[05:59] <hikiko> hi
[09:00] <willcooke> morning desktoppers
[09:00] <willcooke> hopefully back to normal today
[09:03] <flocculant> hi willcooke
[09:03] <willcooke> morning flocculant
[09:04] <Laney> hi!
[09:05] <willcooke> morning Laney
[09:06] <Laney> wb willcooke
[09:06] <Laney> you ok?
[09:07] <willcooke> Laney, I'm good thanks.  Boy #1 had some concerns last week, but the mighty NHS swung in to full speed and he's back at school today (for now, anyway :) )
[09:07] <willcooke> rather, we had some concerns with...
[09:07] <davmor2> willcooke: who's the best person to ping repeatedly till upgrading from 14.04.4 works?
[09:08] <willcooke> davmor2, depends on what the problem is.  If it's an upgrader or an app issue
[09:08] <willcooke> davmor2, if the upgrader, then I think foundations own that still
[09:08] <davmor2> https://bugs.launchpad.net/ubuntu/+source/ubuntu-release-upgrader/+bug/1550741
[09:10] <willcooke> yeah, pretty sure thats foundations.  Laney - is that right?  ^
[09:10] <seb128> good morning desktopers
[09:10] <willcooke> hey seb128!
[09:10] <seb128> hey willcooke, wb!
[09:10] <willcooke> thanks seb128
[09:25] <willcooke> davmor2, yeah, it's a foundations thing.  bdmurray is your go to guy
[09:26] <Laney> willcooke: bdmurray has already posted on the bug so yeah
[09:26] <Laney> hi seb128 !
[09:26] <davmor2> willcooke: ah cool I know he started commenting on it and then it all kinda went quiet so was double checking it hadn't been batted over to your team :)
[09:26] <seb128> hey Laney!
[09:26] <seb128> no
[09:26] <seb128> but it's worth giving him a ping to know if he's still looking at it
[09:27] <seb128> or maybe mvo knows off hand what could trigger a such error
[09:27] <mvo> hm?
[09:27] <mvo> davmor2: context is  https://bugs.launchpad.net/ubuntu/+source/ubuntu-release-upgrader/+bug/1550741?
[09:28] <seb128> mvo, yes
[09:31] <seb128> Laney, could you look at bug #1553156 / comment? The change makes sense but I wonder if that means we need to fix all packages that ship a .desktop in their -common or we can somehow workaround that for xenial
[09:33] <ksamak> hey all
[09:35] <mvo> davmor2: what do I need to do to reproduce? a chroot seems to be enough
[09:36] <davmor2> mvo: well I reproduce it in kvm and on hardware, install of 14.04.4 on kvm update and run update-manager -d and it bulks out on step 2 of the upgrade process
[09:40] <mvo> davmor2: nice, I can reproduce now
[09:42] <davmor2> \o/
[09:45] <Laney> seb128: what kind of workaround?
[09:45] <seb128> Laney, I don't know, I don't understand what part of the stack get confused by the -common
[09:46] <seb128> or why moving it to the main binary fixes things
[09:46] <seb128> I though maybe you would
[09:46] <seb128> if not I need to grab robert_ancell tonight
[09:49] <Laney> it is at least listed in Debian
[09:49] <seb128> so maybe an issue with the apt backend?
[09:51] <willcooke> Laney, seb128 - you know that Firefox extension you guys have which lets you query LP with a URL like lp:12345 - do you know if there is such a thing for Chromium?
[09:52] <Laney> that's just visiting pad.lv/123456
[09:52] <Laney> for me anyway
[09:52] <willcooke> niiiiiiiice
[09:52] <seb128> willcooke, it's not an extensions, it's just shortcuts to bookmarks
[09:52] <willcooke> thanks Laney
[09:58] <Sweet5hark> moin
[09:59] <seb128> hey Sweet5hark
[09:59] <Sweet5hark> mvo: can you have a look at bug 1554015. it looks suspiciously like it could be fallout from your recent changes ...
[09:59] <Sweet5hark> seb128: heya!
[10:00] <mvo> Sweet5hark: that is more something for sergiusenus, I don't do snapcraft myself
[10:01]  * Sweet5hark is not feeling too well today. Running nose and a slight headace ...
[10:03] <Laney> seb128: I commented, need to see what others say
[10:04] <Laney> I think his analysis of why it's not showing is a bit off if it shows on Debian too
[10:04] <Laney> but is buggy anyway
[10:04] <seb128> Laney, thanks
[10:05] <seb128> willcooke, Laney, in chromium just right click on the url bar, pick the "edit search engines" and you can add a line
[10:05] <seb128> like
[10:05] <seb128>  "launchpad package" "lpc" "https://launchpad.net/ubuntu/+source/%s"
[10:05] <seb128> then you do "lpc gedit"
[10:06] <seb128> and you get the gedit source package
[10:06] <Laney> I have that for "lp pkg" already
[10:06] <Laney> ok with pad.lv for other stuff
[10:06] <Laney> http://pad.lv/mps/ubuntu-system-settings
[10:07] <Laney> looking a bit backed up ;-)
[10:07]  * seb128 learns about pad.lv
[10:07] <seb128> but I've also things for other websites
[10:08] <seb128> like "bzg <source>" list the gnome bugs for source
[10:08] <Laney> nod
[10:08] <Laney> those are useful
[10:09] <seb128> anyway that was just replying to willcooke's question ;-)
[10:11] <mvo> davmor2: I commented in the bug, I think I know what is going on
[10:13] <davmor2> mvo: \o/
[10:14] <mvo> davmor2: I did not say its easy to fix ;) but at knowing is better than not knowing
[10:15] <davmor2> mvo: yeah it's only critical for final beta so there is like a week or so :)
[10:15] <Laney> how did we get an uninstallable module-init-tools?
[10:16] <davmor2> Laney: magic
[10:16] <Laney> there is no magic only bugs and humans
[10:16] <seb128> somebody's finder slipped?
[10:16] <mvo> Laney: I have no idea, module-init-tools is in universe right now
[10:17] <Laney> NBS
[10:17] <davmor2> Laney: combine bugs and humans and you get magic :P
[10:17] <seb128> that package isn't there since precise?
[10:17] <seb128> or am I looking at the wrong thing?
[10:17] <Laney> https://launchpad.net/ubuntu/+source/kmod/22-1ubuntu1
[10:17] <seb128> I was :p
[10:17] <seb128> thanks
[10:17] <mvo> Laney: if you could wait a tiny bit with the fix that would be great
[10:18] <mvo> Laney: its such a convinient test case for the apt issue
[10:18] <Laney> I wasn't going to fix it :P
[10:18] <Laney> but now I really won't!
[10:18] <seb128> lol
[10:19]  * mvo hugs Laney and seb128
[10:21] <mvo> good thing that I wait for a snappy image build and can hopefully fix the apt thing before the build is done :P
[10:23]  * seb128 hugs mvo back
[10:25] <davmor2> mvo: well if you didn't break it in the first place ;) hugs mvo too
[10:25] <Sweet5hark> seb128: I unfortunately wasnt very helpful on bug 1510198 so far due to other business (mea culpa), but gunnarhj is asking for sponsoring now and it looks good to me. Could you maybe have a look?
[10:26] <seb128> Sweet5hark, what is needed? is the ffe granted? if not you need somebody from the r-t which I'm not
[10:27] <Sweet5hark> seb128: hmm, I considered this a FFe grant: https://bugs.launchpad.net/ubuntu/+source/libreoffice-dictionaries/+bug/1510198/comments/18 by pitti?
[10:28] <seb128> Sweet5hark, indeed, sorry I didn't bother reading all 26 comments
[10:28] <seb128> so what action do you need? ;-)
[10:28]  * seb128 lazy
[10:29] <Sweet5hark> seb128: as per comment #20, gunnarhj wants the stuff in https://launchpad.net/~gunnarhj/+archive/ubuntu/lo-dicts2 sponsored.
[10:29] <seb128> Sweet5hark, thanks
[10:30] <Sweet5hark> (and yeah, thats a lot of noise on the issue, sorry about that)
[10:44] <Sweet5hark> libreoffice snap build at 9 hours and 11 GB  and counting ....
[10:46] <seb128> is that slower than building the deb? if so why?
[10:47] <Sweet5hark> seb128: building l10n and product in one build isnt what we are doing in *.deb these days
[10:47] <seb128> right, well it's the equivalent of the 2 source packages in deb
[10:48] <Sweet5hark> (also building in a xenial VM on big bertha, I have no good estimates on the performance of that. I still assume it to be faster than e.g. building on a quad-core natively ...)
[10:53] <Sweet5hark> (VM got 16 cores and still has RAM free)
[11:01] <desrt> happy tuesday
[11:02] <willcooke> morning desrt
[11:02] <seb128> hey desrt!
[11:04] <desrt> hello eŭropanoj
[11:05] <Laney> desrt: THAT'S MORE LIKE IT!
[11:05] <desrt> :)
[11:06] <desrt> now i understand the tsk
[11:06] <desrt> i went to bed nice and early last night =)
[11:07] <davmor2> desrt: yeah but bed and sleep are to totally different things :D
[11:07] <willcooke> mia ŝvebŝipo plenas de angiloj.
[11:08] <desrt> s/de/per/ probably, but otherwise quite good :p
[11:08] <willcooke> Google translate ftw :)
[11:08] <Sweet5hark> dudes, Im still clinging on to learn french in case seb128 does a coup in the desktop team and it will be the official team language. do I have to learn esperanto too now, just in case?
[11:09] <willcooke> Surely you *only* need to learn Esperanto, ever.
[11:09] <Sweet5hark> willcooke: xkcd927
[11:10] <willcooke> :@D
[11:10] <willcooke> :D
[11:27] <desrt> willcooke: esperanto helps one learn french faster.  it's scientific fact :p
[11:27] <willcooke> :)
[11:27] <desrt> Sweet5hark: this comic is more accurate than you realise
[11:29] <desrt> Sweet5hark: back in the day esperanto was being considered as adoption by the international association of academies as their working language... but it had some "issues"
[11:29] <desrt> so they decided "well... it's a constructed language... we'll just fix those issues, and within a few years we'll all be speaking reformed esperanto!"
[11:29] <desrt> you can imagine how well that went
[11:30] <desrt> mterry actually learned the 'reformed' version...
[11:32] <Sweet5hark> desrt: heh, awesome.
[11:32] <desrt> if you think the libreoffice/openoffice fight was epic, you have no idea... :p
[11:42] <Sweet5hark> desrt: krkr, "Im writing my _reformed_ esperanto in the one true editor, the vi!"
[11:43] <Sweet5hark> seb128: the good news is: I now why the build took ages. the bad news is: make was endless looping over translations ...
[11:43] <Sweet5hark> s/now/now know/
[11:43] <seb128> good
[11:44] <seb128> well "good" ;-)
[11:44] <Sweet5hark> seb128: possibly GNU make 4.1 regressed for this scenario vs. older versions ...
[11:45] <seb128> downgrade and see in another 9 hours ;-)
[11:46] <Sweet5hark> seb128: hah!
[11:46]  * Sweet5hark took precautiouns.
[11:47] <Sweet5hark> I had some "make || true; make || true; make" in the recipe. So I could snip the make process and it continued ...
[11:47] <Sweet5hark> s/snip/snipe/
[11:48] <Sweet5hark> obviously not a solution, but at least might get me a result in less than 9 hours ...
[11:50] <Sweet5hark> an hey, this is a snap -- no need to be rebuildable AFAI understand it. I could also generate binaries by piping from /dev/urandom until something runnable results.

[12:00] <Laney> good luck with the GPL :P
[12:03] <davmor2> is true; make just like true lies?
[12:05] <Sweet5hark> Laney: most of LibreOffice (and this part in particular) are MPLv2 ..
[12:09] <Laney> Sweet5hark: good, you have to distribute the source for that too :D
[12:12] <Sweet5hark> Laney: meh
[12:13] <Laney> I guess the source for a /dev/urandom binary is just the binary itself though...
[12:14] <Sweet5hark> Laney: no, you have to ship the entire environment that resulted in urandom to be seeded ...
[12:14] <Laney> the universe?
[12:15] <Sweet5hark> Laney: that'll do for a start.
[12:16] <Sweet5hark> Laney: might not be enough depending on how nondeterministic quantum states really are though ....
[12:16] <Sweet5hark> Laney: to be save, ship a multiverse.
[12:26] <Laney> good job there's a 15tb ssd out now
[12:38] <willcooke> Trevinho, nice work on:  https://code.launchpad.net/~3v1n0/unity/gtk-border-radius-support/+merge/288358
[12:39] <willcooke> Trevinho, is desrt a good person to do a review?
[12:40] <willcooke> also; desrt could you review?  ^
[12:41] <Trevinho> willcooke: I think so, or seb128 if he wants to.
[12:41] <desrt> Trevinho: do you think this is destined for upstream?
[12:41] <Trevinho> desrt: no
[12:41] <desrt> why not?
[12:41] <desrt> because 'unity' is in the name?
[12:41] <seb128> Trevinho, desrt is probably better placed to review it
[12:41] <Trevinho> desrt: it's all about unity...
[12:41] <Trevinho> desrt: well, I thought they wouldn't care about that
[12:41] <desrt> i can review it for technical aspects, certainly -- and am happy to
[12:41] <desrt> i can't land it, obviously :p
[12:42] <desrt> erm.  this is a unity patch
[12:42] <Trevinho> desrt: indeed if you think upstream could benefit from that I can drop the unity names, but I guess it's something quite specific
[12:42]  * desrt expected gtk patch
[12:43] <Trevinho> desrt: gtk patch is https://code.launchpad.net/+branch/~3v1n0/gtk/unity-border-radius-support/+merge/288331
[12:43] <desrt> thx
[12:43] <Trevinho> desrt: don't mind about the unity side, andyrock can review that
[12:44]  * desrt loves reading double-diff
[12:44] <desrt> i'm not enjoying your use of XDG_CURRENT_DESKTOP :(
[12:45]  * desrt keeps the review to the MP
[12:48] <Trevinho> desrt: I've removed that, in fact :)
[12:48] <Trevinho> double-diff tricks...
[13:09] <Trevinho> wow I dint' think this would hit a news http://www.omgubuntu.co.uk/2016/03/ubuntu-fix-black-corners-around-csd-windows
[13:11] <Trevinho> seb128: since I'm there, do you also want me to work on the thing about disabling the headerbar window buttons for maximized windows?
[13:11] <seb128> Trevinho, that would be nice yes
[13:12] <Trevinho> I think that can be upstreamed also...
[13:12] <didrocks> Trevinho: ah, you got the required feedback: http://www.omgubuntu.co.uk/2016/03/ubuntu-fix-black-corners-around-csd-windows#comment-2557595803 ! :)
[13:12]  * didrocks is always scrolling until reading this kind of comment :)
[13:12] <Trevinho> desrt: a gtk setting for that i sthe way to go?
[13:12]  * Trevinho has still to read the comment, but generally avoids that :-D
[13:13] <didrocks> Trevinho: quoting it for your lazyness then: "Finally! This should've been implemented 5 *years* ago!
[13:13] <Trevinho> ... and in fact.. :)
[14:38] <Trevinho> seb128, desrt: soo... in order to make the headerbar to look properly when maximized (i.e. with inverted gradient) in unity only, I think that we need some custom theme class as well... Or people using Ambiance/Radiance in gnome/mate would get a wrong result. Is there anything nicer than that I could use?
[14:39] <seb128> not that  I know
[14:39] <Trevinho> At this point, probably adding the "unity"class to any gtk widget running in unity might be somewhat useful... Or, better te XDG_CURRENT_DESKTOP value, or something similar
[14:40] <Trevinho> Not that I like that, but themes could adapt do DE sometimes
[14:41] <alexarnaud> Trevinho: hi!
[14:41] <Trevinho> alexarnaud: hi
[14:41] <alexarnaud> Do you know what's which part of Compiz manage "usr/lib/compiz/migration/" ?
[14:42] <alexarnaud> I would like to create profile for specific needs but relacatabled in my ".override" seem not working
[14:43] <Trevinho> alexarnaud: debian/<pkg>.migrations file
[14:43] <Trevinho> alexarnaud: such as debian/compiz-gnome.migrations
[14:44] <alexarnaud> Trevinho: OK, thanks, I'll check that.
[15:18] <mvo> Trevinho: hey, does https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1554563 sound ok or should I add more details? its about adding the new /var/lib/snappy/desktop directory to unity so that snaps that install desktop files will get picked up
[15:20] <Laney> FFS
[15:20] <Laney> my desktop rang again
[15:21] <qengho> Do we have any font wizards? User reported a bug and someone who knows about font rendering should be able to say where the problem is likely to be. Bug is in Chromium in X but not in V, same version.
[15:21] <qengho> https://bugs.launchpad.net/ubuntu/+source/ubuntu-font-family-sources/+bug/1554508
[15:21] <willcooke> Laney, FYI: Community wallpaper comp is closed and the choosing has started
[15:22] <Laney> I know, I am in the middle of voting :-o
[15:22] <willcooke> Laney,Ah, nice one.
[15:22] <willcooke> qengho, known issue.  We're going to revert that version of the font if the fix isn't forthcoming
[15:23] <qengho> Ah, okay.
[15:30] <willcooke> oh
[15:30] <willcooke> well
[15:30] <willcooke> #startmeeting Desktop Team Weekly Meeting - 2016-03-08
[15:30] <meetingology> Meeting started Tue Mar  8 15:30:40 2016 UTC.  The chair is willcooke. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
[15:30] <meetingology> Available commands: action commands idea info link nick
[15:30] <willcooke> Roll call:  andyrock, attente, desrt,  dgadomski, fjkong, happyaron, hikiko, laney, qengho, seb128, sweet5hark (sick), themuso (out), tkamppeter (out), trevinho, robert_ancell (out)
[15:30] <desrt> word.
[15:31] <seb128> _o/
[15:31] <dgadomski> o/
[15:31] <Sweet5hark> aye
[15:31] <qengho> hi, y'all
[15:31] <hikiko> hello
[15:31] <andyrock> Hey
[15:32] <willcooke> I'll just give others a couple of mins
[15:32] <willcooke> in the meantime, I should be back to normal now
[15:32] <desrt> willcooke: welcome back :)
[15:32] <qengho> We were starting to worry!
[15:33] <hikiko> :D
[15:33] <willcooke> Had a day of being ill and then Boy #1 was sick involving trips to hospital.  We're on the mend now though
[15:33] <willcooke> thanks to Seb for minding the shop
[15:33] <Trevinho> o/
[15:33] <FJKong> hi~
[15:34] <willcooke> Let's get this show on the road....
[15:34] <willcooke> #topic andyrock
[15:34] <andyrock> Hey
[15:34] <andyrock> I ve been working on the suspend lockacreen bug
[15:34] <andyrock> Kind of blocked
[15:35] <andyrock> It still happens after my fix
[15:35] <willcooke> booo.  Anything I/we can do to help?
[15:35] <andyrock> I m trying to understand if we are using login1 api in the wrong way
[15:35] <andyrock> Not yet
[15:36] <willcooke> oki, let me know if that changes
[15:36] <andyrock> I m studying gnome screensaver code right now

[15:37] <willcooke> thanks andyrock
[15:37] <willcooke> #topic attente
[15:37] <attente> hi all, not much from me
[15:37] <attente> just fixing some easy bugs in gnome-software, and debugging some other issues like menus but without much luck with reproducing
[15:38] <attente> (eof)
[15:38] <Laney> you want a gnome-software bug to work on?
[15:38] <Laney> :)
[15:38] <seb128> attente, I filed some bugs this week ..
[15:38] <seb128> Laney snapped me it seems :p
[15:38] <willcooke> :)
[15:38] <willcooke> please prioritise g-s bugs for now attente
[15:38] <attente> sure
[15:38] <Laney> https://bugs.launchpad.net/ubuntu/+source/gnome-software/+bug/1554023
[15:38] <willcooke> \m/
[15:38] <willcooke> thanks
[15:39] <willcooke> #topic desrt
[15:39] <desrt> hey
[15:40] <desrt> had a lot of administrative stuff on my plate this week (and a couple of doctors appointments), so not a full plate of interesting stuff... mostly reviewing and some small patching for (not too serious) security issues
[15:40] <desrt> looking at the gtk patches for the headerbar shadows in unity today
[15:41] <desrt> that's all, really
[15:41] <willcooke> thx desrt
[15:41] <willcooke> #topic dgadomski
[15:41] <dgadomski> hey, not much in the desktop area this week:
[15:41] <dgadomski> * reported bug 1554004, backported a fix. If I get a confirmation that it works I will prepare a debdiff
[15:42] <dgadomski> eof, thanks
[15:42] <willcooke> cheers dgadomski
[15:43] <willcooke> dgadomski, will reply to your mail soon
[15:43] <willcooke> #topic FJKong
[15:43] <FJKong> hi
[15:43] <FJKong> bug1545910
[15:43] <FJKong> Cannot input Chinese characters into TextField in a QML application
[15:43] <FJKong> verify in 15.10
[15:43] <FJKong> sougou IM
[15:43] <FJKong> 1 can't login to desktop after install IM
[15:43] <FJKong> resolved
[15:43] <FJKong> 2 After install sougou IM, it not shown in  panel
[15:43] <FJKong> resolved
[15:43] <FJKong> eof
[15:44] <willcooke> cool, thanks FJKong
[15:44] <willcooke> #topic happyaron
[15:46] <willcooke> happyaron, you forgot to send me your updates again.  Please do it in the morning.
[15:46] <willcooke> #topic hikiko
[15:46] <hikiko> * fixed black dots bug: https://i.imgur.com/yn1sfLd.png?1
[15:46] <hikiko> * adding shadows for windows with alpha which are not shaped (gtk) - close to done
[15:46] <hikiko> EOF
[15:46] <willcooke> \o/
[15:46] <willcooke> good work hikiko, thanks
[15:46] <hikiko> :)
[15:46] <hikiko> thanks
[15:46] <willcooke> #topic Laney #
[15:46] <willcooke> #topic Laney
[15:46] <Laney> #
[15:46] <Laney> • Upload new default wallpapers
[15:46] <Laney> • appstream:
[15:46] <Laney> ∘ Worked with IS on 'split' solution, stuff is being created now
[15:46] <Laney> ∘ 'Time' support is merged into output, allowed me to confirm that LP is buggy and not serving up current data. https://bugs.launchpad.net/launchpad/+bug/1554535
[15:46] <Laney> ∘ Search through dependencies for icons, going round feedback with ximion on that one, hopefully be fixed soon.
[15:46] <Laney> ∘ Add a count to the output so you can guess at how much of your life is going to be over by the time it finishes
[15:46] <Laney> • Updates: evolution (dropped some delta) gtk+2.0 gst-*
[15:47] <Laney> ∘ todo last I checked: gvfs gnome-terminal
[15:47] <Laney> • Fix emacs24 activity-log-manager for appstream, try to fix banshee but that is more widely broken
[15:47] <Laney> • A few FFe reviews
[15:47] <Laney> • Do some rebuilds for poppler transition - calligra is ICEing, need to fwd that upstream and no doko to help :(
[15:47] <Laney> • gnome-calendar: Fix a crash, doing the rounds upstream now
[15:47] <Laney> • Performance review for self & 360s for others
[15:47] <Laney> • A man came to fix a broken radiator and it was a stuck pin that required 2 seconds and 0 tools to fix #diy #fail
[15:47] <Laney> • Window cleaner is currently cleaning round the back but I didn't unlock the gate for him...
[15:47] <Laney> ⅆ
[15:47] <willcooke> :)
[15:47] <willcooke> superb, thanks Laney
[15:47] <willcooke> #
[15:47] <willcooke> #topic qengho
[15:48] <qengho> - working on upstream Cr release 49.0.2623.75
[15:48] <qengho> - ^ should fix default-browser prompt too
[15:48] <qengho> - snap packaging. working on security model with #security. seccomp sandbox conflict and snappy features.
[15:48] <qengho> EOF
[15:49] <willcooke> thanks qengho
[15:49] <willcooke> #topic seb128
[15:49] <seb128> • working mostly on gnome-software & gnome-calendar bugs (reports&fixing)
[15:49] <seb128> • NEW reviews for ubuntu-snappy/new a11y component/some other items
[15:49] <seb128> • some extra playing with nautilus/gtkplacesidebar, still no luck
[15:49] <seb128> • desktops bugs triage & easy updates
[15:49] <seb128> • snappy discussions

[15:51] <willcooke> thanks seb128
[15:51] <willcooke> #topic Sweet5hark
[15:51] <willcooke> - Bumping LibreOffice snap from 5.0 to 5.1
[15:51] <willcooke> - Bumping LibreOffice snap from developer build to something more releasish
[15:51] <willcooke> - found, triaged and worked around bug 1554015
[15:51] <willcooke> - Raised questions for snappy/snapcraft:
[15:51] <willcooke> -- How do we ship Frameworks like JVMs?
[15:51] <willcooke> -- How will we support debugging/stacktraces etc. of code in snaps?
[15:51] <willcooke> -- Will a11y work?
[15:51] <willcooke> -- Will printing work?
[15:51] <willcooke> -- Developer story: How do people do incremental builds use IDEs etc. on code in snaps?
[15:51] <willcooke> -- snapcraft parts and buildtime deps
[15:51] <willcooke> - Library snaps as a solution for extensions is blocking on the Frameworks question
[15:51] <willcooke> #topic TheMuso
[15:51] <willcooke>  * Filed a main inclusion report for a11y-profile-manager, which has since been approved: bug 1552507.
[15:51] <willcooke>  * Started looking into bug 1386379. I have a few different combinations of hardware here I can try with, need to set them up and see if I can try and reproduce this and go from there.
[15:51] <willcooke>  * Dug out my nexus 4 to flash it with latest touch to attempt to help debug some audio related bugs on the phone.
[15:52] <willcooke> #topic tkamppeter
[15:52] <willcooke> - system-config-printer: Drop hard dependency on Samba client components as they depend on Python2, added facility to install these components on-the-fly when needed.
[15:52] <willcooke> - Finalized trip booking for OpenPrinting Summit 2016
[15:52] <willcooke> - Google Summer of Code 2016
[15:52] <willcooke> - Tests for possible new concept for auto-downloadable driver packages.
[15:52] <willcooke> - Bugs.
[15:52] <willcooke> #topic Trevinho
[15:53] <desrt> too busy taking pictures of windows
[15:53] <Trevinho> no, laptop hanging
[15:53] <Laney> hanging out of the window
[15:54] <Trevinho> · Gtk patch for exposing windows borders radius via an atom if supported by the WM (unity): under review
[15:54] <Trevinho> · Unity patch for properly handling these clientside decorations by shadowing them: under review (andyrock and hikiko please check it)
[15:54] <Trevinho> · Gtk patch for hiding window buttons when maximized, and looking at getting better headerbar theming in unity
[15:54] <Trevinho> · Meeting with Kylin guys, reviewing their bottom launcher / dash patches
[15:54] <Trevinho> · Prepared a new unity landing
[15:54] <Trevinho> · Some awesome skiing in the Alps.
[15:54] <Trevinho> 
[15:54] <willcooke> hah
[15:54] <willcooke> good stuff, thanks Trevinho
[15:54] <willcooke> #topic robert_ancell
[15:54] <willcooke> GNOME Software... as usual..
[15:54] <willcooke> #topic willcooke
[15:54] <willcooke> So yeah, been off and stuff
[15:55] <willcooke> Reviews are nearly doner
[15:55] <willcooke> done
[15:55] <willcooke> about half way through
[15:55] <willcooke> I will finish them, check them, then "reveal" them
[15:55] <willcooke> dunnn dunnn dunnnnnnnnnn
[15:55] <willcooke> then we can talk it through in our 1:1s next week
[15:55] <willcooke> Also, I really need to get on with finishing those theming issues
[15:56] <seb128> +1
[15:56] <willcooke> and snappy work of course
[15:56] <seb128> I've some theming bugs to bounce your way as well if interested
[15:56] <seb128> like the gnome-calendar calendars list popover selection color
[15:56] <willcooke> seb128,  sure assign them to me
[15:56] <seb128> cool, thanks
[15:56] <willcooke> I might as well do it all in one go
[15:57] <willcooke> I'm going to try and set aside a day or two next week to do them, so I might need to move some 1:1s around to accomodate
[15:57] <Laney> what about linked buttons? :)
[15:57] <willcooke> accommodate
[15:57] <Laney> hope that's on The List
[15:57] <willcooke> Laney, as in do we want to keep them?
[15:58] <Laney> fix them to look good
[15:58] <willcooke> yeah, I think there is already a bug for that
[15:58] <willcooke> but I'll check when I get to it
[15:58] <seb128> yeah, that's buggy me often in gedit
[15:58] <seb128> bugging
[15:58] <Trevinho> desrt: about theming, did you read the ghing I was saying about a way to theme things based on DE, by adding a specific class or something?
[15:58] <desrt> yes
[15:58] <desrt> you said GtkSetting rather, right?
[15:58] <seb128> Trevinho, I wouldn't bother about making the unity themes look good out of unity
[15:59] <desrt> i'm thinking it makes more sense to rely on the supported atoms as advertised by the WM when it comes to shadow handling... as for everything else... ya.. what seb says
[16:00] <Trevinho> seb128: isn't ambiance also used by mate and ubuntu gnome?
[16:01] <seb128> Trevinho, I don't know and if it is it's not our problem/shouldn't stop us to make things better for us
[16:01] <seb128> they can fork and have their version
[16:01] <seb128> ubuntu gnome is using adwaita I think
[16:02] <willcooke> lemme just end the meeting..... any other business... going in 5, 4.....
[16:03] <willcooke> #endmeeting
[16:03] <meetingology> Meeting ended Tue Mar  8 16:03:00 2016 UTC.
[16:03] <meetingology> Minutes:        http://ubottu.com/meetingology/logs/ubuntu-desktop/2016/ubuntu-desktop.2016-03-08-15.30.moin.txt
[16:03] <willcooke> thanks all
[16:03] <willcooke> good meeting
[16:03] <Trevinho> desrt: Yeah, I agree with that... GtkSettings was mentioned for hiding window buttons on maximized headerbar, since mcatanzaro in a blogpost said that it was something we should do in that way...
[16:03] <desrt> ya.  for that sort of thing it makes some amount of sense.  there is a lot of precedent for that sort of thing being handled in that way.
[16:04] <Trevinho> desrt: as there's "gtk-shell-shows-app-menu" I would add "gtk-shell-shows-maximized-window-buttons" or better worded
[16:04] <desrt> more or less, yes
[16:05] <desrt> i sort of like these things to be descriptive rather than instructive
[16:05] <desrt> "just FYI... i'm gonna be showing buttons... so uh... now that you know that, please act in a way that you feel is appropriate.... (hide yr buttons!!)"
[16:06] <Trevinho> desrt: and maybe... based on that value I could also add a style class? Without bothering of checkin the DE or anything else?
[16:07] <desrt> interesting
[16:10] <desrt> i guess that's the easiest route for you to get the theme to be responsive to this sort of thing
[16:11] <desrt> but is that truly controlled by the theme?
[16:12] <Trevinho> No, well I don't want to control things from the theme... But when the shell shows maximized window buttons, the theme should be aware of that and in case act differently
[16:12] <desrt> like less border, or what?
[16:12] <Trevinho> yes
[16:12] <Trevinho> in our case also we change the gradient, so that it looks more like a toolbar
[16:13] <desrt> that makes sense, then
[16:13] <desrt> but like... what's the real gain here?
[16:13] <desrt> do you expect to patch the light themes to look good when the wm doesn't draw the buttons?
[16:14] <desrt> or do you plan to add adwaita support for unity?
[16:14] <Trevinho> better looking stuff?
[16:14] <desrt> both of those seem a bit.... unlikely
[16:14] <Trevinho> No, they already do that.. I want to patch them to look good in unity.
[16:14] <desrt> 'them' = ?
[16:14] <Trevinho> the light themes... maximized apps with headerbar in unity are currently not looking good.
[16:15] <desrt> right... but why not just make them look good?
[16:15] <desrt> i mean, why do we have to worry about the non-existing case when this setting is not set?
[16:15] <desrt> (for the buttons, sure... that's one thing...)
[16:15] <desrt> but i don't see a lot of people using light-themes on fedora
[16:16] <Trevinho> Aren't these used by mate?
[16:16] <Laney> forked
[16:16] <Trevinho> or ubuntu-gnome?
[16:16] <Laney> adwaita
[16:16] <Trevinho> Ah, ok... Well, then I can just do it in any case...
[16:16] <seb128> yeah, that's what I told earlier
[16:16] <desrt> it almost calls into question if we need the gtksetting at all
[16:16] <Trevinho> But since it's just one line...
[16:16] <seb128> and even if derivates were to use those
[16:17] <desrt> we could almost base hiding the buttons on the selection of the theme
[16:17] <seb128> if they don't like our changes they can fork and have their theme
[16:18] <Trevinho> I mean, the biggest change for us is introducing "gtk-shell-shows-maximized-window-decoration" or whatever is called, and is something upstream said would accept... Adding a style class is just one line then. Well two.
[16:18] <Laney> it's a codepath which nobody tests though
[16:18] <Laney> or a csspath
[16:18] <Trevinho> If we don't mind of that either, I can fix the thing with three or four lines
[16:18] <desrt> smaller patches = good
[16:18] <desrt> also, adding the gtksetting would also require changes to settings-daemon
[16:19] <desrt> so better if we can avoid that, too
[16:19] <Trevinho> Oooook, right
[16:19]  * Trevinho goes the ubuntu way then
[16:19] <attente> Laney: how can i test the case where there's no appstream data? is there a directory that i'm supposed to blow away?
[16:20] <desrt> Trevinho: it's not that i don't think that we shouldn't do this in a more pure/flexible/etc. way...
[16:20] <Laney> attente: the first time you install it this will happen
[16:20] <desrt> Trevinho: but it's that i actually doubt that anybody will bother to put in the effort to actually make it work in the intended way
[16:20] <desrt> ie: who is going to adapt adwaita for unity or light-themes for gnome-shell?
[16:20] <Laney> attente: I'm not sure but maybe you can kill /var/lib/app-info/ ?
[16:20] <alexarnaud> Trevinho: Is possible to change the profile like this (http://paste.ubuntu.com/15328581/), it is the replacement of "/Default/" or "/Unity/" by the name of the profile I want to create. I want to use relocatable path in GSettings context. I've looked python file, I see it make "dconf load PATH" but nothing about the convert file directly and the link between profile creation and convert file and the link between convert file and the
[16:20] <alexarnaud> compiz binary
[16:20] <desrt> so we'd be adding complexity for literally no reason at all
[16:20] <attente> Laney: ok, thanks, will try it
[16:21]  * alexarnaud is sorry fot the long message but I need to be very clear to give time
[16:21] <Laney> attente: I'd expect an 'apt update' to bring it back
[16:21] <Laney> so if it does then that seems good
[16:21] <Laney> & if you get no results then apart from installed packages
[16:22] <Trevinho> alexarnaud: I'm not sure I got that... Is that a compiz config migration file or what?
[16:23] <attente> Laney: hmm... doesn't re-populate it
[16:24] <attente> desrt: any idea what could be causing this stack trace? https://errors.ubuntu.com/problem/23ba6db9a4c30881ade44cef107a6dd5d76a41ef
[16:25] <desrt> yes
[16:25] <desrt> already talked to robert about this one
[16:25] <desrt> someone is trying to show() the window after it was destroyed
[16:25] <desrt> look at gs_application_activate in gs-application.c:683
[16:25] <Trevinho> desrt: my concern was more related to 3rd party themes though... I see many people using numix for example, so adding a class would just give them an help, and no problems to us.
[16:25] <Laney> attente: try this: sudo rm -r /var/lib/app-info/* /var/lib/apt/lists/*Components* /var/lib/apt/lists/*icons*
[16:26] <desrt> this could be related to that hide-on-delete stuff we were talking about before......
[16:26] <Laney> sudo give-laney-all-your-money
[16:26] <desrt> "just type sudo rm - r ..."
[16:28] <attente> Laney: thanks :D
[16:29]  * Laney goes to buy a porsche... bicycle
[16:29] <attente> bank account is 0, but the app-info dir came back
[16:29] <Laney> ok, so that brings you back to the case where you have no dep-11 available
[16:29] <seb128> attente, desrt ah, I manage to reproduce the g-s segfault
[16:29] <Laney> which is what happens right after you install appstream for the first time before you've upgraded
[16:29] <Laney> s/upgraded/updated/
[16:30] <Laney> like when you've done a release upgrade from trusty or wily
[16:31] <seb128> attente, try to "gnome-software --gapplication-service" then start/stop/start gain gnome-software (e.g from the dash)
[16:31] <seb128> -> segfault
[16:31] <attente> Laney: ok, so i assume at that point, there's still no appstream data, so i can just pop up a progress dialog and call org.debian.apt.UpdateCache
[16:32] <Laney> something like that
[16:32] <Laney> assuming that UI looks nice :P
[16:33]  * desrt pops out to lunch. bbiab.
[16:33] <alexarnaud> Trevinho: I need to create an multiple gsettings configuration profiles, nothing more. If it's necessary to use migration or other, well. The format used in the .convert file look good with [org.*.*:Path/_PROFILE_NAME] but I don't know how to create a new profile and use it in the gsettings context. The final goal is to have a profile to sighted person, for blind person or for visual impaired person
[16:34] <seb128> Laney, unsure if you noticed, but the gst updates are blocking on vaapi ... does that just need a no change rebuild?
[16:34] <seb128> desrt, enjoy!
[16:34] <attente> seb128: thanks, that did it!
[16:34] <Laney> I did notice and yes
[16:34] <seb128> attente, :-)
[16:34] <Laney> need to fix -bad for some missing files first
[16:34] <seb128> k
[16:35] <seb128> Laney, also, do you plan to do a gnome-calendar update? I've a fix for the high cpu usage and I'm going to do an upload but I don't want to conflict with you if you have something being worked on for the segfault
[16:37] <seb128> speaking of which, a commit for Laney is landing ;-)
[16:37] <Laney> I just pushed for that thing
[16:37] <Laney> feel free to grab it
[16:38] <seb128> thanks
[16:38] <Laney> good that they agreed in the end
[16:39] <seb128> yeah
[16:39] <Trevinho> alexarnaud: mhmh.. I think those .convert files are left-over... They were used to migrate from gconf to gsettings, and we should have removed them when dropping gconf I think
[16:40] <Trevinho> alexarnaud: what you want to do is programmatically doable by using compizconfig, also with python bindings
[16:40] <Trevinho> alexarnaud: as it's done by ccsm
[16:41] <Trevinho> If you go in preferences, you can create and load profiles on the fly... So I think you can do the same
[17:24] <attente> Laney: not sure how to determine if the appstream data actually needs a refresh... i can't just scan through the list of apps because there are some firmware packages which are always listed and i can't reliably tell whether they're installed or not
[17:25] <Laney> attente: You only need to do this if there is *no* data
[17:26] <attente> Laney: but this means we can get into a state where we have data only for installed apps, in which case we would not do an update
[17:27] <Laney> attente: I think...
[17:27] <Laney> (xenial-amd64)root@nightingale:~# appstreamcli search .
[17:27] <Laney> No component matching '.' found.
[17:27] <Laney> (xenial-amd64)root@nightingale:~# apt update
[17:27] <Laney> [...]
[17:27] <Laney> (xenial-amd64)root@nightingale:~# appstreamcli search .
[17:27] <Laney> <lots of stuff>
[17:27] <Laney> although to be fair I didn't have any desktop comopnents installed
[17:27] <Laney> let me do that
[17:28] <attente> hmm... when i remove /var/lib/app-info and friends, appstreamcli search . still lists a whole lot of apps
[17:33] <Laney> attente: ok add /var/cache/app-info to the list
[17:33] <Laney> and then sudo appstreamcli refresh-index --force
[17:35] <attente> Laney: that's the ticket, thanks
[17:35] <Laney> good stuff
[18:28] <alexarnaud> Trevinho: I've looked into CCSM interface and I don't find any button in "preference pane" that make it possible to deal with profile creation
[18:33] <Trevinho> alexarnaud: see compizconfig/ccsm/ccm/Pages.py:892
[18:48] <alexarnaud> Trevinho: thanks again, well I'm really so short-sighted ^^
[18:55] <willcooke> g'night all
[19:15] <Laney> me too! night!
[20:01] <Sweet5hark> heh.
[20:02] <Sweet5hark> libreoffice 5.1.1~rc3 snap with l10n UI -- but without dictionaries, without help/user documentation: 1.1GB
[20:04] <seb128> it's beefy ;-)
[20:24] <seb128> robert_ancell, hey!
[20:24] <robert_ancell> seb128, hello
[20:25] <seb128> robert_ancell, what issues are you looking at? I pondered looking a bit at the get_changelog ones but I saw you had some commits in git
[20:26] <robert_ancell> seb128, looking at the get_changelog issue and the crash removing one.
[20:26] <seb128> k
[20:27] <robert_ancell> I can avoid the crash removing one by changing the memory allocation, but I think I might just be hiding the issue
[20:27] <seb128> robert_ancell, I was going to look at replacing the '"%s (", source' by just reading the first word of the changelog
[20:27] <seb128> that should fix the binary/source mismatch issue
[20:27] <seb128> yeah, I didn't understand the invalid read
[20:28] <seb128> what is freeing that memory?
[20:29] <robert_ancell> seb128, so it's gvariant freaking out when writing the string to that pointer (which is a stack variable). If I change it to a heap variable it works, but I get warnings about loop being NULL (even though it doesn't seem to be)
[20:29] <robert_ancell> I thought the g_autofree is interacting badly, but when I removed it it still occurred.
[20:30] <robert_ancell> So it makes me think there's some threading issue there I haven't found.
[20:43] <seb128> hum, k
[20:43] <robert_ancell> seb128, what did you think about the evince + .desktop issue?
[20:43] <robert_ancell> To me, the .desktop file should *always* be with the binary
[20:43] <seb128> I got Laney to comment ;-)
[20:43] <seb128> right
[20:43] <seb128> that's the theory
[20:43] <robert_ancell> seb128, delegation :)
[20:44] <seb128> in practice it means that we might have to fix half the archive
[20:44] <seb128> which I don't like
[20:44] <robert_ancell> seb128, I don't think so, I think just evince and evolution are the worst offenders
[20:44] <seb128> language-selector
[20:44] <seb128> like I know of 3 on the 7 updates I've atm
[20:44] <seb128> emacs24 was the same iirc
[20:44] <robert_ancell> seb128, I think we should fix those anyway, cause they're wrong
[20:44] <robert_ancell> (And I'm happy to do it)
[20:44] <seb128> I wouldn't bet on that to not be common
[20:45] <seb128> yeah
[20:45] <seb128> I'm happy to do it also
[20:45] <robert_ancell> seb128, I wanted someone to check the Breaks/Replaces field, cause that always confuses me / I stuff it up
[20:45] <seb128> until we have to spend a week full time on that
[20:45] <seb128> and create a ton of merges over Debian
[20:45] <seb128> just B/R on (<< change_version)
[20:46] <robert_ancell> seb128, and the B/R is on the binary the file is moving to? What about the binary the file is moving from?
[20:46] <seb128> Replaces mean "you can overwrite from"
[20:46] <seb128> B means "don't let that be installed with me"
[20:46] <seb128> so you add those to the new one
[20:47] <seb128> and they point to the old one
[20:47] <seb128> like evince can overwrite files from evince-common <<
[20:47] <seb128> and they can't be co-installed, which force them to be updated together
[20:49] <robert_ancell> mkay
[20:49] <seb128> so yeah your change looks good
[20:53] <seb128> robert_ancell, other topic, which one looks better to you? http://people.canonical.com/~seb128/change.png
[20:53] <robert_ancell> The right one
[20:54] <robert_ancell> seb128, you been fighting the markup?
[20:54] <seb128> not "fighting" so much than deciding to do without
[20:55] <robert_ancell> seb128, I vote do without
[20:55] <robert_ancell> Did you use a method that could be upstreamed?
[20:55] <seb128> just the * to •  change seems a regression
[20:55] <seb128> no
[20:55] <robert_ancell> seb128, the issue is G-S wants markup, and debian/changelog is pseudo-markup
[20:56] <robert_ancell> So they'll never really work together
[20:56] <robert_ancell> seb128, also, do you find the updates UI confusing? The way clicking on an app shows the changelog and not the app page?
[20:56] <seb128> I changed gs-update-dialog.c
[20:56] <seb128> yes
[20:56] <seb128> also I think we should remove the description from the list
[20:57] <seb128> it eats newlines which just doesn't work out with our changelogs
[20:59] <robert_ancell> seb128, yeah, it looks a mess
[20:59] <robert_ancell> seb128, if the changes are good - just commit them to the wip/ubuntu-changes branch on GNOME git
[21:00] <robert_ancell> That's the grab-bag of everything that's not currently upstreamable
[21:00] <seb128> I think I would prefer peer reviews, quite some of those are a matter of taste
[21:01] <seb128> do we have a workflow for that?
[21:01] <seb128> we should maybe put it in git on launchpad and use merge requests?
[21:02] <seb128> robert_ancell, http://people.canonical.com/~seb128/evo.png ... the evo line looks less good right?
[21:02] <robert_ancell> seb128, yes, cramming debian/changelog into that space doesn't seem to work
[21:02] <seb128> I tried to remove the \n mangling
[21:03] <seb128> but then the entries take half of the screen
[21:03] <robert_ancell> seb128, no workflow currently other than just asking eachother for reviews
[21:16] <attente> is there a way to get the license name out of a .deb?
[21:17] <cyphermox> attente: it should contain a copyright file
[21:19] <attente> cyphermox: kind of was hoping for a way to do without extracting it entirely
[21:20] <qengho> attente: ick. ar x, tar xf - /usr/share/doc/pkgname/copyright.gz |zgrep '(bunch|of|different|license|names)'  ?  It can have several licenses, yes?
[21:21] <qengho> I think "ar" is too dumb.
[21:22] <qengho> attente: I don't think it's possible in one step. Sorry.
[21:22] <attente> oooh boy... if that's the best way, i am not a happy camper...
[21:23] <cyphermox> fwiw: https://wiki.debian.org/Debtags/FAQ#Any_reason_why_there_are_no_.60.60license::.60.60_tags_in_debtags.3F
[21:23] <attente> cyphermox: ah, ok. thanks for the info
[21:26] <seb128> attente, I think what robert_ancell made by pointing to the ubuntu website under an "open source" label makes sense
[21:26] <seb128> well for archive main/universe softwares at least
[21:27] <seb128> if it's for local deb we might just set it a unknown or hide it
[21:27] <cyphermox> attente: qengho: otherwise; dpkg-deb --fsys-tarfile foo.deb |tar -xvO ./usr/share/doc/foo/copyright
[21:27] <robert_ancell> I'd be hesitant to put anything for files not in the core archive. It's complicated and getting it wrong seems like a bad idea.
[21:27] <seb128> +1
[21:27] <attente> yeah, i was more concerned about that case. but it always seems to state non-free if there's no license provided
[21:28] <cyphermox> yeah
[21:28] <cyphermox> the copyright file can lie if it's not in the archive
[21:28] <cyphermox> (well, truthfully it can lie even if it's in the archive, but that's why archive admins review the packages)
[21:52] <seb128> robert_ancell, I've attached a suggested change for the changelog formatting on https://bugs.launchpad.net/ubuntu/+source/gnome-software/+bug/1554167
[21:53] <robert_ancell> seb128, thanks
[21:53] <seb128> yw!
[22:13] <attente> i wonder if the gtk_stack_set_visible_child_name() bug is the same as the LIM g_menu_model_get_n_items() bug...