[08:42] <mzanetti> pete-woods: ping
[08:42] <pete-woods> mzanetti: hi
[08:43] <mzanetti> pete-woods: I was looking into the colors of the infographics. seems that comes from the libusermetrics model.
[08:43] <mzanetti> pete-woods: can you point me to the right place?
[08:44] <pete-woods> mzanetti: yes, basically there's an interface (http://bazaar.launchpad.net/~unity-team/libusermetrics/trunk/view/head:/src/libusermetricsoutput/ColorThemeProvider.h)
[08:44] <pete-woods> mzanetti: and a dummy implementation with hard-coded colours (http://bazaar.launchpad.net/~unity-team/libusermetrics/trunk/view/head:/src/libusermetricsoutput/HardCodedColorThemeProvider.h)
[08:45] <mzanetti> pete-woods: so that's the only place were we get the colors from?
[08:45] <pete-woods> mzanetti: at the moment, yes
[08:45] <mzanetti> pete-woods: we'd need to make that configurable.
[08:45] <pete-woods> mzanetti: obvious
[08:45] <pete-woods> ly
[08:45] <mzanetti> pete-woods: not by the user, but for the vendor
[08:45] <mzanetti> pete-woods: no... I mean right now :D
[08:46] <pete-woods> mzanetti: didn't we have this conversation last week?
[08:46] <mzanetti> pete-woods: yes... I wanted to get it done right now
[08:46] <pete-woods> tell me what configuration to read / or if I should just make an API and I'll do it
[08:46] <mzanetti> pete-woods: I guess we can just read some dconf key in there
[08:47] <mzanetti> pete-woods: it's probably not something with a big future. but we need to have that configurable somehow for iteration1
[08:47] <pete-woods> okay
[08:47] <pete-woods> mzanetti: should I just invent some configuration format to stick into dconf then?
[08:48] <mzanetti> pete-woods: I'd say yes... keep it simple... 3 color strings (or whatever amount of different colors we need)
[08:49] <pete-woods> mzanetti: basically we need 3 colours for each "theme", and should probably support multiple themes
[08:50] <mzanetti> pete-woods: I think that's not even needed
[08:50] <pete-woods> mzanetti: I don't want to have to do this twice, though, and it really won't make it much more difficult
[08:51] <mzanetti> pete-woods: ok. in that case... feel free
[08:51] <pete-woods> mzanetti: do we have a "canonical" way of accessing dconf from Qt?
[08:51] <mzanetti> pete-woods: gsettings-qt
[08:51] <pete-woods> mzanetti: okay, thanks, will use that
[08:52] <mzanetti> pete-woods: QGSettings gSettings("com.canonical.Unity.Infographics", "/com/canonical/unity/infographics/");
[08:53] <pete-woods> mzanetti: cool, writing my code for me! :)
[08:53] <mzanetti> just copy pasted it from my code and replaced launcher with infographics :D
[08:53] <mzanetti> in the hope it saves you 5 minutes
[08:53] <pete-woods> probably will!
[08:53] <pete-woods> :D
[09:27] <mzanetti> mhr3: ping
[09:27] <mhr3> mzanetti, pong
[09:27] <mzanetti> mhr3: hey, how hard is it to do the FIXME in scope.cpp:172 ?
[09:27] <mzanetti>  // FIXME: Change to use row index.
[09:28] <mzanetti> I'd need that
[09:28] <mhr3> mzanetti, why would you need that?
[09:29] <mhr3> also pstolowski wrote that code, he'll know better :)
[09:29] <mzanetti> mhr3: because we need to change the previews to be able to swipe them left/right
[09:29] <mzanetti> mhr3: and increasing/decreasing the index would be better that somehow finding all the data from a delegate from the outside
[09:31] <mhr3> i do remember someone mentioning the index when the branch was being reviewed, but there was some reason why we kept it this way
[09:31] <pstolowski> mzanetti: the fixme shouldn't be there anymore, at some point I had a branch that changed it to use index, but it was decided not do so, but unfortunately I don't recall what was the reasoning anymore...
[09:31] <mzanetti> mhm...
[09:31] <mzanetti> interesting
[09:32] <mzanetti> pstolowski: was it Saviq?
[09:32] <pstolowski> mzanetti: let my try to dig up in my logs..
[09:33] <mzanetti> cool, tahnks
[09:34] <pstolowski> mzanetti, mhr3 : ok, I found out what was the reason
[09:35] <pstolowski> mzanetti, mhr3 : https://code.launchpad.net/~mhr3/unity8/use-dee-filtermodel/+merge/171846 made it no longer possible to support index-based activation
[09:36] <mhr3> oh right, cause the activation is on the scope level and not on the category level
[09:36] <pstolowski> mzanetti: in short, the indices doesn't make sense anymore internally afair, mhr3 can explain..
[09:37] <mhr3> pstolowski, you home scope fixes didn't do the category sorting, right?
[09:37] <mhr3> pstolowski, would be nice to get that in with the rest of the fixes
[09:38] <pstolowski> mhr3: right, I forgot about that.. let me add it to the next one
[09:38] <mhr3> cause now the device shows music first in home
[09:38] <mhr3> instead of apps
[09:38] <mhr3> people will complain :P
[09:40] <mzanetti> pstolowski: so if I'm in GenericScopeView, how would I get to the information required to open another preview?
[09:41] <pstolowski> mzanetti: I'm not familar with qml side of things, but essentialy you just pass another row, unless it's the last row in the model
[09:42] <pstolowski> mzanetti: and you get the preview via onPreviewReady signal
[09:42] <asac> 11:40 < asac> anyone can look at [09:42] <mzanetti> pstolowski: that's the thing... I don't have access to that row
[09:42] <asac> 11:40 < asac> and tell me if that changes anything in the ui?
[09:42] <asac> 11:40 < asac> those are staged and just wanted to punt them in the archive... just double checking because of the freezes
[09:46] <mhr3> mzanetti, can't you just pass the model + index to your thing and then do model.get(index+1)?
[09:47] <mhr3> dee models do expose get iirc
[09:47] <mzanetti> I'm afraid Saviq won't approve that :D
[09:47] <pstolowski> hmm
[09:48] <mhr3> mzanetti, imo it's not that terrible if you just do that when you're changing the item
[09:48] <mhr3> ie on user action
[09:50] <mhr3> mzanetti, ultimately, why does this even matter? if you do have the index (which you wanted in the first place) you can just call the get() when activating the item
[09:50] <pstolowski> mzanetti: essentialy, you want to "advance" to the next/prev item in the model, and perform scope.preview(..) on it; can't it be achieved via a signal from delegate to the parent list?
[09:51] <mzanetti> pstolowski: no... the delegate is hidden doesn't have anything to do
[09:51] <mzanetti> pstolowski: the current preview is open and that's the one that needs to trigger loading for the next/previous preview
[09:52] <pstolowski> mzanetti: ah, ok, got it
[09:52] <mzanetti> but yeah... exposing the whole model is probably the best we can... holding the whole model in the preview and call get on next/previous. discarding it when the preview is closed
[09:52] <mzanetti> ok... I'll give that a shot
[10:02] <mhr3> mzanetti, if you're touching the preview code it'd also be nice to have immediate response :)
[10:02] <mhr3> cause right now you see preview after clickandhold + whatever time the scope/network takes to return it
[10:03] <mhr3> it'd be cool to open the preview window right after clickandhold and show a spinner
[10:03] <mzanetti> mhr3: ack
[10:03] <mhr3> mzanetti, then again designs say that preview should be open on tap, not on clickandhold
[10:03] <mzanetti> mhr3: huh? but not for everything, or?
[10:04] <mhr3> mzanetti, right, installed/recent apps are exception
[10:04] <mzanetti> mhr3: i.e. launching an installed app would be tedious if having to go through the preview
[10:04] <mzanetti> mhr3: but what about the other scopes? i.e. music? should we always go through the previews there?
[10:05] <mhr3> according to designs, yes
[10:05] <mzanetti> mhm...
[10:05] <mzanetti> well... anyways... doesn't change the fact that it should open right away and show a spinner while loading
[10:05] <mhr3> indeed
[10:05] <mzanetti> doesn't really matter to which handler it is connected
[10:06] <mhr3> mzanetti, oh but there's a catch - double tap should "launch" stuff right away
[10:06] <mhr3> no preview
[10:06] <mzanetti> ok... I guess that's a different merge than the one with the preview list
[10:07] <mhr3> sure, just something to keep in mind
[10:07] <mhr3> hard to say whether a tap is going to be double one when you receive it :P
[10:07] <mhr3> unless sdk added Oracle {}
[10:32] <easd> do you know how to open window spread effect at first click on the launcher, and not the second one?
[10:40] <mhr3> Saviq, got the lock up again :( same trace as in the bug i reported
[10:40] <mhr3> this time it happened as i was closing lots of open apps
[10:40] <mhr3> yey for deadlocks
[10:48] <asac> mzanetti: Saviq: can you pleasea look for the unity changes
[10:48] <asac> and tell me if they change UI on the desktop?
[10:48] <asac> otherwise we cant push the button and thostrsscopes etc. are blocked
[10:49] <mzanetti> asac: unity7?
[10:49] <asac> http://people.canonical.com/~platform/cu2d/results
[10:49] <asac> search for [10:49] <asac> mzanetti: well check the changes in there
[10:49] <asac> quite a lot desktop changes... if that changes UI i need to find an exception
[10:50] <mzanetti> I have no idea tbh
[10:50] <asac> mzanetti: can you check please?
[10:50] <asac> mzanetti: i have even less of an idea
[10:51] <asac> or tell me who in unity team can check
[10:51] <mzanetti> asac: I wouldn't be so sure about that
[10:51] <asac> we want to get that flushed in
[10:51] <asac> who is doing this stuff? if we kick it out and it doesnt make saucy is that also fine?
[10:51] <mzanetti> asac: better ask Trevinho
[10:51] <asac> :)
[10:51] <asac> kgunn: ^^
[10:51] <asac> kgunn: who knows about unity desktop?
[10:51] <mzanetti> asac: this is unity7... I don't think even kgunn know about it...
[10:51] <asac> kgunn: i need someone to check the changes pending and confirm that it doesnt change UI etc. for the freezes
[10:52] <asac> mzanetti: if he doesnt care, then we just dont take it and get no more updates in saucy ... i doubt thats the case  though :)
[10:52] <asac> Trevinho: hi
[10:52] <asac> Trevinho: your changes... if you need them better tell us now that they are good :)
[10:53] <mzanetti> ChrisTownsend: ^^
[10:53] <mzanetti> mhr3: do you know someone from unity7 that is around at this time?
[10:54] <mhr3> mzanetti, Trevinho andyrock ^^
[10:54] <andyrock> hola :D
[10:54] <mzanetti> andyrock: hi... can you answer asac's questions?
[10:55] <asac> mhr3: look for [10:55] <asac> andyrock: ^^
[10:55] <asac> http://people.canonical.com/~platform/cu2d/results
[10:55] <asac> tell me if anything in there changes the UI
[10:55] <asac> (i dont think so)
[10:55] <andyrock> ok let me check
[10:58] <andyrock> asac, no changes in UI
[10:59] <asac> thanks!!
[11:07] <andyrock> np!
[11:24] <dednick> Cimi: ping
[11:24] <Cimi> dednick, pong
[11:25] <dednick> Cimi: howdy. is ubuntu-settings-components in the archive yet?
[11:25] <Cimi> dednick, apt-cache search :)
[11:25] <Cimi> dednick, dunno
[11:25] <dednick> Cimi: mh. no, it's not
[11:26] <dednick> Cimi: we should probably get that in.
[11:27] <dednick> sigh. and i choose the day that didrocks is not in...
[11:28] <mzanetti> pstolowski: why does preview() require all those parameters and not just the uri?
[11:30] <pstolowski> mzanetti: 2 reasons - 1) we would need to iterate over the model to find the rest (all need to be passed the backend); 2) that assumes uri is unique
[11:30] <mzanetti> pstolowski: yes... I think the url is unique
[11:31] <mzanetti> pstolowski: afaics, when I get the previewReady() signal I have no chance it knowing which preview it actually is
[11:31] <pstolowski> mzanetti: I think we don't want to make that assumption, it bite us in unity7
[11:31] <mzanetti> pstolowski: so how can I know which preview this actually is? it seems the Preview class contains none of the identifying items
[11:31] <dednick> Cimi: you doing any work on that branch still?
[11:32] <Cimi> dednick, it needs to be integrated
[11:32] <Cimi> dednick, becoming real widgets...
[11:32] <mzanetti> pstolowski: and given that we don't have unique id's I would need to have all that stuff in the preview again and compare everything again
[11:36] <dednick> Cimi: hm. i want to move all the indicator widgets into there.
[11:36] <Cimi> dednick, ok...
[11:37] <Cimi> dednick, then import from this package?
[11:37] <dednick> Cimi: yeah
[11:37] <pstolowski> mzanetti: unitycore has the result row already in the preview object, but it's not exposed in the bindings, so it needs to be fixed if we need to match on that
[11:38] <mzanetti> pstolowski: what's the result row? the index?
[11:45] <pstolowski> mzanetti: no, that would be a struct or variant that carries actual values (same values you pass when calling activate/preview)
[11:45] <mzanetti> pstolowski: another possiblity would be to introduce some sort of cancelPreview(...)
[11:45] <mzanetti> pstolowski: so I would know that the incoming signal is the correct one
[11:46] <mzanetti> if always cancelling previous ones
[11:49] <pstolowski> mzanetti: indeed. that would be handled completly in the bindings (transparent for you)
[11:50] <pstolowski> mzanetti: ie. preview() request would automatically cancel the previous request
[11:50] <mzanetti> which bindings?
[11:50] <asac> mzanetti: Saviq: unity8 crashes in automation
[11:50] <asac> https://jenkins.qa.ubuntu.com/job/saucy-touch_ro-mako-smoke-unity8-autopilot/103/?
[11:50] <mzanetti> pstolowski: ah... yeah... even better.
[11:50] <asac> look in build artifacts for the crash
[11:50] <asac> file
[11:50] <mzanetti> pstolowski: do you think you can add that for me?
[11:50] <asac> please file a bug
[11:51] <asac> we will forget in case it doesnt happen on retry
[11:52] <mzanetti> asac: where is the crash happening?
[11:52] <asac> mzanetti: https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1228097
[11:52] <asac> mzanetti: during unity8 autopilot runs
[11:53] <asac> all is in the bug
[11:53] <asac> and artifacts
[11:53] <asac> the tests succeeded, just systemsettle failed because it apparently was processing the crash files
[11:53] <mzanetti> ah... see it
[11:55] <asac> in case it doesnt go away
[11:55] <asac> i will let you know
[11:55] <asac> and get QA escalate it :)
[11:55] <asac> lol
[11:56] <mzanetti> "lol" is not evil enough... you should uas "muahahaha"
[11:56] <mzanetti> s/uas/use/
[11:57] <mzanetti> :)
[12:06] <asac> mzanetti: so its happening again
[12:06] <asac> mzanetti: seems a real issue... just install 57 image on mako
[12:06] <asac> and run unity8 autopilots... i am sure you will hit the crash on 1 or 2nd try
[12:07] <mzanetti> asac: ack
[12:07] <Cimi> mzanetti, still keeping cdimage-touch or we can switch to ubuntu-system^
[12:07] <Cimi> ?
[12:08] <mzanetti> Cimi: not sure... haven't tried ubuntu-system lately
[12:08] <asac> thostr_: ^^ i think home scope is also yours... 57 build always crashes that when running unity8 autopilot
[12:11] <pstolowski> mzanetti: I will. can you in the meantime rely on the previewReady signal as it is; it shouldn't block you, right?
[12:11] <mzanetti> pstolowski: no, shouldn't block me... it will "just" cause the wrong preview to show up until this is in
[12:11] <mzanetti> when the user scrolls too fast, that is
[12:11] <pstolowski> mzanetti: exactly
[12:18] <asac> mzanetti: any news :)?
[12:18] <asac> sorry... just need info if this issue a) is making the phone not dogfoodable and b) whether it is understood to make an assessment if the current image can go out
[12:18] <mzanetti> asac: regarding the crash? no, not yet. I'm kinda in the middle of something
[12:19] <asac> mzanetti: well, this one is most important of all :)
[12:19] <asac> when can you loko at it?
[12:20] <mzanetti> asac: ok... give me a few minutes to bring this to a state where I can interrupt it
[12:20] <mzanetti> and I'll look into the crash
[12:20] <asac> awesome
[12:39] <asac> mzanetti: do you already know enough to give an assessment about impact of dogfoodability in the bug? https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1228097 ? if not, please do once you know and ping me :)
[12:39] <asac> thx
[12:55] <Cimi> maguro doesn't lock anymore here with cdimage of 18 sept
[12:57] <Cimi> well, now works :)
[13:01] <mzanetti> pstolowski: hey, about that crash... this is the closest I could come so far: http://paste.kde.org/pf84d9e23/
[13:01] <mzanetti> pstolowski: any ideas?
[13:02] <pstolowski> mzanetti: I suppose it's random? it's in home?
[13:02] <mzanetti> pstolowski: yes
[13:07] <mzanetti> cwayne: ping
[13:08] <pstolowski> mzanetti: do I need to be running today's image? any pattern to reproduce?
[13:09] <mzanetti> pstolowski: yes, todays image and apparently it only happens with the latest unity build
[13:23] <pstolowski> mzanetti: I'm flashing and will try to repro. can you install dbg symbols to get better stacktrace?
[13:23] <mzanetti> pstolowski: I don't even manage to flash the device...
[13:23] <mzanetti> phablet-flash bailed out and I can't recover any more
[13:23] <mzanetti> but yes... once I have it back up I will try
[13:31] <Saviq> dednick, standup
[13:31] <Saviq> mzanetti, ↑
[13:43] <Cimi> another thing, directionaldragarea uses abbreviations, like maxSpeed instead maximumSpeed
[13:43] <Cimi> I think we wanted always to go with full words, right?
[13:45] <Cimi> Saviq, dandrader|afk, ok was the "maxSilenceTime" causing the drop of the gesture...
[13:46] <Saviq> Cimi, cool
[13:46] <Cimi> wondering what a sane value is...
[13:46] <Cimi> default is 200ms
[13:49] <dednick> Saviq, mzanetti: jenkins ci builds seem to be taking a very long time today... any idea?
[13:49] <dednick> #1054 has been going for 5 hours
[13:49] <mzanetti> dednick: which job?
[13:50] <Saviq> dednick, mzanetti we've switched to real hardware from the vm
[13:50] <Saviq> http://10.97.2.10:8080/job/autopilot-testrunner-otto-saucy/
[13:50] <Saviq> looks pretty backed up
[13:50] <dednick> http://10.97.2.10:8080/job/generic-mediumtests-saucy/
[13:51] <Saviq> dednick, yeah, the hardware is backed up - we seem to be down to two machines that can run autopilot tests
[13:51] <Saviq> fginther, any word on that ↑ ?
[13:51] <dednick> Saviq: ic.
[13:52] <dednick> Saviq: if you get a chance today can you look at the indicator messages merge? https://code.launchpad.net/~nick-dedekind/unity8/unitymenumodel.messages/+merge/181552
[13:52] <dednick> there are a few bugs waiting on it
[13:53] <Saviq> dednick, not gonna happen, I'm afraid, flying home in ~2hrs
[13:53] <dednick> Saviq: ah ok. I'll bug someone else :)
[13:55] <fginther> dednick, Saviq, yes, it's at two machines at the moment. The long queue is the result of a failed machine and jenkins not rescheduling the jobs on the other system correctly
[13:56] <fginther> dednick, Saviq, both machines are up and running now, and tests run in about 1/3 the time they did on the VMs
[14:03] <tedg> Hey folks is greyback out today or just at lunch/rebooting ?
[14:06] <dandrader> tedg, haven't seem him today.
[14:12] <ChrisTownsend> mterry: Hi
[14:13] <mzanetti> tedg: he was at the sprint, but didn't show up in the standup. I assume he's on his way home
[14:13] <Saviq> fginther, sounds awesome, looking forward toit
[14:13] <Saviq> tedg, flying
[14:13] <tedg> Cool, thanks guys.
[14:13] <Saviq> tedg, actually should be around later today
[14:13] <mterry> ChrisTownsend, hello!
[14:13] <ChrisTownsend> mterry: Hey, I would like to know if distro agrees with this for Unity7: https://code.launchpad.net/~jbicha/unity/recommend-telepathy-indicator/+merge/186392
[14:15] <Saviq> tedg, well, actually he was on http://flightaware.com/live/flight/BAW202/history/20130920/0250Z/KBOS/EGLL - so assuming he survived he should be online some time soon
[14:16] <mterry> ChrisTownsend, I'm probably not the right person to approve that these days.  When did we drop telepathy-indicator?  I'm not in-tune with Desktop enough to know whether we want it in or not
[14:17] <ChrisTownsend> mterry: Honestly, I'm not really sure what the history is behind this.  I'd usually ask sil2100 or didrocks, but they are both out, so I thought maybe you could help.  I'll just wait for one of them to be available.  Thanks for looking!
[14:18] <mterry> ChrisTownsend, seb128 isn't here either...  :-/
[14:18] <ChrisTownsend> mterry: Right, he's another I didn't consider, but...
[14:19] <mterry> ChrisTownsend, sorry  :(  I have been working on unity8 this cycle and haven't been paying attention to Desktop movement
[14:19]  * tedg finds it a little weird Saviq was tracking greyback's plane
[14:19] <Saviq> tedg, ;)
[14:19] <ChrisTownsend> mterry: No worries, thanks for taking a peek anyways.
[15:11] <cwayne> mzanetti: hi
[15:12] <mzanetti> cwayne: can you please test/review this: https://code.launchpad.net/~pete-woods/libusermetrics/color-themes-gconf/+merge/186770
[15:12] <mzanetti> cwayne: btw... pete-woods is your hero... I didn't do anything in the end
[15:17] <cwayne> mzanetti: pete-woods: :D thanks guys, will test that out today
[15:17] <Saviq> mzanetti, N900 and N9 being examples of "this can be done" is a rather broken argument ;)
[15:18] <mzanetti> no its not
[15:18] <Saviq> mzanetti, remember we're not normal people - normal people will leave dozens of apps open at any given time
[15:18] <Saviq> mzanetti, try doing that on the N9
[15:18] <mzanetti> I do that too
[15:18] <Saviq> well, for me N9 slows down quite a lot, and quickly, when I leave too many apps open
[15:19] <mzanetti> Saviq: the only real issue I have with that is GPS apps drain battery. but for that I'd say the solution wouldn't be too hard
[15:19] <mzanetti> just stop gps in the qt backend
[15:20] <Saviq> mzanetti, there's also misbehaving apps that are "taken care of" when they're stopped
[15:20] <Saviq> mzanetti, and TBH, I see no point in the argument about us "not doing multi-tasking well" - the user doesn't even know
[15:20] <Saviq> they can do multitasking as they usually would
[15:20] <mzanetti> Saviq: oh yes he does. ever used android?
[15:21] <mzanetti> everything that's more comples than a fart-app will just behave weird once in the background and brought back
[15:21] <Saviq> mzanetti, remember we won't be just killing them for no reason
[15:22] <Saviq> mzanetti, while there's memory available, they'd just be stopped
[15:22] <mzanetti> games start from the beginning without having your high score saved, every network related app drops connections
[15:22] <Saviq> mzanetti, that's just bad apps, not bad multi tasking
[15:22] <mzanetti> Saviq: bad apps come from bad developers
[15:22] <mzanetti> Saviq: its waaaay more complex to get the state saving correct than not to waste resources
[15:23] <Saviq> mzanetti, sure, the gain we get is that bad apps from bad developers won't destabilize your system
[15:23] <mzanetti> Saviq: shutting down some task on a signal minimized() is easier to do than storing a uncountable amount of possible combinations for anything
[15:23] <mzanetti> at the cost of having much more bad apps
[15:24] <mzanetti> nearly all, that is
[15:24] <Saviq> mzanetti, I disagree, but even so, it's still better for the user experience
[15:24] <dednick> Cimi: ping
[15:26] <mzanetti> Saviq: I disagree there...
[15:26] <MacSlow> anybody ever ran into this one "WireProtocolVersionMismatch: Wire protocol mismatch at <session bus :1.355 /com/canonical/Autopilot/Introspection>: is 1.3, expecting 1.4" (trying to run autopilot-tests)
[15:26] <MacSlow> just pulling updates again, in the hope that will fix the issue...
[15:27] <thomi> MacSlow: yes
[15:27] <thomi> MacSlow: that means somehow you installed python-autopilot 1.4 on top of saucy
[15:27] <Saviq> MacSlow, ppa:autopilot maybe?
[15:28] <thomi> MacSlow: there was a period of about 8 hours where the release team accidentally released 1.4 into the archive
[15:28] <thomi> MacSlow: so you may need to manually remove it, update, and reinstall
[15:28] <MacSlow> thomi, Saviq: *sigh* I don't know... or can't remember... to be honest
[15:28] <thomi> MacSlow: if you run 'dpkg -l | grep autopilot' the version numbers should all start with '1.3' or '1.4' if you're running T
[15:30] <MacSlow> thomi, http://pastebin.ubuntu.com/6133175
[15:30]  * MacSlow looks for the PPA now...
[15:31] <thomi> MacSlow: you shouldn't have ppa:autopilot on your system
[15:31] <thomi> MacSlow: we release into the archive now
[15:32] <Saviq> pstolowski, could you reproduce the unity8/home scope crashes?
[15:33] <pstolowski> Saviq: yes (with ap tests)
[15:33] <Saviq> mzanetti, ↑
[15:34] <pstolowski> Saviq: and currently prepping a possible fix
[15:34] <MacSlow> thomi, in the last weeks I've been deep into ext. snap-decision, thus didn't follow ap at all
[15:34] <Saviq> pstolowski, awesome, thanks
[15:34] <pstolowski> Saviq: btw, are AP tests restarting unity8 a lot?
[15:34] <Saviq> pstolowski, yes
[15:35] <Saviq> pstolowski, for each test, more or less
[15:35] <pstolowski> Saviq: just once for a test? and there are roughly ~240 tests?
[15:35] <Saviq> pstolowski, no, just 50 tests or so
[15:36] <pstolowski> hmm
[15:36] <pstolowski> mhr3: ^
[15:36] <Saviq> pstolowski, http://10.97.2.10:8080/job/generic-mediumtests-runner-saucy/3060/testReport/?
[15:37] <Saviq> pstolowski, and actually ~half of those are gallery app tests
[15:37] <Cimi> dednick, pong
[15:37] <Saviq> pstolowski, we're at 36 total tests
[15:37] <pstolowski> Saviq: ok, thanks
[15:38] <MacSlow> thomi, that doesn't look to good... http://pastebin.ubuntu.com/6133208
[15:38] <dednick> Cimi: hey. was just wondering if you had any specs for the menus in ubuntu-settings-components. Just trying to merge mine into there now.
[15:39] <Cimi> dednick, nope...
[15:39] <dednick> Cimi: ie. desgins
[15:39] <dednick> Cimi: ok
[15:39] <thomi> MacSlow: ok, it seems that those packages made it into the next PPA. I'll ask around and see if I can get that fixed.
[15:39] <MacSlow> thomi, how long does that take?
[15:40] <MacSlow> thomi, because this is blocking me from working on the ap-tests for the extend snap-decisions
[15:41] <thomi> MacSlow: I just asked in #ubuntu-ci-eng
[15:42] <thomi> MacSlow: it's not really something I can control
[15:42] <MacSlow> Saviq, ^
[15:42] <thomi> MacSlow: I suggest you ask there as well, so we get some traction sooner rather than later
[15:43] <MacSlow> Saviq, I can't/should not do an MR for the working ext. snap-decisions without AP-tests, I guess...
[15:44] <MacSlow> Saviq, although these not being in trunk blocks pete-woods on some tasks, because he is waiting on the pin-pad support in notifications
[15:45] <fginther> MacSlow, hello, I'm trying to help with the autopilot version problem you are having
[15:45] <pete-woods> MacSlow: it's not the pin pad support I need, it's just the plain old dialogues I need right now, the wifi selection dialogue would also be nice, but can come later
[15:46] <fginther> MacSlow, are you seeing the 1.4 version on your local development machine or somewhere else?
[15:47] <MacSlow> fginther, thanks
[15:47] <MacSlow> fginther, also just put all my questions in #ubuntu-ci-eng for reference
[15:47] <fginther> MacSlow, ok, moving there
[15:47] <MacSlow> pete-woods, "plain old dialogs" what do you mean by this?
[15:48] <pete-woods> MacSlow: as in the simple "password prompt" type dialogue
[15:48] <MacSlow> pete-woods, those are all depending on the same feature added to notifications and need the same ap-testing... which currently doesn't work.
[15:49] <pete-woods> MacSlow: np, I just didn't want you to think I needed more than I actually do
[15:49] <MacSlow> pete-woods, the password-dialog, login/password-dialog and pin-pad entry extended snap-decisions all work... and I've python-examples to play around with...
[15:50] <MacSlow> pete-woods, it just that without any ap-tests I cannot make a MR for those hoping it'll get approved and merged.
[15:50] <MacSlow> pete-woods, and I don't think you'd want to work with non-trunk branches for notifications, or do you?
[15:50] <mhr3> pstolowski, that totally explains it
[15:51] <mhr3> and also makes it test-only failure
[15:51] <mhr3> thostr_1, ^
[15:51] <davmor2> davidcalle: Is there a way to tell if the "themoviedb.org" scope is working, On the phone and desktop I have typed Iron Man 3,  I figured that should trigger a response but all the info options are from wikipedia
[15:51] <mhr3> thostr_1, so it's crashing because the tests are doing what unity doesn't normally do
[15:51] <pete-woods> MacSlow: I'm already working with non-trunk unity8 for my development (I'm using your extended-notitications-2 branch)
[15:51] <mhr3> thostr_1, ie restart
[15:52] <MacSlow> pete-woods, well then you have password, login/password and pinpad already available to you
[15:52] <davmor2> davidcalle: nevermind I lowered it down on desktop
[15:53] <MacSlow> pete-woods, just let me check the revisions... not that I might be sitting on some, I've not pushed yet
[15:54] <davidcalle> davmor2, to check the state of a scope, the easiest way is to activate it directly  : eg. themoviedb:iron man 3
[15:54] <davmor2> davidcalle: ah nice thanks
[15:54] <MacSlow> pete-woods, all good... you have what you need with the currently available revisions of lp:~macslow/unity8/extended-snap-decisions-2 and lp:~macslow/unity-notifications/extended-snap-decisions-2
[15:56] <pete-woods> MacSlow: yes, my development has not been slowed, all that is blocked is the landing :)
[15:56] <MacSlow> pete-woods, ok... yeah... sorry about that... I didn't expect autopilot being still such an issue...
[15:57] <MacSlow> pete-woods, I try to see if I can get around it with the tips from fginther
[15:57] <MacSlow> pete-woods, really wished myself I'd be able to get these into trunk by today
[15:58] <thomi> MacSlow: you really shouldn't be using the next PPA
[15:58] <pete-woods> MacSlow: no worries, as long as you don't break the API, I am very happy!
[15:58] <davidcalle> davmor2, np
[15:59] <MacSlow> pete-woods, well I try to avoid breaking API... as much as possible :)
[15:59] <davmor2> davidcalle: hmmm phone is showing no result but the desktop is I'll try it again in bit :)
[15:59] <MacSlow> pete-woods, and then the "API" is really just a bunch of hint-names
[16:00] <MacSlow> pete-woods, libnotify and UnityMenuModel won't change that easily
[16:04] <davidcalle> davmor2, the terrible wifi doesn't allow me to test now on my Nexus7, but there shouldn't be any difference, afaik
[16:08] <davmor2> davidcalle: got it remove the .org and then it works.  So just themoviedb:iron man 3 and it shows the same on both :)
[16:11] <davidcalle> davmor2, oh cool
[16:38] <cwayne> pete-woods: im trying your libusermetrics theming branch, but it FTBFS for me
[16:52]  * dednick thinks it a good time to end the day when he spills a glass of water onto desk with about £3000 worth of laptop/phones/tablets on it...
[16:53] <dednick> plus the multi-adapter plug which has another few grands worth of stuff plugged in...
[16:53] <dednick> sigh... have a good weekend all.
[18:10] <thomi> dandrader: got a second?
[18:10] <dandrader> thomi, yes
[18:11] <thomi> dandrader: so for autopilot, I need to be able to see where the shell thinks the mouse pointer is. FOr configurations that are mir/unity8, kgunn suggested you're the person to ask.
[18:11] <thomi> dandrader: can we have some sort of simple interface to make that possible?
[18:12] <thomi> the current thinking here is that some sort of simple dbus interface might work...
[18:15] <dandrader> thomi, yes, it's possible to do such thing
[18:15] <dandrader> I mean, it's possible to implement such interface for autopilot to query
[18:15] <thomi> dandrader: heh, that's a very carefully worded reply :)
[18:16] <dandrader> :)
[18:16] <thomi> Before we can do application testing for non-touch platforms on top of mir, we'll need something like this.
[18:17] <thomi> since that's likely to happen in the next cycle, I think it might be wise to schedule this for sometime in the next few months. What do you think?
[18:19] <dandrader> thomi, you mean testing unity8-mir for the desktop form factor?
[18:19] <thomi> dandrader: not just unity8, but anything running within it.
[18:20] <dandrader> thomi, because I didn't hear a word about getting it implemented yet...
[18:20] <thomi> dandrader: my understanding is that that is supposed to wkr (although opt-in) in 6 months time.
[18:20] <dandrader> thomi, ah, you mean xmir
[18:20] <thomi> no
[18:20] <thomi> I mean mir + unity8
[18:20] <thomi> opt in for 14.04
[18:21] <dandrader> thomi, interesting. that's news for me.
[18:22] <thomi> I'm sure kgunn will confirm that - sorry you had to find out this way :P
[18:23] <kgunn> dandrader: its always been that way...but hey for unity8, you can already run no desktop...so the real key is rrunning with mir :)
[18:23] <thomi> dandrader: kgunn: So I'll add a bug task to unity8, and assign it to.... someone?
[18:23] <dandrader> kgunn, it will just look like a huge tablet.
[18:23] <thomi> dandrader perhaps?
[18:24] <dandrader> thomi, fine by me
[18:24] <kgunn> dandrader: yeah ;)
[18:25] <thomi> dandrader: consider yourself pressganged: https://bugs.launchpad.net/unity8/+bug/1169362
[18:25] <dandrader> :)
[18:26] <dandrader> thomi, so the milestone would be 14.04?
[18:26] <dandrader> (for the unity8 side of that bug)
[18:27] <thomi> dandrader: well, hopefully before 14.04 is released, but yeah
[18:27] <thomi> if I had my way it'd be in the next 12 weeks
[18:31] <dandrader> thomi, do your way
[18:32] <dandrader> s/do/done
[18:32] <thomi> :)
[18:37] <mzanetti> dandrader: ping
[18:37] <dandrader> mzanetti, pong
[18:37] <mzanetti> dandrader: this just happened when I wanted to drag down indicators: http://paste.ubuntu.com/6133905/
[18:37]  * dandrader realizes it's Friday night in Germany
[18:42] <mzanetti> dandrader: yeah... it is :/
[18:43] <dandrader> mzanetti, hmm, at least it seems to be easy to avoit this crash. check if DirectionalDragArea::fetchTargetTouchPoint returned a valid touch point instead of null
[18:43] <dandrader> before using it
[18:43] <dandrader> null meaning that it din't find it
[18:43] <mzanetti> dandrader: yeah... I've seen that.
[18:43] <dandrader> mzanetti, the hard/curious part is how did we get into such state
[18:44] <dandrader> where the touch that's doing se gesture seemingly vanished...
[18:44] <dandrader> s/se/the
[18:44] <mzanetti> dandrader: yep. that's when I pinged you... I figured in a method called touchEvent_recognized() the touch event actually shouldn't be 0
[18:45] <dandrader> mzanetti, is it hard to reproduce?
[18:45] <mzanetti> dandrader: not sure... I realized more often crashes when trying to reveal something... but this is the first time I actually got a trace (because was in the middle of deugging something else)
[18:47] <mzanetti> dandrader: I didn't do much actually... unity started up, opened the phone app, did a phone call, and when that ended I tried dragging down the indicators and bam
[18:48] <dandrader> mzanetti, can you file a bug report with the info you got?
[18:48] <mzanetti> dandrader: if only I would be able to reproduce :/
[18:49] <dandrader> maybe the theme here is "making DirectionalDragArea resilient about disappeasing/inconsistent touch events"
[18:50] <dandrader> but it could also be that touch events are coming fine but DDA messed up in a different way but ended up crashing just there
[18:51] <mzanetti> dandrader: yeah... I believe more in the second one
[18:51] <dandrader> mzanetti,  nah, DDA is perfectly engineered!
[18:51] <mzanetti> haha
[18:51] <mzanetti> famous last words
[18:55] <mzanetti> dandrader: bug reported
[18:55] <mzanetti> have  a nice weekend
[18:59] <dandrader> mzanetti, you too!
[20:37] <mterry> Can anyone explain if there's anything special I have to do as a shell to have the OSK work?
[20:38] <mterry> (I'm implementing the first-boot welcome screens, which runs outside of unity, and I'm not seeing it come up)