/srv/irclogs.ubuntu.com/2016/09/22/#ubuntu-desktop.txt

=== JanC is now known as Guest5754
=== JanC_ is now known as JanC
pittiGood morning05:24
hikikohi05:42
seb128good morning desktopers07:06
=== hikiko is now known as hikiko|bbiab
TheMusoHey seb128.07:12
seb128hey TheMuso07:25
flexiondotorgMorning07:48
willcookemorning all07:58
Laneyyo ho ho08:04
didrocksgood morning guys08:07
seb128hey Laney flexiondotorg willcooke08:08
flexiondotorgo/08:08
seb128one again bitten by IRC not handling switching from wifi to cable, grrr08:08
seb128took me a few minutes to think about it08:08
seb128and of course got the timeout even after going back to wifi08:08
Laneydoh08:11
pittibonjuor didrocks et seb128, comment allez-vous ?08:14
pittimorning Laney!08:14
seb128salut pitti, ça va bien ?08:14
didrocksbonjour pitti, ça va, et toi ?08:14
pittije vais bien, merci !08:15
Laneyguten morgen pitti!08:15
TheMusoLater folks.08:22
willcookenight TheMuso08:22
* willcooke rebooting in to Y - brb08:22
=== hikiko|bbiab is now known as hikiko
davmor2willcooke: hey dude I'm hitting and interesting issue on Y if you enable  vpn to auto connect on wired and wireless connections it clashes horribly and you lose routing when you plugin in the wire I assume the wireless connect controlled by nm connects to it first08:36
willcookedavmor2, speak to cyphermox and happyaron I think08:38
davmor2willcooke: on Y do you see your indicator sound?08:44
willcookedavmor2, in U8 or U7?08:46
davmor2willcooke: u708:46
willcookebooting...08:46
willcookedavmor2, on live session - yes08:47
davmor2willcooke: this is full work desktop with rhythmbox and spotify in the indicator let me check it on the installed system and see if either of those apps take it out08:48
davmor2hmmm http://paste.ubuntu.com/23214973/ that might be a problem09:01
davmor2seb128, Laney: just got that from the .cache/indicator-sound.log any ideas?09:02
LaneyI don't think it matters09:03
Laneyhave that here too09:03
willcookedavmor2, is this is u8 installed as well?09:03
Laneyit's used for the unity 8 greeter I think09:03
davmor2willcooke: nope09:04
davmor2willcooke: this is my work machine so just unity709:04
Laneydid the indicator die or was it never there?09:05
davmor2Laney: let me check ps aux but from what I see I started the session this morning and it wasn't there09:05
davmor2Laney: davmor2   3296  0.0  0.1 670716 11732 ?        Ssl  09:02   0:00 /usr/lib/x86_64-linux-gnu/indicator-sound/indicator-sound-service09:06
davmor2hmmmm09:06
Laneyweird09:06
Laneytry killingi t09:06
seb128davmor2, check the settings if the checkbox at the bottom is enabled09:08
seb128the one about displaying the indicator09:08
davmor2seb128: ha nice catch yeap it was unchecked but how the hell did that happen?09:09
seb128you clicked it?09:09
seb128or used something that changed it for you09:09
Laneyoh dear09:10
davmor2seb128: not that I know too09:10
davmor2seb128: I can't think what I might of used different to normal that might of changed it09:10
davmor2seb128: I did a big batch of upgrades yesterday 135-ish if I recall then this morning it was missing but I didn't use the sound settings yesterday09:11
seb128dunno, something changed the config09:12
seb128did you open the sound panel? maybe it was focussed and you hit space or enter09:12
seb128or did you try unity8? they maybe change the config?09:12
seb128in any case dunno09:12
davmor2seb128: ah that is a possibility09:12
seb128it's a gsettings key so I doubt it changed by itself09:13
davmor2might of hit space when it was focused09:13
seb128either some action did it for you or some program09:13
seb128but you are not the first one to hit that issue09:13
seb128I wonder if that setting is more confusing that useful09:13
davmor2seb128: thanks anyway mystery solved09:14
seb128yw!09:14
willcookeseb128, I spoke to TheMuso this morning about terminal app.  He said that we probably need an FFe to get Terminal in to Universe before the MIR can be processed - do we need to do that or is the MIR ok as is?  (cc popey)09:17
seb128willcooke, need a FFe to get a new component in the archive now09:18
willcookeDoes that need a new bug, or can we reuse the MIR one?09:18
seb128reusing the MIR should be alright be better to check with somebody from release team how much they dislike that, e.g Laney09:19
Laneydunno what you're talking about09:21
seb128Laney, reusing a MIR bug for a FFe09:21
seb128cool or uncool09:21
Laneymight be messy09:21
Laneyplus they want different information09:21
seb128like is recycling acceptable or preferred to open a new one09:21
Laneynot sure why you would want to09:22
seb128well, once the MIR is acked09:22
seb128Lazyness09:22
seb128willcooke, sounds like better to open a new one09:22
Laneybut you have to write new stuff for the ffe anyway09:22
Laneyso it's this box or that box09:22
willcookeoki, thanks seb128 Laney09:22
willcookepopey, ^ can you do the honors?09:22
popeyok09:22
Laneymerci!09:23
seb128willcooke, you are going to ask next so replying to that now, if/when the ffe/mir are +1 you just need somebody to get unity8-desktop-session to recommends it so it gets installed09:23
willcookethanks seb12809:24
seb128yw!09:24
Laneymerci!09:29
Laneypopey: Lemme know the number when you have it - happy to be easy with new stuff for this session at this point09:32
Laneyhttp://www.bbc.co.uk/news/entertainment-arts-3743790709:33
Laneyit's over~~~~09:33
popeywill do, just reading docs to make sure I do it right09:35
willcookeLaney, good on her!  And Mel & Sue09:35
davmor2willcooke: here here09:36
Laney\o/09:39
popeyLaney: filed bug 1626457, as per the wiki09:40
ubot5`bug 1626457 in Ubuntu "[FFE] Ubuntu Terminal App request" [Undecided,New] https://launchpad.net/bugs/162645709:40
* popey ninjas in a link to the source which he missed :)09:40
popeyHas Paul said whether he's staying on? Feels like he has more to lose.09:41
Laneypopey: I would also say that it is going to be added as a recommends to unity8-desktop-session, so installed for the new session only09:41
Laneydo these applications have Only/NotShowIn in their desktop files btw?09:42
popeywill do09:42
popeyLaney: no09:43
popeyhttp://bazaar.launchpad.net/~ubuntu-terminal-dev/ubuntu-terminal-app/reboot/view/head:/com.ubuntu.terminal.desktop.in.in09:43
Laneydo you get two things called Terminal then?09:44
popeyIn which session, unity 7 or unity 8?09:44
Laney709:44
Laney8 uses that X-Ubuntu-Touch=true thing to whitelist apps doesn't it?09:44
popeyi do on my xenial laptop, yes09:44
Laneyso you don't get gnome-terminal there09:44
popeycunning, not used that09:44
Laneycould be talking BS09:45
Laneybut... is it desired to have the new applications appearing in the unity 7 dash?09:45
popeythat's a question for willcooke (or other unity8 peeps I think)09:48
willcookeAt this point it would be nice, but not essential09:48
willcookeIf there is a neat packaging trick, like "only show in" that would be nice09:49
Laneywell, is it acceptable to have two things called Terminal in the unity 7 session?09:49
willcookeat this point in time, yes09:50
willcookebut of course, if it doesn't have to, and its easy to stop it, then that would be better09:53
Laneymmm10:02
LaneyI think this is the only one with a collision10:02
popeyooh, Paul signed with Channel 4 https://twitter.com/BBCNewsEnts/status/77889667321705267210:59
pittihappyaron: NM 1.2.4 is in the packaging master, is that ready for upload? I need to adjust the tests to the changed isc-dhcp11:00
pittihappyaron: I can also upload just that and rebase master, but in general it would be nice to put longer developments into branches; stuff in master should always be uploadable11:01
popeywillcooke: Laney https://bugs.launchpad.net/ubuntu/+source/ubuntu-app-launch/+bug/160854611:01
ubot5`Ubuntu bug 1608546 in ubuntu-app-launch (Ubuntu) "NotShowIn and OnlyShowIn in .desktop files are ignored" [High,In progress]11:01
willcookeso until that lands we can either have it in U7 & U8 or nowhere11:02
willcookeis that right?11:02
popeyI tested adding OnlyShowIn=Unity8 and it doesn't appear in either, right willcooke ?11:02
willcookeya11:02
willcookeso I guess that answers our question about where it should appear :)11:02
Laneypopey: huh, I didn't expect ual to be involved in displaying the grid11:03
happyaronpitti: yes ready11:03
popeyme either11:03
Laneyanyway11:03
pittihappyaron: hmmm, you don't use UNRELEASED? or you do and forgot to push the tag?11:04
LaneyNotShowIn=Unity will be better than nowt11:04
willcookegodo call11:04
willcookegood11:04
pittihappyaron: as d/changelog says (1.2.4-0ubuntu1) yakkety but I don't see it uploaded11:04
Laneynot nice for people on other desktops11:04
Laneyor get tedg's branch to go in11:05
seb128Laney, popey, willcooke, the issue is that unity7 and unity8 are called "Unity" for their xdg session so we can't filter them using OnlyShowIn etc11:19
seb128pitti, n-m was up for sponsoring, I guess he's just not using unreleased but it's as well because usually I do the sponsoring for him and I've no clue about the git workflow, does debcommit -r works there?11:20
LaneySounds like it should get "Unity8" in the list then11:21
popeyah11:21
seb128Unity8?11:23
seb128like add a new session type?11:23
seb128that would mean updating most of the existing .desktop which have OnlyShowIn=Unity to have Unity;Unity8 no?11:23
LaneyXDG_CURRENT_DESKTOP can be a list11:24
LaneySo something like Unity:Unity811:24
seb128right11:25
seb128it's a bit tricky but I guess that would better than what we have now11:25
pittiseb128: that's unrelated to git -- that's dch vs. dch -r; but yes, debcommit -ar works there11:25
seb128the side effect is that if you install one of those by choice on unity7 it wouldn't get listed by the dash11:25
seb128pitti, well in bzr that does the tagging no?11:26
pittiseb128: yes, same with git11:26
Laneyof course, but that's better than everyone having confusing duplicates11:26
seb128unsure if there similar concept of tagging uploads in git gbp11:26
pittiseb128: so, let me fix the two test regressions before upload11:27
seb128Laney, seems we are in agreement I meant that by "that would better than what we have now"11:27
pittione is for the changed isc-dhcp behaviour in -proposed, the other for kernel 4.8 (the fake module fails to build)11:27
Laneygood11:27
seb128Laney, can you try to chase people to get that changed next week? everybody is still busy trying to land the session this week and t_ed also has indicators/systemd to solve so I guess better to not have too many conversations at the same time11:28
seb128could be good to open a bug to start11:28
Laneyhang on, some guy has turned up to fix the boiler11:32
Laneybrb11:32
Laneyok, he's beavering away11:37
Laneyseb128: do you know where it is set?11:39
seb128Laney, I think it's lightdm but let me check11:42
seb128right, it is11:43
LaneyI don't actually see XDG_CURRENT_DESKTOP set in unity8's environment11:43
Laneyyou need to set DesktopNames in the .session file11:45
seb128lightdm set it using the DesktopName from /usr/share/gnome-session/sessions/<name>.session I think?11:45
Laneythat wasn't set for unity811:45
seb128k11:45
Laneyso where's the fallback coming from?11:45
seb128what fallback?11:46
Laneyusing Unity as you said11:46
seb128I though we had XDG_CURRENT_DESKTOP=Unity in unity811:46
seb128seems like I was wrong if you say it's unset11:46
Laneyindeed11:47
Laneyso if that behaviour exists it has to come from soemwhere else11:47
Laneyor maybe it just doesn't11:47
Laneyin which case we are lucky11:47
seb128unsure how the unity8 dash decide what to show11:47
seb128question for Saviq I guess11:47
seb128or maybe dobey11:47
seb128unity-scope-click-0.1.1+16.04.20160330/libclickscope/click/interface.cpp:static const std::string ONLYSHOWIN_UNITY("Unity");11:49
seb128Laney, ^ could be that11:49
Saviqyeah, click scope's responsible for that11:56
Saviqand ubuntu-app-launch may need this, too11:57
Laneymeh, this is buggy11:57
Laneyit should look at XDG_CURRENT_DESKTOP11:57
LaneySaviq: is the scope-click responsible for .deb applications too?11:57
Laneyual does it right11:58
Laney(ish)11:58
SaviqLaney, yeah, anything that shows up in u8's Apps list goes through unity-scope-click11:58
Laneyok11:58
Laneywell I think we can make it Unity;Unity811:58
Laneys/;/:/11:58
Laneyand then fix scope-click to consider it properly, and get that UAL merge proposal landed11:59
Laneythen things can have OnlyShowIn=Unity812:00
Laneyseb128: is there a trello or other tracking place?12:07
seb128Laney, https://trello.com/b/mab4G8UQ/unity-8-in-16-1012:16
seb128that was mostly for the MIRs though12:16
seb128but that's because it was most of the work, I guess it's fine to add another card12:17
Laneylike "Fixes outstanding" or something?12:17
seb128yes12:18
seb128we should probably add a column12:18
seb128there are some recommends to drop and such as well12:18
seb128willcooke, ^ wdyt?12:18
Laneycard or list?12:18
Laneylist probably12:18
seb128list?12:19
seb128I would add a column/stack whatever that's called12:19
Laneythat's called a list :P12:19
seb128k12:19
seb128so yeah, list12:19
seb128then a card for the session thing12:19
seb128and we can add some others for other issues we might hit once the session gets more testing12:19
seb128or the recommends that need to be demoted to suggests still12:19
Laneythere12:21
seb128thanks12:22
Laneyjbicha: https://bugzilla.gnome.org/show_bug.cgi?id=771553 and https://bugs.webkit.org/show_bug.cgi?id=162317 fyi12:24
ubot5`Gnome bug 771553 in Backend: Wayland "Shrinking window generates a black patch when gl is used" [Normal,New]12:24
ubot5`bugs.webkit.org bug 162317 in WebKit Gtk "Regression 2.14.0: WebKitGtk+ based apps (like yelp, evolution) bahave strange on resizing" [Normal,Resolved: invalid]12:24
Laneymentioned on #debian-gnome earlier12:24
jbichaLaney: thanks, maybe that's just gtk 3.22? since I don't see a problem with resizing with wayland here12:29
Laneydunno12:29
ximionLaney: any asgen OOM issues on the 32G machine? ;-)12:44
Laneyit did actually :/12:47
Laneyximion: we should do lazy extraction of translations12:50
Laneythis is relatively slow12:50
ximionLaney: what!?12:50
ximionokay, so something big must be leaking then12:50
Laneyhaha12:50
ximionif I would looks for something, the first place to look would be zarchive.d12:51
Laneyit just grows12:51
ximionbecause that's the only place which handles a lot of unsafe code and also a lot of data12:51
Laneyso by the time of the last arch it's pushing 32 G12:51
ximionLaney: I don't see this in Debian12:52
Laneydo a full run12:52
Laneyit's okay on incremental ones12:52
ximionbut I also don't process the whole archive anymore, it's just doing incremental updates12:52
ximionexactly12:52
ximionok, so temporary solution is to just run the incremental updates12:52
=== hikiko is now known as hikiko|lm
=== hikiko|lm is now known as hikiko|ln
ximionin the long run, we need to find the leak12:52
Laneyya12:54
ximionwhen I am home, I will take a look at the zarchive.d code12:54
ximionthe weird thing is that almost all memory is GC-allocated, so we really shouldn't leak12:55
ximionLaney: I will make an asgen release with the Ubuntu changes as soon as I implemented fonts support13:04
Laneyfonts, that'll be fun13:04
ximionalso, I might do the arch:all special-casing I talked about in the last bug report13:04
ximionit might be relatively easy13:04
ximionnot fonts though ^^ - fonts support is 80% complete in asgen, but the last bit will be the hardest13:05
ximionI will also need to depend on Pango, otherwise i18n doesn't work at all for fonts (I thought about getting around it, but then I would need to implement too much of functionality that Pango already has)13:06
dobeyLaney, seb128, Saviq: huh?13:09
ximionLaney: fonts will require a metainfo file, so if we want them we should poke the font package maintainers to add those files13:11
ximionbecause font upstreams usually don't do that ^^13:11
Laneyximion: hmm, did hughsie already do that @ fedora?13:17
ximionLaney: for some bits, yes - but we can do better ;-)13:18
ximionhughsie pushed files into Fedora when he was just experimenting with fonts - since we can build on hughsie's experiences, we can do the metainfo files right the first time13:19
ximionbecause once theyy're there, it's hard to motivate people to update them13:19
tedgpopey: willcooke: Laney: We can't support only show in until they fix trusted prompt overlays for the media player :-(13:27
tedgThey're using it to "hide" and application, so if we implement it correctly it'll break the media player.13:28
tedgI'm really not a fan of it, but that's the reason that got pulled out.13:28
Laneyximion: http://appstream.staging.ubuntu.com/13:35
=== hikiko|ln is now known as hikiko
Laneytedg: ...13:39
ximionLaney: it's beautiful!13:50
ximionLaney: your AppStream copy isn't up-to-date :P13:50
ximion(at 0.10.0 instead of 0.10.1)13:51
Laneyneed to backport it13:51
Laneysome good bug fixes in there13:51
Laneyfrom Quality Contributors13:51
ximionaka you ^^13:52
ximionyou can tell the AppStream release by the issues it emits ^^13:53
Laneyximion: https://paste.debian.net/835005/14:07
vigoLaney, I think I'm hitting a bug related to graphics in yy after fresh installation, do you know which logs should I take a look at?14:08
Laneyvigo: xorg logs / dmesg maybe14:09
* Laney is not very clued up on that, try #ubuntu-x for more experts14:09
vigobasically, waking up after suspend leads to a black screen and "the system is running in low graphic mode" window appears14:09
vigowith some errors14:09
vigoLaney, thank you :)14:10
pittiseb128: finally fixed the NM tests and pushed, so /me unlock14:13
seb128pitti, thanks!14:13
pittiargh, this is such a "everything breaks at once" day14:13
* pitti goes back to cloud fixing and kernel regression untangling, etc.14:14
pittiseb128: I'll just force-badtest the current version for now, but would be great to have the fix in -proposed soon so that we don't risk further regressions and not noticing them14:14
pittiseb128: (not sure what the status of 1.2.4 is, didn't just wnat to upload blindly)14:15
pittiI can also just upload the two fixes and rebase git if that needs more time to mature14:15
seb1281.2.4 should be good/ready for yakkety14:15
seb128happya_ron has been waiting for sponsoring for some days14:16
seb128I just got too busy with other things this week to look at it14:16
pittiseb128: right, just microversion, so hopefully "mostly harmless" </hitchhiker>14:17
vigohttps://pastebin.canonical.com/166279/14:20
vigoLaney, line 93814:21
Laneyvigo: points you to a further file14:24
Laneymaybe its contents help you find a bug report14:24
vigoLaney, yeap I'm pulling it to take a look :)14:27
* Laney lunches - suggest you take it to #ubuntu-x and/or the bugtracker mentioned14:28
seb128jbicha, dunno if saw that but totem ftbfs in the test rebuild, seems something wrong with the nautilus extension, could be have to do with the nautilus multiarch changes (why did that land so late without a ffe?)15:22
seb128hum, yakkety daily env doesn't have any configure keyboard layout in the indicator/settings, unsure what component to blame there :-/15:23
seb128xorg?15:23
jbichaseb128: nautilus was half-multiarchified so I finished the job, didn't think about a ffe for it...15:25
seb128unsure if we should revert that or look at the extensions and fix those15:26
jbichaI plan to check the nautilus extensions15:26
seb128thanks15:29
seb128k, time for dinner and somesport16:59
seb128have a nice evening desktopers16:59
willcookecya seb12816:59
Laneyme too17:12
LaneybyeeeeeEEeeeeEEEeee17:12
willcookenight Laney17:17
willcookecalling it a day here too17:33
willcookenight all17:33

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!