[08:39] <Saviq> Mirv, hey, can we have a silo for row 68 please?
[08:39] <Saviq> sil2100, looks like our tmp app bug is back ;|
[08:40]  * didrocks hugs Saviq
[08:40] <didrocks> Saviq: thanks for looking and not having to poke you :p
[08:41] <Saviq> didrocks, no worries, I'll have a look at the flo and manta fails, too, they look like they can be actual fails
[08:41] <didrocks> Saviq: yeah, do you think there is hope for this fail?
[08:42] <didrocks> I really wonder what magnetic magic happens on the lab…
[08:42] <Saviq> didrocks, not without having eyes on the device, or maybe being able to reproduce 100% what smoketesting does
[08:44] <Saviq> didrocks, while I have you here, can we have a silo for row 68?
[08:44] <didrocks> Saviq: it's firedrill for now, so you should wait for Mirv/sil2100
[08:44] <Saviq> didrocks, ok
[08:44] <didrocks> no image testing for the past 2 images
[08:44] <sil2100> Saviq: yeah, I already noticed it on -ci-eng... ;/
[08:44] <Mirv> Saviq: doing
[08:44] <Mirv> Saviq: just don't reping three people in 5 mins :)
[08:45] <sil2100> Saviq: I'm waiting for new test results from latest image to make sure it's reproducible this time as well
[08:45]  * Saviq wonders how didrocks does it so that both sil2100 and Mirv show up 2s after he pung them ;D
[08:45] <tsdgeos> Saviq: grid-see-more  is unfinished, shouldn't delay merging
[08:45] <Saviq> tsdgeos, hence the ?
[08:45] <tsdgeos> Saviq: looking at the history new-scopes_do_not_request_previews_on_activate was already merged it seems
[08:45] <sil2100> Saviq: that's just uh... coincidence!
[08:45] <Mirv> Saviq: landing-013
[08:45] <didrocks> Saviq: you don't need to ping us theorically btw ;)
[08:45] <Saviq> Mirv, thanks!
[08:46] <Saviq> mhr3, ↑↑
[08:46] <tsdgeos> Saviq: though it still has no tests, so you can add that to the list i guess
[08:47] <Mirv> Saviq: so how does this relate to line 13, can that be removed now?
[08:47] <Saviq> Mirv, ah indeed, yes, please
[08:47] <Saviq> Mirv, thought we had removed it already
[08:48] <Mirv> Saviq: done, thanks. ok.
[08:52] <Saviq> mhr3, tsdgeos https://code.launchpad.net/~saviq/unity8/fix-ratinginput-signal/+merge/211455 please
[08:53] <tsdgeos> Saviq: is that spec'ed somewhere?
[08:54] <Saviq> tsdgeos, http://developer.ubuntu.com/api/devel/ubuntu-14.04/cplusplus/unity-scopes/previewwidgets.html#rating-input
[08:55] <tsdgeos> Saviq: add that into the commit log?
[08:55] <Saviq> tsdgeos, k
[08:55] <Saviq> tsdgeos, done
[08:56] <tsdgeos> Saviq: need to wait for CI to pass to please the checklist
[08:56] <Saviq> tsdgeos, sure
[09:10] <tsdgeos> hmmm
[09:10] <tsdgeos> we have not fixed CardFilterGrid in clean-to-trunk ?¿
[09:11] <tsdgeos> it still references template :/
[09:11] <tsdgeos> this having two branches is a bit of a pain
[09:12] <tsdgeos> mzanetti: ping
[09:12] <mzanetti> tsdgeos: o/
[09:13] <tsdgeos> mzanetti: i can't reproduce "flick one scope to down, so that the header disappears. Then switch to another scope. The header comes back in, but not fully", can you still? Which branch was that? desktop or phone?
[09:13] <mzanetti> tsdgeos: phone. clean-to-trunk
[09:15] <tsdgeos> mzanetti: so works on desktop?
[09:15] <mzanetti> let me try....
[09:16] <mzanetti> tsdgeos: yeah, couldn't reproduce it on desktop trying a couple of times... was happening about 50% of the times on the phone
[09:16] <tsdgeos> mzanetti: next time it'd be worth mentioning i guess :)
[09:17] <mzanetti> tsdgeos: sorry...
[09:25] <tsdgeos> mzanetti: which preview did you open that gave you white text header?
[09:25] <mzanetti> tsdgeos: all of them
[09:25] <tsdgeos> mzanetti: again phone only?
[09:26] <mzanetti> I didn't really test extensively on the desktop, so yes, all those issues are on the phone
[09:26] <mzanetti> but let me check
[09:26] <mzanetti> tsdgeos: desktop too: http://i.imgur.com/5trCs3W.jpg
[09:27] <tsdgeos> i guess we have a different definition of what white means
[09:27] <tsdgeos> what is white there?
[09:27] <mzanetti> below the screenshots
[09:27] <mzanetti> I guess you're looking at the wrong header
[09:27] <mzanetti> the preview cnotent header
[09:27] <mzanetti> the thing with the rating stars
[09:28] <tsdgeos> ah
[09:28] <tsdgeos> that
[09:28] <tsdgeos> i don't see it
[09:28] <tsdgeos> needed to twist my lcd to see stuff there
[09:28] <sil2100> mhr3: ping!
[09:28] <tsdgeos> mzanetti: yeah E_TOO_MANY_HEADERS :D
[09:29] <mzanetti> tsdgeos: so yeah, its not white, but whitish
[09:29] <mzanetti> on a somewhat whitish background too :)
[09:29] <mzanetti> :)
[09:31] <MacSlow> got to ask again... is anybody getting these QWARN-messages "QQmlComponent: Created graphical object was not placed in the graphics scene." when running the unity8-qmltests?
[09:32] <mzanetti> MacSlow: I've seen that with your tests only
[09:32] <mzanetti> MacSlow: probably related to that workaround where you specify an y value and have a comment "position it before the Column takes over"
[09:32] <mzanetti> that's something I've never seen before. usually columns work immediately
[09:33] <mzanetti> so you probably do something weird which triggers that warning and requires that y: workaround
[09:33] <MacSlow> mzanetti, only started to show up after the switch to Qt 5.2
[09:33] <mzanetti> might well be that 5.0 just didn't warn you about it...
[09:35] <MacSlow> mzanetti, hm... also happens when I comment out that line
[09:37] <mzanetti> MacSlow: I didn't say its caused by that
[09:37] <mzanetti> I said its likely that this y: workaround wouldn't be required once you get rid of the reason why this warning appears
[09:37] <mzanetti> MacSlow: but still only a guess. but I would start searching from there
[09:38] <MacSlow> mzanetti, with the ComboButton I hope this will all become a thing of the past
[09:38] <MacSlow> and I've still the wait() to get rid of
[09:41] <tsdgeos> mzanetti: the "\n" issue you discussed with Saviq, right?
[09:41] <Saviq> tsdgeos, mzanetti, yeah, they got a fix in a silo already
[09:41] <tsdgeos> ok, so i can ignore it in unity8's side
[09:43] <Saviq> seb128, pfft! you and your bugs that I file duplicates of ;P
[09:43] <seb128> Saviq, that's right!
[09:43] <seb128> ;-)
[09:49] <MacSlow> mzanetti, got rid of the wait()... but now something the height-asserting fails
[09:49] <mzanetti> tsdgeos: yep, ignore that. and depending on what of the header will be rewritten soon, you might want to ignore some of the header issues too. You decide which ones as I don't really know what is here to stay and what isn't
[09:51] <tsdgeos> mzanetti: i'm installing it into the phone to see if i can reproduce them
[09:52] <tsdgeos> Saviq: https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/5917/console ?¿
[09:58] <Saviq> tsdgeos, that's the split greeter
[09:58] <mzanetti> MacSlow: so you still need help on the second wait()?
[09:59] <Saviq> tsdgeos, I'll check it out in a bit
[09:59] <MacSlow> mzanetti, no... I got rid of that... but now I've to chase the initialHeight assertion (to verify the notification correctly collapsed)
[10:00] <mzanetti> yeah. so the question if you think you can manage that or if I should help you with that
[10:00]  * MacSlow loves to redo tests, which worked before
[10:00] <MacSlow> mzanetti, I still try own my own for the next half hour...
[10:00] <mzanetti> ok
[10:01] <MacSlow> and the fact that roughky 8 out of 10 runs sefault makes me really look forward to the attempt to get it through jenkins again
[10:16] <Saviq> tsdgeos, huh... it's a good question where did unity8-greeter come from in there??
[10:17] <Saviq> tsdgeos, it should only be there in the split branch, somehow the device got "infected" with that or something
[10:17] <tsdgeos> ^_^
[10:17] <tsdgeos> mzanetti: i can't reproduce the dash header getting lost at middle height
[10:17] <tsdgeos> on the phone either
[10:18] <tsdgeos> nor the fact that you say the text is missing
[10:18] <tsdgeos> :/
[10:18] <mzanetti> hmm... really... the missing text was 100% reproducable here
[10:18] <mzanetti> tsdgeos: once directly after startup the header text is missing here
[10:18] <tsdgeos> mzanetti: are you up to date?
[10:19] <mzanetti> tsdgeos: and the other is that it disappears when installing an app
[10:19] <tsdgeos> mzanetti: ah found how how to reproduce the header thing i think
[10:19] <mzanetti> I did a fresh flash yesterday and installed that branch, yes
[10:19] <tsdgeos> ok, no
[10:19] <tsdgeos> happened once but not again :/
[10:19] <mzanetti> which one? the getting stuck half way when coming in from the top?
[10:19] <tsdgeos> yeah
[10:20] <mzanetti> yeah, that's the trickies one, but it did happen regularly here. let me try again
[10:21] <mhr3> Saviq, can we please land the scope-tool changes with the transition to new-scopes?
[10:22] <Saviq> mhr3, sure, probably
[10:22] <Saviq> /we need reviews
[10:22] <mhr3> all our docs assume those already
[10:28]  * Saviq reboots
[10:31] <Saviq> greyback, how's the head?
[10:32] <mzanetti> tsdgeos: so. I've started it up, didn't touch the display except swiping away the greeter yet => header text is missing
[10:32] <tsdgeos> mzanetti: can't reproduce at all, are you sure your're up to date?
[10:32] <mzanetti> I just pulled that branch again, yes
[10:33] <greyback> Saviq: it's fine, thanks for asking? You wouldn't be falling into the assumption that all Irish people get plastered on March 17, would you?
[10:33] <tsdgeos> mzanetti: and do you have icons? i.e. is only the header missing?
[10:33] <Saviq> greyback, NEVER
[10:33] <greyback> :)
[10:33] <mhr3> tsdgeos, for me the "Apps" header text is centered after swiping away greeter
[10:33] <mhr3> which is... weird
[10:33] <mzanetti> tsdgeos: yes, I had to restart scope-registry and and proxy but yes, I have icons
[10:34] <mzanetti> just the header is missing
[10:34] <tsdgeos> can't reproduce neither yours not mhr3
[10:34] <tsdgeos> but anyway that's the tabbar in the sdk
[10:34] <tsdgeos> which is an amazing pain
[10:35] <tsdgeos> and is going away somewhere in the future
[10:35] <Saviq> and is going away
[10:35] <tsdgeos> i've made like 10 workarounds to get it to work
[10:35] <tsdgeos> and none of them seem to be reliable
[10:35]  * greyback popping to shops
[10:36] <mzanetti> tsdgeos: so yeah. I'm not diapproving it because of that... altough the missing text after app installation is more critical imo
[10:36] <mhr3> tsdgeos, one more that i just noticed, when closing last app from the running apps category, i need to click the close button twice and at that point it closes and reopens the app at the same time
[10:37] <mhr3> and yea, preview header widget has white text
[10:37] <tsdgeos> mhr3: not anymore
[10:37] <mhr3> tsdgeos, it does in 013
[10:37] <tsdgeos> i have no clue what 013 is
[10:38] <mhr3> silo with new-scopes
[10:38] <tsdgeos> well, update the silo then
[10:38] <mhr3> i'll leave that to Saviq
[10:39] <Saviq> mhr3, tsdgeos is building now
[10:39] <Saviq> mhr3, with the extend scope tool change
[10:39] <mhr3> Saviq, cheers!
[10:39] <mzanetti> tsdgeos: in any case - here's how I reproduce it: make sure apps and amazon scopes are at the top. go to amazon scope, flick it to the bottom. go back to home scope.
[10:40] <Saviq> mzanetti, there is no home scope, you mean apps?
[10:40] <mzanetti> yeah, apps, sorry
[10:41] <mzanetti> anyways, I only had it like once that it was really badly misplaced. mostly its just a few pixels off...
[10:41] <tsdgeos> can't see it wrong
[10:42] <mzanetti> tsdgeos: ok... can you reproduce the missing header after app isntallation?
[10:42] <tsdgeos> haven't tried
[10:42] <mzanetti> thats happening 100% here
[10:42]  * Saviq just noticed that double-clicking on a hunk in QtCreator's diff viewer takes you to that hunk in the file, AWESOMES
[10:43] <mzanetti> heh, yeah
[10:43] <Saviq> mhr3, what's the FIXME about in the scope tool?
[10:44] <mhr3> Saviq, need a way to discover the endpoint for the ss-proxy
[10:44] <mhr3> not possible atm
[10:44] <Saviq> mhr3, also, 120 columns please (yes, I'm trying to bring astyle over from -api)
[10:46] <mzanetti> noooooo :D
[10:46] <Saviq> mzanetti, yes, it's really easier to not have to discuss that *ever*
[10:46] <Saviq> or well, only once ;P
[10:46] <mzanetti> :D
[10:46] <mzanetti> jk
[10:46] <Saviq> mhr3, also... you're mixing camelCase with unders_score
[10:47] <mhr3> Saviq, :/ qt uses one, scopes the other... it's confusing
[10:48] <tsdgeos> mzanetti: i'm not saying there's no weirdness in the header, i mean you can see it, otoh the header handling has not really changed, it was probably just there
[10:48] <Saviq> mhr3, indeed
[10:48] <tsdgeos> let's see app install
[10:48] <Saviq> mhr3, I won't block, but I'd gladly see it fixed sometime
[10:48] <tsdgeos> hmmm
[10:48] <tsdgeos> i don't have more suggestions in my apps scope
[10:48] <tsdgeos> should i have that?
[10:49] <mzanetti> yes
[10:49] <mzanetti> at least I do
[10:49] <tsdgeos> only my apps and available here
[10:49] <mzanetti> tsdgeos: network connected?
[10:49] <tsdgeos> well i have stuff in the amazon scope
[10:49] <tsdgeos> so i guess yess
[10:49] <mzanetti> hmm...
[10:49] <mhr3> Saviq, so got the astyle script that i can just run on the src?
[10:50] <Saviq> mhr3, not yet, no, the astyle + clang=format combination you guys have in -api breaks a few things in Qt style, so I have to sit down and flesh it out
[10:51] <Saviq> mhr3, I can get you a branch
[10:51] <mhr3> Saviq, if you're fine with doing it later, i am too :)
[10:51] <Saviq> mhr3, yeah
[10:51] <davidcalle> mhr3, didrocks :  summary : the old photos lens from unity7 needs a bug fix and it 1) doesn't use CI train 2) doesn't have a testing infrastructure. What's the best course of action to release that bug fix?
[10:52] <didrocks> davidcalle: they have to get a lander for CI Train, the previous infra isn't supported anymore
[10:52] <didrocks> I think given who worked on the lens, it should be either mhr3 or thorst and I already raised that some weeks ago…
[10:52] <didrocks> when I saw there was no lander
[10:53] <tsdgeos> mzanetti: ouch it crashed on the app isntall thing _S
[10:53] <mzanetti> :/
[10:53] <mzanetti> haven't seen that here
[10:54] <mhr3> didrocks, so let's say i'm the lander, what's next?
[10:55] <tsdgeos> ASSERT: "!isEmpty()" in file /usr/include/qt5/QtCore/qlist.h, line 287
[10:55] <mhr3> didrocks, how do we deal with the lack of test infra?
[10:55] <didrocks> mhr3: just use a silo, as always
[10:55] <didrocks> to land the .debs
[10:56] <mhr3> davidcalle, mind writing a test plan for those scopes?
[10:56] <mhr3> davidcalle, basically bunch of manual testing i guess
[10:57] <davidcalle> mhr3, yeah, they have that already, but you probably mean the new testplanwiki pages
[10:58] <mhr3> davidcalle, right
[10:58] <mhr3> link?
[10:58] <mhr3> tsdgeos, there's something weird with the apps scope page, it allows me to scroll way past beyond the content
[10:59] <mhr3> tsdgeos, i have a feeling it has something to do with the running apps renderer
[10:59] <mhr3> mostly cause the other scopes work fine
[10:59] <davidcalle> mhr3, ok, let me see how they are done and I'll create one for it. The mp is https://code.launchpad.net/~davidc3/unity-lens-photos/fb-deleted-photos-previews/+merge/211481 and the old test plan is in tests/manual
[11:00] <tsdgeos> mhr3: can you reproduce the problem all the time?
[11:00] <tsdgeos> it works fine here :/
[11:00] <mhr3> tsdgeos, no :/
[11:00]  * tsdgeos is getting tired of that it works for him
[11:00] <mhr3> tsdgeos, try installing the 013 silo on the device
[11:01] <mhr3> so we at least run the same thing for sure
[11:02] <tsdgeos> mhr3: i'm fixing another issue at eht moment
[11:02] <tsdgeos> i just commented on yours because i happened to pass by there
[11:04] <Saviq> /food
[11:06] <mzanetti> tsdgeos: got the crash now too. if you click "open" after installing an app, right?
[11:06] <tsdgeos> mzanetti: no, i press back in the preview header
[11:07] <mzanetti> ok... that didn't crash here, but left the header empty. clicking "open" crashed unity
[11:08] <mzanetti> rofl... the stallboard app
[11:09] <tsdgeos> and i'm back into the debugging hell loop in the phone
[11:09] <tsdgeos> sigh
[11:10] <tsdgeos> reboot, debug, mir won't start, debug, mir won't start, reboot, rinse and repeat
[11:12] <mzanetti> is it intentional that new scopes previews for apps don't show the publisher of the app any more?
[11:12] <tsdgeos> awesome, and startin scopes doesn't refresh the dash
[11:12] <tsdgeos> so another reboot loop!
[11:16] <tsdgeos> and another mir crash!
[11:19] <tsdgeos> and another mir crash!
[11:20] <tsdgeos> i forgot how happy i was this week because i didn't have to debug anything in the phone
[11:23] <tsdgeos> mzanetti: you'll have to ask the scopes people about that, we only show what we're asked to via json
[11:26] <mhr3> mzanetti, looks like bug in click scope, it's providing the description instead of the publisher
[11:27] <mhr3> alecu, ^
[11:27] <mzanetti> ok.
[11:27] <mzanetti> mhr3: I also noticed, that the apps I've installed today appeared in the "installed apps" section, but did *not* disappear in the "more suggestion" sections
[11:31] <mzanetti> mhr3: more details: I've installed stallboard using the new-scopes branch. I have that app now in both sections. If I restart the released unity8, its only in the installed apps section. then if I restart the new-scopes unity again, that app shows up in both sections again
[11:32] <mzanetti> so seems the new-scopes backend stuff doesn't seem to filter properly
[11:32] <mhr3> mzanetti, mind opening a bug about it?
[11:32] <mzanetti> ack
[11:32] <mhr3> mzanetti, afaict it's supposed to be working
[12:02] <Cimi> dednick, in case you missed it https://code.launchpad.net/~nick-dedekind/unity8/lp1283191/+merge/211307
[12:12] <MacSlow> tsdgeos, can you look over https://code.launchpad.net/~macslow/unity-notifications/multiple-snap-decision-example/+merge/210638 when you've  a free moment... it's a example to better show off the visual snap-decision queue feature.
[12:14] <MacSlow> tsdgeos, it's the script driving this https://www.youtube.com/watch?v=LVryb_-gRR0
[12:25] <Cimi> dednick, I've been picky, but it's one line https://code.launchpad.net/~nick-dedekind/unity8/lp1283191/+merge/211307/comments/499052
[12:26] <tsdgeos> Saviq: mzanetti: so the crash we're seeing going back from previews is not new-scopes related, it's just a LVWPH bug, do you want me to fix it in new-scopes-clean-to-trunk or as a separate request for trunk?
[12:28] <Saviq> tsdgeos, separate, we'll land it together
[12:29] <mzanetti> works for me
[12:30] <tsdgeos> ok
[12:48] <mzanetti> Saviq: https://code.launchpad.net/~mzanetti/unity8/make-xvfbtestSomething/+merge/211522
[12:48] <mhr3> Saviq, when i unlock  the greeter the first time, the dash is kinda stuck
[12:48] <mhr3> known?
[12:49] <mzanetti> Saviq: it works fine but has one side effect I'm unsure of: it adds the xvfbtestTarget also to qmluitests. Do we want that or not?
[12:53] <Saviq> mzanetti, why the -l? also, please use long options in scrips
[12:53] <Saviq> t
[12:54] <Saviq> mzanetti, and well, I'd say we should make qmltests only run in xvfb
[12:54] <Saviq> mzanetti, with an option to run them outside manually
[12:54] <mzanetti> Saviq: ok
[12:55] <Saviq> mzanetti, there's a but
[12:55] <mzanetti> Saviq: so: make qmluitests runs xvfb, make testSomething runs normal, make xvfbtestSomething runs xvfb. ok?
[12:56] <tsdgeos> i know how to fix the bug and what causes it but can't reproduce it in our testcases
[12:56] <tsdgeos> grrrr
[12:56] <tsdgeos> lunch
[12:56]  * mzanetti is waiting for the but
[12:56] <Saviq> mzanetti, I'm starting to think we should start a python helper that will do those things for us, instead of relying on cmake targets
[12:56] <Saviq> mzanetti, especially since we want to record the runs and store them on failure
[12:57] <mzanetti> Saviq: ok, but that sounds like a bigger task
[12:57] <Saviq> mzanetti, but yeah, that'd be a first step
[12:57] <Saviq> (that == the xvfbTest)
[12:57] <mzanetti> ack
[12:57] <Saviq> mzanetti, also, add --auto-servernum
[12:57] <Saviq> mzanetti, we might be able to parallelize the tests in that case
[12:57] <mzanetti> nice one :D
[12:58] <Saviq> mzanetti, any reason for the -l, though?
[12:58] <mzanetti> no... sometimes it says here "failed to start xvfb" when I start it manually
[12:59] <mzanetti> and dunno why, but I thought the -l fixed that
[12:59] <mzanetti> removed it, still works
[12:59] <Saviq> mzanetti, I doubt it would
[12:59] <mzanetti> ah... now I remember
[12:59] <mzanetti> I was fiddling around with getting some visible output from it
[12:59] <Saviq> mzanetti, it probably used the same server num, and --auto-servernum makes that go away
[12:59] <mzanetti> and that's what was left in my bash history
[12:59] <Saviq> mzanetti, right
[12:59] <mzanetti> its gone now
[13:00] <Saviq> mzanetti, didn't you try xwud? no worky?
[13:00] <Saviq> http://stackoverflow.com/questions/12050021/how-to-make-xvfb-display-visible
[13:01] <mzanetti> Saviq: nope. that only takes screenshots.
[13:01] <mzanetti> no video
[13:01] <mzanetti> unless I missed something
[13:01] <mzanetti> and before doing an endless loop of screenshot I tried to get the video otherwise, but gave up eventually
[13:03] <Saviq> mzanetti, there's one more thing
[13:03] <Saviq> LD_PRELOAD=/usr/lib/x86_64-linux-gnu/mesa/libGL.so.1
[13:03] <Saviq> this will be needed on non-free drivers (at least nvidia)
[13:03] <Saviq> mzanetti, and you need to add some dependencies to ./build, at least, and to qmluitests jenkins job
[13:04] <Prf_Jakob> tjaalton: might be better to talk about bug 1293384 here :)
[13:05] <mzanetti> Saviq: do I need to figure if its nvidia & amd64 or can I unconditionally set the LD_PRELOAD? afaics it doesn't break anything even if not needed
[13:05] <Saviq> mzanetti, yeah, should be fine everywhere
[13:08]  * greyback joining party late, but yay for qt5.2 being released!
[13:10] <mzanetti> greyback: :D
[13:10] <tjaalton> so the bug above suggests that the default session is harder on the cpu with software rendering, but why?
[13:11] <tjaalton> and could that be improved
[13:11] <Cimi> this is a simple branch, why no one approved? https://code.launchpad.net/~bfiller/unity8/fix-password-predictive-text/+merge/210676
[13:11] <Saviq> Cimi, read the bug
[13:12] <Cimi> Saviq, you're right, SDK
[13:12] <Saviq> Cimi, so well, we should merge this anyway with a TODO mentioning the bug
[13:12] <Saviq> Cimi, can you take care of that with bfiller?
[13:12] <Cimi> sure
[13:15] <Saviq> mterry, hey, any idea how unity8-greeter could show up on the -ci mako testrunners? i.e. http://s-jenkins.ubuntu-ci:8080/job/generic-mediumtests-runner-mako/5924/console
[13:15] <mterry> ??
[13:15] <tjaalton> Prf_Jakob: crickets ;)
[13:16] <Prf_Jakob> tjaalton: dang :(
[13:16] <Prf_Jakob> :)
[13:16] <Saviq> mterry, it's like the device wasn't cleaned or something
[13:16] <mterry> Saviq, that is using my split branch for sure, has unity8-greeter and uses version 7.85
[13:17] <Saviq> mterry, yeah, but it's a totally unrelated MP...
[13:17] <mterry> Saviq, I'm assuming that job isn't attached to my split
[13:17] <Saviq> mterry, nope
[13:17] <mterry> Saviq, yeah, sounds like poor cleaning
[13:17] <Saviq> mterry, yeah, what I thought, will pester fginther
[13:44] <Saviq> mhr3, you and your broadcom
[13:45] <mhr3> Saviq, kill it!
[13:45] <mhr3> kill it with chemical fire!
[13:45] <Saviq> mhr3, you asked about some stuck dash on unlock, what was that about?
[13:45] <mhr3> Saviq, right, happened a couple of times, when i start the phone and unlock the animation doesn't finish and it's kinda stuck
[13:46] <mhr3> i have to do right/left swipe to unstuck it
[13:47] <mhr3> to unstick it?
[13:49] <mhr3> Saviq, but can't rep reliably :/
[13:51] <Saviq> mhr3, wash your finger
[13:58] <Saviq> mterry, no, I'm afraid there's no spec for the swipe in greeter, you'll just have to ping John and dig from there
[13:58] <Saviq> mterry, maybe simply a +comment on the design doc would be enough
[13:59] <mterry> Saviq, don't have permissions.  will poke
[14:01] <Saviq> mterry, yeah, me neither
[14:02] <mterry> JohnLea, hello!  Did design want the greeter infographics to switch to a different "fact" on a swipe instead of a double tap?
[14:13] <MacSlow> What would cause a qmltest to not get executed? I get a 0 file-length for the visual-queue and notification-renderer tests.
[14:13] <MacSlow> Sofar I've not found any hint in the console-output
[14:22] <mzanetti> MacSlow: what file?
[14:22] <Saviq> MacSlow, wrong syntax usually
[14:22] <Saviq> MacSlow, check the console log
[14:24] <mzanetti> Saviq: are we ok with this one? https://code.launchpad.net/~mzanetti/unity8/make-single-testfunction/+merge/205035
[14:25] <Saviq> mzanetti, looks OK, I never tested that it works, though
[14:25] <MacSlow> mzanetti, Saviq: https://jenkins.qa.ubuntu.com/job/unity-phablet-qmluitests-trusty/1404
[14:25] <Saviq> mzanetti, btw, didn't you find out that ninja wants $$ or something and that it's still possible to make it happen?
[14:26] <Saviq> MacSlow, Segmentation fault
[14:26] <mzanetti> Saviq: yeah, ninja wants $$ to not bail out on it, however, it passes both $ on to the makefile wich causes make to fail on it
[14:26] <Saviq> mzanetti, mhm
[14:27] <mzanetti> I asked Satoris. He said only way would be env variable. but I think that would require our own qmltestrunner edition
[14:27] <Saviq> MacSlow, and the other:
[14:27] <Saviq> '/tmp/buildd/unity8-7.84+14.04.20140317.2bzr553pkg0trusty1404/tests/qmltests/Notifications/tst_VisualQueue.qml' does not exist under '/tmp/buildd/unity8-7.84+14.04.20140317.2bzr553pkg0trusty1404/obj-x86_64-linux-gnu/tests/qmltests'.
[14:28] <MacSlow> Saviq, *sigh* ... I renamed it... totally forgot about the CMakefile...
[14:28] <Saviq> elopio, hey, do we have an established way to check what device (as in mako vs. manta vs. flo) the tests are run? we need to skip the bluetooth indicator test on manta, as it's not available there
[14:31] <elopio> Saviq: autopilot.platform.model() should return a string that identifies manata.
[14:31] <elopio> *manta.
[14:31] <Saviq> elopio, oh good
[14:31] <Saviq> Cimi, standup
[14:32] <Cimi> ouch!!!
[14:32] <elopio> Saviq: a better way to do it would be to check if the bluetooth service is available, but if it works with model(), we can use that.
[14:32] <elopio> Saviq: I'll skip the stand up again, I'm sorry.
[14:34] <Saviq> elopio, k, the thing is the bluetooth service is available afaict
[14:35] <Saviq> elopio, yeah, but it just doesn't show the indicator, 'cause there's no hardware
[14:36] <Cimi> dednick, https://code.launchpad.net/~nick-dedekind/unity8/lp1283191/+merge/211307
[14:38] <Saviq> elopio, can I check which scenario is run, or should I just make the condition based on the dict() from the scenario?
[14:39] <elopio> Saviq: not sure what you mean. I think the scenarios are not involved here, you should just add a self.skipTest statement at the start of the test
[14:39] <elopio> but I might be missing something.
[14:39] <Saviq> elopio, http://bazaar.launchpad.net/~unity-team/unity8/trunk/view/head:/tests/autopilot/unity8/indicators/tests/test_indicators.py
[14:40] <elopio> oh, I see.
[14:40] <Cimi> mzanetti, https://code.launchpad.net/~mzanetti/unity8/make-xvfbtestSomething/+merge/211522
[14:40] <Saviq> elopio, so I can check for self.indicator_name, if the name of the scenario is not available
[14:41] <elopio> Saviq: yes, that will work.
[14:42] <elopio> Saviq: in the future it might get ugly, so a cool improvement for when we have time is to add a parameter with the names of the platforms where each test should run.
[14:42] <Saviq> elopio, I think we had something like this initially ;)
[14:43] <Saviq> elopio, but then thomi and veebers came in and ripped it all apart :D
[14:43] <elopio> off with their head.
[14:44] <mzanetti> Cimi: can you elaborate on that comment?
[14:44] <Cimi> mzanetti, not sure we have X on a nexus 4
[14:44] <Cimi> mzanetti, so why installing xvfb?
[14:44] <Cimi> mzanetti, when we do ./build inside adb
[14:45] <mhr3> Saviq, oh btw you're resetting the searchString i set :P
[14:45] <Cimi> mzanetti, I'd leave this to desktop only
[14:45] <Saviq> mhr3, oops
[14:46] <Saviq> mhr3, GenericScopeView.qml:onIsCurrentChanged
[14:46] <Saviq> mhr3, if (!isCurrent) should help there
[14:46] <Saviq> mhr3, assuming you're talking about canned queries
[14:47] <mhr3> i am indeed
[14:48] <mhr3> Saviq, hm, actually isn't it the timer?
[14:48] <mhr3> Saviq, GenericScopeView.qml:63
[14:49] <Saviq> mhr3, that could be the case for non-favourite ones, yeah
[14:49] <Saviq> mhr3, and well, that timer should probably go away anyway
[14:51] <mhr3> Saviq, right, i was testing with non-favourited only
[14:51] <mzanetti> Cimi: https://code.launchpad.net/~mzanetti/unity8/make-xvfbtestSomething/+merge/211522/comments/499151
[14:51] <mzanetti> Cimi: besides, you can install xvfb on the device just fine if you want
[14:52] <mzanetti> which I guess in the long run is what we're targeting, running qmltests on our phones with xvfb too
[14:52] <Cimi> mzanetti, why xvfb?
[14:52] <Cimi> mzanetti, we have mir, not x
[14:52] <mzanetti> so what?
[14:52] <Cimi> xvfb isn't xorg stuff?
[14:53] <mzanetti> once we have a mirvfb I'm happy to use that one
[14:53] <Saviq> Cimi, it's virtual
[14:53] <Saviq> Cimi, it doesn't care
[14:53] <Saviq> mzanetti, that's planned, too, btw
[14:53] <mzanetti> yeah, makes definitely sense for the full package
[14:53] <mzanetti> probably not highest prio tho I guess
[14:53] <Cimi> Saviq, what you mean virtual? isn't lke using xorg nested or such?
[14:53] <mzanetti> Cimi: no
[14:53] <Cimi> I don't know, asking
[14:54] <mzanetti> its emulating x - sort of
[14:54] <mzanetti> Cimi: well, it does use the X libraries, but not ends up being sent to the hardware
[14:55] <Cimi> do we have xserver-common on the phone?
[14:55] <mzanetti> yes
[14:55] <Cimi> ok
[14:55] <mzanetti> its ubuntu :) the full thing. you can also install full X11 if you want. it will complain it can't find hardware drivers tho
[14:55] <mzanetti> Cimi: ^
[14:56] <Saviq> Cimi, xvfb != xephyr
[14:56] <Cimi> good guys
[14:56] <Cimi> I shut up :)
[14:57] <mzanetti> :)
[14:57] <Saviq> mzanetti, oh, I crashed xbmcremote, did you rebuild for 5.2?
[14:57] <Saviq> (or have to)
[14:58] <Saviq> mzanetti, crashes reliably on long-press on album
[14:58] <mzanetti> Saviq: yeah, I guess I have to :/
[14:58] <Saviq> mzanetti, yeah, you probably do
[14:58] <Saviq> due to ABI breakage
[14:58] <mzanetti> I noticed some weirdness too but didn't get to it yet. was first updating the ones that didn't even start any more
[14:59] <mzanetti> hey guys: here's a review of new scopes + right edge already: http://www.v3.co.uk/v3-uk/v3-co-uk-labs-blog/2333221/ubuntu-mobile-hands-on-review
[15:18] <tsdgeos> mzanetti: created a MR for the back crash
[15:19] <tsdgeos> now i only need to find out why the empty hader
[15:19] <tsdgeos> +e
[15:21] <mzanetti> tsdgeos: awesome. eating right now. will review in a bit
[15:23] <tsdgeos> mzanetti: dude but it's almost dinner time ^_^
[15:23] <mzanetti> yeah, eatin linner :P
[15:24] <tsdgeos> linner rocks, much better than brunch
[15:25] <mzanetti> :D
[15:25] <tsdgeos> mzanetti: hmmm, i can't reproduce the missing header anymore, wonder if the same fix for the crash also fixed it, please have a nice linner and see if you can still get it after merging the MR into clean-to-trunk
[15:25] <mzanetti> ack
[15:26] <Saviq> lol, just got an email "You hurt my car on the road. Look at these photos in the attached archive and contact me as soon as possible. Otherwise you'll get legal action.", with a US phone number
[15:27] <Saviq> with a image-as-screensaver attachment, too ;)
[15:27] <tsdgeos> he's a tourist :)
[15:28] <mzanetti> the best I ever got was someone pretending to be a hatched man and saying that he would not hurt me if I pay more than the guy who hired him
[15:29] <Saviq> tsdgeos, kicked the silo build with your lvwph fix branch
[15:29] <tsdgeos> Saviq: oki
[15:30] <Saviq> mzanetti, "hatched man"?
[15:30] <mzanetti> meh... I knew dict.leo.org fooled me on that one
[15:30] <mzanetti> a hired killer
[15:30] <mzanetti> not sure what's the correct term
[15:30] <Saviq> should've been hatchet man, if anything
[15:30] <Saviq> hitman
[15:30] <Saviq> is probably the most common one
[15:31] <Saviq> mzanetti, "hatched man" would be one that came out of an egg :)
[15:31] <mzanetti> :D
[15:31] <mzanetti> see. that's why I asked you to proof read the website
[15:32] <Saviq> yeah, sorry didn't get to it yet...
[15:32] <mzanetti> np. not really in a hurr
[15:32] <mzanetti> y
[15:32] <JohnLea> mterry; hyia, in answer to your earlier questions, not atm, lets leave it as double tap for now (swipe has possible conflicts with the unlock gesture)
[15:32] <Saviq> had a tough weekend, friends came over with wedding invitations (and 1.5l of tequila, not like that mattered)
[15:32] <mterry> JohnLea, ok
[15:32] <Saviq> JohnLea, ack
[15:33] <tsdgeos> hatchet man would be scarier
[15:33] <tsdgeos> Saviq: which silo number is it btw?
[15:38] <Saviq> tsdgeos, 13
[15:39] <Saviq> ppa:ci-train-ppa-service/landing-013
[15:42] <sil2100> Saviq: hi! Do you know anything about the unity8 test running migration to click? :)
[15:44] <Saviq> sil2100, no
[16:03] <dednick> Saviq: are tabs staying for indicators?
[16:04] <mzanetti> tsdgeos: is it this one? https://code.launchpad.net/~aacid/unity8/lvwph_big_item_resize/+merge/211554
[16:04] <Saviq> dednick, I don't think so
[16:04] <tsdgeos> mzanetti: yes
[16:04] <Saviq> dednick, but there's no one told me this or otherwise
[16:05] <Saviq> mzanetti, it's building in the silo already
[16:05] <mzanetti> ah, cool
[16:05] <dednick> Saviq: ok. thanks
[16:05] <Saviq> mzanetti, actually should be built already
[16:06] <mzanetti> 13, right?
[16:06] <Saviq> mzanetti, yes
[16:10] <Saviq> elopio, hey, do you have a few minutes?
[16:11] <elopio> Saviq: in a meeting now. What about in 30 minutes?
[16:11] <Saviq> elopio, sure, please ping
[16:12] <Saviq> mzanetti, non-ui qmltests still run under xvfb, that expected?
[16:12] <Saviq> mzanetti, I don't think that should be the case, it's running under minimal QPA for a reason
[16:16] <mzanetti> Saviq: only when calling make qmltests I think. make test should still be QPA
[16:17] <Saviq> mzanetti, fails in jenkins
[16:17] <Saviq> mzanetti, complaining about xvfb-run
[16:17] <mzanetti> yeah, started to update the job but had to wait to see this debug output, will continue in a bit
[16:17] <mzanetti> Saviq: just merged the single-test-function as a prereq. its so awesome :)
[16:17] <mzanetti> calling: make xvfbtestShell FUNCTION="Shell::test_background"
[16:18] <Saviq> mzanetti, now imagine when we put those into autopkgtest, we'll be able to get rid of the qmluitests job from -ci and do it properly :)
[16:18] <Saviq> mzanetti, and you'll actually be able to run them locally without killing your machine in the mean time ;)
[16:18] <mzanetti> yeah
[16:19] <mzanetti> actually I started to clone the nromal ci job and added the hook in there
[16:19] <Saviq> mzanetti, yeah, that'll be the temporary thing, until -ci can do autopkgtests
[16:19] <Saviq> mzanetti, but http://packaging.ubuntu.com/html/auto-pkg-test.html
[16:19] <mzanetti> yep, but one step closer already
[16:20] <Saviq> mzanetti, add to that recordmydesktop, and we'll be golden
[16:20] <Saviq> mzanetti, we need to wrap it all up into a shared package, too, so that others can benefit, not only unity8
[16:20] <Saviq> mzanetti, actually SDK, probably
[16:22] <mzanetti> Saviq: and then we'd need to get rid of AP tests in unity and create the one small-but-good system test suite that will walk through every silo
[16:23] <Saviq> mzanetti, not sure what you mean
[16:23] <mzanetti> just dreaming
[16:23] <Saviq> mzanetti, we'll still need some AP tests that test the AP helpers that we maintain, but that's it
[16:23] <mzanetti> yeah. true
[16:23] <Saviq> mzanetti, integration tests will go away from the unity8 suite
[16:24] <mzanetti> oh is that planned?
[16:24] <Saviq> sure
[16:24] <mzanetti> nice
[16:26] <mzanetti> Saviq: resubmitted: https://code.launchpad.net/~mzanetti/unity8/make-xvfbtestSomething/+merge/211573
[16:26] <Saviq> mzanetti, saw that
[16:27] <Saviq> mzanetti, but do make sure non-ui qmltests don't run under xvfb, because they definitely do now
[16:27] <mzanetti> they do with make qmluitests, just like before they ran with the x11 qpa
[16:28] <mzanetti> they only run with minimal in make test, which is what dpkg executes
[16:28] <mzanetti> unless I'm mistaken, but so far that's what I figured
[16:28] <tsdgeos> mzanetti: any luck reproducing the missing header when going back after an install?
[16:29] <mzanetti> tsdgeos: will test in a minute... stuff piled up during linner :D
[16:30] <Saviq> mzanetti, just look at jenkins builds ;) they all fail due to xvfb-run missing
[16:34] <mzanetti> tsdgeos: ok. just booted the device after installing silo 13. The apps header is not empty any more, but horizontally centered :D
[16:34] <mzanetti> looks funny
[16:34] <tsdgeos> mzanetti: that's what mhr3 was getting
[16:34] <tsdgeos> didn't touch that at all
[16:37] <Saviq> mzanetti, tsdgeos, that's broken tabbar header again
[16:37] <tsdgeos> so
[16:38] <tsdgeos> i get nothing in my dash with silo13
[16:38] <tsdgeos> is that known?
[16:38] <mzanetti> worksforme
[16:38]  * Saviq dupgrades
[16:40] <mzanetti> what tells which scopes are visible? sometimes I have the amazon scope, sometimes I don't
[16:40] <mzanetti> right now there's only scopes apps music video
[16:40] <Saviq> we've been getting http://ci.ubuntu.com/smokeng/ubuntu/touch/manta/243:20140318:20140304/7224/unity8/909181/ more and more these days
[16:40] <Saviq> mzanetti, UNITY_SCOPES_LIST
[16:40] <Saviq> mzanetti, in unity8.conf
[16:41] <mzanetti> Saviq: is that a different one for run_on_device and upstart?
[16:41] <tsdgeos> going to reflash and re-distupgrade
[16:41] <Saviq> mzanetti, run_on_device doesn't do upstart (it should, btw)
[16:41] <Saviq> mzanetti, on desktop you'll have all of them, unless launched via upstart, or exported that above with semicolons
[16:41] <tsdgeos> see if i get anything in the dash this way
[16:41] <mzanetti> ah, understood
[16:41] <mzanetti> thanks Saviq
[16:45] <mzanetti> Saviq: tsdgeos: https://code.launchpad.net/~unity-team/unity8/new-scopes-clean-to-trunk/+merge/209459
[16:45] <mzanetti> Saviq: you may go ahead merging stuff together
[16:46] <Saviq> mzanetti, you for real? :D
[16:46] <mzanetti> why?
[16:46] <Saviq> jk :)
[16:46] <Saviq> and happy
[16:46] <Saviq> I won't land today, though, we need extensive testing
[16:46] <mzanetti> sure
[16:46] <Saviq> will ask for design reviews
[16:47] <mzanetti> Saviq: meh. just when I wrote this, I have that thing that mhr3 mentioned earlier
[16:47] <mzanetti> can't really unlock it
[16:47] <elopio> Saviq: how can I help you?
[16:47] <mzanetti> appears frozen
[16:48] <mhr3> mzanetti, apparently you need to wash your fingers
[16:48] <mhr3> right Saviq? :P
[16:48] <tsdgeos> well
[16:48] <tsdgeos> so i had that
[16:48] <tsdgeos> and it seems it's mir gone crazy
[16:48] <tsdgeos> i press the lock button
[16:48] <tsdgeos> and then i can unlock
[16:48] <tsdgeos> it's as if it was locked but still showing stuff on screen
[16:48] <tsdgeos> ymmv
[16:49] <mzanetti> nope...
[16:49]  * mzanetti reboots
[16:49] <mhr3> i'm not going to deny that it's mir
[16:49] <mhr3> might be
[16:51] <Saviq> elopio, hey, so I've been looking into bug #1294214
[16:51] <Saviq> elopio, the reason why the latter one fails is that the test launches them in main stage, where right edge doesn't work
[16:52] <Saviq> elopio, also, it uses real apps instead of fixtures, so I started looking into switching to that
[16:52] <elopio> Saviq: I see.
[16:52] <Saviq> elopio, but then I looked that there's plenty of things that should be abstracted
[16:52] <Saviq> elopio, like launching an app, killing it, verifying it's focused
[16:53] <Saviq> elopio, and now I'm not sure how all of that should really work, on one hand it makes sense for a rich fixture like that to be in UITK, but then some of it might end up very unity8-specific
[16:54] <Saviq> elopio, so what I thought we should do at least was extending the FakeApplication fixture in unity8 for use in the app lifecycle tests
[16:55] <elopio> Saviq: I generally start doing it specific to the project, until I need it in another one and then I move it to the toolkit. That's how FakeApplication ended up there.
[16:55] <elopio> how would you like to extend it?
[16:55] <Saviq> mhr3, good news, Jouni will take new scopes for a round of design reviews tomorrow
[16:55] <elopio> sounds good, but I'm not sure what you have in mind.
[16:55] <Saviq> elopio, so, looking at test_swipe_out_application_started_by_url_dispatcher
[16:55] <Saviq> elopio, the "extract appname from desktop file name" needs to be common
[16:56] <mhr3> Saviq, yea, i think i'm in that meeting
[16:56] <Saviq> elopio,         self.addCleanup(os.system, 'pkill qmlscene')
[16:56] <Saviq> elopio, maybe        self.assertThat(
[16:56] <Saviq>             self.main_window.get_current_focused_app_id,
[16:56] <Saviq>             Eventually(Equals(application_name)))
[16:56] <Saviq> mhr3, oh, didn't know there was a meeting ;D
[16:56] <Saviq> mhr3, I meant pre-landing design review
[16:56] <elopio> Saviq: so I think what you want is a base test case for life cycle management
[16:57] <elopio> on the set up, you can make sure that all the qmlscenes are killed, and we can make helper methods from the other things you mentioned. I like that.
[16:58] <Saviq> elopio, yeah, that was my first idea, but then... wouldn't it be clean for the fixture to launch/close itself on destruction?
[16:58] <Saviq> *cleaner
[16:58] <Saviq> I'm sure it's possible to add a cleanup for a fixture :)
[17:00] <Saviq> so that's why I started to think about wrapping the FakeApplication in a more elaborate fixture
[17:01] <Saviq> tsdgeos, mzanetti, so everything's in either clean-to-trunk or cleanup, right?
[17:01] <Saviq> no more hanging branches?
[17:01] <tsdgeos> Saviq: did oyou merge my lvwph fix?
[17:02] <Saviq> tsdgeos, or well, apart from that one, that's targeted to trunk and merged separately
[17:02] <tsdgeos> Saviq: that should be all i know yes
[17:02] <elopio> Saviq: we could do a fixture to launch things with url dispatcher.
[17:02] <elopio> it would use the FakeApplication fixture.
[17:02] <Saviq> elopio, yeah, that's what I thought
[17:03] <elopio> but if you would like to launch the application with the launcher, that would be a different fixture
[17:03] <elopio> but that's ok.
[17:03] <elopio> do you need a hand with this?
[17:04] <Saviq> elopio, depends, if you have the time and want to take it on, great
[17:04] <Saviq> elopio, otherwise I'll manage tomorrow
[17:05] <mzanetti> Saviq: yep
[17:06] <tsdgeos> Saviq: clean flash + silo -> nothing in the dash
[17:06] <tsdgeos> phablet@ubuntu-phablet:~$ /sbin/status scope-registry
[17:06] <tsdgeos> scope-registry stop/waiting
[17:06] <tsdgeos> who is supposed to start that?
[17:06] <elopio> Saviq: I don't have time today. Tomorrow maybe we can split some tasks, like updating the current tests so they don't use real apps, fixing that right-edge test, and creating the new fixture.
[17:06] <Saviq> tsdgeos, unity8 emitting a signal
[17:07] <Saviq> tsdgeos, but apparently it doesn't work
[17:07] <tsdgeos> Saviq: doesn't seem to be working for me :/
[17:07] <Saviq> tsdgeos, yeah, me neither
[17:07] <Saviq> tsdgeos, but if you reboot, they should be there
[17:07] <Saviq> mhr3, are the jobs reacting to the upstart signal yet?
[17:07] <tsdgeos> i did reboot
[17:07] <tsdgeos> but let me reboot again
[17:07] <mzanetti> I can reproduce the unlock issue by rebooting, unlocking it, leaving it time out for a while and trying again
[17:07] <mhr3> Saviq, yea, works fine
[17:07] <Saviq> tsdgeos, ah
[17:08] <Saviq> tsdgeos, do you maybe have a unity8.conf in ~/.config/upstart?
[17:08] <mzanetti> didn't we have that in the mwc image too?
[17:08] <mzanetti> maybe a fix that got lost in between the branches
[17:09] <Saviq> tsdgeos, hmm, I think it doesn't work on restart
[17:09] <Saviq> tsdgeos, but if you go stop unity8, start unity8, is fine
[17:09] <tsdgeos> Saviq: i do have lots of stuff there, kill'em'all?
[17:09] <Saviq> tsdgeos, lots? what kind?
[17:09] <tsdgeos> .override too
[17:10] <tsdgeos> it's gone now :D
[17:10]  * tsdgeos reboots again
[17:10] <Saviq> ;d
[17:10] <Saviq> kgunn, I started seeing the issue you reported, too
[17:11] <Saviq> kgunn, the "browser launches", aka "first app launches"
[17:11] <mhr3> +1
[17:11] <Saviq> kgunn, I think it has to do something with the fact that apps don't close straight away
[17:11] <tsdgeos> ok
[17:11] <tsdgeos> that's better
[17:11] <mhr3> Saviq, it looks like the screen doesn't update after closing the app
[17:11] <Saviq> mhr3, or yeah, that
[17:11] <tsdgeos> and now i'm getting the middle centered header on reboot
[17:11] <tsdgeos> ok
[17:12] <tsdgeos> gonna eod, moar tomorrow
[17:12]  * tsdgeos waves
[17:12] <mzanetti> Saviq: something is really bad now with the stuff I installed from the silo
[17:13] <mhr3> Saviq, yea, cause scrolling fixes it
[17:13] <Saviq> mhr3, yeah
[17:14]  * Saviq says mir/unity-mir is the cause :?
[17:14] <Saviq> mzanetti, wdym?
[17:15] <mzanetti> it behaves really nasty... all the time it hangs for seconds
[17:15] <mzanetti> didn't notice this stuff when I tested only the branch alone
[17:15] <Saviq> mhr3, ok, now I understand what's happening, "initctl restart unity8" doesn't re-read the job file, so it doesn't emit the signal
[17:15] <Saviq> on restart
[17:15] <Saviq> only stop/start does
[17:16] <Saviq> stoopid
[17:16] <mhr3> Saviq, but the registry isn't stopped on unity exit, so why would that matter?
[17:17] <Saviq> mhr3, only first install
[17:17] <Saviq> mhr3, I mean dist-upgrade, restart unity8
[17:17] <Saviq> no scopes!
[17:17] <mhr3> ah, right
[17:17] <mhr3> you need to reboot more :P
[17:18] <Saviq> indeed ;P
[17:18] <Saviq> mhr3, which, TBH, was the whole point of the event ;)
[17:19] <mhr3> Saviq, well.. sounds like an upstart bug to me :)
[17:19] <Saviq> but yeah, useful anyway, makes sure scopes are started when unity8 is
[17:20] <Saviq> mhr3, feature, probably
[17:20] <mhr3> bug
[17:20] <mhr3> restart == stop + start
[17:23] <Saviq> yay, 9.3K diff
[17:23] <Saviq> and we tried to keep it small...
[17:27] <Saviq> mhr3, bug #1294230
[17:27] <mhr3> Saviq, thx
[17:55] <karni> Saviq: Do you know why written instructions are gone from the demo-stuff ppa?
[17:55] <karni> Saviq: adding the ppa and dist-upgrade was enough, right?
[17:55] <karni> to get the new scopes
[17:59] <mhr3> karni, Saviq, i deleted the pkgs and the instructions, it didn't have 5.2, so was useless, use the landing silo with new scopes (in landing-013)
[17:59] <karni> mhr3: ack
[18:10] <Saviq> Trevinho, hey, do you know where is the "I saw the shortcut hints, don't show them any more" stored?
[19:20] <alecu> mzanetti, mhr3: I think the bug with duplicated icons in "Available" and "Installed" happens for some apps only, and it's http://pad.lv/1279481
[19:21] <mzanetti> alecu: ah, interesting. yeah, that would make sense
[19:22] <mhr3> "uniquify"? is that a real word? :P
[19:32] <Trevinho> Saviq: ~/.cache/unity/first_run.stamp
[20:30] <Saviq> Trevinho, thanks
[20:44] <Saviq> kgunn, the "browser launched when you tap on close" bug is actually a rendering / compositing bug https://bugs.launchpad.net/unity8/+bug/1293896/comments/1
[20:44] <Saviq> kgunn, caused by the latest unity-{,mir} landing
[20:44] <kgunn> Saviq: ack....
[20:45] <kgunn> nothing changed in unity-mir afaik
[20:45] <kgunn> i'm about to test the mir fix...
[20:45] <kgunn> sounds the same
[20:45] <Saviq> kgunn, ah, you're on it already
[20:46] <kgunn> yep...just updating from silo now
[20:46] <Saviq> mhr3, told you so https://bugs.launchpad.net/ubuntu/+source/upstart/+bug/1294230/comments/2 ;P
[20:54] <mhr3> saviq, hopefully systemd make sane design choices :P
[20:54] <mhr3> made*
[20:54] <Saviq> ;)
[20:57] <rsalveti> Saviq: updated bug 1293478
[20:57] <rsalveti> Saviq: seems to be an issue with qt + unity8
[20:58] <Saviq> rsalveti, or well, qmenumodel needs to cope
[20:59] <rsalveti> Saviq: why would -1 be double now?
[20:59] <Saviq> rsalveti, we can't really rely on JS doing anything sane wrt typing
[20:59] <rsalveti> right
[20:59] <Saviq> rsalveti, possibly the v8→v4 change or os
[20:59] <Saviq> so
[21:00] <rsalveti> hm, right, that's annoying
[21:02] <Saviq> rsalveti, could you check with silo 13 if it solves the "I played with slider and it doesn't work no more" issues?
[21:03] <rsalveti> Saviq: sure, the interesting thing is that I wasn't even getting the dbus messages anymore (with dbus-monitor)
[21:03] <Saviq> rsalveti, yeah, the binding is broken
[21:03] <Saviq> rsalveti, in some cases
[21:04] <rsalveti> right
[21:04] <Saviq> rsalveti, silo 13 should have that fixed (via bug #1283191)
[21:04] <rsalveti> great, let me give that a try
[21:05] <Saviq> rsalveti, as for the int vs. double, I'll have a chat with dednick tomorrow morning if we can think of something better than u8 having to know the type
[21:05] <rsalveti> yeah
[21:22] <rsalveti> Saviq: yeah, not perfect but better
[21:22] <rsalveti> it seems that after you change the value in the indicator itself, it the up/down limit changes as well, you basically can't reach 0 anymore with the hwkeys
[21:23] <rsalveti> probably a scale issue or similar
[21:23] <Saviq> rsalveti, probably because 0.5 - 1 gets you -0.5 that gets rejected
[21:23] <rsalveti> yeah
[21:24] <Saviq> rsalveti, should get clamped
[21:24] <Saviq> rsalveti, comment on the bug/MP please, we'll get that fixed before landing it
[21:24] <rsalveti> sure