[00:16] <Saviq> sergiusens, it's green!
[00:16] <Saviq> (almost, armhf is just finishing up)
[00:17] <sergiusens> Saviq: ack
[00:19]  * sergiusens opens tab for build
[00:21] <sergiusens> publishing
[00:25] <asac> Saviq: good stuff :)
[00:25] <asac> if thats what we see on phones as well, +1
[00:26] <asac> have you tried?
[00:26] <asac> real phone? real image? :)
[00:26] <Saviq> asac, 20 successful subsequent runs on maguro
[00:26] <asac> Saviq: locally?
[00:26] <asac> nice
[00:26] <asac> :)
[00:26] <Saviq> asac, 1 in 50 tests on manta crashes unity8
[00:26] <Saviq> asac, but the crash is in Qt
[00:26] <Saviq> asac, need to try with Qt 5.1
[00:29] <asac> Saviq: where is unity8 latest atm? which ppa?
[00:29] <Saviq> asac, it's not released into no ppa
[00:29] <Saviq> asac, 'cause we stopped it
[00:29] <Saviq> asac, so it doesn't get into the image
[00:29] <Saviq> asac, will enable tomorrow
[00:29] <asac> Saviq: but its daily built somewhere still? or not even that?
[00:30] <Saviq> asac, it's built, but not published
[00:30] <Saviq> asac, so we'll see the daily release results in the qa dashboard, for example
[00:30] <Saviq> asac, but it's not publishing automatically
[00:31] <Saviq> asac, there's still https://bugs.launchpad.net/platform-api/+bug/1203698 that needs to land before we can test on devices, though
[00:33] <Saviq> sergiusens, beat you to it http://s-jenkins:8080/job/ubuntu-touch-phablet-image-saucy-mir/54/
[00:34] <sergiusens> Saviq: was in the process of...
[00:34] <sergiusens> :-P
[00:34] <Saviq> sergiusens, ;)
[00:34] <sergiusens> Saviq: good thing you kept your semi god mode :-)
[00:34] <Saviq> sergiusens, indeed :D
[00:35] <sergiusens> Saviq: don't make too much of a fuss about it just in case :-)
[00:35] <Saviq> HAHAHAHAHAHAHAHAHAHAHAHAHAHA
[00:36] <sergiusens> Saviq: you should get some sleep ;-)
[00:37] <Saviq> sergiusens, sleeping is overrated
[00:37] <Saviq> HAHAHAHAHAHAHAHAHAHAHAHHAHAHAHAHAHA
[00:37] <sergiusens> Saviq: during sprints perhaps
[00:37] <Saviq> sergiusens, and what is that we're doing, I ask you?
[00:38] <sergiusens> latey I have no idea what you guys are doing!
[00:38] <sergiusens> I sort of miss pre MWC
[00:38] <sergiusens> it's light a turning point for me
[00:38] <sergiusens> light/like
[00:38] <Saviq> sergiusens, yeah, there's just too many things
[08:19] <tsdgeos> Saviq: https://codereview.qt-project.org/#change,60809
[08:19] <tsdgeos> what now?
[08:20] <Saviq> tsdgeos, we need to harass Mirv to distro-patch it for us, nothing else we can do :/
[08:20] <Saviq> tsdgeos, some tests would be nice, though
[08:21] <tsdgeos> sure
[08:22] <Saviq> tsdgeos, I assume all the current ListView tests pass?
[08:22] <Saviq> tsdgeos, as well as our LVWPH ones?
[08:23] <tsdgeos> i haven't tried
[08:23] <tsdgeos> but see no reason why they shouldn't
[08:23] <tsdgeos> the new code is just bypassed by default
[08:23] <Saviq> tsdgeos, indeed
[08:29] <larsu> dednick: the unitymenumodel MR is up: https://code.launchpad.net/~larsu/qmenumodel/add-unitymenumodel/+merge/176857
[08:30] <larsu> dednick: it requires a few changes in the indicator code in unity8
[08:30] <dednick> larsu: whoop. i'll get on it asap
[08:31] <larsu> dednick: I'm around, you'll most likely have questions
[08:32] <larsu> dednick: maybe I should help porting?
[08:32] <hikiko> #ubuntu-gr
[08:32] <hikiko> sorry :)
[08:33] <dednick> larsu: yeah, probably. i'm just working on something else at the moment. I'll let you know once i start taking a look
[08:33] <tsdgeos> Saviq: so what should i be doing next, add the tests to help the distro-patching or continue helping greyback on the unity-mir thing?
[08:33] <Saviq> tsdgeos, mir for now
[08:33] <tsdgeos> ok
[08:34] <greyback> tsdgeos: there's a review there, if you could take it please
[08:35] <tsdgeos> sure
[08:35] <tsdgeos> greyback: what should i be looking at regarding functional testing?
[08:40] <greyback> tsdgeos: use this with qml-demo-shell: http://pastebin.ubuntu.com/5910449/
[08:40] <greyback> tsdgeos: when surface created, I can (with a delay) grab the applicatoin screenshot.
[08:40] <greyback> tsdgeos: also get notified of surface being destroyed
[08:43] <tsdgeos> greyback: do i need to update the phone?
[08:43] <tsdgeos> bazaar.launchpad.net/~mir-team/unity-mir/trunk/revision/19
[08:44] <tsdgeos> ?
[08:44] <greyback> tsdgeos: yes
[08:44] <tsdgeos> apt-get update enough? or android parts flash required?
[08:44] <greyback> tsdgeos: apt-get is enough
[08:49] <tvoss> Saviq, ping
[08:49] <Saviq> tvoss, pong
[08:50] <tsdgeos> !kill
[08:50] <tsdgeos> is a bad thing to run
[08:50] <tsdgeos> may kill not what you wanted :D
[08:52] <mhr3> greyback, the dbus appmanager api is already exposed?
[08:54] <greyback> mhr3: yes. I hope to be able to point you to a working phablet image in a few hours
[09:02] <mhr3> jamesh, ^
[09:03] <jamesh> awesome.
[09:17] <sil2100> jibel: hello!
[09:17] <sil2100> jibel: https://code.launchpad.net/~sil2100/otto/fix_1203809/+merge/176692 <- could you take a look later?
[09:17] <sil2100> ;)
[09:18] <jibel> Hey sil2100
[09:18] <jibel> sil2100, yes I had a look but I am not super fan of removing -e which IMO should be the default for any shell script :)
[09:19] <jibel> sil2100, I'll comment on the MP
[09:19] <sil2100> jibel: in this case it breaks the error handling we have
[09:19] <sil2100> jibel: since sometimes you for instance expect a command to return false when something wrong is going on
[09:20] <sil2100> jibel: I can, of course, change that and simply do a set -e and then set +e for the specific piece of code I need to work
[09:20] <jibel> sil2100, yes, maybe that's the bit that must be fixed, but honestly I didn't really have time to look closely at the problem
[09:21]  * sil2100 doesn't like -e
[09:22] <sil2100> It's as if every function in C that returns non-0 would cause the program to bail out
[09:25] <tsdgeos> wy do i need to login in the ubuntu pastebin
[09:25] <tsdgeos> ...
[09:26] <Saviq> tsdgeos, +1000000000000000000000
[09:27] <larsu> tsdgeos: use fpaste.org
[09:27] <tsdgeos> breaks my wget
[09:27]  * larsu hides
[09:27] <tsdgeos> larsu: well it was greyback that did the paste, i just copied it to paste.kde.org so i could wget it from the phone :D
[09:27] <Saviq> tsdgeos, https://bugs.launchpad.net/ubuntu-website-content/+bug/784485
[09:28] <tsdgeos> greyback: about that test shell you said, shall qmlscene foo.qml work?
[09:28] <greyback> tsdgeos: nope, you'll need lp:~gerboland/+junk/qml-demo-shell/
[09:28] <tsdgeos> greyback: ah, so replace the qml from there with that new one?
[09:28] <tsdgeos> Saviq: pfff
[09:29] <greyback> tsdgeos: that has small main.cpp to set up the mir server. Yep then replace main.qml
[09:29] <Saviq> tsdgeos, indeed
[09:29] <greyback> tsdgeos: I thought I'd pointed you to qml-demo-shell before, sorry for confusion
[09:29] <tsdgeos> greyback: sure you had
[09:30] <tsdgeos> was just unsure if this was supposed to work standarlone or not
[09:30] <greyback> gotcah
[09:31]  * Saviq just wasted an hour of his life to realize there was a missing ! in the shebang...
[09:31] <Saviq> tsdgeos, btw, tried 5.1, got worse results and less usefull tracebacks
[09:32] <Saviq> tsdgeos, but I expect this to be caused, at least in part, by the fact that I haven't rebuilt stuff
[09:32] <tsdgeos> ouch
[09:32] <tsdgeos> may be
[09:32] <Saviq> tsdgeos, btw the support-qt51 branch, if you want the things you commented with in, I'm game
[09:32] <Saviq> tsdgeos, and why didn't you merge-propose before already? ;)
[09:32] <tsdgeos> because ifdefs suck :D
[09:32] <Saviq> tsdgeos, better idea?
[09:32] <tsdgeos> not relaly
[09:33] <Saviq> tsdgeos, we could reduce them with typedefs
[09:33] <tsdgeos> i mean what i had was https://code.launchpad.net/~aacid/+junk/lvwph_51 and https://code.launchpad.net/~aacid/+junk/lvwph_50
[09:33] <tsdgeos> just did a diff and compared to your patch
[09:33] <Saviq> tsdgeos, ah got it
[09:34] <tsdgeos> Saviq: i think the suggested changes don't hurt, so if you don't disagree adding them++
[09:34] <tsdgeos> greyback: something went wrong somewhere
[09:34] <Saviq> tsdgeos, so the only real change was the QQuickItem/QObject for QChangeSet, was it?
[09:34] <Saviq> tsdgeos, but that we should actually be able to not ifdef
[09:35] <tsdgeos> greyback: http://paste.ubuntu.com/5910586/ when compiling the demo shell
[09:35] <greyback> tsdgeos: the best kind of bug report :D
[09:35] <Saviq> tsdgeos, anyway, not this week ;)
[09:35] <greyback> tsdgeos: yay the Mir API change
[09:36] <tsdgeos> Saviq: there's three changes, the changeset class renamed, the model class renamed and the model functions returning QObject instead of QQuickItem
[09:36] <tsdgeos> greyback: but i dist-upgrded? is not yet there? or i miss some ppa?
[09:36] <greyback> tsdgeos: what package version have you for libmirserver0?
[09:36] <tsdgeos> 0.0.8bzr883saucy0+834~saucy1
[09:37] <mhr3> Saviq, do you know if there's a bug for all the missing themed icons?
[09:37] <Saviq> mhr3, https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1192948 ?
[09:37] <Saviq> mhr3, that's at least one
[09:37] <Saviq> tsdgeos, the renames we could handle with typedefs, not sure it's better, though
[09:38] <mhr3> thanks
[09:38] <greyback> tsdgeos: ok, that's correct.
[09:38] <greyback> tsdgeos: right, need to fix that... Wanna try? I need to be afk for 40 mins
[09:38] <tsdgeos> Saviq: i was hoping at one point we'd just switch to 5.1 and don't need any define/typedef
[09:38] <Saviq> tsdgeos, the s/QQuickItem/QObject/ could probably be handled better, too, if we assumed QObject for 5.0 as well
[09:39] <tsdgeos> greyback: any pointer?
[09:39] <Saviq> tsdgeos, sure, being ready doesn't hurt :)
[09:39] <tsdgeos> Saviq: i was ready with the code in the other repo ;-)
[09:39] <Saviq> :P
[09:40] <greyback> tsdgeos: actually I'm confused now. unity-mir doesn't directly reference any of those symbols, but does inherit mir::DefaultServerConfiguration (see shellserverconfiguration.h)
[09:41] <tsdgeos> greyback: ok, i'll try to rebuild everything
[09:41] <tsdgeos> maybe something got borked
[09:41] <greyback> tsdgeos: I've kicked off a unity-mir rebuild, will be a new package available in 30, if everything is ok.
[09:47]  * greyback back in 40
[10:01] <Saviq> asac, ok, so how do we proceed with unity8 for smokes?
[10:02] <asac> Saviq: so i think we could start with having unity8 published again and then ensure that our dashboard does not regress
[10:02] <Saviq> didrocks, ↑
[10:02] <Saviq> sil2100, ↑
[10:02] <asac> once thats the case we can land our tests (or at the same time)
[10:03] <asac> Saviq: didrocks: would be good to understand what to do when it breaks our results through
[10:03] <asac> e.g. can we backout? do we have a mitigationm strategy?
[10:03] <nic-doffay> Saviq, got a moment?
[10:03] <asac> how can we plan so we can say 100% certain that if there is regression it won't be in for longer than 4 hours or so
[10:03] <asac> mayube we cannot answer these perfectly, but we should start answering those to exercise the idea
[10:03] <Saviq> asac, you mean for manual testing, as there was no automated one?
[10:03] <asac> of guaranteed max. downtime
[10:04] <asac> Saviq: well. you have stuff ou prepare, and you need a plan how to get stuff reverted in case you break something that you didnt see in yuour testing before
[10:04] <asac> so you can investigate without being on a firedrill
[10:04] <asac> 1. test and prep
[10:04] <asac> 2. land
[10:04] <asac> 3. revert/backout in case there is any regression
[10:04] <asac> 4. reland once fixed
[10:05] <asac> the unity8 tests we can shovel in at samme time if you want
[10:05] <asac> or just say today we see if unity8 breaks other stuff
[10:05] <asac> your call
[10:05] <Saviq> asac, the only thing I'm afraid of now is the occasional Qt crash on manta
[10:05] <Saviq> asac, I say let's land it, and push everything through asap (so release, image, test)
[10:06] <sil2100> Saviq: ok
[10:06] <Saviq> asac, if we regress, I revert everything between now and previous release
[10:06] <Saviq> asac, and we push through again
[10:06] <sil2100> Saviq: didrocks is having internet issues today - and we have problems with the cu2d tools right now
[10:06] <sil2100> Saviq: but once he's back and we deal with those, we get it published
[10:07] <Saviq> sil2100, thanks
[10:07] <Saviq> asac, for the general case I'd say we need a "cutoff" time, so if there's a regression - we revert it if there isn't a fix within x hours
[10:07] <Saviq> asac, can even be automated
[10:11] <asac> Saviq: right. lets give folks 12 hours to land stuff and get zero regressions on other tests
[10:11] <asac> so talk to didrocks if and how we could revert at atll
[10:11] <asac> if so we can see if we stick to 12 hours
[10:11] <asac> or give you a bit more time :)
[10:12] <Saviq> asac, only thing I'm afraid of is the occasional Qt crasher
[10:12] <asac> Saviq: if that only shoots your own tests its not as bad as if it causes others to crash
[10:12] <Saviq> asac, wasn't able to verify it's fixed with Qt 5.1, but let's see how problematic it is?
[10:12] <asac> i guess we should just try
[10:12] <asac> and see :)
[10:13] <Saviq> asac, +1
[10:13] <asac> i think its important that we are also open to throw stuff in
[10:13] <asac> without knowing 100% what will happen
[10:13] <asac> as long as we are also happy to throw it out again :)
[10:13] <asac> without much emotional feeling
[10:22] <Saviq> asac, yeah, no hurt feelings
[10:23] <mzanetti> Saviq: is there a review I can do without stepping on someones toes?
[10:26] <tsdgeos> mzanetti: https://code.launchpad.net/~aacid/unity8/updateFirstVisibleIndexOnEmptyList/+merge/176146 :D
[10:27] <mzanetti> tsdgeos: cheers
[10:28] <tsdgeos> mzanetti: this goes with http://bazaar.launchpad.net/~unity-team/unity8/trunk/revision/122
[10:28] <tsdgeos> i'll restart the CI job
[10:28] <dednick> larsu: ping
[10:29] <dednick> larsu: http://pastebin.ubuntu.com/5910706/
[10:30] <larsu> dednick: oopsie. What wee you running?
[10:30] <larsu> *were
[10:31] <dednick> larsu: just started modding unity8 for a UnityMenuModel
[10:31] <larsu> dednick: any output on stderr?
[10:32] <dednick> larsu: (process:29040): GLib-CRITICAL **: g_sequence_get: assertion '!is_end (iter)' failed
[10:32] <dednick> larsu: hang on, there's actually quite a bit
[10:36] <dednick> larsu: actually nothing else interesting i dont think
[10:36] <dednick> larsu: but your example seems to be working for me, so i'm not sure what's happening
[10:36] <larsu> dednick: ya, I'm trying to make it fail right now, with no success so far
[10:37] <dednick> larsu: have 2 models pointing at the same dbus
[10:37] <larsu> dednick: ooh! Shouldn't be a problem, but let me try that
[10:38] <larsu> dednick: ya, I can reproduce
[10:38] <larsu> dednick: will fix it right away
[10:38] <dednick> larsu: cool. thanks
[10:39] <dednick> larsu: although i should really fix the need for 2 in unity...
[11:04] <larsu> hm dednick is gone. I'm going to lunch, can someone please tell him I've fixed his issue if he's asking?
[11:48] <Cimi> need some c++ help :)
[12:02] <Cimi> mzanetti, did you have chance to look back at the review of the settings components?
[12:02] <mzanetti> Cimi: hmm... don't remember what was the last state there
[12:02]  * mzanetti checks
[12:04] <mzanetti> Cimi: according to this it seems I did: https://code.launchpad.net/~cimi/ubuntu-settings-components/calendar-newmodel/+merge/172598
[12:05] <Cimi> will have a look later… I'm knackered by an headache since I woke up and I am giving up now
[12:22] <mhr3> larsu, do you know where are all the gvariant<>qvariant copies?
[12:22] <mhr3> larsu, i just found a bug there and fixing it in 3 projects doesn't sound too appealing
[12:22] <mhr3> larsu, so i'm just thinking about exposing it from dee-qt
[12:23] <mhr3> larsu, would that work for your uses?
[12:27] <larsu> mhr3: you mean make gsettings-qt link against dee? I'd rather not to be honest...
[12:27] <larsu> mhr3: what's the bug?
[12:28] <mhr3> larsu, and that's why i was asking where was the additional copy
[12:28] <mhr3> larsu, it didn't handle variants
[12:28] <mhr3> as in "v"
[12:30] <larsu> mhr3: I don't know where all the copies are. There's definitely one in gsettings-qt and one in qmenumodel (I'm on my way to deprecate that one, though)
[12:32] <mhr3> doesn't look like the gsettings-qt one is more complete than dee-qt
[12:33] <mhr3> it's different though :)
[12:33] <larsu> ya. desrt wrote it :)
[12:34] <larsu> this problem is not really solvable in a satisfactory way anyway. I'd love to just ignore it...
[12:56] <jbicha> MacSlow: could you look into why the notify-osd autopkgtest is failing? it's keeping gtk+3.0 from migrating out of proposed
[12:57] <MacSlow> jbicha, I'm busy dealing with ap-test atm... has to wait a bit, if that's ok
[12:59] <seb128> jbicha, notify-osd didn't change in years, are you sure it's just not finding an issue with the gtk patches?
[13:00] <jbicha> seb128: well https://jenkins.qa.ubuntu.com/view/Saucy/view/AutoPkgTest/job/saucy-adt-notify-osd/ didn't start failing this week
[13:02] <jbicha> but I've no idea if the gtk upload made things worse
[13:02] <seb128> jbicha, can you reproduce the issue?
[13:03] <seb128> jbicha, I doubt waiting on MacSlow is a good strategy here, he's busy enough on other thing and notify-osd didn't change in ages, so not likely an issue there
[13:06] <greyback> tsdgeos: any chance to look at https://code.launchpad.net/~gerboland/unity-mir/surface-notifications/+merge/176719
[13:07] <jbicha> seb128: I'm finally able to reproduce tests that fail but I don't understand what needs to be done to fix them
[13:07] <MacSlow> seb128, jbicha: from what I can take form those error-logs the deprecation warnings trigger the failure due to the use of gtk_window_set_opacity() and gdk_display_get_pointer()
[13:08] <MacSlow> seb128, jbicha: later this evening I can try to see how fast I can come up with a fix and MP for this.... but no promise as the autopilot-tests for phone-shell notifications have higher prio
[13:08] <mzanetti> dandrader: ah.. sorry. the mailman took a bit longer for the list
[13:08] <seb128> jbicha, from the log it seemed that there was a segfault, having a stacktrace would be a good first step
[13:08] <mzanetti> dandrader: ignore my last mail to you
[13:09] <seb128> MacSlow, I doubt it's a depreciation issue, tests were working on that same version of gtk earlier in the cycle
[13:09] <dandrader> mzanetti, no, you were right. when I click "reply" on thunderbird it replies to the sender only. I've to click on "reply to list" to get it where I wanted
[13:09] <dandrader> so I had to reply it again indeed
[13:10] <mzanetti> dandrader: yeah, but you figured and corrected before I complained :)
[13:10] <dandrader> mzanetti, yes :)
[13:10] <MacSlow> seb128, when searching for "gtk_window_set_opacity" in https://jenkins.qa.ubuntu.com/view/Saucy/view/AutoPkgTest/job/saucy-adt-notify-osd/37/ARCH=i386,label=adt/consoleFull I find numerous warnings
[13:10] <seb128> MacSlow, that shouldn't make the test fail though
[13:10] <MacSlow> seb128, ah ok
[13:10] <tsdgeos> greyback: approved, we have no automerger, right?
[13:11] <greyback> tsdgeos: correct, could you merge please?
[13:11] <tsdgeos> i can try :D
[13:17] <jbicha> MacSlow: here's what I got so far: http://paste.ubuntu.com/5911129/
[13:22] <MacSlow> jbicha, hm... that should certainly not happen... the name/string wasn't changed
[13:24] <MacSlow> jbicha, can it be that notify-osd isn't started on that VM where the test is run for some reason... or there's a timing issue with starting notify-osd and running the test?
[13:24] <jbicha> yeah maybe that's not the root issue, I don't normally even have notify-osd installed
[13:25] <MacSlow> jbicha, you ran that locally on your machine?
[13:25] <jbicha> yes, the jenkins output wasn't helpful
[13:26] <MacSlow> jbicha, well the withlib tests need notify-osd to be running
[13:38] <greyback> mterry: loicm would be able to help you
[13:38] <mterry> greyback, ok, thanks
[13:39] <sil2100> greyback: hi!
[13:39] <greyback> sil2100: hey
[13:39] <sil2100> greyback: could you take a look and check if the deps look ok?
[13:39] <sil2100> https://code.launchpad.net/~sil2100/unity-mir/packaging_split/+merge/176935
[13:40] <greyback> sil2100: sweet, thanks! Will look immediately
[13:44] <mzanetti> dandrader: for the notes... I didn't understand which MP you finished
[13:44] <dandrader> mzanetti, https://code.launchpad.net/~dandrader/unity8/lp1116207/+merge/175163
[13:45] <mzanetti> ok, thanks
[13:49] <jbicha> the automerger runs autopkgtest, right?
[13:51] <jbicha> MacSlow: could you try approving https://code.launchpad.net/~jbicha/notify-osd/tests-depend-on-at-spi2-core/+merge/176943 then?
[13:54] <MacSlow> jbicha, how does that fix the assertion-failure?
[13:55] <jbicha> MacSlow: which assertion failure?
[13:55] <mzanetti> Saviq: I've seen you resubmitted https://code.launchpad.net/~unity-team/unity8/app-preview-data/+merge/176941
[13:55] <mzanetti> is it ready to be re-reviewed?
[13:55] <Saviq> mzanetti, I didn't touch the code
[13:56] <Saviq> mzanetti, I had to resubmit because I resubmitted the prerequisite (that I touched
[14:00] <MacSlow> jbicha, did you paste this http://paste.ubuntu.com/5911129 earlier?
[14:00] <MacSlow> jbicha, line 8
[14:00] <mhr3> sil2100, btw what's up with unity? seems like it's stuck in proposed?
[14:00] <jbicha> MacSlow: uh, I had to install notify-osd, log out of GNOME Shell and log into Unity :|
[14:01] <MacSlow> jbicha, so that's solved, right?!
[14:02] <jbicha> MacSlow: my test environment isn't quite the same as the run autopkgtest runs in but I have confirmed that the tests fail if at-spi2-core isn't installed
[14:02] <jbicha> maybe something that used to depend on it a few weeks ago no longer does
[14:12] <sil2100> mhr3: hm, dunno, will ask around in a moment
[14:12] <Saviq> greyback, can you do me a scratchpad please?
[14:12] <Saviq> greyback, erm, sketchpad?
[14:13] <greyback> Saviq: http://studio.sketchpad.cc/WEz742SNCE
[14:13] <Saviq> greyback, ta
[14:28] <mzanetti> Saviq: hmm... weird... autopilot tests don't work for me
[14:28] <mzanetti> AttributeError: 'NoneType' object has no attribute 'isLoaded'
[14:28] <mzanetti> do they work for you?
[14:39] <mhr3> tsdgeos_, could you https://code.launchpad.net/~mhr3/dee-qt/variant-conversions/+merge/176949 ?
[14:42] <tsdgeos_> mhr3: why the statics in deemodel?
[14:43] <mhr3> tsdgeos_, there's a comment explaining that
[14:43] <tsdgeos_> right
[14:44] <tsdgeos_> but why not simply namespace variantconversions.h ?
[14:46] <mhr3> tsdgeos_, didn't seem like the qt-way
[14:47] <sil2100> bschaefer: hi!
[14:47] <Saviq> mzanetti, yeah
[14:47]  * sil2100 has been waiting for bschaefer 
[14:47] <bschaefer> sil2100, hello!
[14:48] <sil2100> bschaefer: can I nag you with some XIM SRU questions?
[14:48] <bschaefer> sil2100, sure, ive a meeting in 12 min :)
[14:48] <tsdgeos_> mhr3: tbh the statics feel  a bit weird for me, i'd go for the namespace, but maybe gather some other opinions?
[14:49] <tsdgeos_> mzanetti: Saviq: greyback: ↑↑↑ https://code.launchpad.net/~mhr3/dee-qt/variant-conversions/+merge/176949
[14:49] <mhr3> tsdgeos_, i'm fine eitherway really...
[14:52]  * mzanetti would probably use the statics... but no real reason... just personal preference
[14:52] <tsdgeos_> mhr3: sure, just trying to gather some more opinions, api design is like colors, everyone likes one :D
[14:53] <mhr3> tsdgeos_, no, it's like cars, not colors :P
[14:53] <tsdgeos_> mzanetti: mhr3: thing is that to be it'd feel weird having to use DeeListModel::VariantForData in something that may not have anything to do with a Dee or DeeModel
[14:53] <tsdgeos_> that's why i suggest a more "neutral" namespace
[14:54] <mhr3> tsdgeos_, agreed, but then it'd ideally be a completely separate lib
[14:54] <tsdgeos_> yep D:
[14:54] <mzanetti> tsdgeos_: right... in that case yeah. I assumed it would be only used internally
[14:54] <mhr3> it feels somewhat natural to me cause GVariants is Dee's "data"
[14:55] <mzanetti> dunno... I wouldn't mix that stuff. but then again, I have no clue about Dee
[14:55] <tsdgeos_> ok, let's go for this
[14:55] <mzanetti> but if its a D-Bus protocol I would probably just parse it in Qt instead of mixing in G stuff
[14:55] <tsdgeos_> and maybe once people start using this externally
[14:56] <mhr3> mzanetti, it's not so much about dee really, it's just we need this in one place and we don't want to create completely new lib for the two statics
[14:56] <tsdgeos_> we can decided on the naming :D
[14:57] <tsdgeos_> mhr3: about the "Pawel Stolowski <pawel.stolowski@canonical.com>"
[14:57] <tsdgeos_> the code in there is mostly Florian's
[14:57] <tsdgeos_> with a bit of mine and Pawel's
[14:57] <tsdgeos_> tbh i have those Authors: lines
[14:57] <tsdgeos_> it always gets out of sync
[14:57] <tsdgeos_> and that's why we have bzr blame :D
[14:59] <mhr3> tsdgeos_, i just copied it from unity8's plugins :P
[14:59] <dandrader> tsdgeos_, yes, they're are not the supreme truth but still pretty handy. specially when code is moved between repos etc
[14:59] <tsdgeos_> well thing is the code is already in that repo :D
[15:00] <tsdgeos_> anyway looks good
[15:00] <tsdgeos_> i'd suggest improvements
[15:00] <tsdgeos_> but they are unrelated
[15:01] <Cimi> have issues with my plugin
[15:01] <mhr3> tsdgeos_, shoot if i'm going to change it anyway
[15:01] <Cimi> I tried mocking gsettings
[15:01] <Cimi> this is the code
[15:01] <tsdgeos_> mhr3: no idea why pawel added the QVariantHash thing as first thing of QVariantFromGVariant for example
[15:01] <Cimi> http://paste.ubuntu.com/5911433/
[15:01] <Cimi> but the error that qml prints is
[15:01] <tsdgeos_> instead of inside case G_VARIANT_CLASS_ARRAY where probably mkes more sense
[15:01] <Cimi> file:///home/cimi/Development/unity/unity8.wallpaper/Shell.qml:120:9: Invalid grouped property access
[15:01] <Cimi>              schema.id: "org.gnome.desktop.background"
[15:01] <tsdgeos_> and then he added no test for in in the conversiontetst.cpp
[15:02] <tsdgeos_> but as said nothing to do with pure "let's expose this method"
[15:02] <mhr3> tsdgeos_, we were talking about it, generic a{??} is just too generic
[15:02] <hakermania> Hello there. I saw the logs and I saw that there was a discussion here about the Qt Creator's startup error (Cannot mix incompatible Qt library (version 0x50001) with this library (version 0x50002))
[15:02] <hakermania> Was anyone able to solve it?
[15:02] <tsdgeos_> hakermania: recompile the thing that uses 5.0.1 to use 5.0.2
[15:03] <tsdgeos_> or viceversa
[15:03] <tsdgeos_> dirty trick is editing the .so and changing the version manually
[15:03] <tsdgeos_> but you'll explode if you try to do that :D
[15:03] <tsdgeos_> mhr3: ok, so i'm approving it, ok?
[15:03] <mhr3> tsdgeos_, ok
[15:04] <hakermania> tsdgeos_, thanks! Any more details? I just tried adding the ubuntu-sdk and Creator didn't start.
[15:04] <hakermania> Removing the SDK, and completely purging & reinstalling Creator and the libs didn't work either.
[15:04] <tsdgeos_> hakermania: you have some ppa mix probably, can't really help more
[15:04] <hakermania> tsdgeos_, thanks, anyway ! :)
[15:05] <dandrader> Cimi, I've hit this "Invalid grouped property access" before. worked around it by using a different api. didn't get to pin-point the cause :(
[15:05] <Cimi> dandrader, but the api works for the real gsettings plugin, this is a mock with less code...
[15:05] <Cimi> Saviq, you master, have idea? ^
[15:06] <dandrader> Cimi, maybe you have to register the types or somehting (QML_REGISTER_TYPE, if I'm not mistaken)
[15:06] <Cimi> mmm ok
[15:06] <Saviq> Cimi, not really
[15:07] <Saviq> larsu, maybe you can help ↑ - it's your API after all :)
[15:07] <sil2100> kenvandine: hi!
[15:07] <sil2100> kenvandine: do you have a moment for a packaging merge review?
[15:07] <Cimi> Saviq, larsu : dandrader was right
[15:07] <Cimi> it's the register type thing
[15:08] <Saviq> Cimi, good
[15:08] <kenvandine> sil2100, sure
[15:08] <sil2100> kenvandine: https://code.launchpad.net/~sil2100/unity-mir/packaging_split/+merge/176935
[15:09] <larsu> Cimi: woah, why are you mocking it from c++?!
[15:09] <larsu> this is crazyness
[15:10] <kenvandine> sil2100, why the -dbg package?
[15:10] <Cimi> larsu, because I cannot set a property to use id
[15:10] <mhr3> tsdgeos_, https://code.launchpad.net/~mhr3/unity8/no-variant-conversion-copies/+merge/176964 :)
[15:10] <Cimi> larsu, it's reserved to qml
[15:10] <Cimi> larsu, so schema.id is not doable
[15:11] <sil2100> kenvandine: Gerry wanted to have debugging separately
[15:11] <larsu> Cimi: lol
[15:11] <larsu> Cimi: sorry about that :)
[15:11] <Cimi> no worries
[15:11] <Cimi> larsu, I would do anything to avoid using C++, trust me :)
[15:12] <kenvandine> sil2100, and libunity-mir1 should conflict/replace unity-mir
[15:12] <tsdgeos_> mhr3: magic!
[15:12] <sil2100> kenvandine: hm, can add that, but unity-mir wasn't released yet, so I thought it's not needed?
[15:12] <sil2100> kenvandine: since it's a fresh package we don't daily yet, not even preNEWed
[15:12] <mhr3> tsdgeos_, and since i got hold of reviewer :) https://code.launchpad.net/~mhr3/unity8/scopes-real-category-ids/+merge/176139
[15:12] <sil2100> Do we need to handle the older cases too?
[15:12] <kenvandine> sil2100, not even in the PPA?
[15:13] <sil2100> kenvandine: true, maybe in some PPA it was published, but not in daily-build etc.
[15:13] <kenvandine> yeah, but the mir-team ppa was publicized a bit
[15:13] <sil2100> kenvandine: it doesn't have autolanding even yet, but it was in some phablet-team PPA I guess - should I add that anyway?
[15:13] <sil2100> Ok
[15:13] <sil2100> Sounds reasonable
[15:13] <kenvandine> it would be nice to people :)
[15:14] <kenvandine> i know i installed it
[15:14] <sil2100> :D
[15:14] <mzanetti> Saviq: do you perhaps have a hint for me?
[15:14] <mzanetti> Saviq: the tests just don't work
[15:14] <mzanetti> introspection fails badly
[15:14] <Saviq> mzanetti, did you try in the VM?
[15:14] <mzanetti> Saviq: no
[15:15] <kenvandine> sil2100, actually unity-mir wasn't in that ppa
[15:15] <Saviq> mzanetti, http://10.97.0.1:8080/view/cu2d/view/Head/view/Unity8/ says it works
[15:15] <mzanetti> Saviq: yeah, I know... still I'd like to run them locally too
[15:16] <Saviq> mzanetti, yeah I know
[15:16] <mzanetti> It seems as it wouldn't load any scopes
[15:16] <sil2100> kenvandine: I pushed a modification, added also Provides: unity-mir, not sure if it's needed though
[15:16] <mzanetti> but if I just run it all the scopes are here
[15:16] <mzanetti> at least the home scope is
[15:17] <mzanetti> which is the one that fails
[15:18] <mzanetti> Saviq: and I just realized that the lockscreen scenarios are messed up since the refactor
[15:18] <mzanetti> Saviq: the lockscreens are not supposed to be there on the Nexus10 for example... I wonder how it actually can work
[15:18] <kenvandine> sil2100, i'm having second thoughts now :)
[15:19] <kenvandine> it wasn't in the team that was publicized for testing out mir
[15:19] <kenvandine> so maybe best to just leave the cruft out
[15:19] <Saviq> mzanetti, like never?
[15:19] <kenvandine> i don't even see a source for it
[15:19] <mzanetti> Saviq: no... the tablet form ractor is supposed to have the pinlock on the greeter
[15:19] <Saviq> mzanetti, sounds like we have a bug, then
[15:20] <mzanetti> yeah
[15:20] <Saviq> mzanetti, as the lockscreens work just fine
[15:20] <mzanetti> Saviq: well, it might be the combination of various lightdm mocks
[15:21] <mzanetti> but yeah... need to check
[15:21] <kenvandine> sil2100, ah, it's in ppa:phablet-team/mir
[15:21] <kenvandine> sil2100, i looks like a ppa just for transitioning unity8 to mir?
[15:25] <Saviq> kenvandine, sil2100 yes it is
[15:25] <kenvandine> Saviq, do you care if we provide an easy transition for people that have that installed?
[15:25] <Saviq> greyback, ↑
[15:26] <kenvandine> i suspect that isn't really used by real users
[15:26] <greyback> reading..
[15:26] <mzanetti> Saviq: hmpf... found it
[15:26] <Saviq> mzanetti, some environment issue?
[15:26] <mzanetti> Saviq: we can run them locally any more... I had some unity8 package installed and it launched that one...
[15:27] <mzanetti> Saviq: uninstalled it and calling autopilot run unity8 locally fails now
[15:27] <tsdgeos_> guys
[15:27] <tsdgeos_> can't build the shell
[15:27] <tsdgeos_> complains about missing /usr/include/unity/shell/launcher/LauncherItemInterface.h
[15:27] <tsdgeos_> what am i missing?
[15:27] <Saviq> tsdgeos_, ppa:ubuntu-unity/next
[15:28] <Saviq> tsdgeos_, ./build -s; ./build -c
[15:28] <greyback> kenvandine: unity-mir only appeared in that PPA yesterday. No real user has picked it up, it's not worth the transition IMO
[15:28] <mzanetti> Saviq: get_binary_path() seems borked
[15:28] <kenvandine> greyback, thanks!
[15:28] <kenvandine> sil2100, ok... drop that last change :)
[15:28] <kenvandine> sil2100, sorry for the noise
[15:29] <mzanetti> what was wrong with my version of the ap tests? except the failures on the phone which turned out to be platform bugs everything worked fine :/
[15:29] <mzanetti> now its totally complicated and breaks all sorts of cases except the installed run
[15:30] <greyback> sil2100: I approve of the changes, but haven't tested it. If that lands, will need an update the build deps of  lp:~unity-team/unity8/unity8-integrate-mir/ it builds ok
[15:30] <greyback> +so that it builds ok
[15:31] <Saviq> mzanetti, we have QEs for that now ;)
[15:33] <Saviq> mzanetti, but indeed get_binary_path seems broken
[15:33] <jbicha> MacSlow: could you Approve the status too? https://code.launchpad.net/~jbicha/notify-osd/tests-depend-on-at-spi2-core/+merge/176943
[15:33] <Saviq> mzanetti, the autopilot target works, though
[15:33] <MacSlow> jbicha, sorry... always forget that
[15:34] <Saviq> mzanetti, aaah
[15:34] <Saviq> mzanetti, you need to make install first
[15:34] <Saviq> mzanetti, /my fault for not advertising that
[15:34] <Saviq> mzanetti, but that's needed so that we only pick up the mocks that we really need
[15:34] <Saviq> mzanetti, not all of those that are there
[15:35] <MacSlow> mzanetti, I think I mentioned that ... the "make -C buildir install"? Or did I forget that?
[15:36] <mzanetti> MacSlow: yeah... you did... but between a ton of other useless information so my brain skipped it
[15:36] <mzanetti> sorry... /me is pissed of with ap
[15:37] <MacSlow> mzanetti, yeah... I should have made that point stand out stronger
[15:37] <Saviq> mzanetti, anyway, that's why you have the autopilot target
[15:38] <Saviq> mzanetti, make autopilot and it'll work
[15:38] <mzanetti> Saviq: can't choose single tests there
[15:38] <mterry> Cimi, you are working on the first-boot setup screens?
[15:38] <Saviq> mzanetti, which we should try and fix
[15:38] <Saviq> mzanetti, but yeah
[15:38] <mzanetti> and I won't wait 34 * 40 secs to see the first of MacSlow's failing tests
[15:38] <Cimi> mterry, will work
[15:38] <Cimi> mterry, if you want we can work together
[15:38] <mterry> Cimi, ok.  Let me set the blueprint item to you (from xnox)
[15:38] <Cimi> mterry, think I will start monday
[15:38] <tsdgeos_> mhr3: i'd prefer if you find someone to review that one
[15:38] <mterry> Cimi, do you have all the design documents you need?
[15:39] <Cimi> mterry, didn't look yet, I know I have wireframes for everything, and maybe some visuals
[15:39] <tsdgeos_> mhr3: my phone is in unity-mir mode and it's hard to test stuff like that is not breaking in the phone since the phone is "broken" already :D
[15:39] <Cimi> mzanetti, how would you test the background changed?
[15:39] <mhr3> tsdgeos_, works on desktop too ;)
[15:39] <mterry> Cimi, OK, design stuff should exist
[15:39] <Cimi> mzanetti, I now have a fake working plugin
[15:40] <mterry> Cimi, thanks, bug me or katie if you need more
[15:40] <tsdgeos_> mhr3: somehow i get no data in the installed thing on the destkop
[15:40] <Cimi> ok
[15:40] <mzanetti> Cimi: did you write a c++ plugin for that?
[15:40] <Cimi> yes
[15:40] <tsdgeos_> mhr3: any idea why that may happen?
[15:40] <mzanetti> Cimi: oh... now it makes sense (your comment in the standup)
[15:40] <Cimi> mzanetti, http://paste.ubuntu.com/5911557/
[15:40] <Saviq> kgunn, http://studio.sketchpad.cc/WEz742SNCE
[15:40] <Cimi> mzanetti, was terribly killed by my headache during the standup
[15:41] <mhr3> tsdgeos_, hmm, let me check, did that branch a few days ago
[15:41] <mzanetti> Cimi: Should have been possible to just create file GSettings.qml with this content Item {property imageUrl}
[15:41] <Cimi> mzanetti, feeling fine now
[15:41] <tsdgeos_> mhr3: i mean i get nothing even without the patch
[15:41] <mzanetti> Cimi: ah no...
[15:41] <mzanetti> Cimi: you need the plugin...
[15:41] <mhr3> tsdgeos_, ah, yea i had that at some point as well, but somehow it magically fixed itself
[15:41] <tsdgeos_> ah no wait
[15:41] <tsdgeos_> without it i get stuff
[15:41] <Saviq> kgunn, anything else you had in mind?
[15:41] <mzanetti> Cimi: Register a second class with the plugin, probably as a singleton with a setImageUrl() method
[15:42] <mzanetti> Cimi: in there you just change the url from the other class
[15:42] <mhr3> tsdgeos_, ehm, so the branch is actually broken?
[15:42] <mhr3> then i really do need to look at it again
[15:42] <mzanetti> Cimi: then you can use something like this in your test: GSettingController.setImageUrl("/foo/bar/baz.png")
[15:42] <tsdgeos_> mhr3: maybe not, let me try again
[15:42] <mzanetti> Cimi: and all the created instances of the GSettings would have onUrlChanged
[15:43] <mzanetti> Cimi: makes sense?
[15:43] <mzanetti> MacSlow: anyways... can finally run tests now. will try to sort your issue
[15:44] <tsdgeos_> mhr3: it works fine, somehow the "stuff not loading" also automagiclly fixed itself
[15:45] <mzanetti> mterry: I discovered that we have cases where the greeter has a password field (like on the tablet mode) and additionally the phone lockscreen pops up behind that. Do you think you can have a look at that? must be something how we handle the lightdm mocks
[15:45] <MacSlow> mzanetti, great
[15:45] <mhr3> tsdgeos_, for one, there's nothing displayed if i don't have the patch :)
[15:45] <Cimi> mzanetti, i need visually
[15:46] <Cimi> mzanetti, how do I check the shell really changes wallpaper?
[15:46] <MacSlow> mzanetti, I'll be away for dinner... but back in 1-2 hours (any findings via eMail I suggest)
[15:46] <mzanetti> MacSlow: I'll be online
[15:46] <MacSlow> mzanetti, d'accord
[15:46]  * kgunn reads saviq's sketchpad
[15:47] <mzanetti> Cimi: ah... I'd say you check for statusChanged on the CrossFadeImage and when it goes to Image.Ready you check the current source of the crossfadeimage
[15:47] <Cimi> makes sense
[15:47] <mzanetti> :)
[15:47] <mterry> mzanetti, on call, hold on
[15:48] <kgunn> Saviq: ...post review with rick, would be good to have the lockscreen on in one of those
[15:48] <Cimi> smspillaz, think your twitter account got hacked :p
[15:48] <sil2100> kenvandine: reverted and pushed!
[15:48] <Saviq> kgunn, that we can enable at any time
[15:48] <Saviq> kgunn, adding
[15:49] <kgunn> Saviq: and can we add some dummy data back in for infog...so boring when it says "no data"
[15:49] <Saviq> kgunn, we can just make some photos ;)
[15:49] <kgunn> cool
[15:49] <Saviq> kgunn, and you'll have real data
[15:49] <kgunn> that ought to do it
[15:50] <Saviq> pete-woods, are there other sources available for infographics?
[15:50] <kenvandine> sil2100, thx
[15:51] <pete-woods> Saviq: at the minute, only the camera has been done, we're waiting on approval from security, then we'll make an announcement of the input API
[15:51] <Saviq> pete-woods, cool
[15:51] <pete-woods> Saviq: it's only a couple of lines of code to add a source to your app/service, so I'm hoping they will appear in short order then
[15:52] <Saviq> pete-woods, yup
[16:04] <Cimi> mzanetti, so with the controller class
[16:04] <Cimi> mzanetti, I don't understand how to write it
[16:04] <MacSlow> bbl
[16:04] <Cimi> I registered the gsettings class
[16:05] <Cimi> now I want this controller to control exactly the gestating component initialised by the qml file
[16:05] <Cimi> not sure how this could work
[16:06] <Cimi> would be better to simply change the pictureUri of that gsettings mock
[16:06] <Cimi> if I could be able to get this component from the test file
[16:09] <Cimi> mzanetti, don't think findChild works...
[16:10] <Saviq> mterry, heh, managed to trick the lockscreen :D
[16:10] <Saviq> mterry, when you delete the .unity8-greeter-demo file
[16:10] <Saviq> mterry, after authenticate() has been called
[16:11] <Saviq> mterry, you're screwed ;)
[16:11] <mterry> Saviq, don't delete the demo file!
[16:11] <Saviq> mterry, ;)
[16:12] <Saviq> mterry, but also, on phone authenticate() should probably be called on-unlock
[16:12] <mterry> Saviq, hm?
[16:13] <Saviq> mterry, although I know that's tricky (or maybe even not useful in the real life, since you'd be logged in when changing the auth method, so you'd have to lock first - so it would authenticate)
[16:13] <Saviq> mterry, never mind, just some ramblings
[16:13] <mterry> Saviq, detecting auth method changes is tricky indeed
[16:15] <mzanetti> Cimi: Write a C++ class for that Controller thingie. make it a singleton
[16:15] <mzanetti> Cimi: in the ctor of the GSettings mock register the mock with the controller
[16:15] <Cimi> è arabo per me :D
[16:15] <mzanetti> :D
[16:16] <Cimi> I think I understood a bit
[16:16] <mzanetti> Cimi: do you know how to make singletons in C++?
[16:16] <Cimi> mzanetti, that's my difficulty, I know what they are I don't know the sintax
[16:16] <Cimi> I'm googling
[16:17] <mzanetti> Cimi: make a private constructor and a public static method like this:
[16:17] <mzanetti> static GSettingsMockController *instance()
[16:17] <mzanetti> Cimi: then a private static variable: GSettingsMockController *s_instance
[16:18] <mzanetti> Cimi: in instance() do this:
[16:18] <mzanetti> Cimi: if (s_instance == 0) { s_instance = new GSettingsMockController(); } return s_instance
[16:18] <mzanetti> Cimi: so noone can do new GSettingsMockController because of the private ctor
[16:18] <Cimi> mzanetti, is this valid? http://subinsebastien.tumblr.com/post/15822604498/singleton-classes-in-c
[16:19] <mzanetti> Cimi: yeah... thats pretty much what I'm saying except that this example does lots of other stuff we don't need
[16:19] <Cimi> ok
[16:19] <Cimi> I'll shrink it
[16:20] <mzanetti> Cimi: anyways, because calling "new GSettingsMockController()" does not work (because of the private ctor) if you want an instance of that class, you need to get it through a call to instance()
[16:20] <mzanetti> Cimi: and that one only calls new GSettingsMockController() for the very first time it is called. all the other times it returns the one existing instance
[16:20] <Cimi> mzanetti, I'll copy and adapt that code
[16:22] <mzanetti> Cimi: do you have your branch with the plugin somewhere? I can quickly add that controller for you
[16:22] <mzanetti> Cimi: if you promise you study the code afterwards and can do it yourself next time
[16:22] <Cimi> mzanetti, I wanna try it
[16:22] <mzanetti> ah ok. fine with me
[16:22] <Cimi> mzanetti, I ail never learn otherwise :)
[16:22] <Cimi> *will
[16:22] <mzanetti> thats the spirit
[16:28] <nic-doffay> Saviq, good news, after a lot of fiddling around that Z order has been solved.
[16:29] <dandrader> Saviq,  how can I find the available lenses in a system and where are they registered/located? ie. what decides the lenses that show up in unity8?
[16:29] <Saviq> dandrader, com.canonical.Unity.Dash/scopes gsetting
[16:29] <Saviq> nic-doffay, cool
[16:30] <dandrader> Saviq, ok, thanks!
[16:30] <dandrader> will try that after lunch
[16:41] <mhr3> sil2100, did a moment pass? :)
[16:41] <jbicha> is notify-osd set to auto-publish?
[16:41] <Cimi> mzanetti, lp:~unity-team/unity8/unity8.background_gsettings_plugin
[16:42] <Cimi> mzanetti, some weird c++ errors...
[16:42] <sil2100> mhr3: it passed long ago ;p But didrocks passed as well!
[16:42] <sil2100> And I will have to pass away as well
[16:42] <mhr3> sil2100, so you don't know?
[16:43] <sil2100> mhr3: well, ekhm, it's not in -proposed anymore
[16:44] <sil2100> mhr3: so it's hard to know what happened and why it was stuck
[16:44] <sil2100> Would have to poke didrocks
[16:44] <mhr3> sil2100, hmm, i see, but at least it's released now
[16:44] <sil2100> mhr3: (it's in main since like 30 minutes)
[16:44] <sil2100> Right
[16:45] <sil2100> Too bad it took so long though
[16:45] <sil2100> See you tomorrow!
[16:55] <mzanetti> Cimi: /me looks
[16:59] <mzanetti> Cimi: Yeah... this is because of the singleton
[17:00] <mzanetti> Cimi: If you just register it normally, QML wants to call new GSettingsControllerQml everytime you use it in QML
[17:00] <Cimi> so how do I register it?
[17:00] <mzanetti> Cimi: but it can't because of the private constructor
[17:00] <mzanetti> Cimi: I fixed it... one sec, I'l push
[17:01] <mzanetti> Cimi: pushed
[17:01] <mzanetti> Cimi: should I simplify the controller a bit?
[17:01] <Cimi> mzanetti, go for it
[17:02] <Cimi> mzanetti, the lesson is learnt :)
[17:02] <mzanetti> Cimi: ok... I'll also Qt'ify the coding style a bit
[17:05] <mzanetti> Cimi: do you really need that pimpl and that schema stuff?
[17:05] <mzanetti> Cimi: seems total overkill to me. especially for a mock
[17:06] <mzanetti> Cimi: ok. pushed the simplified singleton
[17:06] <Cimi> mzanetti, not really
[17:06] <Cimi> mzanetti, it's not needed, just to be similar to the original code
[17:06] <mzanetti> Cimi: what I did is to drop the bool and just use the instance variable to check if its created already
[17:06] <Cimi> ok
[17:07] <Cimi> I have to leave now, will continue the work later :)
[17:07] <mzanetti> Cimi: for that I initialized the s_controllerInstance with = 0 and set it back to 0 when deleting
[17:07] <Cimi> looks better!
[17:07] <Cimi> yeah smart
[17:07] <mzanetti> Cimi: ok. ping me if you have troubles continuing. I should be mostly online but EODing now... so not sure how responsive
[17:08] <Cimi> mzanetti, I'll work later in the evening/night
[17:08] <Cimi> mzanetti, have gym and dinner with friends...
[17:08] <mzanetti> have fun!
[17:08] <Cimi> but I want to see if I can fix this before tomorrow
[17:08] <Cimi> and helps me falling asleep :)
[17:14] <dednick> larsu: ping
[17:23] <jbicha> kenvandine: can you try publishing notify-osd?
[17:38] <kenvandine> jbicha, you mean in the daily release stack?
[17:39] <jbicha> yes
[17:40] <jbicha> but I don't know if the autolanding test runs the autopkgtest so this may still not be done
[17:41] <jbicha> I'll try to get someone else to do the gtk upload next time so they can enjoy the fun ;)
[17:41] <kenvandine> :)
[17:41] <kenvandine> ok, it's running now
[17:43] <jbicha> if it works, it's nice that the fix was simple but it's still been a headache
[18:16] <mzanetti> MacSlow: so... the reason is that run_loop_with_kill_switch is blocking
[18:17] <mzanetti> MacSlow: that one waits for 10 secs but blocks the autopilot test
[18:17] <mzanetti> MacSlow: when the loop is killed everything is shut down and the test fails
[18:19] <MacSlow> mzanetti, hm...
[18:19] <MacSlow> mzanetti, still seeing this "DBus.error.serviceUnknown..." never the less
[18:19] <geekette> hi
[18:19] <mzanetti> MacSlow: yeah... that one makes sense now
[18:19] <geekette> any one is here
[18:19] <MacSlow> mzanetti, but tap happens now... at least some progress
[18:20] <geekette> i have question
[18:20] <geekette> actually am new to the DE thing :-p
[18:21] <mzanetti> geekette: so you're a command line guru? :D
[18:21] <geekette> xD
[18:21] <geekette> not really
[18:22] <mzanetti> geekette: if you have a question, just ask. Don't ask for asking.
[18:22] <geekette> all i wanna ask is i wanna make my own DE but then modifying some existing DE projects will be better
[18:23] <geekette> but i cant really undrestand from where i can start
[18:23] <geekette> for example
[18:24] <geekette> i dont wanna see letters and alpha
[18:24] <geekette> i wanan convert it to how to say it
[18:24] <geekette> to pictures special caracter that i made it
[18:24] <mzanetti> MacSlow: yeah... its 100% that issue. the loop is blocking and by the time your test conrinues, all the stuff is gone already because the notification server is killed
[18:25] <geekette> did you get it sorry i cant explain more am not good in english
[18:25] <geekette> sorry
[18:25] <mzanetti> geekette: so you want a Desktop with no text, just icons?
[18:25] <MacSlow> mzanetti, even without the loop being created/touched  it fails
[18:25] <geekette> and not normal text it will be like dots
[18:25] <geekette> you know the braill
[18:25] <mzanetti> MacSlow: yeah well... in that case the notification won't appear, right?
[18:25] <geekette> *braille
[18:25] <geekette> for blind people
[18:25] <geekette> i wanan something like that
[18:26] <mzanetti> geekette: ah... I see
[18:26] <mzanetti> geekette: I guess that should be doable by switching the system to a font that does that
[18:26] <geekette> emm i thought about it
[18:27] <mzanetti> Saviq: regarding the passphrase test... we can enable it but I think/hope it'll be obsolete soon again: https://bugs.launchpad.net/unity8/+bug/1204984
[18:27] <geekette> but i wanna something more strong and more basic i was thinking abtout making something with xlib and gtk+
[18:27] <mzanetti> mterry: btw ^^
[18:28] <mterry> mzanetti, yeah, I saw that fly by, but didn't dig into it
[18:28] <mzanetti> mterry: ok... no rush. just wanted to make sure you didn't miss it
[18:29] <mzanetti> geekette: that sounds like a whole lot of work...
[18:29] <geekette> yep
[18:29] <geekette> i know that is why am here
[18:29] <geekette> to advice me
[18:29] <geekette> ;-)
[18:30] <mzanetti> geekette: probably it'll be a good idea to check out RazorQt. Thats quite minimalistic and should be a rather easy starting point to dig into how desktops work.
[18:30] <Saviq> mzanetti, yeah
[18:30] <geekette> thank you mzanetti
[18:30] <Saviq> mzanetti, it's passing reliably, so let's leave it at that - and have another test to verify the correct behaviour and fix it
[18:31] <mzanetti> geekette: not trying to get you away from ubuntu of course. but unity is probably not the easiest one to start with in terms of coding
[18:31] <mzanetti> geekette: especially if you want to "fork" and do your own thing
[18:31] <mzanetti> Saviq: fine with me
[18:32] <geekette> yes i get it am on their website thank you ..i will give you some feedback soon thank you
[18:33] <mzanetti> MacSlow: did you understand the issue or do I need to explain more?
[18:34] <MacSlow> mzanetti, I'm still missing something obviously... as the callback is not triggered... but the tap finally works (seeing the mouse move)
[18:34] <mzanetti> MacSlow: you mean your hardocded pixel move?
[18:35] <mzanetti> MacSlow: yeah... because the loop quits. then the test continues. if you have some hardcoded pixel where the mouse should move to, thats ok for autopilot and it happens
[18:36] <mzanetti> MacSlow: but if you try to tap_object(notification), that wont work as soon as the loop dies because when the notificationserver dies, the notification gets destructed too
[18:38] <mzanetti> MacSlow: so what you need is either a non blocking run_loop() method or you start your own thread where you run the loop.
[18:43] <MacSlow> mzanetti, *sigh*
[19:02] <kenvandine> jbicha, published
[19:29] <jbicha> kenvandine: thanks, it looks like it still failed the autopkgtext though
[19:30] <jbicha> can you look up the full tests/test-suite.log or can we change how the tests run so that we can see what exactly is failing?
[19:32] <kenvandine> ah, the daily release stack does run the tests for that
[19:33] <kenvandine> https://jenkins.qa.ubuntu.com/view/Saucy/view/AutoPkgTest/job/saucy-adt-notify-osd/lastCompletedBuild/testReport/
[19:33] <kenvandine> jbicha, ^^
[19:34] <kenvandine> doesn't look very helpful
[19:54] <gotwig> please remove all amazon scopes  from  the default unity setup
[19:54] <gotwig> they just suck, unity has more potential than that
[20:01] <Saviq> kgunn, I think that's all from me on the IoM doc
[20:01] <kgunn> Saviq: more than i could ask for....go have an evening, would you?
[20:01] <Saviq> kgunn, it's still early! there's stuff to do! ;)
[20:02] <Saviq> kgunn, yeah, I'll definitely actually go to sleep tonight, damn lawnmowers woke me up after 4hrs of sleep this morning...
[20:03] <Saviq> the best sleeping aid - a clean conscience, though - so there must be something here ;)