kgunnlpotter: whachya need ?00:20
lpotterjust saying hello :)00:21
lpotteractually kind of wondering how that InputDevice thing is working00:25
kgunnlpotter: been working quite well afaict, we're using it pretty heavily for switching in/out of our windowed mode00:32
lpotterya I noticed the windowed mode when I connected a bluetooth keyboard recently00:33
lpotteram I correct that mako has hdmi support?00:35
lpottervia usb->hdmi adaptor00:36
=== pete-woods1 is now known as pete-woods
tsdgeosdednick: you there?09:40
dednicktsdgeos: I am there.09:41
dednickI am everywhere09:41
tsdgeosdednick: you' ve been doing "frame painting" things lately right? I am investigating this issue in unity8-dash on the desktop in which it gets stuck without painting anything until you restart it. I've added some debug and i've confirmed the items are being created so it may be a "stuff is not getting put on screen" thing. Any idea where i'd put a debug to know if stuff is being asked to be painted?09:42
tsdgeosdednick: "All is All" re-zoolander 2 trailer by cumberbatch :D09:42
dednicktsdgeos: you can use the qt render loop debugging09:43
dednicktsdgeos: QSG_RENDERER_DEBUG=render gives shed loads. cant remember the simple one..09:44
dednicktsdgeos: are you wanting to know if specific items are painting?09:45
tsdgeosit just seems that something got stuck09:45
tsdgeoseither the qt render loop09:45
dednicktsdgeos: whole shell?09:45
dednicktsdgeos: ah. k09:45
tsdgeosdednick: unity8-dash09:45
dednicktsdgeos: shell still responding?09:46
tsdgeosdednick: shell is fine09:46
dednicktsdgeos: right. you can test the client using  'restart unity8-dash QSG_RENDERER_DEBUG=render"09:46
tsdgeosdednick: nope, as said if i restart it, all is fine D:09:47
tsdgeosis just the first time that fails09:47
dednicktsdgeos: ah. k :/09:47
tsdgeosi'll put the QSG_RENDERER_DEBUG=render in the upstart file09:47
dednickput it in the upstart09:47
tsdgeosand see what's up09:47
dednicktsdgeos: it could be qtmir surface item related if you're still seeing client render logs while you're giving the dash surface touch events..09:48
dednickwe've been having a few issues getting it in states where it stops rendering recently.09:49
tsdgeosdednick: so it would seem from the output of QSG_RENDERER_DEBUG=render that it is actually rendering things10:03
dednicktsdgeos: what version of unity8 you running?10:03
tsdgeosdednick: xenial latest10:04
dednicktsdgeos: sorry, qtmir10:04
tsdgeossame :D10:04
dednicktsdgeos: ah. looks like updates havent landed in xenial.10:05
dednicki mean been released to image.10:05
dednicktsdgeos: 0.4.6+16.04.20151113-0ubuntu1 is latest.10:06
tsdgeoswhat's the point of having a development version10:06
tsdgeosthat is outdated10:06
dednicktsdgeos: 1102 still has the occlusion in it.10:06
tsdgeosdednick: can i just recompile qtmir and install it or will i need more stuff?10:06
dednicktsdgeos: i have no idea why things aren't getting updated in xenial10:07
dednicktsdgeos: yup10:07
tsdgeossure not blaming you10:07
tsdgeosmore than a wish setence :D10:07
dednicki think it's still compatible.10:07
tsdgeosk, let me recompile and see if then it's gone10:07
dednicktsdgeos: remember -DNO_TESTS=1 . much faster10:08
tsdgeostoo late :D10:09
tsdgeosdamn the tests failed :S10:16
tsdgeosmaybe that's why is still not on xenial ?10:17
tsdgeosdednick: greyback__: do you know if tests should succeed on xenail10:17
tsdgeosi have10:17
greyback__tsdgeos: hmm, there's no reason I'm aware of that they shouldn't. Lemme check10:18
greyback__ltinkl: hey, did you see https://code.launchpad.net/~lukas-kde/qtmir/fixWheel/+merge/276252/comments/70327610:19
dednicktsdgeos: they should do. i always do tests on laptop.10:21
tsdgeosanyhow compiling without tests now D:10:22
dednickalthough the shared wakelock used to be a bit flaky i havent seen an issue lately10:22
=== alan_g is now known as alan_g|test
=== alan_g|test is now known as alan_g
tsdgeosdednick: yeah so the new qtmir fixes it10:41
tsdgeos*BUT* now i get a crash every time i log otu10:41
tsdgeosi didn't use to10:41
tsdgeosthe crash backtrace is a bit unexplanatory, just quickitems destructors one after other10:41
tsdgeosgreyback__: dednick: do you guys want it?10:41
greyback__tsdgeos: sure10:42
=== greyback__ is now known as greyback
tsdgeoslet me repeat10:42
greybackdid it look like this: https://errors.ubuntu.com/bucket/?id=/usr/bin/unity8%3A11%3AQObject%3A%3AstaticMetaObject%3Acall%3AQMetaObject%3A%3Aactivate%3AQMetaObject%3A%3Aactivate%3AQQuickWindow%3A%3AframeSwapped10:43
tsdgeosdon't think so10:44
tsdgeosand now i can't make it crash :S10:48
tsdgeosi should have kept the backtrace when i got it10:48
tsdgeosit is still crashing10:49
tsdgeosbut the backtrace is different10:50
tsdgeoslog says10:50
tsdgeos*** Error in `unity8': free(): invalid size: 0x00007fd674119950 ***10:50
tsdgeoslet's try valgrind maybe10:52
pstolowskimzanetti_, ping10:59
mzanetti_pstolowski, hey ho11:00
=== mzanetti_ is now known as mzanetti
pstolowskimzanetti_, hey, i've just found out you're going to land a bunch of changes with silo 60, namely the diff updates of unity-scopes-shell11:00
mzanettihmm, you just found out? :)11:00
mzanettipstolowski, didn't you ping me over those branches last week?11:00
pstolowskimzanetti, no i don't think so11:01
mzanettihmm... I could swear you did :) anyhow, hope it's not an issue11:01
pstolowskimzanetti, well11:01
pstolowskimzanetti, it's great to land it (i had a silo myself with it), but it needs to wait for the release of all agregator scopes with ota811:01
mzanettithis silo has been waiting for QA since friday already11:02
pstolowskimzanetti, that's why i held back my silo some time ago11:02
mzanettipstolowski, what is the issue? in my testing everything seemed to work11:02
pstolowskimzanetti, i know, and i see it's under testing now11:02
pstolowskimzanetti, with slow network the experience will be suboptimal in aggregator scopes11:03
mzanettipstolowski, but wait, OTA-8 is being released this very moment11:03
mzanettiwould that unblock it?11:03
pstolowskimzanetti, yes, but i was told that aggregator scopes will be upgraded via store11:03
mzanettipstolowski, any ETA on that?11:04
mzanettiI mean, it's rc-proposed, a slightly suboptimal experience for a day would be ok I guess if the fixes are already in the queue11:04
mzanettiI'd hate to recompile it and wait for another week to get QA on it11:04
pstolowskimzanetti, probably11:04
pstolowskimzanetti, yes that' probably too much11:04
pstolowskimzanetti, forwarded you the email, that's all i know11:05
mzanettiok, thanks.11:05
mzanettipstolowski, so, sorry for the miscommunication, last week when I created the silo I had one that depended on a unity-scopes branch. I asked if I can land it and you gave me two more that should go together. But I might have misunderstood...11:06
pstolowskimzanetti, yeah, looks like a misunderstanding ;)11:08
tsdgeosgreyback: dednick: i can't really reproduce the crash other than that line on the unity8.log11:08
greybacktsdgeos: ok, thanks for trying anyway11:09
mzanettipstolowski, reading through that mail I get the impression that rc-proposed is fine... and that's where the silo is going. I guess if QA doesn't block on the suboptimal experience we're good. Just need to make sure all those scopes are released by OTA-9, which is january11:09
tsdgeosgreyback: dednick: do you guys have a bug for that occlusion rendering thing i can duplicate the one for the dash not painting to it?11:12
tsdgeos1515356 ?11:12
pstolowskimzanetti, i know where the misunderstanding came from11:12
pstolowskimzanetti, https://code.launchpad.net/~aacid/unity8/dash_reset_instead_of_fatal/+merge/27436311:13
tsdgeosit's weird because that one says "until touched"11:13
pstolowskimzanetti, it lists my branch as required :)11:13
tsdgeosmaybe 151455611:13
mzanettiyep, and I remeber I asked if I can/should land those11:13
dednicktsdgeos: https://bugs.launchpad.net/ubuntu/+source/qtmir/+bug/151535611:13
ubot5Ubuntu bug 1515356 in qtmir (Ubuntu) "After a boot the dash doesn't display until touched" [High,In progress]11:13
dednickor yeah, either on...11:14
tsdgeosdednick: but in my case was even if i touched it? (well clicked on it, there's no touch on my laptop)11:14
tsdgeosi'll put it onto https://bugs.launchpad.net/ubuntu/+source/qtmir/+bug/151455611:14
ubot5Ubuntu bug 1514556 in qtmir (Ubuntu) "With occlusion detection, apps stop responding when switching from another app." [Critical,Confirmed]11:14
dednickor there is https://bugs.launchpad.net/qtmir/+bug/151713911:15
ubot5Ubuntu bug 1517139 in QtMir "unity8 crash on running hellogl2 in qtbase5-examples" [Critical,In progress]11:15
greybacktsdgeos: that occlusion detection stuff was reverted11:15
dednickmight be related.11:15
pstolowskimzanetti, the scopes will be released very soon, so ota9 will be fine for sure11:15
tsdgeosgreyback: yes, and as i said, not landed into xenial11:15
tsdgeosgreyback: so I still see it on the desktop unless i compile my own qtmir11:15
greybacktsdgeos: if any interaction with the dash (mouse or touch) makes the dash contents appear, then it's this occlusion bug11:16
greybacki.e. any single event to make it redraw11:16
tsdgeosgreyback: no it does not make the contents appear11:16
tsdgeosgreyback: but i have confirmed the new qtmir does not reproduce the bug anymore11:16
greybacktsdgeos: ok11:16
tsdgeosi.e. i can't reproduce the bug anymore when using 0.4.6+16.04.20151113-0ubuntu1 of qtmir while the bug is there with 0.4.6+16.04.20151102-0ubuntu111:17
tsdgeosof course if someone else ahs time to double verify it's always a nice thing :)11:18
tsdgeosgreyback: i think it would be good if we could get the lastest qtmir in xenial, no idea why it's old11:20
greybacktsdgeos: the issue I was seeing was different: I could interact with dash, but all I ever got was the + thing at the bottom of the screen. If I expanded that + thing, I could select a scope, and then it worked11:20
greybacktsdgeos: it's blocked in release due to a library problem11:21
mzanettitsdgeos, hey, the bug about the dash not showing up in desktop11:59
mzanettitsdgeos, that already happened before we had the occlusion branches landed11:59
tsdgeosmzanetti: define "not showing up"11:59
mzanettitsdgeos, window is there, background image loaded, you can see the bottom edge hint, no content tho12:00
tsdgeosmzanetti: that goes away with the latest qtmir for me12:00
mzanettihmm, ok12:00
tsdgeosi don't know if it happened before or not12:00
tsdgeosbut i can't reproduce it anymore with my self compiled qtmir to the latest version12:01
tsdgeosmzanetti: if you can still reproduce it would be nice to know12:01
mzanettitsdgeos, ack, will watch out for it and let you know12:02
tsdgeosdandrader: https://code.launchpad.net/~aacid/unity8/fix_cursor_noise/+merge/27798212:04
dandradertsdgeos, wonder why you were getting the builtin cursor though12:07
dandradertsdgeos, you should be getting the xcursors from DMZ white theme12:07
dandradertsdgeos, (the ones Unity 7 use)12:08
tsdgeosdandrader: is that dependent on the user config?12:08
dandradertsdgeos, no12:08
dandradertsdgeos, what do you have in /usr/share/icons/default ?12:08
tsdgeosdandrader: http://paste.ubuntu.com/13344574/12:09
dandradertsdgeos, ahh... so you're a KDE user?12:10
dandradertsdgeos, so index.theme is a symlink to KDE's breeze theme12:10
dandradertsdgeos, instead of DMZ-White12:10
tsdgeosdandrader: what does "being a KDE user" mean :D12:10
tsdgeosdandrader: still i should get that theme and not the builtin cursor, no?12:11
dandradertsdgeos, yes and no12:11
dandradertsdgeos, xcursor naming is a big hot mess12:11
dandradertsdgeos, as is there's not much of a standardization to it12:12
dandradertsdgeos, so code has to come up with a whole bunch of fallback names and what not for every single cursor shape12:12
dandradertsdgeos, as every theme like to give different names for the cursor shapes12:12
tsdgeosi guess one can't configure the cursor theme in unity7, right?12:13
tsdgeosthat'd be too much flexibility12:13
dandradertsdgeos, I've a branch that improves the fallback list of xcursor names quite a bit12:13
tsdgeossince you can't configure it, why not hardcode it to the one we want to use?12:13
Trevinhotsdgeos: it can be changed using external tools (such as gnome/unity tweak)12:13
dandradertsdgeos, ltinkl is about to report a unity8 bug about specifically this issue. that when he switches to breeze cursor theme unity8 doesn't pick up its cursors for some reason12:14
dandradertsdgeos, because even with this naming mess, we still should get most if not all of them12:14
ltinkltsdgeos: yup, just writing the BR (found out yesterday too)12:23
ltinkltsdgeos, dandrader: https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/151787812:30
ubot5Ubuntu bug 1517878 in unity8 (Ubuntu) "Non-default cursor themes fail to load" [Medium,New]12:30
dandraderltinkl, let someone else set the importance. reset it to Undecided12:32
dandraderltinkl, usually the team lead or manager sets it12:33
ltinkldandrader, greyback_ sorry :)12:43
dandradergreyback_, what's the silo number?12:43
greyback_dandrader: 4812:44
dandradergreyback_, fix will require a unity-api and  a unity8 branch12:45
dandradergreyback_, pretty simple ones, though12:45
greyback_dandrader: then we'll land separately12:46
dandradergreyback_, ok, pull it back from the silo then12:46
greyback_dandrader: for my info, what needs to be fixed?12:46
dandradergreyback_, mir mouse events are being sent from qtmir to the qml mouse pointer, so it decides what to do and synthesizes the Qt event, feeding the window where it lives with that event12:48
dandradergreyback_, wheel event should follow the same path12:48
dandradercutting corners didn't work out so well12:49
=== davidcalle_ is now known as davidcalle
=== dandrader is now known as dandrader|afk
=== alan_g is now known as alan_g|lunch
=== dandrader|afk is now known as dandrader
=== alan_g|lunch is now known as alan_g
=== nudtrobert1 is now known as nudtrobert
tsdgeosmzanetti: any idea what's up with the ci links not working?14:43
mzanettitsdgeos, jenkins completely down atm14:44
tsdgeosis it? i read something about being down for like 30 min only14:44
tsdgeosmust have misread14:44
mzanettihmm. let me check14:44
tsdgeoshttps://code.launchpad.net/~mzanetti/unity8/move-screenshots-to-tests/+merge/276798 just "finished" 34 min ago14:45
tsdgeosand all the links but the qmluitests ones work14:45
tsdgeoswhich is a bit fishy14:45
mzanettitsdgeos, for me all the links give a 40414:46
tsdgeosmzanetti: https://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-vivid-touch/5259/ works14:46
mzanettitsdgeos, this happens for me since 2 days now. I've pinged cihelp already but no reply so far14:46
tsdgeosit's the first on https://code.launchpad.net/~mzanetti/unity8/move-screenshots-to-tests/+merge/276798/comments/70359014:46
tsdgeosbroken CI is not very useful14:46
=== dpm is now known as dpm-afk
=== dandrader is now known as dandrader|afk
mterrymzanetti, sudo gdbus call --system --dest org.freedesktop.Accounts --object-path /org/freedesktop/Accounts/User32011 --method org.freedesktop.DBus.Properties.Get com.canonical.unity.AccountsService DemoEdgesCompleted15:41
mzanettimterry, (<['', 'left', 'bottom', 'top', 'right']>,)15:41
mzanettiis the '' intentional?15:41
mterrymzanetti, hrm... no?  depends, when setting it with tools, you need to specify an empty string, since none of the tools can handle an empty array!15:42
mterrymzanetti, but I doubt you did that15:42
mterrymzanetti, so that's A.  I'll look at that15:42
mterrymzanetti, but B, it looks like the tutorial THINKS you opened the right edge at some point15:43
mzanettiyeah... will try again15:43
mzanettimterry, is it enough to set the other key to false?15:44
mzanettior do I manually need to clear this one too?15:44
mzanettiqdbus --system com.canonical.PropertyService /com/canonical/PropertyService com.canonical.PropertyService.GetProperty edge15:44
mzanettithat one ^15:44
mzanettierm... /me tries again15:44
mterrymzanetti, uh no, not if you use my instructions15:44
mterrymzanetti, that is a meta property that will clear the other properties, if you use it15:45
tsdgeosmzanetti: https://code.launchpad.net/~mzanetti/unity8/move-screenshots-to-tests/+merge/276798 fails15:45
mzanettimterry, well, I've read through through the phablet-tools merge and saw that one being reset15:45
mterrymzanetti, yeah.  phablet-tools calls that, which then sets the two AS fields15:45
mzanettitsdgeos, ack, thanks, will check it15:45
mterrymzanetti, but my instructions don't rely on you having my phablet-tools branch, so they include instructions for resetting everything manually even if you have old tools15:46
mterrymzanetti, but if you do use those tools, you'll get an empty string in your list.  So maybe that's where it's from -- I'm guessing you cleared your settings before testing15:46
mzanettimterry, but the new phablet-tools branch doesn't do anything magic, it just drops the check, doesn't it?15:46
mterrymzanetti, i.e. my instructions leave you with an empty string15:46
mzanettimterry, I used the old phablet-tools (before I saw your branch) and it worked to enable it15:46
mzanettiat least for 3 eges :D15:47
mterrymzanetti, but there's another patch, not in VCS but in that silo (patch for dbus-property-service) that adds the extra clear for DemoEdgesCompleted15:47
mterrymzanetti, but if you also used my instruction for manually clearing DemoEdgesCompleted, you'll end up with an empty string15:47
mzanettioh, I see15:47
mzanettiok, so I did:15:48
mterrymzanetti, between the three tools, dbus-send, gdbus, and qdbus, none of them worked perfectly for setting an empty array15:48
mzanettiqdbus --system com.canonical.PropertyService /com/canonical/PropertyService com.canonical.PropertyService.SetProperty edge true15:48
mterrymzanetti, some couldn't do arrays, and those that could, couldn't do an empty one15:48
mterrymzanetti, did you install silo 33?15:48
mterrymzanetti, then the empty string is expected15:48
mterrymzanetti, dbus-property-service uses gdbus with a empty-string-array15:49
mzanettimterry, will that patch land too? or is that just for the silo?15:49
mterrymzanetti, because I couldn't find a better way to do it with shell!15:49
mterrymzanetti, that will land too15:49
* mzanetti reboots and hopes to see a right edge too :)15:49
mterrymzanetti, so normal users won't see the empty string.  But anyone that uses phablet-tools will15:49
mzanettimterry, ack, wfm15:50
mzanettimterry, ok. got the right edge this time... still a bit puzzled why I didn't see it at first15:53
mterrymzanetti, I hope it was just that you used the right edge and forgot...  :-/15:54
mterrymzanetti, it's easy to accidentally do15:55
mterrymzanetti, especially when you want to launch more apps -- I often go to the dash15:55
mterrymzanetti, and them hit my forehead  :)15:55
mzanettiprobably... yes15:55
tsdgeoscimi: maybe you can review https://code.launchpad.net/~aacid/unity8/listviewforreviews/+merge/277428 ?16:10
tsdgeoscimi: and https://code.launchpad.net/~aacid/unity8/unfavorite_scope_test/+merge/27715716:10
cimitsdgeos, ok16:10
tsdgeoscimi: also please remind me what's the satatus of https://code.launchpad.net/~cimi/unity8/new-shadows-1.3/+merge/27161116:12
tsdgeosis it good to go? or are we waiting on something?16:12
cimitsdgeos, just review... design is approved16:13
mterrymzanetti, greyback: is it expected that u8 seems to restart when plugged into a monitor?  it does the dot-progress thing, so it looks intentional, since it's not a reboot.  I just didn't expect that16:17
greybackmterry: known issue16:17
mterrygreyback, ok16:17
ubot5Ubuntu bug 1513909 in mir (Ubuntu) "unity8 crash on unplug of monitor" [Critical,In progress]16:18
mterryoh duh, the dot-progress is what I see when it crashes  :)16:20
mterryI forgot we had dots on that screen, and I helped make that screen16:20
tsdgeosmzanetti: now that ota8 is out are we planning a landing? have like a zillion approved branches16:21
mterrymzanetti, if you're in a reviewing-mterry's-branches mood, https://code.launchpad.net/~mterry/unity8/warn-on-xapp/+merge/277915 might be up your alley, since it's basically a copy of your dialog branch16:22
mzanettitsdgeos, silo 60 has just been approved by QA16:28
mzanettitsdgeos, once it's merged I'll put together the next one16:28
mzanettimterry, heh, ack16:28
tsdgeosmzanetti: cool16:29
tsdgeoscimi: can you merge unity8 into your branch? it merges fine but that way it'll also compile just after downloading :D16:29
* mterry has a mental image of mzanetti shoveling coal into a furnace to keep the silos pumping16:30
tsdgeosoh so we're landing the diff updates for scopes16:30
tsdgeosvery cool16:30
mzanettiwe are16:31
mzanettialthough pawel seemed to be happy that I'm landing it. he said it'll my fault if it blows up, which gives me a slightly uncomfy feeling about it16:32
mzanettiit was working great in my testing tho, and apparently QA didn't find issues either16:32
=== dandrader|afk is now known as dandrader
mzanettimterry, weird... the diff seems to show the prereq's prereq's changes, doesn't it?16:55
mterrymzanetti, no, that's all my changes16:55
mzanettithe isTouchApp too?16:56
mterrymzanetti, yeah -- that's how we know whether to show the dialog16:56
mterrymzanetti, isTouchApp was added for ApplicationInfoInterface16:56
mzanettimterry, yes, but didn't you add thos to the no-touch-no-lifecycle branch?16:56
mterrymzanetti, but this is for LauncherInterface (i.e. before app is launched)16:56
mterryLauncherItemInterface rather16:56
mzanettiohhh. right16:56
mterrymzanetti, I have an associated unity-api branch too16:57
mterrylinked in description16:57
mzanettimterry, I see... how would that apply to dash launches then?16:57
mterrymzanetti, good question, this branch doesn't address that right now, since dash never shows legacy apps (yet), right?16:58
mterrymzanetti, how does app scope launch apps now?  by itself?  or does it ask unity8 to launch them for it?16:58
mzanettiyeah, doesn't atm...16:58
mzanettimterry, Qt.openUrlExternally()16:58
mterrymzanetti, that's what I feared16:58
mterrymzanetti, so that means u8 isn't involved at all, and it would be difficult to show this same dialog for them16:58
mterrymzanetti, when we add support for showing legacy apps in dash, maybe we want to add a call to dashcommunicator?16:59
mterrymzanetti, but ultimately, any app can cause another app to open16:59
mterryvia openUrlExternally16:59
mterrynot just a dash problem16:59
mzanettiI'm wondering if we shouldn't rather add something to allow as rejecting an app with ApplicationManager16:59
mzanettimterry, will talk to gerry when he's on tomorrow17:00
mzanettimterry, however, your branch would still apply if we need to show the icons somehow differently17:01
mterrymzanetti, so putting logic in qtmir, u8 sees denial code from AppManager, and shows dialog?  could work.  But would have to be careful to be able to open the app up again when docked (so save arguments and such)17:01
mterrymzanetti, yeah LauncherItemInterface changes and all that still make sense17:01
mterrymzanetti, as does the dialog.  Just the connection of the two might need adjusting17:01
mzanettimterry, I'm thinking rather of qtmir emitting something like "iWantToStartApp(appId)" and we accept or reject it17:02
mzanettibut need to be careful to not slow down app startup time17:02
mzanettineeds a word with gerry definitely17:02
mterrymzanetti, yeah17:02
mterrymzanetti, right, of course qtmir couldn't accept/deny on its own17:02
mterrymzanetti, also... does qtmir even get involved before the exe is run?17:03
mterrymzanetti, we probably want to prevent the exe from being run at all if possible17:03
mterryto avoid any side effects17:03
mzanettiyep, u-a-l already asks qtmir for permission before launching the binary17:03
mterrymzanetti, oh neat17:04
mzanettiwe even have means of knowing that... the application item will be added to the model at that point already17:05
mterrymzanetti, I'm going to change the tutorial branch -- keep track of a separate bottom field for each app -- so "[top, bottom-dialer-app, bottom-messaging-app]" etc17:05
mterrymzanetti, oh ah.  So the LauncherItemInterface change will be purely potential-future-proofing17:05
mzanettiyep. but I'm quite confident we'll need it17:05
mterrymzanetti, would be nice to get alexm to weigh in17:06
mzanettiyes. putting a design chat regarding this on my todo list too17:06
mterrymzanetti, I also just made up the text for the dialog: "Dock your device to open this app"17:06
mzanettiok, seems not so straigth forward as a review either :D17:07
mterryAh well17:07
mzanettiwhat's the reason for changing the list (re tutorial)17:07
mterrymzanetti, design -- they want separate text for each app and they want the coach mark for each app (instead of once for any of them)17:08
mzanettiah, /me likes17:08
mterrymzanetti, they had been happy with just the dialer-app, but I had to open my big mouth  :)17:08
dandradermzanetti, https://code.launchpad.net/~dandrader/unity8/mouseEdgePush/+merge/276306 is not following the design spec17:38
dandradermzanetti, s/not/now17:38
mzanettidandrader, awesome!17:39
dandradermzanetti, it boiled down to just adjusting the gradient values. width was already right17:40
=== alan_g is now known as alan_g|EOD
=== dandrader is now known as dandrader|afk
=== dandrader|afk is now known as dandrader
=== charles_ is now known as charles
mzanettimterry, heh, regarding the width/height values. I didn't even realize they were swapped. The question was really why it's those huge integers?20:39
mzanettimterry, I guess it's the image file dimensions, but do we really want to set that for the sourceSize?20:39
mterrymzanetti, maybe I don't grok sourceSize or whatever.  But those are the sizes of the underlying png files20:39
mterrymzanetti, I thought it was an optimization to specify those20:39
mzanettimterry, no, it's the size you want to keep in memory. when it loads the file, it scales it to sourceSize in memory20:40
mzanettiit is used to prevent constant rescaling of the image20:40
mterrymzanetti, yeah, so that's good right?  like, screen might become bigger, so you keep the original size around20:40
mzanettimterry, I'd set it to the screensize then20:40
mterrymzanetti, but if they plug in a new monitor20:41
mzanettiif the image is bigger, we can save some memory20:41
mterrymzanetti, actually those images should ideally be rotated if they plug in a landscape monitor...20:41
mterrymzanetti, but design specified crazy specific orientations for those bg images (each one is technically different)20:42
mzanettimterry, right, I'd say plugging a monitor is seldom enough to rescale in that case, but save memory in the normal case20:42
mterrymzanetti, so I don't have guidance on the landscape issue20:42
mterrymzanetti, yeah.  OK so I drop the lines?20:42
mzanettiwell, they will be dropped from memory anyways after the wizard20:42
mzanettion a second thought it might even be better to keep the full size then20:43
mzanettinot sure. you have a point there20:43
mzanettiok, works for me then. makes sense now :)20:44
mterryyay for not making any changes  ;)20:46
mzanettigood I cought the wrong aspect ratio then :D20:46
mzanettipure luck20:47

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