missing a line to mount the ceph volume
[09:46] <seb128> pitti, sorry I probably asked that in the past, but are autopkgtests supposed to autoretrigger on new versions uploads?
[09:47] <seb128> pitti, like sphinx has failed pyresample and a new pyresample version has been uploaded
[09:47] <seb128> is it going to retry by itself with the new pyresample?
[09:47] <pitti> seb128: we want them to, yes, but it's not currently working (there's a bug)
[09:47] <seb128> or does it need to be manually nudged?
[09:47] <seb128> k
[09:47] <seb128> so let me manually retry it then
[09:47] <seb128> danke
[09:52] <Laney> https://git.launchpad.net/~ubuntu-desktop/+git/appstream-cloud/commit/?id=055ab0887fedd8bbfef82751d6c75f106700b964 fix0r3d
[09:53] <tjaalton> seb128: turns out it's the same commit that broke display scaling,  and lightdm starting with nvidia prime
[09:54] <seb128> tjaalton, ah, good
[09:54] <tjaalton> dunno what unity-greeter et al do, but it doesn't bend to randr-1.5
[09:56] <hikiko> hi
[10:02] <willcooke> https://bugs.launchpad.net/ubuntu/+source/gtk+3.0/+bug/1547423
[10:02] <seb128> willcooke, thanks
[10:03] <seb128> hikiko, hey, happy friday!
[10:03] <willcooke> hikiko, did you fix your computer?
[10:06] <hikiko> yes willcooke :) in case you have rge same problrm the solution is: boot from somewhere else, run update-gtub and reboot (or copy the conf files and reboot if you boot from flash/live cd)
[10:06] <hikiko> the*
[10:07] <hikiko> (becayse it happened after dist-upgrade)
[10:07] <willcooke> aiieee
[10:10] <seb128> willcooke, want to try http://people.canonical.com/~seb128/software-properties-gtk_0.96.17_all.deb ?
[10:11] <willcooke> seb128, sure.  Did you just patch in the padding again?
[10:11] <seb128> willcooke, no, that's the proposed source thing
[10:11] <seb128> software-properties
[10:11] <willcooke> ohhhh
[10:11] <willcooke> sorry, didnt read that
[10:11] <seb128> sorry, I didn't give you context ;-)
[10:11] <willcooke> I should have read it
[10:11] <willcooke> :)
[10:13] <willcooke> seb128, perfect :)
[10:14] <tjaalton> is Trevinho out today?
[10:14] <willcooke> tjaalton, no, just starting later
[10:14] <tjaalton> ok
[10:15] <Laney> he'll be working alllllllll night long (all night)
[10:15] <willcooke> \o\ /o/
[10:15] <Laney> (supergreen)
[10:16] <willcooke> TIL: Emeinmen sampled a 70s track for "My name is"
[10:17] <Laney> MULTIPASS
[10:17] <willcooke> https://youtu.be/xKISdd2mKzU?t=124
[10:19] <willcooke> So Ubuntu Firefox bookmarks are done in a patch.  Should I make a patch for that patch, or just a whole new patch file?
[10:20] <Laney> I would branch the bzr repository and make the fix in there, then merge propose that
[10:20] <willcooke> Laney, ack, thanks
[10:21] <Laney> which seems to be lp:~mozillateam/firefox/firefox.xenial
[10:21] <willcooke> yay!
[10:21] <willcooke> I got the right one
[10:22] <Laney> looks like Vcs-Bzr in debian/control.in debian/control says wily though, you could fix that while you're there
[10:22] <Laney> ;-)
[10:22] <willcooke> Laney, just a simple substitution for xenial?
[10:23] <Laney> ya
[10:23] <Laney> it makes apt-cache showsrc firefox show the wrong thing
[10:23] <willcooke> done
[10:28] <seb128> willcooke, https://code.launchpad.net/~seb128/software-properties/newtab-enable-proposed/+merge/286626
[10:29] <willcooke> thanks seb128
[10:29] <seb128> yw
[10:31] <willcooke> chrisccoulson, check it!  https://code.launchpad.net/~willcooke/firefox/bookmark-and-control-changes/+merge/286627
[10:31] <willcooke> http://cdn.meme.am/instances/500x/62614952.jpg
[10:32] <willcooke> ohhh
[10:32] <willcooke> I think I proposed it in the wrong place
[10:33] <willcooke> undo undo
[10:33] <chrisccoulson> willcooke, you proposed it in the right place, but from the wrong branch :)
[10:33] <willcooke> ohhh
[10:33] <willcooke> so I need to take the main firefox branch?
[10:34] <willcooke> not the xenial one
[10:34] <chrisccoulson> willcooke, yeah, that's the one that we build nightlies from and where I pick changes to other branches from
[10:34] <willcooke> chrisccoulson, https://code.launchpad.net/~mozillateam/firefox/firefox-trunk.head ?
[10:34] <chrisccoulson> well, when we're actually able to build nightlies
[10:34] <chrisccoulson> yeah, that's the one
[10:35] <willcooke> thx
[10:35] <Laney> you should put something in README.source
[10:35] <Laney> or fix Vcs-Bzr
[10:35] <willcooke> you = me ?
[10:35] <Laney> the maintainer :)
[10:40] <willcooke> chrisccoulson, fixed: https://code.launchpad.net/~willcooke/firefox/bookmarks-update/+merge/286633
[10:45] <seb128> willcooke, I reported the gtkassistant padding issue as https://bugzilla.gnome.org/show_bug.cgi?id=762315
[10:46] <seb128> I've the feeling they are going to say that all the gtkassistant consumer needs to be "fixed"
[10:46] <seb128> well they at least need to fix gtk-demo in this case
[10:47] <seb128> but another annoying case of slight behaviour change which breaks existing softwares
[10:51] <willcooke> yeah
[10:52] <willcooke> Should I start adding the padding back in now?
[10:56] <willcooke> ignore, I read the bug
[11:04] <Trevinho> tjaalton: hey
[11:06] <tjaalton> Trevinho: display scaling is broken with xserver 1.18 and I bisected the commit to https://cgit.freedesktop.org/xorg/xserver/commit/?id=7e1f86d42b54fb7f6492875e47a718eaeca3069b
[11:06] <tjaalton> Trevinho: so I guess unity needs some fixing?
[11:07] <tjaalton> the same commit also broke nvidia prime (greeter crashes)
[11:08] <tjaalton> or is it u-s-d
[11:08] <tjaalton> that needs something
[11:09] <Trevinho> tjaalton: mh. Well unity handles the scaling by its own, so... What you mean by "broken"?
[11:09] <Trevinho> There's no automatic recognition?
[11:09] <Sweet5hark> seb128: in case you didnt notice the MIR/team-subscriptions thing on bug 1487148 started a nice bikeshed ... justsaying ...
[11:09] <tjaalton> Trevinho: using the slider does nothing
[11:10] <Trevinho> Well, that's pretty weird since that has nothing to do with X....
[11:10] <Trevinho> It's just a GSetting we listen and then we resize the UI based on that
[11:10] <Trevinho> what could be... Is the name of the montitors that have changed
[11:11] <tjaalton> right
[11:11] <Trevinho> what's the output for gsettings get com.ubuntu.user-interface scale-factor ?
[11:11] <Trevinho> once you changed the scaling?
[11:12] <Trevinho> So maybe there's a mismatch between them... I really wanted to change this thing anyway (so using the edid, instead of plug names). So... Maybe it's the right moment
[11:13] <tjaalton> I think the names lost the dash
[11:15] <Trevinho> So the problem might be caused by a mismatch of what GdkScreen returns and the gnome_rr code which we use to set these at U-C-C level
[11:15] <desrt> moin
[11:18] <tjaalton> Trevinho: {'HDMI2': 8, 'eDP-1': 12, 'eDP1': 10, 'eDP-0': 11}
[11:19] <andyrock> morning
[11:19] <willcooke> morning desrt andyrock
[11:20] <desrt> greetings willcooke, andyrock
[11:20] <desrt> happy friday :)
[11:23] <tjaalton> Trevinho: reset to defaults, then munged it a bit and now it's "{'eDP-0': 18}"
[11:23] <Trevinho> mh, ok, so let me see something else
[11:25] <Trevinho> tjaalton: what you get instead with python -c 'from gi.repository import Gdk; print(Gdk.Screen.get_default().get_monitor_plug_name(0))' ?
[11:25] <Trevinho> or change the plug name based on what you're using
[11:26] <tjaalton> "None"
[11:27] <Trevinho> Mh, nice
[11:27] <Trevinho> maybe gdk has to be fixed then?
[11:28] <Trevinho> tjaalton I guess print(Gdk.Screen.get_default().get_n_monitors())' gives you the right output, isn't it?
[11:29] <tjaalton> "1"
[11:29] <Trevinho> yeah, and so python -c 'from gi.repository import Gdk; print(Gdk.Screen.get_default().get_monitor_geometry(0).width)' I guess
[11:29] <Trevinho> I think gdk has some troubles with that then
[11:30] <tjaalton> "3200" so yes :)
[11:31] <tjaalton> where's gdk these days?
[11:32] <tjaalton> ah, gtk
[11:32] <Trevinho> https://github.com/GNOME/gtk/blob/master/gdk/gdkscreen.c
[11:32] <Trevinho> well, actually.. https://github.com/GNOME/gtk/blob/master/gdk/x11/gdkscreen-x11.c
[11:32] <Trevinho> (or in gnome git)
[11:33] <tjaalton> yep
[11:34] <tjaalton> no updates upstream
[11:38] <Trevinho> tjaalton: I think init_randr15 has to be fixed to set plug_name
[11:39] <Trevinho> output_name actually
[11:40] <Trevinho> This is what is set for randr 1.3 http://pastebin.ubuntu.com/15130288/ and this is for randr15... http://pastebin.ubuntu.com/15130291/ Something is missing
[11:48] <tjaalton> hmm I was looking at the wrong file from git..
[11:51] <tjaalton> but all commits from master seem to be there
[11:52] <Trevinho> tjaalton: what you mean?
[11:52] <tjaalton> nevermind, our gtk is patched with the top commit from master to that file
[11:55] <Trevinho> tjaalton: when is new server supposed to land?
[11:56] <tjaalton> Trevinho: when this is fixed :)
[11:56] <tjaalton> it's in ppa:canonical-x/x-staging
[11:56] <Trevinho> tjaalton: so... fixing gdk would be nice...
[11:56] <Trevinho> tjaalton: is there a ppa or something I can use?
[11:57] <Trevinho> without rebuilding x
[11:57] <tjaalton> 13:56 < tjaalton> it's in ppa:canonical-x/x-staging
[12:02] <tjaalton> I need to head off soon
[12:02] <tjaalton> Trevinho: should I file a bug upstream?
[12:02] <tjaalton> I don't mind if you have time to look at it :)
[12:03] <Trevinho> tjaalton: well, I can look at it, but not sure I can today...
[12:31] <willcooke> http://www.wired.co.uk/news/archive/2016-02/19/virgin-galactic-spaceshiptwo-space-unveiling
[12:39] <seb128> Sweet5hark, we subscribed desktop-bugs so it should be good anyway, right?
[12:42] <andyrock> seb128: where should i propose a debian patch for gnome-software?
[12:42] <seb128> andyrock, Debian or Ubuntu? what does it do
[12:42] <seb128> Debian BTS & launchpad is probably best
[12:42] <andyrock> Ubuntu
[12:42] <andyrock> integration with unity-launcher on the apt plugin
[12:42] <andyrock> *in the
[12:42] <tjaalton> Trevinho: ok, I'll file it upstream
[12:43] <andyrock> i cannot find the branch in launchpad
[12:44] <Sweet5hark> seb128: yeah
[12:44] <Trevinho> tjaalton: ok, post it here as well the
[12:44] <Trevinho> then*
[12:45] <tjaalton> sure
[12:45] <seb128> andyrock, launchpad then
[12:45] <Laney> robert_ancell has a branch on gnome git
[12:45] <seb128> andyrock, ^
[12:45] <Laney> so probably supply a format-patch on top of that
[12:45] <andyrock> oki
[12:45] <andyrock> thanks
[12:45] <Laney> https://git.gnome.org/browse/gnome-software/log/?h=wip/rancell/apt
[12:46] <Laney> or I guess just push your own branch if you have commit there and ask him to merge it
[12:47] <seb128> woot, new gstreamer tarballs!
[12:47] <seb128> and slomo already uploaded to Debian :-)
[12:55] <Laney> pitti: is there something wrong with the apt update?
[12:55] <Laney> you're trying to get 6.6 but 6.7 is latest
[12:56] <Laney> and the update only shows proposed being fetched
[12:56] <tjaalton> Trevinho: is this accurate :P https://bugzilla.gnome.org/show_bug.cgi?id=762319
[12:57] <Sweet5hark> seb128: FWIW, pushed some minor fixes to the libreoffice .deb to a ppa, which should go to xenial (but not on a friday)
[12:58] <seb128> Sweet5hark, ok
[12:58] <Sweet5hark> seb128: arrgh, still need to do the indic-fonts-foo
[12:58] <seb128> Sweet5hark, right
[12:58] <seb128> Sweet5hark, also you can upload on friday so it builds over the w.e, just tag a bug to block in proposed?
[13:15] <flexiondotorg_> Trevinho, I've updated a merge proposal as you requested - https://code.launchpad.net/~ubuntu-mate-dev/compiz/marco-gsettings/+merge/282882
[13:16] <Trevinho> flexiondotorg_: thanks
[13:17] <flexiondotorg_> Trevinho, I aslo have another small one for MATE - https://code.launchpad.net/~ubuntu-mate-dev/compiz/fix-1067951/+merge/283039
[13:29] <willcooke> Trevinho, andyrock hikiko, seb128 - just updated the meeting with a hangout
[13:29] <Trevinho> willcooke: ah i  was about to ask that
[13:29] <seb128> willcooke, k
[13:41] <Sweet5hark> .oO(There is a special place in hell for people who xkcd927 "abbreviations" for l10n .... "nono, in addition to the full name and a two letter l10n code, we need to have a 4 letter abbreviation. this is important.")
[13:53] <flexiondotorg_> Trevinho, Thanks for the Approvals.
[13:55] <Trevinho> flexiondotorg_: yw
[14:19] <ricotz> hello desktopers
[14:20] <ricotz> is this a known gtk3 bug in trusty? -- a GtkStack loosing its background -- https://launchpadlibrarian.net/240539411/g.png
[14:20] <seb128> ricotz, hey
[14:21] <seb128> did GTK has GtkStacj in trusty?
[14:21] <seb128> it might be due to the theme or the overlay-scrollbars
[14:21] <seb128> try to LIBOVERLAY_SCROLLBAR=0
[14:21] <seb128> and see if that fixes it
[14:21] <ricotz> yeah, it is a GtkStack (since 3.10)
[14:22] <ricotz> I didnt see it myself here, I assume it is compiz related
[14:22] <seb128> I doubt it is
[14:22] <ricotz> https://bugs.launchpad.net/ubuntu/+source/gtk+3.0/+bug/1478175
[14:22] <seb128> right, that bug is known
[14:23] <ricotz> this one claims similar is fine on trusty but happens with 3.14.x
[14:23] <seb128> right
[14:23] <seb128> which matches what I remember
[14:23] <seb128> we had those issues after trusty
[14:23] <ricotz> I see
[14:23] <seb128> maybe your user/report has a new gtk from a ppa or such?
[14:24] <ricotz> could be a usual it lacks information
[14:24] <ricotz> https://bugs.launchpad.net/plank/+bug/1547473
[14:24] <seb128> yeah :-/
[14:27] <ricotz> alright :\ , thanks :)
[14:27] <seb128> yw!
[14:59] <xnox> Laney, pitti, desrt - i can't remember who else was with us at systemd sprint, but we fixed highdpi unity at that point and login screen. However my lock screen is now partially correct - indicators and logo on the bottom left are good, but the password input is 1x dpi.
[15:00] <xnox> is lock screen hiding somewhere separate? different user and/or codepath? to fix that too?
[15:05] <seb128> xnox, it's part of unity
[15:05] <xnox> unity7 i take it, ok. will poke some time.
[15:05] <seb128> and should scale
[15:06] <xnox> i think it's just a one off miss-calculation, or a missed place.
[15:06] <xnox> or some such.
[15:06] <Trevinho> xnox: lockscreen or unity-greeter?
[15:06] <xnox> Trevinho, lockscreen.
[15:07] <xnox> Greeter is good.
[15:07] <Trevinho> Weird...
[15:07] <Trevinho> the lockscreen scaling is there since forever...
[15:07] <Trevinho> except that...
[15:07] <xnox> and it does indicators and background right, not the input field.
[15:07] <seb128> xnox, you changed the scaling factor from the control center?
[15:07] <xnox> seb128, yes.
[15:07] <xnox> (and manually hackery to twiddle greeter scaling factor)
[15:08] <seb128> that shouldn't be needed in xenial
[15:08] <xnox> by the way, it takes too long for scaling factor to kick on the greeter. it loads, small and then resizes.
[15:08] <seb128> :-(
[15:08] <xnox> seb128, autodetection was totally disabled in greeter at fosdem time.
[15:08] <seb128> I guess it takes too long for xorg to provide the infos about your screens and their resolution
[15:09] <seb128> yes, I re-enabled it since
[15:09] <seb128> we fixed the gtk menu positionning issue that made us disable it
[15:11] <Trevinho> seb128: I think it might be the u-s-d that takes longer to load..
[15:11] <Trevinho> seb128: like it gets loaded later than the greeter, might be the case?
[15:11] <Trevinho> (while it should be other way around)
[15:12] <seb128> it could be
[15:12] <Trevinho> xnox: can you check that^?
[15:12] <seb128> it shouldn't take more than 1s to load on xnox's modern laptop though
[15:12] <Trevinho> Yeah... In fact no. But then readapting the UI, rescaling... Might be visible
[15:12] <seb128> yeah
[15:12] <seb128> I don't have a good solution to that
[15:13] <seb128> out of having a file on disk with a static config that could be read
[15:13] <seb128> doing the dynamic computation is always going to be slower
[15:13] <xnox> i used to have a scale x2 theme in plymouht, but latest merges from some sales guy broke it unnoticed. I guess sales people don't have high dpi.
[15:13] <xnox> and to be honest i'm now thinking a kernel command-line parameter something like
[15:14] <xnox> dpi.scale=2x or dpi=<integer> -> such that everything keys on to that: plymouth, greeter, lockscreen, user desktops.
[15:14] <xnox> and we calculate that at install time, and record it in /etc/defaults/grub, or in one of the generators.
[15:15] <xnox> this stuff should work by default.
[15:27] <ricotz> Laney, hi, could you adjust the desktop-file patch while processing the current update of evolution? X-Ayanta-* should really be replaced with FDOs DesktopActions
[15:37] <seb128> xnox, yeah, it's a bit more complex than that though, unity has scaling by monitor and it's sort of an user preference as well, some people like tiny UI elements some don't (depend of how good your eyes as well I guess)
[15:37] <xnox> and how high the dpi is.
[15:38] <xnox> cause e.g. on 1.2x screens -> 2x is too big, and 1x can be too small. =(
[15:38] <xnox> per monitor you say, on jodh's laptop with external screen it was not per screen.
[15:38] <xnox> oooh, i can test that!
[15:39]  * xnox totally forgot that i have hdmi output on this brand new laptop, instead of mini display port.
[15:40] <seb128> well, I said "unity"
[15:40] <xnox> seb128, right, i see what you mean now.
[15:40] <seb128> that applies to the Unity shell
[15:40] <seb128> GTK does int scaling and for all monitors
[15:41] <xnox> yeah.... cause e.g. my cursor is either "normal" or "XXL Pimp My Ride" size
[15:42] <xnox> indicators look good -> but then you open the menu it's "Pimp my Ride size sub-boofers pull out"
[15:42] <xnox> seb128, why does GTK not have float scaling or per screen scaling?
[15:42] <xnox> did they really think, people will plugin "high-dpi" projects for many years to come?
[15:43] <seb128> I think the int scale is because using non int value looks less good with pixmaps&co
[15:43]  * bregma questions the sanity of the GTK designers on that point
[15:43] <seb128> the per screen might be resolved under wayland
[15:43] <seb128> I didn't look at it for a while
[15:43] <seb128> I don't remember the reason (if they had one)
[16:01] <Laney> ricotz: got patch?
[16:04] <ricotz> Laney, it should look like https://paste.debian.net/plain/399409
[16:06] <ricotz> those need similar treatment /usr/share/applications/inkscape.desktop /usr/share/applications/unity-activity-log-manager-panel.desktop
[16:12] <Laney> ricotz: and this could go upstream now too ;-)
[16:12]  * Laney wonders why no accept mail for eds
[16:15] <seb128> Laney, forgot to include the orig (-sa)?
[16:15]  * seb128 takes a guess
[16:15] <Laney> nein
[16:15] <Laney> just dak being shit I guess
[16:15] <seb128> weeeird
[16:15] <seb128> I just uploaded thing
[16:15] <seb128> oh, you mean to Debian?
[16:15] <Laney> not to the same place I did :-)
[16:16] <seb128> right
[16:16] <Laney> launchpad is a beautiful place
[16:16] <seb128> :-)
[17:29] <xnox> Laney, seb128 - so if software-centre is gone, does that mean that qt4 sso authenticator is no longer pulled in on the desktop, and it means we can try to kill qt4 from ubuntu-desktop?
[17:29] <xnox> along with gtk2?
[17:29] <xnox> (ISO and the default install)
[17:30] <seb128> xnox, yes, see #ubuntu-desktop backlog from yesterday
[17:30] <seb128> xnox, and no for gtk2, libreoffice&firefox still use it
[17:30] <xnox> ah, darn.
[17:30] <Laney> please reply to my previous ping
[17:30] <xnox> i thought there is firefox with gtk3?
[17:30] <xnox> Laney, hm, where/what?
[17:31] <Laney> xnox: could you send a second cfn for the dmb please?
[17:31] <xnox> Laney, ack.
[17:31] <Laney> thx
[17:31] <xnox> why is libreoffice still in the dark ages? =)
[17:31] <seb128> xnox, there are firefox&libreoffice ongoing gtk3 ports, they are just no LTS quality/stability yet
[17:31] <xnox> seb128, ok.
[17:31] <Laney> is qt4 just for fcitx?
[17:31] <seb128> no, cf log from yesterday :p
[17:31] <xnox> looks like fcitx-frontend-qt4 holding it up.
[17:32] <Laney> I thought you might like to tell us
[17:32] <Laney> but I'll go look up the look
[17:32] <seb128> we have fcitx/sni-qt/appmenu-qt
[17:32] <seb128> sorry was typing
[17:32] <xnox> imho, we should make qt4 core library depenend on those?
[17:32] <seb128> and looking for the seed mp from happyaron to drop fcitx
[17:32] <seb128> right
[17:32] <seb128> that's what we discussed with mitya57 yesterday
[17:32] <xnox> that way, if something needs qt4, qt4 fonts/menu/et.al get pulled in.
[17:32] <Laney> not going to get fcitx off the iso
[17:33] <seb128> Laney, -qt4
[17:33] <Laney> it is installed on demand from there, that's some work we did last cycle
[17:33] <seb128> we currently seed allfrontends
[17:33] <xnox> seb128, i wish there was "WantedBy" equivalent for package dependencies =)
[17:33] <seb128> which depends on gtk2/3 & qt4/5
[17:33] <seb128> Laney, https://code.launchpad.net/~happyaron/ubuntu-seeds/drop-fcitx-frontend-qt4/+merge/286613
[17:33] <seb128> xnox, ^
[17:34] <Laney> ok
[17:34] <seb128> xnox, Laney, it's a bit annoying for sni-qt/appmenu-qt
[17:34] <seb128> we don't have conditional recommends
[17:34] <seb128> so either we recommend them and have other flavor to pull those in
[17:35] <seb128> or we suggests and users are going to miss them and have their qt4 apps not integrated
[17:35] <xnox> but like mumble is qt4, no? and so is like skype.
[17:35] <Laney> does it have loads of deps or something?
[17:36] <xnox> seb128, i thought appmenu-qt is an "inert" package, if one doesn't have appmenus in the right desktop it does nothing.
[17:36] <Laney> seems like fcitx is going to be more annoying to handle
[17:36] <xnox> no idea what sni-qt or fcitx is like.
[17:36] <seb128> Laney, what's the issue with fcitx?
[17:36] <Laney> how do you give people the qt4 thing?
[17:36] <seb128> ah, that's what you mean
[17:36] <seb128> I though you meant dropping the dep from fcitx
[17:36] <seb128> yes indeed :-/
[17:36] <Laney> it's nice now
[17:37] <Laney> that it is only installed if you pick chinese
[17:37] <xnox> i have an idea.
[17:37] <xnox> we have this ubuntu-drivers-common thing, that can totally checkout which packages one has, and pull those qt4 things if a certain set of stuff is installed.
[17:37]  * xnox hides
[17:38] <Laney> you can do this kind of thing with language-selector
[17:38] <xnox> that too.
[17:38] <Laney> I'm not exactly sure when that gets triggered though
[17:38] <Laney> every login?
[17:38] <xnox> some say qt4 is a language, and others say his name is stig.
[17:39] <Laney> # Column 3: dependency package(s). Only install the package in column 4 when this dependency package is already installed.
[17:39] <Laney> GunnarHj is never around when you need him :P
[17:41] <seb128> I don't think it's a service starting at logging
[17:41] <seb128> at least I hope not
[17:41] <seb128> it's python and slow :p
[17:41] <Laney> what triggers it?
[17:41] <seb128> you manually starting language-selector
[17:42] <seb128> I wonder if we have a one time thing autostarted on first logging as well though
[17:42] <seb128> no autostart in the deb
[17:42] <seb128> hum, I would need to test again
[17:43] <xnox> i thought it was an upstart session thing, no?
[17:43] <seb128> not afaik
[17:43] <xnox> hm, don't see it either.
[17:43] <Laney> well that would be good
[17:44] <seb128> what would that do?
[17:44] <seb128> check if you have libqt4-something
[17:44] <Laney> Maybe something like an apt/dpkg post hook thingy to run check-language-support
[17:44] <seb128> and suggest that you install sni-qt/appmenu-qt/fcitx-qt4 (if fcitx is installed) if you do
[17:44] <seb128> ?
[17:45] <Laney> then show the incomplete dialog if needed
[17:45] <seb128> well, check-language-support doesn't have a doesn't ask again I think
[17:45] <seb128> it would need to win that
[17:45] <Laney> a what sorry?
[17:45] <seb128> a way to tell him to shut up :p
[17:45] <seb128> like I don't install the full packs on purpose for some of the locales
[17:45] <seb128> and I don't want to be prompted after every package install or loging
[17:46] <Laney> fine
[17:46] <Laney> but it seems like a feasible problem to solve
[17:46] <seb128> yeah
[17:46]  * Laney wonders if a language guy wants to work on that ;-)
[17:47]  * Laney just ate a fork full of hot sauce
[17:47]  * Laney dies
[17:47] <seb128> I wonder if making some qt4 lib recommends appmenu/sni-qt and having the chinese install pull in fcitx-qt4 would be good enough for this cycle
[17:47]  * seb128 hands some bread to Laney
[17:47] <Laney> depends how much you want qt4 off the iso
[17:47] <Laney> or if just out of the default install is ok
[17:47] <seb128> out of the iso is nice
[17:47] <xnox> seb128, yeah i don't think people will complain about a stray fcitx-qt4 / sni-qt / appmenu-qt too much.
[17:47] <seb128> but we don't need it there to have ubiquity to download it if you are online
[17:48] <xnox> Laney, the size decrease will be lovely, plus less obsolete / CVE buggy packages on the iso.
[17:48] <Laney> meh, undoing our cool work
[17:48]  * xnox wants libqt4core or what not, just recommend those three packages.
[17:48] <Laney> xnox: what's the difference?
[17:49] <Laney> in 'live' on the iso or installed on demand by language-selector
[17:49] <Laney> slash recommended by qt4
[17:49] <Laney> they should be as supported as each other
[17:50] <xnox> Laney, either langauge-selector or depends route, will get the qt4 stack off the ubuntu-desktop iso. And both routes will make sure the three extra packages are there, if and when any qt4 app is installed (e.g. skype).
[17:50] <xnox> however language-selector is quite obscure, and i don't think it will run e.g. straight after "apt install skype"
[17:51] <xnox> in that sense recommends is the most reliable thing to do.
[17:51] <Laney> we were just talking about fixing that
[17:51] <xnox> or we can continue ship it on the CD, and thus force install qt4 by default, when no apps use it anymore by default. and only obscure ones remaining.
[17:52] <Laney> I buy CD size arguments but not support arguments if we have it pulled in by some other mechanism
[17:52] <xnox> even if langauge-selector is fixed, it is still kind of an abuse to install appmenu-qt via language selector =)
[17:52] <Laney> fcitx-qt4
[17:52] <Laney> appmenu-qt and sni-qt would be recommended
[17:52] <xnox> i'd be ok with that.
[17:53] <robert_ancell> seb128, should bug 1547608 say "without actions" ?
[17:53] <robert_ancell> or "uses notifications with actions"
[17:53] <Laney> probably "uses"
[17:53] <seb128> robert_ancell, uses
[17:53] <seb128> robert_ancell, I knew I should have titled it "shouldn't use actions in notifications"
[17:54] <seb128> would have been easier to read :p
[17:54] <robert_ancell> seb128, watch my renaming skillz :)
[17:54] <seb128> lol
[17:55] <robert_ancell> seb128, is the issue the buttons don't work in notify-osd or they're not appropriate?
[17:56] <seb128> robert_ancell, they look like that http://i.stack.imgur.com/FQjdC.png
[17:56] <seb128> it's just not a nice experience
[17:56] <seb128> also "launch" didn't work in this case
[17:56] <seb128> but maybe that's specific to vim and it tried to start the command line utility or something
[17:56] <Laney> you get a cool persistent notification in gnome shell :)
[17:56] <seb128> one with buttons
[17:57] <Laney> no
[17:57] <seb128> no "launch" button?
[17:57]  * Laney screams at the huge Shutdown icon in the dash now
[17:57] <seb128> I wonder why notify-osd acted like that then...
[17:57]  * Laney is getting a screenshot
[17:57] <Laney> if it's still there
[17:57] <seb128> now we know that Laney uses gnome-shell :p
[17:58] <Laney> virt-manager baby
[17:58] <Laney> that's why I saw that grim icon
[17:59] <Laney> http://people.canonical.com/~laney/weird-things/NICE.png
[18:00] <seb128> ah, cool
[18:04] <seb128> ohoh, g-s segfault!
[18:05] <seb128> it doesn't like to be denied polkit auth, every few tries it goes down
[18:05] <seb128> I guess I'm going to report some other bugs next week
[18:05] <seb128> robert_ancell, btw I reported https://bugzilla.gnome.org/show_bug.cgi?id=762314 upstream
[18:06] <robert_ancell> seb128, nice
[18:17] <seb128> hum I've aptdaemon warning in my unity log that could be due to gnome-software as well
[18:17] <seb128>   File "/usr/lib/python3/dist-packages/aptdaemon/client.py", line 623, in _on_property_changed
[18:17] <seb128>     self.emit("progress-details-changed", *value)
[18:17] <seb128> TypeError: could not convert type dbus.Double to (null) required for parameter 4
[18:18] <seb128> going to play more with that next week
[18:18] <seb128> but enough for this week
[18:18] <seb128> have a good w.e everyone!
[18:18] <willcooke> cheers seb128
[18:19] <willcooke> off myself now.  I'll y'all a post card
[18:19] <willcooke> toodles
[18:21] <Laney> have fun!
[18:37] <Trevinho> tjaalton: I've the fix for gdk... Did you open the bug then?
[18:38] <Trevinho> oh, sorry... Missed the link
[18:43] <Laney> happy weekend!
[18:54] <Trevinho> seb128: get ready to backport https://bugzilla.gnome.org/show_bug.cgi?id=762319 this next week 😃
[18:54] <Trevinho> Happy we anyone
[20:01] <tjaalton> Trevinho: whee, great!
[20:06] <robert_ancell> seb128, Is there an appropriate method to check the notification system? GNotification doesn't seem to have any API for that
[20:28] <robert_ancell> andyrock, is there a LP bug for the GNOME Software Launcher integration?
[20:28] <robert_ancell> And is that patch all that is required for G-S or will there be more patches
[20:28] <robert_ancell> I'm writing the debian/changelog for it, looking to attach the change to a bug number
[21:26] <Sweet5hark> re
[21:35] <robert_ancell> andyrock, I opened bug 1547676
[22:05] <robert_ancell> seb128, when do you think we should demote u-s-c in universe? Should we do it now and MIR it back in a disaster situation?
[22:05] <robert_ancell> oh, it is done I see
[22:05] <seb128> robert_ancell, libnotify has an api to query the server capabilities, dunno about gnotification
[22:05] <robert_ancell> Moving your trello card...
[22:06] <seb128> robert_ancell, right, things get semi-auto demoted when nothing keep them in main
[22:06] <robert_ancell> huh, I did not know that.
[22:06] <robert_ancell> Useful
[22:07] <seb128> robert_ancell, rb code example
[22:07] <seb128> 		/* ask the notification server if it supports actions */
[22:07] <seb128> 		caps = notify_get_server_caps ();
[22:07] <seb128> 		if (g_list_find_custom (caps, "actions", (GCompareFunc)g_strcmp0) != NULL) {
[22:07] <robert_ancell> seb128, thanks, that's just what I was looking for
[22:07] <seb128> yw
[22:07] <robert_ancell> Laney, the "set org.gnome.software require-appdata false" action - is that just an override file in the gnome-software package?
[22:08] <seb128> robert_ancell, we tend to use ubuntu-settings for overrides that are ubuntu choices
[22:08] <seb128> rather than patching the source
[22:08] <robert_ancell> ok
[22:08] <seb128> easier to maintain and let derivatives opt out
[22:08] <robert_ancell> yep
[22:10] <robert_ancell> seb128, do you know what the require-appdata option is for/does?
[22:11] <robert_ancell> I'm guessing that shows all apps, even those with broken data?
[22:11] <seb128> I don't, but that seems it could be that