[03:52] <pero> anyone know how to get the icons for unity web apps? my gmail icon in the panel has some settings-looking thing, but in dash it's good
[04:05] <hyperair> pero: what gmail icon in the panel?
[04:05] <hyperair> pero: do you mean in the messaging menu?
[04:05] <hyperair> o
[04:06] <hyperair> try running this command: gsettings get com.canonical.indicator.messages applications
[04:06] <pero> hyperair,  in the left-side panel
[04:07] <hyperair> you should get something like.. [.... 'Gmailsomethingsomething.desktop']
[04:07] <hyperair> O_o
[04:07] <hyperair> oh the launcher?
[04:07] <hyperair> why don't you just remove the icon from the launcher and drag it down from the dash again?
[04:08] <pero> doesn't work
[04:08] <hyperair> what's the output of: gsettings get com.canonical.Unity.Launcher favorites
[04:08] <pero> reverts to some nuts/bolt looking thing
[04:08] <hyperair> odd
[04:08] <pero> $ gsettings get com.canonical.Unity.Launcher favorites
[04:08] <pero> ['application://nautilus.desktop', 'application://gnome-terminal.desktop', 'application://chromium-browser.desktop', 'application://GMailmailgooglecom.desktop', 'application://ubuntu-software-center.desktop', 'application://gnome-control-center.desktop', 'unity://running-apps', 'unity://expo-icon', 'unity://devices']
[04:09] <hyperair> application://GMailmailgooglecom.desktop <-- probably this one
[04:09] <pero> i think so?
[04:09] <hyperair> check the icon field of the ~/.local/share/applications/GMailmailgooglecom.desktop file
[04:10] <pero> is there an easy way to see hidden files or manually input the path in nautilus?
[04:11] <hyperair> ctrl+l
[04:13] <pero> it's some standard colour-pallette icon
[04:13] <hyperair> um
[04:13] <hyperair> what does it say?
[04:14] <pero> in the file? there's nothing
[04:14] <hyperair> grep ^Icon ~/.local/share/applications/GMailmailgooglecom.desktop
[04:14] <hyperair> what?
[04:14] <hyperair> that can't be right...
[04:14] <pero> file:///home/petar/.local/share/applications/GMailmailgooglecom.desktop
[04:14] <pero> that's the entirety of the contents
[04:15] <hyperair> that's the URL of the file.
[04:15] <hyperair> i want the contents of the file
[04:15] <hyperair> where are you looking?
[04:15] <pero> leafpad
[04:15] <hyperair> uh
[04:15] <pero> oh ok
[04:16] <pero> $ grep ^Icon ~/.local/share/applications/GMailmailgooglecom.desktop
[04:16] <pero> Icon=GMail-mail.google.com
[04:27] <hyperair> there we go, an icon name.
[04:27] <hyperair> now check ~/.icons/GMail-mail.google.com.png
[04:28] <hyperair> also try: find /usr/share/icons ~/.icons -type f -name 'GMail-mail.google.com.*'
[04:28] <hyperair> pero: ^
[04:28] <pero> na that first one is there and good
[04:29] <pero> and that's the only one
[04:35] <hyperair> =\
[04:35] <hyperair> well, for some reason, it's not locating the icon properly
[04:35] <pero> =/
[07:09] <mzanetti> o/
[07:15] <Saviq> \o
[07:19] <tsdgeos> ~o~
[07:21] <tvoss> o/
[07:46]  * tsdgeos tests CI to see if it's still unhappy
[07:48] <tsdgeos> mzanetti: is anyone having a look at https://code.launchpad.net/~mzanetti/unity8/launcher-own-quicklist/+merge/191398 ?
[07:48] <tsdgeos> shall i claim it?
[07:48] <tsdgeos> Saviq: ↑↑↑ ?
[07:49] <Saviq> tsdgeos, go for it
[07:49] <mzanetti> tsdgeos: yep, thanks :)
[08:18] <tsdgeos> mzanetti: https://code.launchpad.net/~aacid/unity8/warnings_on_shutdown/+merge/191945 ?
[08:19] <Saviq> tsdgeos, happroved
[08:19] <Saviq> /biab
[08:20] <mzanetti> +1
[08:20] <mzanetti> btw. is autolanding now fully working again?
[08:21] <tsdgeos> i am not sure
[08:21] <tsdgeos> i approved one of Cimi's branch earlier this morning to see how it goes
[08:21] <tsdgeos> on friday there was some problem installing dependencies or something
[08:25] <tsdgeos> greyback: i think it did merge because there was nothing to merge :D
[08:26] <greyback> tsdgeos: but it didn't :)
[08:26] <tsdgeos> greyback: what does merging a branch that has nothing to merge do? :D
[08:27] <greyback> tsdgeos: bzr commit --unchanged ? ;)
[08:40] <tsdgeos> nope
[08:41] <tsdgeos> we're still dependency locked :D
[08:41] <tsdgeos> https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-maguro/2574/console
[08:41] <tsdgeos> any idea who do we complain about that?
[08:41] <tsdgeos> Saviq: greyback: mzanetti: all: ↑↑↑↑
[08:42] <mzanetti> tsdgeos: well, usually fginther, but let me see if I can help
[08:46] <mzanetti> tsdgeos: hmm.. don't really know why this would work, but if it's only happening for the phones, om26er should probably be able to help too
[08:46] <tsdgeos> mzanetti: bad ppas?
[08:47] <mzanetti> tsdgeos: the hooks look ok. however, I don't really know what has changed in the QA infrastructure lately
[08:47] <tsdgeos> ok
[08:47] <tsdgeos> so no landing for now :D
[08:47] <mzanetti> I guess omer will show up soon
[08:52] <tsdgeos> dednick: still lost with the ActionStateParser, where do you use it?
[08:56] <dednick> tsdgeos: Unity.Indicators.RootActionState derives from ActionStateParser. Even though it's built against the proper qmenumodel library the run time linking requires it in the mock.
[08:57] <tsdgeos> ahhhhhh
[08:57] <tsdgeos> i see
[08:57] <tsdgeos> tx :-)
[08:58] <dednick> tsdgeos: no prob :) it's a bit cheeky.
[09:00] <tsdgeos> mzanetti: print("sleected", appId) in the test?
[09:01] <mzanetti> tsdgeos: hmm... missed that I guess. will remove
[09:02] <mzanetti> tsdgeos: pushed
[09:07] <Saviq> tsdgeos, mzanetti, I know what's happening :/
[09:08] <mzanetti> Saviq: tell me more
[09:08] <Saviq> tsdgeos, mzanetti /usr/bin/unity8 is bind-mounted from /run/ so that we can set the capabilities on it for adjusting oom scores
[09:10] <Saviq> which is unfortunate
[09:10] <Saviq> and results in package installation fail, 'cause we're trying to overwrite a bind-mounted file, which fails of course
[09:10] <mzanetti> mhm...
[09:10] <mzanetti> so we're installing into /run or what?
[09:10] <Saviq> mzanetti, no, *we* don't do anything special
[09:11] <Saviq> mzanetti, but there's an upstart job that:
[09:11] <Saviq> - mounts tmpfs in /run/unity8-setcap
[09:11] <Saviq> - copies /usr/bin/unity8 to /run/unity8-setcap
[09:11] <Saviq> - setcaps /run/unity8-setcap
[09:11] <Saviq> - bind-mounts /run/unity8-setcap/unity8 to /usr/bin/unity8
[09:11] <Saviq> it's ugly as hell
[09:12] <mzanetti> seems like raping the OS
[09:12] <Saviq> but the only thing we could think of the night prior to V1
[09:12] <mzanetti> haha
[09:12] <mzanetti> what capabilities do e set then?
[09:12] <Saviq> allows unity8 to set OOM-killer scores on processes
[09:12] <dednick> Saviq: whose in charge of getting packages landed?
[09:13] <mzanetti> ah. I see
[09:13] <Saviq> dednick, no landing for now
[09:13] <Saviq> dednick, until T opens
[09:13] <Saviq> mzanetti, so that apps are killed earlier than other processes
[09:13] <dednick> Saviq: when is that?
[09:14] <Saviq> dednick, tomorrow or so
[09:15] <Saviq> dednick, at which point we'll go back to automatic release and such
[09:15] <dednick> Saviq: ok. thanks
[09:16] <mzanetti> Saviq: I guess the proper long term solution would be to outsource this to some other system component and just provide the information about apps from unity8
[09:16] <Saviq> mzanetti, proper long term solution is to use a sane filesystem instead of ext2 for the system image
[09:16] <mzanetti> and that would make a difference?
[09:16] <Saviq> mzanetti, where we can just set the caps in post-inst
[09:16] <mzanetti> ah, yeah sure
[09:16] <Saviq> mzanetti, setcap is stored in xattr
[09:16] <mzanetti> ext2?
[09:16] <Saviq> mzanetti, yes, / is ext2
[09:16] <mzanetti> :)
[09:17] <Saviq> for what reason I don't know
[09:17] <mzanetti> well, we probably don't want journaling
[09:17] <Saviq> then just disable journalling in ext4 ;)
[09:18] <mzanetti> :)
[09:18] <Saviq> mzanetti, but yeah, we need +xattr anyway
[09:18] <mzanetti> MacSlow: here you go: lp:~mzanetti/unity8/generic-lockscreen
[09:18] <Saviq> which AFAIK ext2 can't do
[09:18] <MacSlow> mzanetti, thx
[09:18] <mzanetti> MacSlow: I guess we merge them separately... makes the diff smaller
[09:18] <mzanetti> MacSlow: but you can use this a prerequisite
[09:19] <MacSlow> mzanetti, sure... I'll base my work on your branch still
[09:19] <mzanetti> MacSlow: here's the merge: https://code.launchpad.net/~mzanetti/unity8/generic-lockscreen/+merge/191951
[09:31] <Saviq> hey om26er, can you disable unity8-autolanding for the time being? mediumtests-touch are broken for us still
[09:31] <om26er> Saviq, I can try
[09:31] <Saviq> om26er, so it beats the purpose to run it and de-approve all the branches
[09:32] <om26er> Saviq, CI is already disabled? you just want me to disable -autolanding ?
[09:32] <Saviq> om26er, just -autolanding
[09:32] <om26er> ok
[09:32] <Saviq> om26er, -ci is somewhat useful still - for results from the other jobs
[09:33] <Saviq> hmm anyone else getting 404 from ports.ubuntu.com when apt-get update'ing the device?
[09:35] <Saviq> hmm maybe that's caused by the release happening
[09:35] <Saviq> didrocks, pointers ↑↑?
[09:36] <Saviq> hrm
[09:36] <didrocks> Saviq: I think cjwatson hasn't finished yet opening the release
[09:36] <Saviq> didrocks, but I'm getting 404s for saucy
[09:36] <didrocks> oh, interesting
[09:36] <didrocks> Saviq: maybe ask on #ubuntu-release?
[09:36] <Saviq> didrocks, i.e there's no "Packages" in http://ports.ubuntu.com/ubuntu-ports/dists/saucy/main/binary-armhf/
[09:37] <Saviq> hmm maybe that's my non-writable image causing it
[09:37] <Saviq> didrocks, will do, thanks
[09:37] <didrocks> Saviq: if you let me ~1h, I'll be able to dig into it
[09:37] <didrocks> (in a meeting now)
[09:37] <Saviq> didrocks, no worries
[09:37] <didrocks> and upgrading otto machines ;)
[09:39] <Saviq> didrocks, false alarm, stupid apt errors when non-writable ;P
[09:41] <tsdgeos> Saviq: yep
[09:42] <Saviq> tsdgeos, ?
[09:42] <tsdgeos> Saviq: yep, the errors suck :D
[09:42] <Saviq> tsdgeos, ah :)
[09:45] <didrocks> ah great ;)
[10:09] <Saviq> mzanetti, there's a small conflict in switching-previews
[10:09] <mzanetti> Saviq: ack
[10:09] <Saviq> oh actually not that small
[10:10] <mzanetti> Saviq: do'h someone added a preview spinner
[10:10] <Saviq> mzanetti, memory problems? :D
[10:10] <mzanetti> seems so
[10:11] <Saviq> mzanetti, it's Diego's code preventing clicking on buttons if an action was triggered
[10:11] <mzanetti> Saviq: ah right... didn't expect that to have landed yet
[10:11] <Saviq> mzanetti, we needed it for v1 - click scope really hates people clicking "install" multiple times
[10:12] <Saviq> mzanetti, I'm fine with you reverting it in your branch and finding a better solution
[10:12] <mzanetti> haha, yeah, I agree
[10:12] <Saviq> mzanetti, at least for the UI part, the backend things seem ok
[10:12] <mzanetti> sure
[10:14] <om26er> Saviq,  I have a branch to disable autolanding for unity8 but I cannot deploy that because that requires an access TOKEN.
[10:14] <om26er> mzanetti, hey! if we just disable the autolanding job manually. that won't hurt, right ?
[10:14] <Saviq> om26er, no, it should be just fine
[10:14] <mzanetti> om26er: nope
[10:15] <mzanetti> om26er: it will be reenabled once someone deploys something... but that's not an issue given that the disabling is only temporary anyways
[10:15] <om26er> mzanetti, Saviq ack. disabled it now ;)
[10:22] <Cimi> Saviq, any bug you want me to work on?
[10:22] <Cimi> or I pick up one? :)
[10:22] <Saviq> Cimi, nothing in particular, no
[10:33] <tsdgeos> Saviq: shall i top re-approve the things that were set to not approved by failed autolanding?
[10:33] <Saviq> tsdgeos, yes
[10:33] <Saviq> tsdgeos, we've disabled autolanding for now, so let's have everything in place for when we re-enable it
[10:34] <tsdgeos> oka
[10:44] <tsdgeos> Saviq: updated https://code.launchpad.net/~aacid/unity8/start_ssh/+merge/189602
[10:44] <Saviq> tsdgeos, thanks
[10:44] <tsdgeos> Saviq: so are we jumping to trusty as soon as it opens?
[10:45] <Saviq> tsdgeos, yup
[10:45] <tsdgeos> oka
[10:45] <greyback> tsdgeos: this be handy? http://pastebin.ubuntu.com/6276004/
[10:46] <greyback> so you only start it if not running. The sleep a bit arbitrary tho, think it needs to be longer for gnexus
[10:46] <tsdgeos> greyback: i only start it if it's not running
[10:46] <tsdgeos> don't i?
[10:46] <greyback> tsdgeos: maybe, I kinda just skimmed :)
[10:47] <greyback> ahh there it is
[10:47] <greyback> tsdgeos: never mind me
[10:47] <tsdgeos> i never do :-P
[10:47] <tsdgeos> j/k
[10:48] <greyback> :|
[10:56] <Saviq> Cimi, did you follow up with the sdk guys about InverseMouseArea?
[10:56]  * greyback going for lunch plus IRL meeting, back in about 2 hours
[11:04] <tsdgeos> Mirv: https://bugs.launchpad.net/ubuntu/+source/qtbase-opensource-src/+bug/1242630 and https://bugs.launchpad.net/ubuntu/+source/qtdeclarative-opensource-src/+bug/1242629
[11:05] <tsdgeos> Saviq: FYI ↑↑
[11:07] <Saviq> tsdgeos, k, Mirv is (should be) on holidays this week
[11:07] <tsdgeos> ok
[11:07] <tsdgeos> it's not *ultra* urgent
[11:07] <tsdgeos> i guess
[11:08] <tsdgeos> but tbh i get some itches by us shipping a patch that's been rejected upstream :D
[11:08] <Saviq> tsdgeos, yeah, we'll SRU it
[11:12]  * Saviq giggles at the fact that "sru" is a polish onomatopoeia for stuff landing hard
[11:16] <tsdgeos> :D
[11:16] <tsdgeos> trusty is open \o/
[11:16] <Cimi> Saviq, no update on the bug
[11:18]  * tsdgeos wonders if he should s/saucy/trusty in his sources and see what happens :D
[11:23] <Saviq> tsdgeos, I just tried  do-release-upgrade but it didn't like me ;)
[11:23] <Saviq> or, didn't like trusty at least
[11:23] <Saviq> there's a trusty image already, too
[11:23] <Saviq> --channel=trusty-proposed
[11:24] <tsdgeos> cool :D
[11:30] <tsdgeos> trusty only wants to give me updates for http://paste.ubuntu.com/6276216/
[11:30] <tsdgeos> i guess nothing can go wrong with that ... :D
[11:31] <mzanetti> famous last words
[11:31]  * mzanetti waits for tsdgeos to drop out IRC
[11:34] <Saviq> tsdgeos, well, nothing much happened yet ya know
[11:34] <tsdgeos> :D
[11:34] <Saviq> tsdgeos, trusty is basically a copy of saucy +some
[11:38] <Saviq> Cimi, did you ping Zsombor about the InverseMouseArea issue?
[11:38] <Saviq> Cimi, also, what's the status of the incorrect overlay shape in carousel-music-video?
[11:43] <tsdgeos> ok, let's reboot and see if i can come back :D
[11:44] <om26er> Cimi, btw bug 1229804 is already fixed. though you may want to fix the 'Applications' to 'Apps' .. (I tried but couldn't find the string)
[11:45] <Saviq> worked here ;)
[11:49] <Saviq> tsdgeos, welcome, how're you feeling? ;)\
[11:50] <tsdgeos> i feel trusty now!
[11:51] <Saviq> tsdgeos, now wait for the sync from debian unstable ;)
[11:51] <tsdgeos> yeah :D
[11:51] <mzanetti> lol
[11:52] <Saviq> ooh, arm64 is enabled now :)
[11:52] <mzanetti> Saviq: merged the switching-previews
[11:52] <Saviq> mzanetti, thanks
[11:52] <mzanetti> Saviq: arm64 for what?
[11:52] <mzanetti> ah
[11:52] <Saviq> mzanetti, trusty is building arm64 too, now
[11:52]  * mzanetti was reading amd64 and didn't see the big news
[11:53] <mzanetti> stupid brain FEC
[11:53] <tsdgeos> what's arm64?
[11:53] <tsdgeos> which hw?
[11:53] <tsdgeos> panda? some phone? some "desktop-cpu"?
[11:53] <Saviq> tsdgeos, there isn't any in real world :D
[11:54] <Saviq> tsdgeos, iPhone 5s
[11:55] <tsdgeos> right
[11:55] <tsdgeos> so how are we building it?¿
[11:55] <mzanetti> very good question
[11:55] <Saviq> tsdgeos, cross- probably ;)
[11:55] <Saviq> https://wiki.debian.org/Arm64Port
[11:56] <Saviq> "This is the first non-x86 self-bootstrapped Debian port: first 150 packages cross-built using build-profiles to untangle cyclic build-dependencies. "
[11:56] <tsdgeos> lunch
[11:56] <mzanetti> so the trusty arm64 build doesn't really contain many packages so far?
[11:57] <mzanetti> assuming we can't easily cross compile the whole thing
[11:57] <mzanetti> yet
[11:57] <Saviq> mzanetti, we should be much better now in that regard
[11:57] <Saviq> mzanetti, and obviously will get even better over the course of this cycle - if only because of arm64
[11:58] <mzanetti> sure, but notthereyet (TM) I guess
[11:58] <Saviq> mzanetti, we're pretty close indeed
[11:58] <Saviq> mzanetti, unity8 cross-builds if you fix a few things in qt's cmake modules
[11:58] <Saviq> and in automoc, AFAIR
[11:58]  * mzanetti can't wait for cross compiling unity8 instead compiling on the Galaxy nexus
[11:58] <Saviq> +1
[11:59] <Saviq> mzanetti, you should compile on manta ;)
[11:59] <mzanetti> true :)
[11:59] <mzanetti> it's faster than gusch's notebook :D
[11:59] <Saviq> mzanetti, I've a "setup-sbuild.sh" script if you want it ;)
[12:00] <mzanetti> Saviq: yeah, now that I can use the latest kernel again it might work
[12:07] <mzanetti> did something change how the fs is made writable?
[12:08] <Mirv> Saviq: nope, last Wed-Fri only, but right now trusty stuff
[12:11] <Saviq> mzanetti, no
[12:11] <mzanetti> yeah... I figured the issue
[12:53] <tsdgeos> greyback: btw i compiled fine with 5.2 on friday, maybe you had some weird path issues or something
[12:53] <greyback> tsdgeos: grrr
[12:53] <tsdgeos> couldn't run it
[12:53] <tsdgeos> though
[12:53] <tsdgeos> because it needs a lot of more packages to be rebuilt
[12:53] <greyback> for a minute, I'd a flash of a nice life where it would just work :)
[12:53] <tsdgeos> since once you start using private stuff
[12:53] <greyback> true
[12:54] <tsdgeos> things don't like getting mixed
[12:54] <tsdgeos> but at least it compiles :D
[12:55] <greyback> Well like yourself, I'm more interested in seeing how v4 performs, or what bugs it might have
[12:57] <greyback> tsdgeos: anything look dodgy in my configure options? http://pastebin.ubuntu.com/6276600/
[12:58] <tsdgeos> everything? :D
[12:58] <tsdgeos> don't know i don't pass almost any option
[12:58] <tsdgeos> missing comma
[12:58] <tsdgeos> don't know, i don't pass almost any option
[12:59] <tsdgeos>   /home/tsdgeos/qt5/qtbase/configure -prefix /home/tsdgeos/qt5_install -developer-build -debug -confirm-license
[12:59] <tsdgeos> i understand why you do the don't build mysql and the nomake stuff
[12:59] <tsdgeos> the rest not sure you need it
[13:00] <tsdgeos> or if they can be bad or not
[13:12] <tsdgeos> Saviq: can you review this one? https://code.launchpad.net/~aacid/unity-notifications/micro_optimizations/+merge/191764
[13:13] <Saviq> tsdgeos, k
[13:18] <tsdgeos> dednick: there?
[13:18] <dednick> tsdgeos: yo
[13:19] <tsdgeos> dednick: how do i test https://code.launchpad.net/~nick-dedekind/unity8/indicator.submenu-reinit/+merge/187885 (besides the autotest?)
[13:19] <tsdgeos> what changes/improvements should i see?
[13:20] <dednick> tsdgeos: um, restarting indicators re-populates the menus correctly.
[13:20] <dednick> tsdgeos: it's a bit hard to test unless you're using the messaging menu, since it's the only one that generally changes.
[13:22] <dednick> tsdgeos: previously it was just "reusing" the old menu items, but they were stale. So any changes to the order/add/removes werent being reflected in the UI.
[13:22] <tsdgeos> i see
[13:23] <dednick> tsdgeos: you have the demo messaging menu?
[13:23] <tsdgeos> nope
[13:23] <tsdgeos> or not that i know :D
[13:29] <greyback> tsdgeos: (oh you replied to me, sorry was distracted) I took those configure options from the 5.0 debian/rules script
[13:30] <tsdgeos> greyback: they don't look bad per-se, just unneeded to me, the problem was maybe somewhere else?
[13:30] <tsdgeos> i remember i had to play a bit with cmake to get it to pick the qt i wanted
[13:30] <greyback> tsdgeos: dunno, just finished a rebuild, trying now. Yeah me too.
[13:35] <Saviq> dednick, standup?
[13:36] <Saviq> nic-doffay, can you hear us?
[13:40] <nic-doffay> Saviq, I could not even after restarting mumble twice. Apparently third time lucky.
[13:40] <Saviq> nic-doffay, cool ;)
[13:42]  * greyback needs to reboot
[13:52] <hikiko|lunch>  /nick hikiko
[13:53] <hikiko> sorry :)
[13:55] <mzanetti> dednick: hey, testing your branch.
[13:55] <mzanetti> dednick: the notifications are much faster :)
[13:55] <mzanetti> dednick: sorry. the indicators
[13:55] <dednick> mzanetti: you mean the loading
[13:55] <dednick> mzanetti: yep
[13:55] <mzanetti> dednick: but: there is a notification staying on my screen
[13:56] <mzanetti> dednick: only happens with your branch
[13:56] <mzanetti> dednick: an empty one, never going away
[13:56] <dednick> huh. weird
[13:56] <tsdgeos> dednick: mzanetti: missing merge with trunk
[13:57] <tsdgeos> mzanetti: we have a ghost notification now thanks to me and some last minute problem with focus and litsviews
[13:57] <tsdgeos> mzanetti: so probably dednick's branch doesn't have the code merged for that
[13:57] <mzanetti> ah
[13:57] <Cimi> Saviq, do-release-upgrade -d doesn't detect T, sure we have to upgrade now?
[13:58] <Cimi> can do sed… but
[13:58] <tsdgeos> Cimi: i hand edited my files
[13:58] <tsdgeos> it's what i always do :D
[13:58] <tsdgeos> old-style
[13:58] <Cimi> ok
[13:59] <mzanetti> dednick: false alarm. a merge with trunk does fix it indeed
[13:59] <dednick> mzanetti: cool
[14:01] <kgunn> mhall119: let's chat here...
[14:02] <kgunn> when was the last time you tried to open an app in unity8 on a simulated desktop ?
[14:03] <mzanetti> dednick: approved
[14:04] <mhall119> kgunn: running it now
[14:05] <kgunn> mhall119: to answer your other query, we did actually have an old bp to do a mir emulator...mir on nux effectively
[14:05] <mhall119> kgunn: I can't get the launcher out with my mouse :(
[14:05] <mzanetti> mhall119: run it with ./run instead of unity8
[14:06] <mzanetti> mhall119: if you've installed it into the system you won't have ./run. you can run it with "unity8 -mousetouch"
[14:06] <kgunn> mhall119: but the "mir emulator" task got "distracted" by xmir & then unity8 had a way to run on desktop...so...shelved it
[14:07] <kgunn> mhall119: yeah...-mousetouch at the very bottom of the instructions on that webpage i sent earlier
[14:11] <mhall119> kgunn: thanks, working now
[14:11] <kgunn> mhall119: cool...good to know
[14:12] <mhall119> kgunn: none of the apps display though
[14:12] <kgunn> mhall119: and following up to the earlier questions...we're feeling like best use of time wrt mir/desktop will be to put effort into rootless x this cycle...and hope to have a nice desktop preview of untiy8 on mir
[14:12] <mhall119> only black screens
[14:13] <kgunn> mhall119: you're so picky
[14:13] <mhall119> :)
[14:13] <kgunn> mhall119: can you navigate back to the shell ?
[14:13] <kgunn> swipe
[14:13] <mhall119> yeah
[14:14] <larsu> Saviq: hi! The system settings guys need simple dbus access from qml. Do we have anything like that?
[14:14] <mzanetti> larsu: no
[14:14] <larsu> mzanetti: any plans?
[14:14] <mzanetti> larsu: don't think so
[14:14] <larsu> if not, I'd be up for spending some time on this
[14:14] <mzanetti> larsu: it's easier to just write a C++ plugin for the thing you need instead to fiddle with DBus arguments in QML
[14:15] <larsu> mzanetti: totally agree, but that's not what I want
[14:15] <larsu> it should be a real object (component?) based on the introspection data
[14:15] <seb128> mzanetti, it would be very handy to be able to bind a qml property to a dbus one
[14:16] <seb128> mzanetti, like the battery % of the settings to upower corresponding value
[14:16] <larsu> arguments can then be matched precisely from whatever js gives us
[14:16] <mzanetti> seb128: sure... for simple stuff like integers and strings yes. but everything else will cause you bad headaches
[14:17] <mzanetti> I personally would vote for not opening that door and use it how it is meant to be used. but sure. nothing prevents you from creating sucha component
[14:17] <seb128> mzanetti, well, those are common enough that it would already be quite useful
[14:17] <larsu> mzanetti: in what way? (also, most things in the system settings app are simple types)
[14:17] <larsu> is the problem that Qt doesn't support dynamic signals and Q_INVOKABLES?
[14:17] <larsu> (or does it?)
[14:18] <mzanetti> larsu: the problem is that you need to qDBusRegisterType for anything non-basic
[14:18] <kgunn> mhall119: how many apps display black ? all of them ?
[14:18] <mzanetti> larsu: and yes. dynamic signals/slots might be a problem too
[14:18] <larsu> mzanetti: if you use QDBus...
[14:18] <larsu> mzanetti: ya, that's what I thought :(
[14:19] <mzanetti> larsu: you need to somehow know the data type without Qt too
[14:19] <larsu> mzanetti: would the qmetaobjectbuilder stuff help me? (yes, I know it's private api)
[14:19] <mzanetti> larsu: I don't really know that one... sorry
[14:19] <larsu> ah okay
[14:20] <larsu> I'll have a look at that. Using some private API in a lib might be worth saving a couple of dozen plugins for simple dbus access
[14:20] <mzanetti> larsu: be prepared for bug reports about "dbus interface org.foo.bar doesn't work with QML plugin"
[14:20] <mzanetti> :)
[14:21] <larsu> tbh, I don't see the problem...
[14:21] <mhall119> kgunn: no, the browser and gallery which are still placeholders show something
[14:21] <mhall119> and some alls don't even try to launch
[14:23] <mzanetti> mhall119: hmm... I think they should work if they are installed
[14:23] <mhall119> mzanetti: I have them installed as .debs from the PPAs
[14:23] <dednick> mzanetti: thanks
[14:23] <mhall119> not as click
[14:23] <mzanetti> mhall119: that should be fine
[14:24] <mzanetti> mhall119: ah no... we're using the fake appmanager on the desktop
[14:24] <mzanetti> mhall119: yes, won't work
[14:25] <mhall119> ok
[14:25] <kgunn> mzanetti: could that change or is the app life cycle too tied up in mir ?
[14:26] <mzanetti> kgunn: so far we have an app manager for SF and one for Mir
[14:27] <mzanetti> so maybe it would even work if running Mir on Desktop with unity8 and the unity-mir qpa
[14:27] <mzanetti> but I haven't tried. greyback might know
[14:27] <kgunn> mterry: ping (all this unity8 on desktop talk makes me think greeter on mir)
[14:28] <mterry> kgunn, you want a status update?
[14:28] <kgunn> mterry: basically...
[14:28] <greyback> kgunn: mzanetti: ultimately we'll use the same app manager on desktop/mobile, but it'll probably disable most app lifecycle stuff on desktop.
[14:29] <mterry> kgunn, I'm fixing some issues with switching between sessions with USC/lightdm.
[14:29] <kgunn> greyback: "disable"...because apps can be wasteful little piggies on desktop ?
[14:29] <greyback> kgunn: mzanetti: the main blocker for using it on desktop right now is indeed the Mir integration
[14:30] <mterry> kgunn, I'm thinking that landing the separate greeter should be two-phase: first phase is just landing the split, then later actually enable lock support
[14:30] <kgunn> mterry: greyback mzanetti ...right, just avoiding a request for "emulate mir" on nux...seems once greeter split gets done, we've got the ability to run a full unity8-mir on desktop (minus the legacy x support)
[14:30] <greyback> kgunn: we can have that conversation when we get there. I'm not decided if we want rigourous app lifecycle management when plugged into mains. But a laptop on battery...
[14:31] <mzanetti> I telly you... if my background apps are suspended on desktop too I'm off :D
[14:32] <greyback> mzanetti: I'm kinda of the same opinion, but at same time, there'd be times if I didn't notice, I wouldn't care...
[14:33] <kgunn> greyback: you'd care in a sense you're battery lasted longer :)
[14:33] <greyback> kgunn: yep, as long as it didn't impede functionality
[14:33] <kgunn> mterry: so, itd be split but basically "autologin"
[14:33] <kgunn> ?
[14:33] <kgunn> on first phase
[14:34] <mterry> kgunn, not autologin, but no password
[14:34] <mterry> kgunn, i.e. you'd see the greeter, but can't set a PIN
[14:34] <mterry> same as today
[14:34] <tsdgeos> greyback: like all the stuff i have on the background that i like running, like music player, irc client, email client, web browser, etc :D
[14:34] <kgunn> mterry: got it
[14:35] <kgunn> mterry: i'm game...and i like the fact its broken apart from a risk perscpective
[14:35] <mterry> kgunn, what's broken?
[14:35] <mterry> ah
[14:35] <mterry> "broken apart"
[14:35] <kgunn> mterry: :) split/divide/staged
[14:37] <greyback> tsdgeos: as I said, if I didn't notice, I wouldn't care :) And with all those, I'd notice if they were suspended. But LibreOffice? Totem? What if music player not playing? Just throwing ideas out there
[14:39] <tsdgeos> greyback: so you want a little smart man inside unity that decides what and what can not be suspended
[14:39] <tsdgeos> let's not build skynet just for that :D
[14:39] <greyback> tsdgeos: I'm just proposing that the idea is worth a conversation
[14:40] <mzanetti> that's going to be so much fun with stale file handles, dropped network connections, broken desktop effects
[14:40] <greyback> even if it's a quick "that's a stupid idea Gerry, no"
[14:41] <greyback> mzanetti: you're being too practical, why not dream of what is possible! :)
[14:41] <mzanetti> because I think it makes application development so much harder
[14:42] <mzanetti> which results in worse applications
[14:42] <mzanetti> and I don't think battery life would improve much. except for the case of misbehaving apps
[14:44] <mzanetti> and in turn I think we would have more misbehaving apps which probably wouldn't consume your battery, but would consume your nerves
[14:44] <mzanetti> well, we had that already in depth :D
[14:45] <greyback> true. But how about phone/tablet apps on the desktop. They (should) behave with the app management stuff anyway.
[14:48] <tsdgeos> lol, the awful png but jpeg file is biting us too :D
[14:48] <tsdgeos> Shell.qml:139:5: QML QQuickImage: Invalid image data: file:///usr/share/backgrounds/warty-final-ubuntu.png
[14:48] <tsdgeos> will someone ever fix that?
[14:56] <greyback> tsdgeos: that's an old one https://bugs.launchpad.net/ubuntu/+source/ubuntu-wallpapers/+bug/296538
[14:56] <tsdgeos> Cimi: https://code.launchpad.net/~aacid/unity8/panelOpenClose/+merge/192000
[14:56] <tsdgeos> greyback: i know i know
[14:56] <greyback> you can see I had to deal with it in unity2d :)
[15:10] <tedg> greyback, Is there anything you need in upstart-app-launch?
[15:10] <tedg> greyback, Besides actually firing the failed signal :-)
[15:10] <greyback> tedg: heh, yes that would be nice :)
[15:13] <greyback> tedg: question: if the OOM strikes an app launched by upstart-app-launch, does upstart consider that app dead?
[15:13] <tedg> greyback, Yeah, anything that removes the PID.
[15:14] <greyback> tedg: ok
[15:15] <greyback> tedg: only other thing I think I need is list of PIDs actually owned by the app. I think you only give me one PID, and I guess it's possible for an app to spawn a sub-app?
[15:15] <greyback> tedg: main example on top of my head is browser app, as it spawns a separate QtWebProcess
[15:15] <tedg> greyback, Yeah, we can't get it until upstart can use cgroups, but I think that's on the table for next release.  Need to verify.
[15:17] <greyback> tedg: ok cool, good to hear it. I guess we can use cgroups to suspend apps too, better than sending sigstop. Will upstart-app-launch offer that API?
[15:17] <greyback> i.e. a "upstart_app_launch_suspend(app_id)" method
[15:17] <tedg> greyback, Uhm, I hadn't planned on it.... I don't see why not.  That'd be nice for removing a callback.
[15:17] <tedg> greyback, I'll put it on the "Figure this out" list
[15:18] <greyback> tedg: ok. I'll admit it's me pushing more of the app management into upstart-app-launch itself, so if that's too much, let me know
[15:19] <greyback> tedg: also do you know, we plan to use the OOM as the primary app killer? Or is it a last-resort tool?
[15:19] <tedg> greyback, Yeah, I'm just not sure where Upstart is going to be there.  If it's cgroups is early that's different than being late in the cycle.
[15:20] <tedg> greyback, I haven't heard of any other.
[15:20] <greyback> tedg: ok
[15:21] <greyback> tedg: one small subtle thing. The strings I'm passing into upstart-app-launch, it ok for them to be utf8? I'm assuming the app id may not be ascii. I wrong?
[15:21] <tedg> greyback, Yeah, I'd assume they're utf8
[15:22] <greyback> tedg: great. Ok, that's all the questions I've got for you. Thanks! :)
[15:22] <tedg> Perhaps I should figure out some tests for that though, just to make sure.
[15:22] <greyback> +1, just in case
[15:26] <Cimi> mterry, https://code.launchpad.net/~unity-team/unity8/fix-1238232
[15:27] <mterry> Cimi, what's the attack there?
[15:27] <Cimi> mterry, attack?
[15:27] <mterry> Cimi, what's the bypass there?
[15:27] <Cimi> mterry, dunno what you mean for bypass :)
[15:27] <mterry> Cimi, oh, reading bug
[15:27] <mterry> Cimi, it's in title of bug "can bypass lockscreen" but now I'm reading
[15:28] <mterry> I see
[15:29] <mterry> Cimi, that might want a comment.  It's an obscure path to bug and the reason for denying the ShowHome request when power is off is not obvious in code
[15:40] <om26er> mterry, question: do you plan to have a version of deja dup for touch ?
[15:40] <om26er> having sms/contacts backup/restore feature will help
[15:44] <kgunn> mzanetti: curious...are you diving into right edge navigation proto ( i know we started chatting on friday....but wasn't sure where we left it)
[15:44] <mzanetti> kgunn: yeah, I didn't start yet, but sure, I'm happy to take that part
[15:45] <kgunn> mzanetti: cool...i'd been pushing it off for the phone freeze....so now i just want to followup with design on it
[15:45] <mzanetti> kgunn: ok. I'll start soon with that
[15:46] <kgunn> they've been fairly patient, just want to follow up and treat them right
[15:46] <mzanetti> kgunn: please let me know if there is any previous discussion on this to follow up. Otherwise I'll find the responsible designer and get it going
[15:48] <mterry> om26er, I've thought about it.  I wouldn't be opposed to it.  Just a matter of time
[15:49] <kgunn> mzanetti: nope that's about it...just look thru the spec attached to that bp...if you have questions, ping JohnLea
[15:51] <om26er> mterry, cool. I plan to write a simple backup/restore for my usecase i.e. sms and contacts.
[15:53] <mterry> om26er, using deja-dup or just a separate thing?
[15:55] <om26er> mterry, I have never written a real qml app so I'll keep it as test bed to get better grip of things, probably 'omer\s app' ;)
[15:55] <om26er> probably something I won't even publish
[15:56] <mterry> om26er, yeah, I've done some qml development, but never an app from scratch.  That's part of what's kept me from tackling it
[15:57] <fginther> om26er, the disable unity8 autolanding MP needs an update
[15:57] <om26er> fginther, pushed the fix
[15:58] <om26er> per your comment
[15:59] <fginther> om26er, thanks
[16:30]  * greyback eod
[17:41] <kdub> is there a faster way to build unity than on-device?
[18:52] <kgunn> Saviq: you still on ?
[19:05] <bjsnider> is there a way to detect if globalmenus are in use?
[19:57] <Saviq> kgunn, wassup?
[19:57] <larsu> bjsnider: when com.canonical.AppMenu.Registrar is on the bus (but apps can elect to circumvent having their menus shown there)
[19:58] <larsu> so... there seems to be no way to find out when a property on a remote object changes when using QDBusInterface
[19:58] <larsu> or am I missing something?
[19:59] <larsu> also, no way to do async?
[20:00] <bjsnider> larsu, this is an issue for an app called gnome-mplayer, and the way it ic currently being dealt with is UBUNTU_MENUPROXY is read at app launch time
[20:00] <bjsnider> but it's set no matter if unity is running or not
[20:03] <larsu> bjsnider: ugh, that sounds wrong... why does this app need to know?
[20:03] <larsu> does it only export a menu when on unity?
[20:03] <bjsnider> larsu, it needs to know because of how the window size is calculated. since it's playing video it needs a precise window size and whatnot
[20:04] <bjsnider> so the best thing to do is to find a foolproof way to always know when globalmenus are being used and when they're not
[20:05] <bjsnider> it exports a menu if UBUNTU_MENUPROXY=1
[20:05] <bjsnider> which it does in gnome-shell and everywhere else i think
[20:05] <bjsnider> so this isn't good enough
[20:05] <larsu> bjsnider: this sounds very wrong... shouldn't the window size be calculated automatically?
[20:06] <bjsnider> the window size is based on the presence of the menubar or lack of it
[20:06] <bjsnider> if it's exported the window size is different
[20:06] <bjsnider> and the size of the video goes into it
[20:06] <larsu> yes, but why do you set it explicitely instead of just setting a minimum size of the video widget
[20:07] <bjsnider> i'm just telling you how the developer is choosing to handle this
[20:07] <bjsnider> and the developer isn't me
[20:07] <larsu> I know. I'm telling you that this isn't a smart way to handle this
[20:07] <kgunn> Saviq: no worries...was just going to ask for approval on https://code.launchpad.net/~kgunn72/unity-mir/bump-mir-dep16/+merge/191551
[20:07] <bjsnider> larsu, how would you handle it?
[20:07] <larsu> bjsnider: set a minimimum size on the video widget and let gtk handle the rest
[20:08] <larsu> bjsnider: alternatively, you could check for the size of the menu bar widget (which should be 0 when the menu is exported)
[20:08] <Saviq> kgunn, won't merge until Mir 0.0.16 is available, which it isn't yet https://launchpad.net/ubuntu/+source/mir ?
[20:08] <larsu> but this only works if the application is not using GtkApplication
[20:09] <bjsnider> larsu, part of the problem is the developer is not interested in using unity or ubuntu and is not as well-versed in how to handle the globalmenu issue
[20:09] <kgunn> Saviq: see rev 1445 at https://code.launchpad.net/~mir-team/mir/trusty
[20:09] <kgunn> Saviq: so...lp:mir now points to that
[20:10] <kgunn> Saviq: but yeah...just want the approval so all the merges/daily build can all be triggered at the same time
[20:10] <kgunn> was gonna bug gerry but he's eod (like a normal euro)
[20:10] <larsu> bjsnider: he won't have to be if he uses GtkApplication. Does he?
[20:11] <bjsnider> i don't know
[20:12] <bjsnider> larsu, https://code.google.com/p/gnome-mplayer/source/browse/#svn%2Ftrunk%2Fsrc
[20:14] <larsu> bjsnider: sorry, I don't have the time to help out with this
[20:15] <bjsnider> well, i will pass on your suggestions
[20:15] <larsu> cool
[20:16] <bjsnider> i keep wondering how totem handles this
[20:17] <bjsnider> it should have the same problem
[20:30] <Saviq> kgunn, they can't be triggered at the same time - mir needs to get released first - then we can merge the new unity-mir dep
[20:31] <Saviq> kgunn, but, do we actually need the new dep? i.e. does unity-mir depend on something that's there in mir 0.0.16 but not in 0.0.15?
[20:31] <Saviq> kgunn, or do we just need a no-change rebuild of unity-mir (ABI break?)
[20:31] <Saviq> kgunn, as now unity-mir depends on mir >= 0.0.15
[20:32] <Saviq> kgunn, so 0.0.16 is fine - *if* unity-mir does not need something that's only there in mir 0.0.16
[20:33] <kgunn> Saviq lemme doubel check
[20:38] <kgunn> Saviq: yeah...we did
[20:55] <Saviq> kgunn, we did... need a new dep? something new in mir since 0.0.15 that unity-mir uses?
[20:59] <Saviq> mterry, hey, some days ago "Light Display Manager" showed up as an entry on my login screen, that known/expected?
[20:59] <kgunn> Saviq: someone changed functions around in default_server_configuration.h
[21:00] <kgunn> like right in the middle...so even the ordinals will be screwed
[21:00] <mterry> Saviq, as a session option?
[21:00] <Saviq> mterry, as a user
[21:00] <Saviq> kgunn, ok, so ABI break, but unity-mir does not care whether it's going to be 0.0.15 or 0.0.16 for the time being
[21:00] <Saviq> kgunn, so we need a no-change rebuild is all
[21:01] <mterry> Saviq, that is unexpected.  On your desktop machine?
[21:01] <Saviq> mterry, yes
[21:01] <Saviq> kgunn, granted, I'm not sure how to get one other than bumping the version anyway :)
[21:01] <mterry> Saviq, meaning that lightdm user is < 1000 uid?  That seems crazy
[21:01] <Saviq> mterry, 105
[21:01] <Saviq> so should be fine
[21:02] <Saviq> ⟫ getent passwd | grep lightdm
[21:02] <Saviq> lightdm:x:105:110:Light Display Manager:/var/lib/lightdm:/bin/false
[21:02] <mterry> Saviq, what is your UID_MIN in /etc/login.defs?
[21:02] <kgunn> oh crap...gotta run grab the man-child
[21:02] <mterry> right, I meant > 1000 uid.  105 is a reasonable number
[21:02] <Saviq> mterry, 1000
[21:03] <Saviq> mterry, it's consistently there, though (i.e. survived a reboot et al
[21:03] <seb128> Saviq, mterry: I've seen that bug as well, it's not happening every time, like one on ten times
[21:03] <seb128> and a vt switch fixes it for me the other day
[21:04] <Saviq> mterry, seb128 it also seems to be "arriving" late
[21:04] <Saviq> mterry, seb128, like it comes in asynchronously from somewhere?
[21:04] <mterry> seb128, Saviq: sounds odd...  only in trusty or also in saucy?
[21:04] <Saviq> i.e. I start typing my own password in... and then it gets screwed 'cause there's a new entry at the top
[21:04] <Saviq> mterry, started in saucy, now in trusty, too
[21:05] <seb128> mterry, saucy, I had it like 10 days ago, I mentioned to robert_ancell/send him a greeter log by then
[21:06] <robert_ancell> seb128, did you get a log?
[21:07] <seb128> robert_ancell, yes, I pastebined it to you before leaving the other day
[21:07] <Saviq> [+1,38s] DEBUG: User /org/freedesktop/Accounts/User105 added
[21:07] <Saviq> [+1,38s] DEBUG: user-list.vala:988: Adding/updating user lightdm (Light Display Manager)
[21:07] <robert_ancell> Saviq, is it still there, can you browse to that user on d-feet?\
[21:08] <Saviq> robert_ancell, no
[21:09] <robert_ancell> hmm, so the question is is it accounts-service getting confused or does the entry in accounts service have a flag set we're ignoring
[21:09] <robert_ancell> as an aside, I think I just found the regression that might trigger it, but regardless a-s shouldn't create a user object unless it's a valid user
[21:11] <robert_ancell> Saviq, seb128, have you seen any pattern/trigger that causes this?
[21:11] <robert_ancell> I haven't seen it locally yet
[21:12] <Saviq> robert_ancell, I think I saw that like every time I log in
[21:12] <Saviq> robert_ancell, so nothing special
[21:12] <seb128> robert_ancell, no, but I've had cases of "loginctl list-users" listing "lightdm" as logged in even with no greeter
[21:13] <seb128> because indicator-session-service is still running
[21:13]  * Saviq tries some things
[21:13] <seb128> and logind lists the session actives when there are still processes running
[21:13] <robert_ancell> seb128, the regression is that we don't tell logind that the greeter session is a greeter not a user session which is probably triggering it
[21:15] <Saviq> right, can't reproduce now
[21:16] <Saviq> but it seems like it came in with the remote-login options last time I saw that (i.e. when network is connected)
[21:26] <Saviq> tedg, I *think* the TTS mentioned is text *to* speach, not the other way 'round
[21:26] <tedg> Saviq, Oops, misread.
[21:26] <Saviq> *speech
[21:26] <tedg> Heh, someone should do that too :-)
[21:26] <tedg> Though, that one is a bit easier just for basic support.
[21:26] <Saviq> yeah
[21:32] <kgunn> Saviq: refresh me...what does resize give us even in the fullscreen/opaque unity-mir on touch ? (animation transitions?)
[21:32] <Saviq> kgunn, ability to rotate the shell
[21:33] <Saviq> kgunn, since the panel changes place, app geometry changes
[21:33] <kgunn> Saviq: oh...just a little old rotation...simple :)