=== _salem is now known as salem_ === salem_ is now known as _salem [07:27] Saviq: hi! I'm testing unity8 trunk, which should be improving the testsuite, but I'm getting the usual 21/22 errors with phablet-test-run -n unity8 - is this correct, ie. are the improvements to be found outside unity8's AP tests, or how is it? [07:28] Mirv, no, no unity8 test improvement, unless you run each test separately [07:28] Mirv, second and subsequent test in a run don't get input for some reason - dunno why yet [07:28] Mirv, but it will get webbrowser app tests to 100% [07:28] Mirv, and probably some others, too [07:28] Saviq: aha, webbrowser_app 100% to be expected, check! :) I'll try out [07:29] Mirv, with the newly released unity-mir, that is [07:29] I like how it nowadays recovers after -n run to normal operation [07:29] Saviq: yep, I've #90 which just build + unity8 only updated [07:30] I still need network over usb since I can't type the password to wlan dialog [07:30] Mirv, another important bug to test https://bugs.launchpad.net/unity8/+bug/1234903 [07:30] Ubuntu bug 1234903 in telephony-service (Ubuntu Saucy) "[mir-only] On incoming calls the greeter remains and dialer-app is not started" [Critical,In progress] [07:31] Mirv, ah, but that requires telephony service and dialer-app [07:35] Saviq: I can't test that, but I'm happy with no regressions and if that webbrowser AP truly works it's awesome [07:36] the latest #90 test shows 36/39 failing on webbrowser, so let's see with updated unity8 [07:48] Saviq: success, great [07:52] Cimi: Saviq: what you say on https://code.launchpad.net/~unity-team/unity8/music-video_renderers/+merge/190198/comments/436671 ? [08:03] oh my, the instructions on https://bugs.launchpad.net/unity8/+bug/1237208 are a huge "what???" [08:03] Ubuntu bug 1237208 in Unity 8 "Launcher can get into an unusable state when attempting to unlock the screen." [Undecided,In progress] [08:03] lots of steps :D [08:14] Saviq: published current trunk. === iahmad is now known as iahmad|afk [08:24] Mirv, awesome, thanks! [08:25] tsdgeos, I agree [08:26] tsdgeos, Cimi, we can have a BaseTileStyle that the others inherit from - even using Loaders like the Preview does, WDYT? [08:27] Saviq: i'd say it makes sense to me [08:28] Saviq: we decided we merged manually, right? [08:29] tsdgeos, yes [08:29] ok [08:29] Saviq: so i'm going to test https://code.launchpad.net/~larsu/unity8/use-volume-action/+merge/190292 now that needs a new indicator-sound, what's up with that? [08:29] i.e. when do we merge that in? [08:30] wait for the indicator-sound to be released? [08:30] tsdgeos, I'm on this, sorry [08:30] tsdgeos, no, we merge [08:30] ah [08:30] ok [08:30] you do it then [08:30] * Saviq claims [08:30] done [08:31] tsdgeos, if you could look at https://code.launchpad.net/~saviq/unity8/fix-frequent-app-ids/+merge/190161 [08:31] sure [08:31] tsdgeos, I have a package for url-dispatcher - but it didn't work for me [08:31] i'll create my own [08:31] just in case [08:31] tsdgeos, i.e. http://pastebin.ubuntu.com/6216036/ [08:31] Saviq, we're still on manual merge-mode, if we top-approve a unity8 branch, right? [08:32] MacSlow, yes [08:34] tsdgeos, OTOH [08:34] tsdgeos, Cimi, we need to take a step back on all the Tiles anyway [08:34] why? [08:34] tsdgeos, 'cause they're a mess ;) [08:34] right [08:35] that's what i was complaining about :D [08:35] tsdgeos, there's no common API, they just all have the same props [08:35] tsdgeos, yeah, but those changes probably don't warrant a rehaul that we need [08:35] especially not on freeze day ;) [08:35] sure [08:36] tsdgeos, Cimi, so go for it - but with reason [08:37] larsu, will you be bumping indicator-sound version for the "volume" action? [08:37] Saviq, anything I can do to persuade someone to look at the ext. snap-decision part2 branches... the wifi-selection is sitting there working for two weeks now :) [08:37] MacSlow, thing is... no one would be using it anyway [08:38] MacSlow, so I'd rather leave it post-release, k? [08:38] MacSlow, basically I'd like to refrain from merging stuff that won't show in 13.10 at all [08:38] Saviq, oh... I thought it was right after simunlock priority-wise... [08:38] Saviq, nm then [08:38] MacSlow, not that I know of - please correct me if I'm wrong [08:39] MacSlow, but I didn't think there's anything that would trigger it currently [08:39] Saviq, well there's been so much up/down/left/right with ext. snap-decisions that I might got things mixed up... [08:42] mzanetti, ping [08:42] Saviq, tsdgeos I don't understand [08:42] Saviq, tsdgeos this is our toolkit in the end [08:42] Cimi, it's not about that [08:42] Cimi, it's the fact that *Style.qml are almost the same [08:43] Cimi, where they can be inheriting from a BaseTileStyle.qml or something [08:43] Saviq, you won't save much [08:44] few are not using ubuntushape [08:44] few have multiple lines [08:44] only reused stuff is the properties [08:45] Cimi, which is one thing we should make common [08:45] which I didn't love it either [08:45] Cimi, either way - I'm not hung up on it, but we do need a rehaul at some point in the near future [08:45] Saviq: com.ubuntu.developer.webapps.webapp-facebook/webapp-gmail ? [08:45] facebook+gmail? [08:45] tsdgeos, wrong, of course [08:45] Saviq, only thing I thought is to rebase them under this base stylebla bla [08:45] Saviq, which is an empty item with a set of properties [08:46] tsdgeos, pushed [08:46] Cimi, that's not necessarily the case [08:47] Saviq: you can have a complete tile with all the stuff in it, and then enable/disable them as needed, or have building blocks that help you building the tile without that much c&p, no? [08:47] Saviq, only common thing always is one label [08:47] Cimi, if we split the Tile into image, title and subtitle [08:47] Saviq, we could make the description follow that label [08:47] Cimi, and use Loaders for all three [08:47] Cimi, we can mix'n'match as needed [08:47] Saviq, isn't too many loaders? [08:48] Cimi, not necessarily worse [08:48] Saviq, the delegate itself is a loader [08:48] Cimi, either way - not now, ok? [08:48] ok [08:49] Cimi, tsdgeos, if you can think of a non-invasive way to abstract the props for now - go for it, but let's not do a whole rework today [08:50] Saviq: same error with the facebook thing [08:50] Saviq: maybe the url is not the right one? [08:51] tsdgeos, that's what ted told me yesterday, but if you can look at http://bazaar.launchpad.net/~indicator-applet-developers/url-dispatcher/trunk.13.10/revision/34 [08:51] tsdgeos, and see what we should use instead [08:51] yeah [08:51] on it [08:52] tsdgeos, url-dispatcher appid://com.ubuntu.calculator/calculator/0.1.3 should work, AFAIK - but doesn't [08:52] greyback, o/ back to work? [08:52] Saviq: yep [09:04] greyback, got one for you: bug #1237850 [09:04] bug 1237850 in Unity 8 "Apps started via upstart don't participate in window management correctly" [Critical,New] https://launchpad.net/bugs/1237850 [09:04] greyback, please confirm the behaviour on latest image + dist-upgrade [09:04] Saviq: yessir [09:05] * Saviq reboots [09:10] Saviq, did you read what I was proposing yesterday for the carousel with small item count? === iahmad|afk is now known as iahmad [09:18] Saviq, that's the same as launching apps from indicator menus... [09:19] system settings opened from the top indicators behave bad [09:22] om26er, yeah - they're launched via upstart [09:22] om26er, same for anything that goes through url dispatcher [09:23] Saviq, right. Important bug indeed. [09:23] hope we get the opening animation thing sorted for this case as well :) [09:23] Cimi, I think I did, but it was too late for me to understand [09:23] Saviq, you have backlog? :) [09:23] I don't :P [09:25] Saviq, getRenderer will choose a root Item that will load the renderers [09:25] Saviq, we will send this item the rendererId [09:25] Saviq, internally it will decide the final renderer to load? [09:26] mhr3_, re your "let's not do count-based things..." carousel behaves really bad when there's not enough items [09:26] Saviq, are these the safest flash flags to use currently? phablet-flash ubuntu-system --channel devel-proposed [09:26] nic-doffay, yes [09:26] Saviq, cool ta [09:27] nic-doffay, well, safest maybe not - but the ones that you should be using ;) [09:27] errrr [09:27] * tsdgeos weirds out [09:28] Saviq: if i compile my own url-dispatcher it works [09:28] tsdgeos, HUH [09:28] so building the package somehow breaks it? [09:28] ¿? [09:28] where's THE TED when you need'im [09:29] hmmm [09:29] actually not that [09:30] if i run url-dispatcher as my user it works [09:30] Saviq, so? :) [09:30] Saviq, do you have better ideas? [09:30] * tsdgeos is confused [09:30] Cimi, you'll have to translate that to me ;) [09:30] Saviq, or let the carousel file load [09:30] * tsdgeos reboots the phone [09:31] Saviq, getRenderer is a function we use in the scope view to load the renderer [09:31] Saviq, but we don't have the count yet, the renderer has it [09:31] Cimi, yes [09:31] Saviq, the renderer doesn't know the rendererId because it's not sent [09:32] Saviq, so if we give the renderer the rendererId property, we can decide later [09:32] o/ [09:32] hey mzanetti [09:32] Saviq, otherwise we could load this from the carousel [09:32] ciao mzanetti [09:32] Cimi, well, don't we know the count? [09:32] Saviq, we know the count only from the renderer I think [09:32] Saviq, model.count is the count of categories [09:32] Cimi, no, we know the count from the *model* [09:33] Saviq, and you said this is not during search [09:33] not dinamic [09:33] Cimi, yes, but each category has a results model [09:33] Saviq, is this dynamic? [09:33] Cimi, how else? [09:34] Cimi, it's the model we take the actual results from [09:34] Saviq, is it worth all the flickering and weird switches when adding results will be slow? [09:34] Saviq: i'm ultra confused now, it works fine after a reboot :D [09:34] tsdgeos, oh, that's interesting [09:34] Saviq: maybe you did not reboot the phone after installing your packages and me neither? [09:34] tsdgeos, right of course... [09:34] Saviq, my point is just that at this point i wouldn't want to open that can of worms [09:34] and the old url-dispatcher was still there? [09:34] tsdgeos, it's a dbus service [09:35] mhr3_, on one hand I know what you mean [09:35] mhr3_, but on the other you can't even activate anything in the carousel if there's only one result :/ [09:35] or count < 5 or something [09:35] Saviq: tbh i'd expect the package insatll to start/stop the service, like the apache packages do [09:35] Saviq, fix carousel then? :) [09:35] tsdgeos, it probably should indeed [09:35] mhr3_, it's a feature, not a bug :P [09:36] Saviq: anyway, works, so approving + mergin [09:36] Saviq, of course it is :P [09:36] tsdgeos, cool [09:36] mhr3_, it was never meant to support a low number of items [09:36] Saviq: did I overhear correctly the other day that there was a branch for unity that allowed an application to be launched so that it pops up over the greeter (i.e. for testing apps)? [09:36] veebers, already merged [09:36] veebers, just launch an app - it'll swipe the greeter away [09:36] Saviq: oh awesome, is there any special magic needed to get it working? [09:36] Saviq: ah, this is using upstart to launch it? [09:37] Saviq, i don't buy that really, just make it static when there's no enough items [09:37] veebers, well, anything, really... should work with --desktop_file_hint, too [09:37] so one can't just do: qmlscene test.qml . . .? [09:37] oh cool [09:37] MacSlow: changed the simunlock example. [09:37] dednick, taking a look now... [09:38] veebers, there's a BUT [09:38] BUTT [09:38] veebers, that's not gonna stay like that - we can't be unlocking the phone just because an app was launched [09:38] double T. thats a big one. [09:38] Saviq, we also said that when searching we never use the carousel [09:39] Cimi, that's right! mhr3_ ↑ [09:39] Saviq: agreed! [09:39] Saviq: oh lol, look at the commit log, we got a release from 420 with changelog commited at 423 [09:39] 424 [09:39] Saviq: thanks for that I'll have a play around now [09:39] wonder if that's going to break later [09:39] Saviq, Cimi, what makes you think surfacing is completely static? [09:39] tsdgeos, yeah, it's tricky [09:40] tsdgeos, no, it's on purpose [09:40] just try to copy media over mtp and you'll see it isnt [09:40] Saviq: ok [09:40] tsdgeos, if stuff merges between merger.start and merger.finish [09:41] tsdgeos, now even more - 'cause it takes longer for manual publishing - and that's when the changelog is synced back [09:42] mhr3_, yeah, I understand what you mean [09:42] but am not convinced that this is "fixing" the carousel at all :/ [09:43] Saviq: hmm, doesn't seem to work for me, perhaps I don't have the right version. I'll ask someone to check it out over my night time. [09:43] veebers, not released yet indeed [09:44] veebers, bazaar.launchpad.net/~unity-team/unity8/trunk/revision/422 [09:44] Saviq, ultimately what i'm saying is that taking count in account when deciding on renderer is going to cause issues, and i'd rather deal with those at some other time than 1week before release [09:44] Saviq: ah ok, I'll let some others know to expect it etc. [09:44] awesome, thanks again Saviq [09:45] dednick, I'll do a test on the device before I merge [09:50] dednick, something is broken... I just merged your unity8 branch with trunk and notifications no longer work... [09:50] Saviq, anyway, can we merge cimi's grid fixes? functionally it's 100% according to design now [09:51] and will leave me with just one gripe - video carousel [09:51] mhr3_, you said not [09:51] mhr3_, not on the tablet [09:51] ? [09:51] mhr3_, tsdgeos and Cimi are on it [09:51] Cimi, i said it's weird in tablet mode [09:51] Cimi, but there's no design for tablet [09:54] MacSlow: what do you mean by "not work"? as in nothing? [09:54] dednick, yes... no notifications at all [09:54] dednick, just checking plain lp:unity8 and lp:unity-notifications ... [09:54] um, well i didnt change that much in unity8. doesnt sound right [09:55] dednick, your set of branches worked before I merged trunk to your unity8 branch... [09:56] Saviq, tsdgeos mhr3_ let me first "fix" the tablet [09:56] Cimi, as you wish, but phone-wise it's as specced [09:56] mhr3_, don't like half baked stuff [10:00] MacSlow: you testing on desktop? trunk merged works for me. You remember to kill notify-osd? [10:00] dednick, no... I was testing too many things at once... [10:01] dednick, had still another unity8-shell window open that was hidden behind other windows... [10:01] dednick, works fine with trunk merged... now testing on thed evice [10:01] * MacSlow wipes sweat away [10:12] ok, now it better on the tablet mhr3_ tsdgeos [10:14] yep, looks better now [10:14] tsdgeos, I'd probably elide right for video too [10:14] label [10:14] Saviq, having issues with read only folders after the flash, what's the fix again? [10:15] nic-doffay, dab shell touch /userdata/.writable_image [10:15] nic-doffay, adb [10:15] nic-doffay, then reboot [10:16] Cimi, anything that could be done about the text wrapping in http://imgur.com/skPdL9M ? [10:16] Cimi, cheers [10:16] mhr3_, yeah, using word-wrap now [10:17] mhr3_, but resident evil screws [10:17] Cimi, yea, that's why i'm asking [10:17] is it possible to word wrap + ellipsize? [10:17] mhr3_, wrap yeah [10:19] mhr3_, Cimi no, we need to force wrap in that case [10:19] Saviq: any bugs floating about? [10:19] Cimi, so Text.Wrap [10:19] Saviq, yeah [10:19] Saviq, wasn't thinking of someone insane to have such long names [10:19] Cimi, we should be using Wrap everywhere , TBH [10:20] dednick, there't this one assigned to me I haven't gotten to yet: https://bugs.launchpad.net/unity8/+bug/1092976 [10:20] Saviq, yeah [10:20] Ubuntu bug 1092976 in Unity 8 "[shell] shouldn't animate lenses switching when opening dash from locked state" [Low,Triaged] [10:20] pushed [10:21] dednick, can't run your unity8 branch merged with trunk on the device right now... getting build-errors regarding to Qt5XmlPetternsConfig... no clue right now... trying to dig into it [10:21] MacSlow: build-dep? [10:21] Saviq, having more issue trying to run on device after flash ,http://pastebin.ubuntu.com/6217416/ [10:21] any clues? [10:21] mhr3_, so, veto from your side on dynamic renderers? ;) [10:21] nic-doffay, guess [10:22] Saviq, for this week, yes [10:22] nic-doffay, why do you usually have network problems? [10:22] nic-doffay, deploy network [10:23] dednick, that exists with errors [10:23] Saviq, he lives in a cave like MacSlow :P [10:23] mhr3_, thing is... "fixing" carousel is not gonna be easy [10:23] mhr3_, it's not about just making it static - you'll be clicking items behind other items :/ [10:23] mhr3_, we'd have to spread it or something [10:23] Saviq, isn't the -s supposed to set everything up? [10:23] nic-doffay, no, not network [10:23] Saviq, ah [10:24] nic-doffay, CODING: - make sure your device is connected to a network [10:24] because the device keeps on switching networks! [10:24] nic-doffay, phablet-network [10:24] nic-doffay, it deploys your laptop network on the device [10:24] Saviq, i don't think it's so critical for now, can remain in current state imo [10:25] mhr3_, ok, it's on you ;P [10:25] my phone seems to love switching to open networks which dont actually work [10:25] dednick, my phone gets stuck in a boot loop when it's plugged in on start. [10:25] £$£"Q$ problems [10:25] Saviq, afterall simple fix - copy more media :) [10:26] dednick, Saviq: any idea http://pastebin.ubuntu.com/6217428/ [10:27] MacSlow: dist-upgrade. [10:27] MacSlow, yeah, update at least [10:27] Saviq: so i guess for the tile's we're good as it is and we'll look for a "better" solution later? [10:27] tsdgeos, yeah [10:27] * greyback moving to office [10:28] Cimi: ok, so what are the differences i should be seeing? [10:28] tsdgeos, where? [10:30] Cimi: with your tile changes [10:30] https://code.launchpad.net/~unity-team/unity8/music-video_renderers/+merge/190198 [10:30] tsdgeos, look at music and video lens [10:30] what's the difference of old vs new i should see on the phone [10:30] is there a list of how it should look [10:30] tsdgeos, https://docs.google.com/a/canonical.com/presentation/d/1X3_G6reCpcFJxzxn1Xtv9wK5n3-wOW-6z2l0qSZ8-sg/edit [10:30] tsdgeos, here is a spec [10:30] ok [10:30] tsdgeos, https://lh5.googleusercontent.com/YJsbwT4BhY1HuNX9vaLwtOruPFxSRVWs8zIcVoqHUTLFdi4ETM3SUDrJKb0sHn3AOoh7TFnGSos7MidzXW5fbgKLwcfYQRDjOXvUU3R8be6s-z4dYfLLSUM1Nts [10:31] tsdgeos, https://lh6.googleusercontent.com/DiFH5zJxUPhXES9dICX4ydlmQzpkKFMNeJlq3NsPhPIUAzbAH8XjX4YlZ0eii1w5csbAX2ZqnW9l6bjLSDzzd8Y1oCJf2qKzJYZ49lDeg5dPDbCiCEI-3eyi8Fs [10:31] tsdgeos, these are the changes - http://imgur.com/qAgx9jD [10:31] Saviq: https://bugs.launchpad.net/unity8/+bug/1092976 . Doesnt seem to reset to home lens anymore. My unlock just stays on current lens. Is supposed to go to home? [10:31] Ubuntu bug 1092976 in Unity 8 "[shell] shouldn't animate lenses switching when opening dash from locked state" [Low,Triaged] [10:32] mhr3_, that wallpaper looks so bad on the phone [10:33] dednick, tap on Dash icon in launcher [10:33] dednick, should go to Home indede [10:33] Cimi, meh, some aspects of convergence are good, some are not [10:33] *indeed [10:33] ahh. i should read... [10:34] dednick, yes! [10:36] Saviq, I think there's still an issue with the setup and permissions: http://pastebin.ubuntu.com/6217448/ [10:37] nic-doffay, that's permissions on *your* host [10:37] Saviq, yeah [10:37] Saviq, will it affect the setup? [10:37] nic-doffay, the script doesn't touch those - you need to fix perms on ~/.ssh/known_hosts [10:37] nic-doffay, it might [10:37] nic-doffay, the "cannot remove..." is fixed by tsdgeos's branch [10:38] hmmm [10:38] now that you say, i'm not sure i removed the delete [10:38] of the file [10:38] nic-doffay: anyway, appreciated if you or anyone else review the branch :D [10:43] tsdgeos, sure [10:43] tsdgeos, if you send me a link I'll get to it today. [10:43] tsdgeos, shall we merge the unrevert and thus, force the fix to Qt? ;) [10:43] Cimi: doesn't it look a bit weird for music with no title (or author, not sure what is missing) [10:44] Cimi: want a photo of the thing? [10:44] Saviq: up to you :D [10:44] tsdgeos, no album [10:44] tsdgeos, well, let's keep it ugly [10:44] Cimi: why? [10:44] tsdgeos, so you learn to add title [10:44] *album [10:44] you're joking [10:44] tsdgeos, nope [10:45] Cimi: ok [10:45] tsdgeos, also [10:45] then get someone to approve your branch [10:45] and get someone to fix the internet [10:45] tsdgeos, I think it's important to keep same position [10:45] tsdgeos, so you know the first row is always album [10:45] tsdgeos, and second row artist [10:45] because i'm not sure how smartness will teach me to fix the "popular online" albums not being there [10:45] shall i hack into whatever is giving me the data to make it look nice? [10:46] or can i have a phone that doesn't give me ugly stuff ? [10:46] i'd prefer the second [10:46] tsdgeos, will still be ugly imho [10:46] tsdgeos, even if you move the text one line above [10:46] mhr3_: hey hey! [10:46] tsdgeos, it's still a mess without the album ttle [10:46] Cimi: so you say everything is in the same line? [10:46] alecu: are you around for a nice regression game as well? ;) [10:47] tsdgeos, I'd say artist is always second line [10:47] Cimi: it's not if the album has two lines, no? [10:48] tsdgeos, it's always second in priority [10:48] tsdgeos, so you learn that the line is empty because the album is empty [10:48] tsdgeos, not because the album title is "Bob Marley" [10:48] didrocks, uh oh [10:48] Cimi: can't i learn that because it's non bold? [10:49] didrocks, what did i do? [10:49] tsdgeos, it's an easy fix though, just putting everything inside a column [10:49] mhr3_: a lot of mean things! just back and you are already in the line for shooting :p [10:49] mhr3_: more seriously, it seems that between image 89 and 90, we have "more suggestions" category randomly disappearing [10:50] (for the apps) [10:50] tsdgeos, but I hate we're using services that don't give us data [10:50] didrocks, fix for that was merged in click scope yesterday [10:50] tsdgeos, it'll always gonna look ugly [10:50] mhr3_: do you have any idea where we should start having a look? I don't have any crash [10:50] oh oh [10:50] let me look [10:50] didrocks, rev73 [10:51] mhr3_: excellent! [10:51] didrocks, but updating click means removing unity-lens-applications [10:51] otherwise there'll be duplicates [10:52] mhr3_: removing, like, removing? [10:52] dednick, Saviq: while testing current branches lp:~nick-dedekind/unity-notifications/actionless-menumodel-snapdecision lp:~nick-dedekind/unity8/simunlock.dialog I saw the osk not receiving any taps... but they all fell through onto the window below. [10:52] or taking latest u-l-a? [10:52] didrocks, yes, like removing from the image [10:52] MacSlow, window below or the shell? [10:52] Cimi: honestly http://picpaste.com/pics/IMG_00000598_edit-HLXvhQjK.1381402335.jpg looks weird as hell to me [10:52] didrocks, click does it's job fully now [10:52] hum [10:52] unity8 deps on it, right? [10:53] Saviq, no... it fell onto the shell... and not been consumed/seen by the osk [10:53] tsdgeos, looks bad [10:53] didrocks, ubuntu-touch does [10:53] tsdgeos, I'd say let's fix visually but not now [10:53] only recommends, so good enough for now [10:53] ok, I can handle it [10:53] didrocks, 135 in landing plan [10:53] tsdgeos, we have to fix the services [10:53] Saviq, taps on notifications work... but osk not [10:53] thanks mhr3_ for those infos :) (still catching up :p) [10:53] tsdgeos, we cannot give those "popular" [10:53] MacSlow, fixed already [10:53] Cimi: ?¿ [10:53] MacSlow, which image number? [10:53] tsdgeos, let's keep the crap in front of the users [10:53] tsdgeos, so the bug is clearly visible [10:54] which bug? [10:54] tsdgeos, Pixies - ... [10:54] Saviq, I flashed yesterday... and just did dist-upgrade... so I guess 89 [10:54] Saviq: it's the internet :P [10:54] tsdgeos, that's wrong - Pixies should be in a separate metadata field [10:54] tsdgeos, 3 out of 6, 50%, have either no album art or no album title! [10:54] Saviq, where do I check the image again? [10:54] tsdgeos, are they kidding? [10:54] tsdgeos, we need to fix the online scopes [10:54] MacSlow, /etc/media-info [10:54] tsdgeos, it can't be that popular songs are so damn broken [10:55] Saviq, "Ubuntu Saucy Salamander (development branch) - armhf (20131009.1)" [10:55] tsdgeos, also, those songs don't seem that popular :P [10:55] MacSlow, that's yesterday's - upgrade [10:55] Cimi: so you're arguing that the ui is fine? [10:55] tsdgeos, yes [10:55] tsdgeos, I said it looks bad [10:55] ok [10:55] tsdgeos, the Artist - Album comes from the scopes [10:56] tsdgeos, where it should come in separate fields [10:56] mhr3_: but "Handle non-click apps correctly in scopes" seems quite large [10:56] tsdgeos, but I think that keeping that ugly will push people to fix the scopes [10:56] Saviq: sure, i'm just saying that we should not have that whitespace in there [10:56] mhr3_: did you test the transition yourself? [10:56] but if you guys agree [10:56] i'm fine [10:56] didrocks, of course i did [10:56] tsdgeos, I think we should - 'cause that's a required field [10:56] tsdgeos, it's fixable if I put the two labels in a column [10:57] tsdgeos, if album is empty, artist goes top [10:57] didrocks, also ultimately it's +800, and -5500 (for ula) [10:57] tsdgeos, super easy [10:57] didrocks, and removed code... yey! :) [10:57] ahah, you want to sell it more! ;) [10:57] Saviq: ok [10:57] ok, let's coordinate that with sil2100 + I [10:57] mhr3_, should we be using metadata['album'], metadata['artist'] for that: http://picpaste.com/pics/IMG_00000598_edit-HLXvhQjK.1381402335.jpg [10:57] I think it's a good one for 92 [10:57] Cimi: it's fine, Saviq says it's a required field so let it be empty + ugly for now [10:58] Saviq, those are from grooveshark, it doesn't fallback to metadata [10:58] MacSlow, any reason why you not top-approved/merged "fit LazyImage"? [10:58] mhr3_, so we need to be fixing the online scopes there [10:58] Saviq, because I'm busy with dednick's branches as they seem more important right now [10:59] Saviq, i suggested removing grooveshark altoghether, they don't have good mobile-friendly site yet [10:59] MacSlow, k [10:59] Saviq, I want to get back to LazyImage once done with that [10:59] Saviq, but the checkboard pattern looks good, no? :) [11:00] mhr3_, ;) [11:00] that default image is ugly, TBH [11:03] Mirv, what do we do to get the Qt fix in? [11:04] Saviq: convince didrocks/asac [11:04] Saviq: I ran all the AP:s on non-Mir without regressions, and marked that down to the sheet [11:05] thats 154? [11:05] (which fix?) [11:05] https://bugs.launchpad.net/ubuntu/+source/qtdeclarative-opensource-src/+bug/1236765 [11:05] Ubuntu bug 1236765 in qtdeclarative-opensource-src (Ubuntu) "[PATCH] Fix infinite loop in QQmlIncubator::forceCompletion" [Undecided,In progress] [11:05] asac, didrocks ↑ [11:05] asac: didrocks yes 154 [11:05] Saviq: anything else committed on that branch? [11:05] Saviq: https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0Au6idq7TkpUUdGNWb0tTVmJLVzFZd0doV3dVOGpWemc#gid=1 row 154 [11:05] i think its a friday thing [11:06] asac, what's "a friday thing"? [11:06] asac, unity8 will be locking up without that [11:06] Saviq: is this is a mir-by-default regression? [11:06] Saviq: what's the condition to trigger this infinite loop? [11:06] didrocks, one we know is open/close/open/close dash plugins in unity8 trunk [11:06] Saviq: we dont land things that is not a mir-by-default regression before we have mir under ccontrol... earliest tomorrow [11:07] aka friday thing [11:07] anyway, give details to didrocks etc. [11:08] k [11:11] Saviq, still... even with image 91 I have the osk not working === MacSlow is now known as MacSlow|lunch [11:12] MacSlow, define "not working" please - and file a bug if you can reproduce reliably [11:13] MacSlow|lunch, but it was fixed in http://bazaar.launchpad.net/~unity-team/unity8/trunk/revision/409 [11:13] MacSlow|lunch, bug [11:13] https://bugs.launchpad.net/bugs/1236773 [11:13] Ubuntu bug 1236773 in Unity 8 "[mir] Need an input trap in shell behind the OSK" [Critical,Fix committed] [11:13] Saviq, any idea who aside from tmoenicke is familiar with the OSK? [11:13] just upgraded to image 90. seems quite sluggish. is that happening for everyone or just me? [11:13] nic-doffay, dandrader, to some extent [11:13] nic-doffay, whassup? [11:14] mzanetti, it gets worse [11:14] didrocks: I clarified that I did not only run unity8 AP:s, but all AP:s we run for images, for qtdeclarative. but obviously with surfaceflinger. [11:14] mzanetti, https://bugs.launchpad.net/ubuntu/+source/mir/+bug/1235190 [11:14] Ubuntu bug 1235190 in mir (Ubuntu Saucy) "[mako] Scrolling and animations are very slow with Mir" [High,Confirmed] [11:14] Saviq, it's regarding the https://code.launchpad.net/~nicolas-doffay/unity8/dismiss-keyboard/+merge/188599 [11:14] There appears to be an issue with the InverseMouseArea and OSK. [11:14] right... yeah. seems like this. (used to be quite snappy when I manually enabled mir in the last few weeks tho) [11:14] Since the behaviour issues only occur on the mobile and not the desktop. [11:15] Saviq, at least that's my assumption. [11:15] mzanetti, yeah, regression [11:15] Mirv: ok, I think we should start running some with Mir as well (once we have the unity8 which is > 0% ;)) [11:15] I'd like to get a second opinion on it. [11:15] nic-doffay, any more details? [11:15] didrocks, unity8 is >0% - 1 test always passes! ;P [11:15] Saviq, other than what's in that MR, not right now. [11:15] Saviq: the "start unity8" one? :p [11:15] didrocks, but for real, in utah it should be fine now - if I run a single test at a time for unity8 [11:15] didrocks, they all pass [11:15] Saviq, this is why I'd like to bring it up with someone who is more familiar with the components. [11:16] nic-doffay, yeah, but what is broken? [11:16] Saviq, check Mirco's comment. [11:16] fourth from the bottom I think. [11:16] Saviq: I tried btw http://pastebin.ubuntu.com/6217575/ but didn't have much joy [11:16] Saviq, it appears that the InverseMouseArea doesn't function well on the mobile. [11:17] But on the desktop it works every time and never runs into the issues Mirco listed in his comment. [11:17] Mirv, stopped powerd first? [11:17] (I had similar issues too) [11:17] As did dednick [11:17] I've tested this pretty extensively on both platforms. [11:17] Mirv, we started clearing the socket late yesterday [11:17] Mirv, before every tests [11:17] -s [11:18] Mirv, I'll try that with unity8 trunk [11:18] Saviq: crossing fingers! [11:18] Cimi: hey, the Hud button now comes in way to easily. are you aware of that/working on it? [11:18] nic-doffay, did you try adding rectangle? [11:18] nic-doffay, colored semi transpaernt? [11:18] Cimi, you can't add a rectangle to an InverseMouseArea. [11:18] nic-doffay, to see where the mouse areas really are? [11:19] It will just show up where the anchors are set. [11:19] nic-doffay, yeah [11:19] Cimi, that's not the problem. [11:19] nic-doffay, anchors fill the inverse mouse are from outside [11:19] Cimi, as I said there are only issues on the mobile. [11:19] Saviq: no, I didn't know stopping powerd [11:19] The desktop functions as expected. [11:19] mzanetti, it works as design [11:19] mzanetti, swipe less :) [11:20] Cimi: no... it comes in when I touch the screen more than 1 cm away from the edge [11:20] that can't be by design [11:20] mzanetti, that's not tru [11:20] e [11:20] mzanetti, threshold is units.gu(8) [11:20] well [11:20] which is 1 cm indeed [11:20] so it's fine [11:20] that's not fine... it's annoying as hell [11:20] mzanetti, it should appear after the toolbar is locked [11:21] Cimi: you can bring in the hud button without bringing in the toolbar [11:21] mzanetti, lets fix the sdk then [11:21] god now [11:21] -w [11:22] Cimi: if you use the dialer app the hud comes in when you press the dial button. [11:22] mzanetti, if you don't like design, complain with designers [11:22] Cimi, there's one more issue - when the HUD button is up, tapping anywhere on the screen will kill the toolbar [11:22] Saviq, as design [11:22] afaik the Hud should come in when swiping from the bottom edge [11:22] Cimi, no, it's not [11:23] Cimi, it's Oren that brought this to me yesterday [11:23] Cimi, the hud button should go away, yes, but not the toolbar [11:23] Saviq, I think he said me the opposite [11:23] Saviq, but ok [11:23] Saviq, I'll change it [11:23] Cimi, https://bugs.launchpad.net/ubuntu-ux/+bug/1130446/comments/7 [11:24] Ubuntu bug 1130446 in webbrowser-app "[Toolbar] Bottom edge swipe is detected 8gu away from edge" [Undecided,New] [11:24] Cimi: Saviq: merged [11:24] Cimi, please sync with him on mzanetti's issue, too [11:24] tsdgeos, thanks! [11:24] * mzanetti noted that people at the DevDays were very confused about the hud button coming in that easily [11:25] swiping it in from the bottom edge is ok, but not when starting the gesture somewhere *inside* the app [11:25] Cimi: ^ [11:25] mzanetti, mmm [11:26] mzanetti, tell me a testcase [11:26] Cimi: open the dialer up [11:26] mzanetti, and I'll fix it [11:26] and start dragging from the dial button upwards [11:26] mzanetti, you're right [11:26] mzanetti, ugly [11:26] yep [11:27] MacSlow|lunch: when you get back could you try opening the u1 app on maguro and signing in please, let me know if you can get the keyboard back [11:30] Saviq: another thing that I noted really badly while using *only* the ubuntu phone for the whole conference is the clocks running out of sync [11:30] Saviq: anything in progress about that already? [11:30] Saviq: ok getting somewhat better results now, the tests are visibly running and some are successing === hikiko is now known as hikiko|lunch [11:30] mzanetti, https://bugs.launchpad.net/unity8/+bug/1194823 [11:30] Ubuntu bug 1194823 in unity8 (Ubuntu) "Standby screen clock lags behind panel clock" [Medium,Triaged] [11:30] mzanetti, I made a beautiful spec / blueprint https://blueprints.launchpad.net/ubuntu-ui-toolkit/+spec/time-component [11:30] mzanetti, which is, as of now, ignored [11:32] Saviq: mhm... I see. thanks [11:34] dandrader, ping [11:34] Saviq: I think I got a hang after 10 tests - maliit-server consumes 100% of CPU and test doesn't proceed [11:35] nic-doffay, pong [11:35] Saviq: correction, 9 tests, 5/9 success (first 4 fail) [11:35] Mirv, yes [11:36] Mirv, first 4 are application lifecycle, right? [11:36] mzanetti, TBH that should be part of QtQuick [11:36] Mirv, maliit-server is looping on exit somewhere when unity8 stops [11:36] Saviq: err. you mean the time thingie? [11:36] mzanetti, yes [11:36] Saviq: life-cycle + unity8.shell.tests.test_hud.TestHud.test_hide_hud_click also failed [11:36] Mirv, might be a fallout [11:37] dandrader, Saviq mentioned you're familiar with the OSK. [11:37] Saviq: I tend to agree. it still requires some platform daemon I'd say [11:37] *somewhat* [11:37] Are you familiar at all with the InverseMouseArea too? [11:37] nic-doffay, kind of, yeah [11:37] mzanetti, it *could*, not sure required, though [11:37] dandrader, there appears to be issues with a large InverseMouseArea and the OSK. [11:37] dandrader, check this out: https://code.launchpad.net/~nicolas-doffay/unity8/dismiss-keyboard/+merge/188599 [11:37] mzanetti, but yeah, it'd have to be in QPA [11:38] In particular MacSlow|lunch 's comments [11:38] Cimi, stop what you're doing ;) [11:38] Cimi, if you're doing HUD [11:38] dandrader, these issues never happen on the desktop. [11:38] Saviq, :P [11:38] Only the mobile with the OSK. [11:38] Saviq, I almost fixed it [11:38] Cimi, Oren will file a new bug, there's confusion everywhere [11:38] dandrader, was wondering if you had encountered something similar or could direct me to someone who might know more. [11:38] Saviq, I'm doing that mzanetti asked [11:38] Saviq, which is a good point [11:38] Saviq: it continues now from the next test after killing maliit-server + autopilot [11:39] Cimi, yeah, that's correct [11:39] Saviq, edgeDragArea was 8gu [11:39] Mirv, yup, sounds like it [11:39] Saviq, one moment [11:39] Saviq, if edgedragearea is 8gu [11:39] Saviq, how can I detect stuff over those 8gu= [11:39] ? [11:39] like till the middle of the screen? [11:39] Cimi, when it starts *within* gu [11:40] Saviq, yes this is what I am doing [11:40] Cimi, it grabs the gesture - it's one touch [11:40] Cimi, and it gets "ownership" of sorts [11:40] Saviq, so edgedragarea detects stuff outside it? [11:40] Saviq, well [11:40] Cimi, continues, not detects [11:40] nope [11:40] yeah [11:40] exactly [11:40] so fix is easy [11:40] don't need to do what I was doing :P [11:40] Cimi, yeah, 2gu high if toolbar hidden [11:40] (using touchStartY [11:41] Cimi, or even - 2gu high *always* [11:41] Saviq, it's what I just did [11:41] Cimi, cool [11:41] nic-doffay, so the InverseMouseArea is not getting any events or the focus is not being cleared from the search field? [11:41] mzanetti, http://paste.ubuntu.com/6217660/ [11:42] Cimi: +1 (haven't tested it, but reads good) [11:42] dandrader, the InverseMouseArea gets events. [11:42] mzanetti, I thought the drag area should be 8 gu to detect 8gu long swipes [11:42] * mzanetti thinks if might be even only one gu... as we want the gesture to start outside the screen anyways [11:42] As I mentioned it's always fine on the desktop. [11:42] mzanetti, my bad [11:42] dandrader, but sometimes it seems to be a bit erratic on the mobile. [11:43] nic-doffay, dandrader so the issue is keyboard gets dismissed when you tap on it? [11:43] Cimi: yeah, the length of the gesture is fine [11:43] nic-doffay, dandrader, that's expected [11:43] Saviq, no. [11:43] nic-doffay, does, here ;) [11:43] mzanetti, Saviq 1 or 2gu for edge swipes? [11:43] in general, within the shell [11:43] Cimi, there's a prop on shell [11:43] Saviq, sometimes on the mobile in certain places the z ordering with the inversemousearea appears problematic. [11:44] Cimi, shell.edgeSize [11:44] Saviq, yep got it already [11:44] thx [11:44] Saviq, did you try out that branch? [11:44] If so does the keyboard disappear correctly for you? [11:44] http://paste.ubuntu.com/6217670/ thn [11:44] then === MacSlow|lunch is now known as MacSlow [11:45] nic-doffay, disappeared when I tried to type, so no - not correctly ;) [11:45] dandrader: do we still make use of the minimum speed for edge gestures? [11:45] nic-doffay, but that might've been an outdated ubuntu-keyboard/unity-mir [11:46] or unity8, for that matter [11:46] Saviq, yeah issues such as those. [11:46] This is what I'm wondering about. [11:46] davmor2, trying now [11:46] Saviq, they never occur on the desktop. [11:46] mzanetti, let me check [11:46] dandrader: we noted at the dev days that it's quite easy to trigger edge gestures by accident as there doesn't seem to be any minimum speed any more [11:46] MacSlow: https://bugs.launchpad.net/mir/+bug/1237957 [11:46] Ubuntu bug 1237957 in maliit-framework (Ubuntu) "maguro: image 90: keyboard dissapears on ubuntuone login" [High,New] [11:46] mzanetti, thanks for noticinf [11:46] Saviq, that's what I meant by "erratic behaviour" [11:47] mzanetti, tested [11:47] mzanetti, much better [11:47] mzanetti, From what I remembered, Pat was complaining about false negatives and so rules have been relaxed significantly [11:47] Cimi: cool :) [11:47] Saviq, with plain image 91 the osk works... but trying dednick's branch (or his branch merged with trunk) the osk does not receive any tap-events and they fall through to the window/surface below... [11:47] MacSlow: I added steps to the bug, apparently it doesn't come back if you manage to start typing before maliit vanishes [11:47] dandrader: mhm... I think a but too much as we have many false positives now... [11:48] mzanetti, check Components/EdgeDragArea.qml [11:48] davmor2: but I agree there as a time when we were too strict [11:48] Saviq, that's all the infor I have right now. [11:48] MacSlow, all, or just shell? [11:48] mzanetti, it's not being used indeed [11:48] err... dandrader, sorry davmor2 [11:48] Saviq, it also happens when a non-shell window is below the osk [11:49] MacSlow, dandrader, our InputFilterArea for the OSK seems broken indeed [11:49] Saviq, I don't feel good with merging anything to trunk until I know what's causing this [11:49] MacSlow, rather unrelated, but yeah, I'm seeing what you're seeing [11:49] Saviq, again? [11:50] dandrader: hmm... hard to say how to proceed here... we (the people at the dev days) felt there needs to be some minimum speed. probably not as much as there used to be tho [11:50] dandrader, maybe the socket connection fails for some reason? [11:50] dandrader, plain image 91 still works fine [11:50] dandrader, if keyboard is restarted, maybe it doesn't try to reconnect again? [11:50] MacSlow, could this possibly be related to the issue I'm having? [11:50] Saviq, yeah, it's IPC. There's room for things to go wrong [11:50] free launchpad karma to the quickest https://code.launchpad.net/~cimi/unity8/hud_use-edgeSize/+merge/190342 [11:51] nic-doffay, maybe [11:51] Cimi, mine [11:51] Saviq, but if you run on the terminal you will see if there are problems with it as it prints out issues [11:51] heh [11:51] Saviq, like failing to connect, or connection lost [11:51] Saviq, what are your thoughts on the matter? If this is an issue should the branch land regardless? [11:51] Saviq, *issue with the OSK [11:51] nic-doffay, let's fix it first and see what's what [11:52] * mzanetti has to go away for a bit. bbl [11:52] mzanetti, you lost time writing +1, Saviq is more efficient :P [11:53] Saviq, right. if keyboard dies and the unity-mir socked gets simply disconnected (as opposed to getting an error), it won't try to connect again [11:53] Saviq, didn't test that scenario [11:53] Saviq: the whole autopilot log of (what equals to) http://pastebin.ubuntu.com/6217659/ at http://pastebin.ubuntu.com/6217698/ - 7/22 [11:53] Saviq, I'm unsure how to proceed with this then. I'm not really sure where exactly the problem lies. [11:54] nic-doffay, you don't need to proceed with that [11:54] Saviq, cool I'll hold off with it then. [11:55] Mirv, yeah, some of them I see failing 'cause they're the second ones in an ap run [11:55] Mirv, lifecycle ones need fixing - we knew that [11:55] Mirv, crash, and then I'm not sure what happened [11:56] ah [11:56] OK happened ;) [11:56] then crash, crash [11:56] Saviq: I think only one had 2 tests, other were individual ones [11:56] Mirv, yeah, possible [11:56] Mirv, yeah, we need to look at them in order [11:57] dandrader, yeah, something's wrong - we don't seem to be doing the OSKController well [11:57] dandrader, that's fully dist-upgraded [12:00] Saviq, so, any steps to reproduce the issue? anything? [12:00] dandrader, just verifying here - works after a reboot [12:00] Saviq, there's nothing like a reboot! just like windows xp [12:00] davmor2, I only see some flickering... but can't get to a state where the osk refuses to come back up... I'm on image 91 with a NexusGalaxy btw [12:01] dandrader, restaring unity8 breaks it [12:01] dandrader, http://pastebin.ubuntu.com/6217718/ I didn't see before [12:01] MacSlow: yes you're right it is 91 [12:02] dandrader, and restarting maliit-server seems to kill unity8 [12:04] dandrader, trying to retrace [12:05] Saviq, hmm, it's getting scary. will test it more thoroughly (failure scenarios) [12:05] * Cimi -> groceries [12:05] dandrader, so yeah, the first time you restart unity8 you can't get maliit anymore [12:06] Saviq, you get a "Failed to connect[...]" message? === alan_g is now known as alan_g|lunch [12:13] dandrader, didn't see one, no [12:14] Saviq, but you run unity8 from the terminal, right? [12:14] (as opposed to "start unity8" [12:14] ) [12:14] dandrader, no, upstart [12:14] dandrader, but didn't see that in the .log [12:15] Saviq, but don't worry. there's certainly room to make that IPC code more robust and I knew that when I proposed it. But that's what we get when things are pushed on a ASAP/firefighter spirit [12:15] dandrader, of course [12:15] dandrader, yeah, got "Failed to connect after 10 attempts" indeed === jhodapp|afk is now known as jhodapp [12:21] om26er, the Time component needs to happen in the SDK [12:22] om26er, while it needs to be used in unity8 - primarily the bug is for the SDK [12:22] om26er, with the related blueprint [12:22] Saviq, is it re: time not being in sync with the datetime indicator ? [12:22] om26er, yes [12:23] dandrader, see if https://bugs.launchpad.net/mir/+bug/1237957 is related please [12:23] Ubuntu bug 1237957 in ubuntu-keyboard (Ubuntu) "maguro: image 91: keyboard dissapears on ubuntuone login" [High,New] [12:23] Saviq, so for 13.10 we can't do anything in unity8 itself (a workaround maybe ?) [12:25] om26er, no, we don't have a real time source [12:25] om26er, we'd have to increase the timer to tick every second or so - we don't want that [12:25] Saviq, ack. yeah that would be resource hungry. [12:26] Saviq, i'll tag the bug accordingly. Thanks [12:33] mzanetti, I think 2gu is too much [12:33] mzanetti, we should have 1.5 or 1 [12:33] in general === hikiko|lunch is now known as hikiko === alan_g|lunch is now known as alan_g [12:41] Saviq, that http://pastebin.ubuntu.com/6217718/ doesn't seem to be related with my patches [12:41] dandrader, it is in the sense that OSKController is doing that [12:42] dandrader, it's coming out of InputArea [12:42] Saviq: I didn't really look, but with latest unity8, we are still having no unity8 tests starting: http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/91:20131010.1:20131010/4658/ [12:42] mind having a look? (I really didn't get deeper) [12:42] didrocks, Mirv did, yes, we never touched them yet [12:43] hum, sorry, wdym by "we never touched them yet"? it's expected? (we will only see webbrowser app getting better with this new unity8?) [12:43] didrocks, "could not unblank display" [12:43] didrocks, display blank [12:43] didrocks, are we talking unity8 tests or webbrowser tests? [12:43] ok, and webbrowser app was supposed to get better from what I heard this morning [12:43] both [12:44] didrocks, so let's not talk unity8 tests, those are busted [12:44] didrocks, I just ran webbrowser and got 100% [12:44] ok, seems it's not the case in the image though [12:44] 7.7%, same pass rate, right? [12:44] sil2100, could we have mediascanner rebuilt in daily ppa when https://code.launchpad.net/~jamesh/mediascanner/ca-certs-dependency/+merge/190333 merges? [12:44] didrocks, I flashed, made writable and ran phablet-test-run, nothing else [12:45] Saviq: you stoppped powerd (at least at first) :) [12:45] didrocks, ok, I did stop powerd, running now with powerd running [12:45] mhr3_: sure, if I miss this being merged, please poke me and I'll rebuild [12:45] ;) [12:45] Saviq: fresh reboot etc.? [12:45] sil2100, thx [12:45] asac, yes - here's a catch [12:45] asac, adb got disconnected for me when MTP connected [12:45] right [12:46] that happens from times [12:46] asac, do we take that into account? [12:46] yeah [12:46] thats not happening in the test cases [12:46] Saviq, who will fix mir bugs? [12:46] if it happened you wouldnt get the fail/pass results [12:46] Cimi, define "mir bugs" [12:46] we sleep long enough to wait [12:46] Saviq, all tests that fail on mir [12:46] and i think we also have it turned off somewhere [12:46] Saviq, do they require changes or it's mir guys? [12:47] Cimi, depends what causes them [12:47] Cimi, some of them we fixed already [12:47] Cimi, between unity8, unity-mir, mir, qtubuntu, platform-api, ubuntu-keyboard, and then some :P [12:48] Cimi: debug is helpful....even if its a mir guy in the end...figuring out the where in that ^ list [12:48] kgunn, yeah, that's why I was asking [12:48] kgunn, Cimi I wouldn't touch unity8 yet [12:49] kgunn, there's something really wrong with input - and crashes here and there - those results are not really interesting [12:49] kgunn, what we should do is make sure those projects that have some tests passing, bring them up to green [12:49] ok [12:49] kgunn, but it can't be us doing everything [12:49] paying off the small debt first [12:49] kgunn, it should be the respective maintainers [12:50] Saviq: for sure...i need to ask bfiller for his team to start trudging [12:51] asac, maliit and unity8 crashed for me now, got 18 failures / 37 tests [12:51] * Saviq retraces the crashes === _salem is now known as salem_ [13:03] Saviq, reproduced. if you "stop unity8" and then start it again that new instance won't connect to ubuntu-keyboard and therefore the OSK input filter will be left with size (0,0) [13:03] Saviq, good news is that it should be easy to plug that hole [13:04] good find dandrader [13:05] kgunn, it's easy for a person to find the issues in the code he wrote himself :D [13:05] :) [13:07] dandrader, cool... as soon as you've a branch with the fix up, I gladly test it [13:10] dandrader, why won't it connect? [13:10] Saviq: fyi...lool just confirmed, starting mallit after unity8 comes up w/o crash [13:11] he's confirming and fast tracking a change to upstart [13:11] * Saviq gotta go, back in 1hr [13:14] Saviq, ubuntu-keyboard accepts only one connection. so it might be that then unity8 is stopped it won't properly recognize that the corresponding connection was lost [13:14] Saviq, and therefore refuse the connection attempt from the new unity8 process [13:14] s/then unity8/when unity8 [13:24] lol [13:24] i just realized the brightness slider doesn't have any indication of what it does [13:25] should we have something that says "Brightness" close to it or something? [13:25] tsdgeos, indeed... one could even think that the slider "relates" to the battery-status ;) [13:26] tsdgeos, like... "How much battery-load do you want?" [13:27] :D [13:28] mzanetti: you back ? [13:28] mhr3_: merge is in - the mediascanner will rebuild itself once other dependent stacks will finish, as we need those to be done first [13:28] MacSlow: ah, there's even a bug about it [13:28] sil2100, any eta for that? [13:28] https://bugs.launchpad.net/indicator-power/+bug/1227250 [13:28] Ubuntu bug 1227250 in indicator-power "Power indicator no longer shows label for screen brightness slider" [High,In progress] [13:28] where does phablet-flash push the files? sideloading? [13:29] mhr3_: let me see [13:29] just finished the drive on my galaxy nexus [13:29] °:° [13:29] ahrd drive [13:29] Cimi: i think in tmp [13:29] which is quite strange to me, having done [13:29] just a fist-upgrade and the shell folder in home [13:29] *dist [13:30] mzanetti: standup? [13:30] kgunn: ^^ [13:31] mzanetti, notes? [13:31] greyback, I might have a potential "stacking" issue... [13:31] have a browser in background [13:32] then trying to install Karma Machine [13:32] not logged into U1, hitting "create account" from the app preview [13:32] but instead of system settings/accounts I get the browser up [13:32] +back up [13:32] want me to file that against unity8? [13:36] mterry: do you know if all the code changes for the "answer phone w greeter" have been added to the ask sheet ?? [13:37] kgunn, no, they haven't at all. /.\ [13:37] kgunn, well, I don't know about the telephony stuff. [13:37] olli: please do [13:37] kgunn, but not the unity8 bit [13:37] * mterry looks [13:38] kgunn, the stuff we need is on line 169 [13:38] kgunn, but the greeter bit (which is arguably a separate fix, just needs those first) isn't listed [13:39] kgunn, added line 172 [13:40] mterry: great....so that's everything? [13:42] mhr3_: it shouldn't take too long, as stacks are anyway not running any check jobs now, I just need to abort some stuff etc. [13:42] kgunn, we had other deps, but those have landed now (upstart-app-launch, unity-mir) [13:42] mterry: ack....bug # ? [13:43] kgunn, 1234903 [13:43] mterry: thanks man...ok...done pestering for the moment [13:45] kgunn, ah man, looks like I wasn't so far off from bug 1234567 [13:45] bug 1234567 in GNU Mailman "Czech catalog bug" [Low,Fix committed] https://launchpad.net/bugs/1234567 [13:45] There's always 12345678 [13:47] :) [13:47] mterry: ok...got one for you...at least got some whiner [13:53] kgunn: hey, I am now [13:53] mzanetti: welcome back!...hope it was a good conference [13:53] greyback, fyi: https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1238020 - low prio to me [13:53] Ubuntu bug 1238020 in unity8 (Ubuntu) "system settings/accounts not being in fron when invoked from app preview" [Undecided,New] [13:53] kgunn: thanks, yes, it was, mostly :) [13:53] mzanetti: was curious if Saviq already has you chasing something [13:53] kgunn: nope [13:54] mzanetti: ok, so life is now...trudging thru AP tests with unity-mir on to make them green :) [13:54] kgunn: unity8 or all? [13:55] mzanetti: unity8 is #1 [13:55] ack [13:55] which i know Saviq is kind of working [13:55] mzanetti: the problem there is running thru the suite i think [13:56] mzanetti: because if/as unity8 crashes/stops you get the "can't blank display" trouble from mir & a stale mir socket issue [13:56] mzanetti: so i'm unsure if the individual tests pass or not [13:56] mhm... I'll check [13:57] ricmm_: curious...was there a powerd code change to address the unity8 AP test (the good ol' unblank from shell being stopped prob) ?? [13:59] mzanetti: there was this change from Saviq that was supposed to remove the stale socket....https://code.launchpad.net/~saviq/unity8/remove-socket-in-ap-tests/+merge/190175 [13:59] mzanetti: which i believe was effecting both the unity8 AP & the webbrowser AP [13:59] mzanetti: and i think Saviq is chasing the webbrowser aspect [14:05] dandrader, do you know how to get around this issue http://pastebin.ubuntu.com/6218123 when trying to build a .deb of a unity8 branch on the device? [14:06] MacSlow, what command did you use to build the package? [14:06] dandrader, bzr bd [14:07] MacSlow, try "dpkg-buildpackage -b" [14:08] MacSlow, that should skip dpkg-source [14:08] dandrader, looks better... thx [14:08] MacSlow, and "dpkg-buildpackage -b -nc" on later build [14:09] to avoid rebuilding from scratch every time [14:10] dednick: pete-woods ...so, are we cool to get a review of https://code.launchpad.net/~nick-dedekind/indicator-network/simunlock.dialog/+merge/185810 and put it into the image ?? can we log a bug to go back after phone v1 and add tests ? [14:10] this one gets a lot of attention [14:11] i mean, we at least have the full integration test of the stack itself right ? [14:11] kgunn: I'm not the right person to review that, I don't know vala at all [14:11] I just made a comment about how I felt it should have unit tests [14:11] pete-woods: got it...so, you're not disapproving in a strict sense ? [14:11] sure [14:11] yeah, I have no intention of blocking it [14:12] kgunn, no, it was unrelated to webbrowser [14:12] kgunn, it was just supposed to help slightly in unity8 tests, but there's much more to be done there [14:12] kgunn: i think tedg should review [14:13] dandrader, right [14:13] greyback, any luck with the bug? confirmed at least? [14:13] Saviq: yep confirmed, fix on its way [14:13] greyback, cool [14:14] dednick, Is the Unity stuff ready there? I thought we were waiting on that? [14:16] * tedg reboots [14:18] Saviq: did you get the status on those webbrowser tests? [14:18] (with powerd) [14:18] didrocks, unity8 and maliit crashed on me [14:18] didrocks, just retraced and filing bugs [14:19] didrocks, https://bugs.launchpad.net/ubuntu/+source/maliit-framework/+bug/1237870 [14:19] Ubuntu bug 1237870 in maliit-framework (Ubuntu) "maliit-server crashed with SIGSEGV in QHostAddress::clear()" [High,Confirmed] [14:21] * Saviq has retracing power [14:22] dandrader, that might be yours ↑ ? [14:23] Saviq: they crashed on startup or later? [14:23] didrocks, somewhen during testing [14:24] tsdgeos, mzanetti tasks incoming [14:24] Saviq: oka [14:24] tsdgeos, verify all the hud ap tests pass on unity8 (one by one) [14:24] tvoss: maliit crash during the tests are running, interested? ^ [14:24] and find out why not, if it does not [14:24] didrocks, hah :) [14:25] https://bugs.launchpad.net/ubuntu/+source/maliit-framework/+bug/1237870 [14:25] mzanetti, application lifecycle tests for you [14:25] Ubuntu bug 1237870 in maliit-framework (Ubuntu) "maliit-server crashed with SIGSEGV in QHostAddress::clear()" [High,Confirmed] [14:25] dandrader, I strongly suspect this is us ↑ [14:25] Saviq: ack, on it [14:27] tvoss, didrocks I say dandrader is on that already [14:27] Saviq, ack [14:27] ah, great! === alan_g is now known as alan_g|tea [14:30] Saviq, what if I manipulate the carousel model and I add empty elements? [14:30] Saviq, that are not visible or clickable? [14:30] Saviq: ok [14:30] tedg: could you elaborate on "stuff" wrt "Is the Unity stuff ready there? I thought we were waiting on that?" [14:30] Saviq, might contribute in making the logic working [14:30] or maybe dednick knows.... [14:30] Cimi, true, true [14:31] Cimi, might be pretty tricky though [14:31] Saviq, don't have other ideas === pete-woods is now known as pete-woods-away [14:31] Saviq, unless unloading the carousel [14:31] Cimi, you'd need a proxy that adds minCount - currentCount items [14:31] Cimi, not sure it's worth it [14:31] Saviq, yeah that's my idea [14:32] tvoss, didrocks got a nice unity8 one https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1238050 [14:32] Ubuntu bug 1238050 in unity8 (Ubuntu) "unity8 assert failure: upstart/com.ubuntu.Upstart.c:3304: Unhandled error from upstart_emit_event_sync: Cannot allocate memory" [Undecided,New] [14:32] Saviq, it will make it work and will look nice [14:32] kgunn, The pin widget in snap decisions [14:33] Saviq: oh! so upstart-app-launch as well? [14:33] Saviq, let me try, I'm curious at this point :) [14:33] didrocks, "Cannot allocate memory" though [14:33] Saviq: urgh, indeed, that doesn't sound… nice [14:34] didrocks, sounds like we're OOM [14:34] Saviq, could you check the size of init? [14:34] Saviq: shall we use latest image or latest image + apt-get update or? [14:35] tsdgeos, dist-upgrade [14:35] ah ok [14:35] tvoss, http://pastebin.ubuntu.com/6218259/ - but it's after a reboot [14:35] Saviq: ok [14:36] tvoss, so kinda useless [14:36] Saviq, ack [14:36] * Saviq got two more for upstart [14:36] Saviq, something is really spamming memory, that's not a leak, that's like a whole ocean of memory being lost [14:36] tedg: ah, thanks...yeah, my understanding that landed some time ago https://code.launchpad.net/~macslow/unity8/extended-snap-decisions-part1/+merge/187312 [14:37] kgunn, correct it's in the image by now [14:37] MacSlow: most definitely already in the image [14:37] Ah, okay. I wasn't sure if that was in the same branch or not. [14:37] tedg, yes, dednick has some small things for it [14:38] tedg: do you mind reviewing dednick's branch/mp today ? === ricmm_ is now known as ricmm [14:39] kgunn, Yeah, I can probably get to it today. I have a couple things I need to get to first, but I hope they're short. [14:39] tedg, kgunn: I'm on dednick's simunlock-related braches for unity8 and unity-notifications... just to avoid any conflicts/duplicate efforts [14:39] tedg: thank you...yeah, crazy times. it'll help reduce some heat if we can merge that today/tomorrow [14:40] MacSlow: ah... dednick just said tedg should review...but you guys can leg wrestle for it [14:40] kgunn, Yeah, we have about 5-6 indicator-network branches coming in.... [14:40] Saviq: is it supposed to work at all right now on the Galaxy Nexus? [14:40] mzanetti, yeah [14:40] kgunn, I'm almost good to merge it... but want actually also see it working _without_ issues on the device (keyboard-related) [14:41] tedg, ^ [14:41] onModelChanged is called anytime something in the model changes? [14:41] MacSlow, go with image 88 or 89 then [14:41] MacSlow, the keyboard breakage is new [14:42] MacSlow, and unrelated, dandrader is working on it already [14:42] Saviq, ok [14:42] mzanetti, tsdgeos mind you, running two unity8 tests in one autopilot run won't work - the second one doesn't get input for some reason [14:42] MacSlow: great...as you probably have a locking sim [14:42] Saviq: ok, still flashing though [14:42] will take a few more mins to download [14:43] tedg is a U.S. guy...us crazy americans don't do sim locking :) [14:43] Saviq: ah ok... was wondering... doesn't seem to work at all here [14:43] kgunn, not anymore.. unlocked my several months ago... I'm just using the stand-alone example for testing [14:43] Saviq: even the first one fails [14:43] mzanetti, what failure? [14:43] mzanetti: you got a locking sim ? [14:43] Saviq: rocessSearchError: Process exited with exit code: -6 [14:44] kgunn: not right now. but I can enable it if needed [14:44] kgunn, You think we could change the world and get rid of the feature by not supporting it in Ubuntu Touch? ;-) [14:44] mzanetti, any stdout / stderr output? [14:44] mzanetti, you made sure your screen is on? [14:44] Saviq: what(): display factory cannot create fb display [14:44] mzanetti, tsdgeos, pro tip: stop powerd when unity8 is working and unlocked [14:44] Saviq: yep, I tried [14:44] mzanetti, right, reboot [14:45] mzanetti, or well [14:45] too late [14:45] rebooting [14:45] but I actually already tried that too [14:45] mzanetti, next time you get that - see if maliit isn't flying the CPU [14:45] ok [14:46] kgunn, still got my regular one... [14:46] kgunn, which I keep "pin-locked" [14:47] mterry, you updating description for a merged branch? ;) [14:47] MacSlow: mind trying with that ? [14:47] that'll keep me from bugging mzanetti [14:47] kgunn, I can... just hope I won't "fry" it in tihe process :) === alan_g|tea is now known as alan_g [14:49] Saviq, for landing testers! [14:49] :) [14:50] Saviq: rebooted, stopped powerd, ran a single test, same issue. maliit not spinning [14:50] mzanetti, make sure you can run unity8 from the console - otherwise seek assistance with mir guys [14:51] MacSlow: if it gets fried, we'll expense a new one :) [14:51] tedg, https://bugs.launchpad.net/ubuntu/+source/upstart-app-launch/+bug/1238068 [14:51] Ubuntu bug 1238068 in upstart-app-launch (Ubuntu) "desktop-exec crashed with signal 5 in main()" [Undecided,New] [14:51] Saviq: that works fine. will dig into it [14:51] kgunn, well the costs are minimal... the data on it is valuable :) [14:52] mzanetti, if your run with -v - you'll see what library and import paths it exports [14:52] mzanetti, try exporting the same for your manual run [14:52] Saviq, That appid is wrong... [14:52] yep [14:53] Saviq, app_id = 0xbefebc79 "/usr/share/applications/webbrowser-app.desktop" [14:53] Saviq, Do you know who's sending that? [14:56] tedg, unclear, but it shouldn't crash anyway, should it ;) [14:56] om26er, re: bug #1237501 [14:56] bug 1237501 in unity8 (Ubuntu Saucy) "[Regression] Expanding Installed category does not show all the installed apps at once" [Critical,Confirmed] https://launchpad.net/bugs/1237501 [14:56] om26er, the suggestions is click scope refreshing on every in/out of Apps scope [14:57] Saviq, Sure, need to figure out how to report the error on the caller ;-) [14:57] om26er, can't reproduce for Installed, though [14:57] mhr3__, there a bug for click scope refreshing all the time? [14:57] Saviq, yep... [14:57] https://bugs.launchpad.net/unity-scope-click/+bug/1233756 [14:57] Ubuntu bug 1233756 in unity-scope-click (Ubuntu) "Possible infinite loop in scope search" [Undecided,New] [14:58] Saviq, dednick_, tedg: the simunlock-related branches for lp:unity8 and lp:unity-notifications are merged now [14:58] although it's still weird when you're not connect to the net [14:58] MacSlow: cool. thanks [14:58] MacSlow, awesome, thanks [14:59] kgunn, so regarding the unlock-testing... [14:59] kgunn, which image or branch you want to see tested? [14:59] Saviq, all models have append function? [15:00] Saviq, if I do in js, proxyModel = model; proxyModel.append(..) [15:00] will it work? [15:00] I'm waiting this thing while compiles.. [15:00] Cimi, doubt it [15:00] hah ok [15:01] Cimi, you'd need a C++ minimal proxy [15:01] Cimi, but really, don't get hung up on this [15:01] Saviq, I liked the idea of adding invisible elements [15:01] Cimi, yeah, but since we wouldn't use it for more than a week, when we stop using carousel for < something [15:02] Saviq, but results can be dynamic [15:02] MacSlow: you should use nick's branch for testing...i suppose run on device could be used [15:02] Cimi, yes, in which case we'd display a grid [15:02] bbiab [15:02] Saviq, how about the animation? [15:03] Saviq, when the carousel disappears? [15:03] kgunn, no... that doesn't work... I've to build a deb on the device [15:03] it will look bad [15:03] MacSlow: sorry dude...but well worth the effort to be certain it works [15:03] Cimi, we can make it not look bad if we cross-fade the old renderer with the new one [15:03] kgunn, or maybe not... nm... for the pinpad I don't need the osk to work... [15:03] Saviq, I think would be nicer a small carousel... [15:04] kgunn, was still mentally with the password/user-auth cases [15:04] Cimi, THEY DON"T WANT TO SHOW CAROUSEL WITH NOT ENOUGH ITEMS ;P [15:04] ok [15:04] :D [15:04] clear enough [15:04] Cimi, I'd rather show one with "inactive" items [15:04] Cimi, as in placeholders - "see - there's more space here, you should add more items to your videos" or something [15:04] Cimi, but the current design is as the bug says - switch between carousel and grid when needed [15:05] Saviq, so I need containers [15:05] Saviq, of carousels/grid [15:05] for each type [15:05] Cimi, a "dynamic" renderer that will load either a carousel or a grid, yes [15:06] Saviq, 2-3 loaders [15:06] music, video [15:06] Cimi, loaders aren't bad, you know ;) [15:06] people in future [15:06] Cimi, why do you hate Loaders so much? :D [15:07] Saviq, I don't hate loaders by themself [15:07] Saviq, I don't like how we load renderers [15:07] Cimi, why is that? [15:07] Saviq, seems a confusing mess [15:08] there are loads of file [15:08] from styles, to delegates [15:08] containers/loaders [15:08] Cimi, sorry, it's a complicated project ;) [15:08] pstolowski: dash plugins and more suggestions are fighting for second position in the Applications scope [15:08] Cimi, if you can find a less complicated way to achieve the same, pray tell [15:08] Saviq, I'm trying to make it simple [15:09] Cimi, no no, you're trying to change the design to make it simple ;) [15:09] Cimi, that's cheating :P [15:09] 8-) [15:09] it's a hattrick [15:09] dednick_, which indicator does trigger the unlock dialog? [15:09] MacSlow: network [15:10] dednick_, and which branch do I need to test this (apart from lp:unity8 and lp:unity-notifications)? [15:10] MacSlow: https://code.launchpad.net/~nick-dedekind/indicator-network/simunlock.dialog/+merge/185810 [15:10] mhr3__: want to top-approve https://code.launchpad.net/~jpakkane/ubuntu-ui-toolkit/video-image/+merge/190138? [15:10] didrocks, i don't have the powers :( [15:10] mhr3__: who should? [15:11] didrocks, bzoltan [15:11] or tim [15:17] tsdgeos, the order of categories is fixed; it looks like this because dash plugins scope has results faster and they are displayed immediately while more suggestions is still empty [15:17] pstolowski: actually i realized [15:17] mhr3__: did you coordinate with him on that one? [15:17] tsdgeos, or did I misunderstand it? [15:18] pstolowski: it was that more suggestions was appearing and disappearing [15:18] not that they were fighting for 2nd position [15:18] :S [15:18] they still are btw [15:19] didrocks, let me ping someone from sdk [15:20] tsdgeos, they disappear as you type, right? but they end up 2nd, and dash plugins 3rd, right? [15:20] pstolowski: no no === pstolowski is now known as pstolowski|bbl [15:20] it disappears while i'm idle [15:20] or at most swtiching left <-> side in the dahs [15:25] tsdgeos, bug in click scope, it's fixed already [15:25] well.. kinda [15:25] awesome :-) [15:25] will need fix #2 [15:25] or kind.. awesome :-) [15:25] but fix #1 makes it better [15:28] kgunn, Saviq, dednick_: I can't test the pin-unlock... my only locked sim-card is a micro-sim card (3FF) and the one that goes into my GalaxyNexus is a Mini-sim card (2FF) [15:34] kgunn, Saviq, dednick_: I'll go downtown and see, if I can get an adaptor for this [15:34] MacSlow, don't, I have one here === MacSlow is now known as MacSlow|afk === MacSlow|afk is now known as MacSlow [15:34] MacSlow|afk, just let me know what to do and I'll tell you if it works [15:34] Saviq, ah ok [15:34] Saviq, well I'm not 100% myself... [15:35] Saviq, quick question before I code, can I access properties of components inside Component () [15:35] ? [15:35] mmm don't think so... [15:35] Saviq, I grabbed and "make install"ed lp:~nick-dedekind/indicator-network/simunlock.dialog and lp:unity-notifications... [15:35] Cimi, no, you need to access item when it's loaded [15:35] better do to in another way statically [15:35] ok [15:35] Saviq, on the device... [15:35] yeah [15:35] Saviq: run the indicator-network from my branch. You will get a entry added to the bottom of the indicator menu to unlock sim. [15:36] Saviq, then switch it off... insert the locked sim-card... switch it on again... [15:36] Saviq, ./run_on_device lp:unity8 [15:36] Saviq, and then... I don't know [15:37] Saviq, dednick_ should know how to proceed from here [15:37] Saviq, not sure if it pops up a dialog/notification by itself upon recognizing the locked sim-card or if you have to open network indicator yourself to trigger it === pete-woods-away is now known as pete-woods [15:38] Saviq, I'm juast about to see what this setup does with an unlocked sim-card [15:40] greyback: can you work with slangasek (or someone on his team) ? they've narrowed down (at least one thing) to unity8 not handling upstart events as a source of mem leak [15:40] i believe he has a live system to debug (all probed up i imagine) [15:40] greyback, if we need to take something over from you - let me know [15:40] greyback: the problem is unity failing to /handle/ the events, resulting in them being queued up in upstart [15:40] kgunn: ok [15:41] dednick, conflict with lp:indicator-network [15:41] Saviq: I should manage [15:42] Saviq: so hud tests ran [15:42] Saviq: they do work *but* i've run into two issues [15:43] Saviq: a) sometimes autopilot hangs waiting for unity to die when it's already dead [15:43] Saviq: b) sometimes autopilot bails out at the very beginning because it says it could not find unity running [15:43] tsdgeos, a) - sure maliit isn't holding it still? [15:44] tsdgeos, b) crash? [15:44] tsdgeos, we have an issue where maliit will sometimes loop on unity8 exit === jhodapp is now known as jhodapp|lunch [15:44] tsdgeos, that's when autopilot waits with "Killing..." indefinitely [15:45] Saviq: you say that maliit keeps running unity8? [15:45] tsdgeos, yes [15:45] hmmm [15:45] ok let me see [15:45] i think it was not there in ps [15:45] but let me try to repro [15:45] tsdgeos, I just SIGSEGV'ed maliit and retracing [15:45] to see where it was looped [15:46] slangasek: where can we talk? [15:46] greyback, kgunn: please work with jodh on this - you should be able to find him on #ubuntu-touch [15:49] Saviq: ok, i'll take a look [15:55] Saviq: this one is interesting [15:55] the test succeeded, but unity8 crashed [15:56] tsdgeos, unity8 crashes on exit sometimes, yeah [15:57] Saviq: do i worry about that or not atm? [15:57] tsdgeos, no [15:57] Saviq: fixed conflicts [15:57] Saviq: i.e. i have to leave now-ish, tomorrow morning want me to continue with this or with the LVWPH bug you assigned me? [15:57] dednick, thanks [15:57] tsdgeos, the bug first [15:57] ok [15:58] tsdgeos, we'll see what's what overnight [15:58] seems there's been some people arguing about how to run AP tests....the official way is here https://wiki.ubuntu.com/Touch/Testing#Testing_your_Ubuntu_Touch_Code_before_submission [15:58] just in case...boom ^ [16:03] * tsdgeos has to leave [16:03] tomorrow more! [16:04] Saviq, is it possible that mir causes icon to get lost? [16:05] mhr3__, you mean like it can't display an icon? [16:05] mhr3__, doubt it, but good timing: bug #1238116 [16:05] bug 1238116 in unity8 (Ubuntu) "unity8 crashed with SIGSEGV in QIcon::~QIcon()" [Undecided,New] https://launchpad.net/bugs/1238116 [16:05] Saviq, eh, any way to screencap on the phone with mir? [16:05] screencap doesn't seem to work [16:06] mhr3__, yes, use your other phone ;P [16:06] mhr3__, no way to do screenshots yet, no [16:06] pff, so complicated [16:07] mhr3__, screencap was talking to sflinger, now we'll need something talking to unity8 [16:07] Saviq, http://imgur.com/t7ulRjx [16:07] super weird [16:07] Saviq, I remove songs [16:08] but it's not dinamic... [16:08] https://code.launchpad.net/~unity-team/unity8/carousel-loader/+merge/190406 [16:08] mhr3__, ~/.cache/upstart/unity8.log says something about loading imageS? [16:08] hmmm [16:08] file:///usr/share/unity8/Components/Tile.qml:39:16: QML QQuickImage: Error decoding: file:///usr/share/applications/webbrowser-app.desktop: Unsupported image format [16:08] Cimi, you mean you remove them from the phone but they don't disappear from the scope? [16:08] Saviq, btw I managed to have a simple loader, not much code [16:08] Saviq, not immediately [16:09] Cimi, there's a bug in mediascanner [16:09] Saviq, if I add them they appar [16:09] Cimi, bug #1238017 [16:09] bug 1238017 in mediascanner "Media removals aren't detected sometimes" [Undecided,New] https://launchpad.net/bugs/1238017 [16:10] Cimi, you sure that's all? [16:10] Saviq, as if the roles changed... can that happen? [16:10] mhr3__, shouldn't [16:10] or well... when/why does that happen [16:10] we see it can :) [16:10] mhr3__, better check what you got from the scope ;P [16:10] Cimi, rendererLoader connects much more to its item [16:11] Saviq, it's the overrided cat [16:11] overriden [16:11] mhr3__, ah interesting :) [16:11] Saviq, seems to work [16:11] Saviq, the loader [16:11] I am searching for songs and I see swapping between the two [16:11] carousel/filter [16:11] very quickly [16:11] surprise surprise [16:12] Cimi, when *searching* you should see no carousel at all [16:12] mhr3__: damn, mediascanner building now in PPA - amd64 succeeded [16:12] Saviq, that's another bug [16:12] sil2100, yey! :) [16:12] mhr3__, do the scopes come up with "dynamic" at all currently? [16:12] Saviq, but it's the most dynamic way I have now to test when model.count changes :P [16:12] Saviq, no, we removed it [16:12] mhr3__: but daaaamn, this took longer than the 'shouldn't take long' I suspected! cu2d has issues right now, many many [16:12] Saviq, I'm basically using the bug to my needs :P === mhr3__ is now known as mhr3 [16:13] Cimi, anyway - sorry, but I can't look at it today [16:13] dandrader: if you get a fix for that, can you please ensure you ping robru to get that in as quick as possible? [16:13] sorry, was looking at an old scrollback :p [16:17] Saviq, mind blown, i'm able to make it toggle between the working and "shifted" states [16:17] mhr3, tricky ;) [16:26] dandrader, you around? === jalcine is now known as jalcine_ [16:26] Saviq, yes [16:26] dandrader, was afraid you disappeared ;) [16:27] dandrader, let me know if/when you need a hand with the keyboard [16:27] Saviq, just proposed the osk ipc improvements [16:27] ah [16:27] Saviq, this time I did a good bit of testing :) [16:27] dandrader, good :) [16:27] Saviq, https://code.launchpad.net/~dandrader/ubuntu-keyboard/improve_kbd_info_ipc/+merge/190418 and https://code.launchpad.net/~dandrader/unity-mir/improve_osk_ipc/+merge/190417 [16:28] dandrader, thanks! will check it out [16:28] Saviq, so you can start/stop unity8 or maliit-server and things still work [16:28] Saviq, unless maliit-server dies due to some mir surface thingy [16:28] which also brings unity8 down [16:28] but then it's unrelated [16:28] to that osk ipc [16:29] greyback: Saviq: https://code.launchpad.net/~mzanetti/unity-mir/fix-appid-parsing/+merge/190419 [16:29] dandrader, yeah, sounds like https://bugs.launchpad.net/ubuntu/+source/maliit-framework/+bug/1238107 ? [16:29] Ubuntu bug 1238107 in maliit-framework (Ubuntu) "maliit-server crashed with SIGSEGV in __GI___pthread_mutex_lock()" [Medium,New] [16:30] greyback: Saviq: fixes application_lifecycle autopilot tests [16:30] mzanetti, \o/ [16:32] https://bugs.launchpad.net/unity8/+bug/1238141 [16:32] Ubuntu bug 1238141 in Unity 8 "Dash, Apps - close icon is top right, not consistent with unity desktop" [Low,In progress] [16:32] I attached a branch to fix it [16:32] was annoying me :) === dandrader is now known as dandrader|lunch [16:33] Saviq: can you look after that? It's removing a hack that might be relied on (appId sometimes a full desktop file path) [16:34] dandrader|lunch, $XDG_RUNTIME_DIR/ubuntu-keyboard-info please [16:35] dandrader|lunch, and chmod 600 it, or jdstrand will have us for dinner [16:35] Cimi, btw there is a design bug for that reported a few days ago. want me to duplicate yours ? (bug 1236285) [16:35] bug 1236285 in Unity 8 "[Dash] Running apps close button on the wrong side " [Medium,Confirmed] https://launchpad.net/bugs/1236285 [16:35] greyback, yes I will [16:35] om26er, yes pls [16:35] I like that bug tag [16:35] Saviq, well, it was already going to /tmp before. I just made it official [16:35] now I'm really of for lunch [16:36] dandrader|lunch, yeah, go [16:42] if i've gotten: [16:42] creating surface at (0, 0) with size (768, 1280) with title 'Qml Phone Shell'terminate called after throwing an instance of 'boost::exception_detail::clone_impl >' [16:42] what(): Invalid surface attribute. [16:42] i've probably installed something where the abi doesn't match? [16:43] kdub, never saw that, so I'd say possible [16:43] greyback, ↑? [16:44] kdub: new to me too, yep I'd suggest a rebuild of platform-api [16:44] okay, will do === alan_g is now known as alan_g|EOD [17:04] mterry, hey [17:04] mterry, what's causing bug 1238159 ? [17:04] bug 1238159 in unity8 (Ubuntu) "last screen is shown for fraction of a second when turning the screen on" [Medium,New] https://launchpad.net/bugs/1238159 [17:05] om26er, Mir [17:06] Saviq, I should move it to Mir then. Mind adding something to the bug report, maybe a comment of what you know ? [17:06] om26er, https://bugs.launchpad.net/mir/+bug/1233564 [17:06] Ubuntu bug 1233564 in Mir "Greeter is seen animating when pressing the side button to wake up" [High,Triaged] [17:06] haah!! [17:07] dednick, wow it looks ugly having the indicators as background... [17:07] Saviq: eh? [17:07] dednick, the PIN dialog [17:07] * mterry notes that Saviq answered already, thanks :) [17:07] dednick, it's the same background the indicators have [17:08] dednick, oh, and it's transparent to touch [17:08] Saviq: yeah. it's transparent [17:08] partially [17:09] dednick, so, what's the deal with not being able to say "you have x tries left" and/or "enter PUK"? mzanetti seems to have been able to display that, at least after the first thing you entered [17:10] dednick, oh, and why are some networks greyed out? [17:10] dednick, well, it's because http://bazaar.launchpad.net/~unity-team/unity8/trunk/revision/432 [17:10] dednick, but why would the backend mark them so? [17:10] Saviq: the ofono dbus property for the type of pin required does not dynamically change on my device. [17:10] dednick, not even after you put something in? daamn [17:11] that sounds so wrong [17:11] Saviq: it is. there is a bug logged [17:11] bugs [17:11] * greyback going afk, back in 3 hours or so === boiko_ is now known as boiko [17:12] dednick, think we should close indicators on notifications open? [17:12] dednick, they look so bad over indicators... [17:13] or - they don't look at all.. [17:13] * Saviq files bugs [17:13] Saviq: yeah, we're working on that for other things [17:13] opening settings/apps [17:13] dednick, mhm [17:13] will file a notifications + indicators conflict bug anyway [17:13] ok [17:14] Saviq: functionally works for you? [17:14] dednick, yeah [17:15] dednick, feck, now all of my networks are greyed out... wth? [17:15] * Saviq reverts [17:16] Saviq: no idea. [17:18] Saviq, it seems to be "this ap is out-of-range" [17:18] mhr3, yeah, how do I care about *all* of the networks? why are they there? [17:19] if they're out-of-range, why am I seeing them at all? [17:19] good questions... all of them [17:22] dednick, so bug #1206941 [17:22] bug 1206941 in ofono (Ubuntu) "[ofono][rild] "Retries" property does not not work" [High,Confirmed] https://launchpad.net/bugs/1206941 [17:24] Saviq: and #1225022 [17:24] Saviq: https://bugs.launchpad.net/touch-preview-images/+bug/1225022 [17:24] mup: bug #1225022 [17:24] Ubuntu bug 1225022 in ofono (Ubuntu) "[ofono][rild] "PinRequired" property does not always work" [Undecided,Confirmed] [17:26] Saviq: i need to go for today. anything else? [17:27] dednick, go [17:27] dednick, thank you [17:27] :) thanks [17:46] kgunn, sim unlock is working, pending approval from API folks [17:48] sweet [17:48] there's all kinds of bugs around it, but at least the functionality is there [17:49] I mean mostly user experience bugs [17:49] /food === jhodapp|lunch is now known as jhodapp === dandrader|lunch is now known as dandrader [18:15] mterry: just fyi...i asked gagnon to put a camera on the devices [18:16] not sure if you had already [18:16] kgunn, oh, thanks. I'm finishing lunch, hadn't asked yet [18:17] tedg: are you ok if we merge dednick's branch ? we at least tested and it worked...just looking for someone from unity api team to say "ok" [18:18] kgunn, I imagine so, that's just poping to the top of my queue [18:18] tedg: i'll give you a bit...just ping me either way, thumbs up or down [18:19] kgunn, Let me have a few minutes with it... alone. [18:19] don't forget maniacal hand rubbing [18:21] unity folks: is there a difference (in terms of application state) when i bring an application back by swiping right-to-left vs when i click on the 'recent apps' thumbnail? [18:22] kdub: i don't think so... Saviq ^ ? [18:22] or rather maybe ricmm ^^ ? [18:23] kdub, I don't think there's a difference [18:23] although it's been a while since I've last looked at this code [18:24] specifically, in my 'make it go fast!' branch... i see ~1-3 frames of white when clicking recent apps on the client [18:24] just trying to figure out if its something i did, or if its normal [18:24] kdub: i would think they would be "aware" in terms of focus [18:25] kdub: e.g. they do know when they get put to the "back" i would think...i learned yesterday, they can render for up to 3 sec when put to back then they stop...so [18:25] they must have some focus awarenesss [18:26] make it go fast is unhinged rendering ? [18:27] if that's the name that's been kicked around for making unity operate more like a triple buffered client again, yes [18:27] :) [18:27] kdub: no.... [18:27] kdub: :) so this is triple buffer huh? [18:28] kdub: i thot you were just testing "make it go fast" == remove honoring fences [18:28] no, i'm shifting where we wait [18:28] so, its a real solution [18:28] ah [18:28] from unity's thread to the compositor's thread [18:28] kdub: i've seen that same white render ur talking about tho [18:28] kgunn, Approved. Can't test ofcourse, but it makes sense to me. [18:28] tedg: thanks much! [18:29] kgunn, We're going to have to pick up locked SIMs next time we're in Europe :-) [18:29] tedg: i liked your first idea...just ignore it and get the industry to quit doing that [18:29] Heh, yeah. [18:37] tedg, tested here - apart from bugs in ofono not letting us do the right thing, worked [18:38] kgunn, kdub, there's all kinds of magic'n'trickery when we focus apps [18:38] kgunn, kdub there's no *intentional* difference, but those are different codepaths when you go dash → app vs. app → app [18:39] Saviq, Cool. You'll have to show me some day :-) [18:39] tedg, lol [18:42] Saviq is there a way to get unity to make unity8.log a unity8.log.old before it just overwrites ?? [18:42] we've hit one of those.... [18:42] total hang...and can't get adb connection [18:43] kgunn, that's upstart [18:43] ah [18:43] kgunn, but, at least on my desktop, I'm getting old logs [18:43] kgunn, so it must be configurable [18:43] tedg, ideas ↑? [18:43] davmor2: ^ [18:44] It does log rotate on login. [18:44] So the old one should be a 1.gz [18:45] Or 2 or whatever [18:45] davmor2: ...have you seen a #.gz ? [18:47] http://paste.ubuntu.com/6219220/ [18:47] kgunn, tedg: ^ [18:47] or am I looking else where? [18:47] And if you shutdown and restart? [18:48] tedg: this is about my 6th restart [18:48] let me reboot though for a confirmation [18:50] Hmm, that's odd. I've tweaked with my phone a lot. Perhaps logrotate isn't installed? [18:50] (it is for me currently) [18:50] tedg: http://paste.ubuntu.com/6219240/ [18:51] root@ubuntu-phablet:/# dpkg -l | grep logrotate [18:51] ii logrotate 3.8.3-3ubuntu2 armhf Log rotation utility [18:52] tedg: I see logrotate [18:52] Yeah, how odd. I definitely have .gz logs in my cache dir [18:53] I'm guessing you have a /usr/share/upstart/sessions/logrotate.conf as well? [18:54] tedg: http://paste.ubuntu.com/6219259/ [18:55] just to double check it is the same as yours [18:55] Yeah, looks the same. [18:55] Not sure what else to say. Perhaps just do a "start logrotate" and see if they work? [18:55] It'll take a minute [18:56] root@ubuntu-phablet:/# start logrotate [18:56] start: Unknown job: logrotate [18:56] tedg: do I need to be phablet? [18:57] yes [19:00] tedg: right I have a logrotate process id, so in theory if I reboot now I should have a gz log correct? [19:00] davmor2, No, it's on a timer. It takes literally a minute. [19:01] tedg, maybe it logrotates on logout, and we're never actually logging out on devices? [19:01] Yay I see gz's [19:01] hmm no, startup [19:01] Yes, it's a minute after start [19:01] But, BTW, it doesn't delete log files, just appends. [19:01] Saviq: oh yes that could be I'm having to pull the battery to get a working system again [19:02] So different runs will just be appended to the same file. [19:04] kgunn, davmor2, right, so you don't need the old one - just look before the current one [19:05] kgunn, davmor2 unity8 on startup says "__pthread_gettid -2" [19:05] kgunn, so anything above that will be the previous run [19:05] Saviq: that's a great way for it to say "hello i'm starting" [19:05] let me have a look here [19:06] kgunn, indeed! ;) [19:06] kgunn, truth be told, upstart should add a header [19:06] tomorrow i plan to roll over in bed and say to my wife "pthread get id negative 2" [19:08] kgunn: is that how you get so many black eyes? I thought it was all sports related ;) [19:08] :) [19:09] kgunn, ;) [19:11] morning [19:11] mornin' thomi [19:18] Saviq: curious....i turned on qml renderer timing....and when i reveal/navigate to the app lens in dash....it renders _alot_...like for a few seconds...even tho the screen isn't changing [19:18] is that just a render per icon ? [19:19] kgunn, no, QML generally has a timeout of sorts [19:19] kgunn, it will render for a few secs and then settle down [19:19] another minor peformance nit...it seems to render several-to-dozens of frames when i hit the power button both to idle & to wake [19:22] kgunn, yup, it does [19:27] * kgunn really really hates this crap usb cable...but can't find the good one.... [19:30] I have lost so many usb cables over the last year [19:30] it's not even funny [19:32] Saviq, ok for me to take out the tap to preview from switching-preview and mp it as a separate branch? [19:32] i have a bad hunch that switching previews isn't going to make it [19:38] tvoss: the one i'm using is just absolute crap....i'm frustrated enough to go get a new one [19:41] kgunn, Saviq has got a cool one [19:41] he might lose it in oakland [19:41] :)) [19:44] mhr3, mzanetti is back around - he'll fix it [19:45] tvoss, no comment! [19:45] mhr3, https://code.launchpad.net/~vorlon/unity8/lp.1235649/+merge/190443 [19:55] http://www.aviiq.com/products/ready-clips [19:58] Saviq, http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/92:20131010.2:20131010/4662/ [19:58] tedg, the indicators upstart signals are sent just twice right? [19:58] mhr3, Which ones? [19:58] tedg, https://code.launchpad.net/~vorlon/unity8/lp.1235649/+merge/190443 [19:59] That's not entirely true [19:59] It's a shared connection object [19:59] By libdbus [19:59] So when upstart-app-launch creates the same connection, it gets reused. [20:00] So the indicator signals are only each sent once. [20:00] But the connection is used more than that. [20:00] Saviq: want me to top approve/add to ask this one ? https://code.launchpad.net/~vorlon/unity8/lp.1235649/+merge/190443 [20:00] tvoss, tvossnice! [20:01] tedg, it's believed to be source of the upstart mem balooning [20:01] Uhm, weird. [20:01] So the other way is that we'll reconnect each time. [20:01] I mean, that's not the end of the world, but seems like it shouldn't be a problem to have a long lived connection. [20:02] kgunn, dunno that code enough - but if slangasek did it and tvoss approved, I have no questions [20:02] Saviq, CI revoked [20:02] tedg, it's that noone's reading that connection, that's really the problem [20:02] I don't think it'll break anything [20:02] tvoss, "revoked"? [20:02] Ah, perhaps, not getting into a mainloop. [20:02] Saviq, marked needs fixing [20:02] tedg, right [20:03] mterry: hey any news...what do you see in the lab on the bottom bar reveal issue ? [20:03] tvoss, ah we're not looking at those at all now [20:03] tvoss, we're expecting all kinds of hell there ;) [20:03] libdbus has a way to connect it to the loop, no? [20:03] Or is that just dbus-glib. [20:03] tedg, nope, I looked through, no easy way [20:03] tedg, so you're basically saying that even if it's fixed here the app-launch lib will hit anyway [20:04] Yes, but it'll ref and unref it, which will cause connect/disconnect [20:04] Basically each app start/stop/pid check will cause a connection. [20:05] tvoss, we have a fix/workaround for the lifecycle tests, but need to verify it's ok [20:05] Saviq, shoot [20:06] tedg, where is that code? [20:06] tvoss, https://code.launchpad.net/~mzanetti/unity-mir/fix-appid-parsing/+merge/190419 [20:06] tvoss, but it's not entirely safe - we need to make sure everything works still [20:06] mhr3, http://bazaar.launchpad.net/~indicator-applet-developers/upstart-app-launch/trunk.13.10/view/head:/libupstart-app-launch/upstart-app-launch.c [20:07] tvoss, and there's a suspicion it won't without some love [20:07] tvoss, the only other failure is input not always getting to the shell [20:08] tedg, yea, so that has the same issue [20:08] tvoss, ah actually [20:08] tvoss, the only other failure *is* two tests in one go - 'cause autopilot just does string comparison [20:09] tvoss, so if you have tests "foo" and "foo_bar", running "foo" will run both [20:09] * thomi ducks [20:09] mhr3, Well, it should drop the connection when the last ref drops, no? [20:10] Saviq, ? [20:11] tedg, mhr3: right, we believe the same problem affects the app-launch lib; jodh just reached that conclusion before he hit EOD, I haven't had a chance to look at that part yet [20:11] tvoss: Saviq: that's my fault - sorry. Promise I'll fix that soon :) [20:11] thomi, I know it is ;) [20:11] but since I'm told unity only uses this socket to send one event on startup and one on shutdown, I think it's pretty clear that we don't need to keep /that/ socket open long-term [20:11] tedg, i meant same issue == opening+closing [20:12] slangasek, Well, it's shared through libdbus. So it'd be using the same socket. [20:12] slangasek, mhr3, tedg I will top-approve then [20:12] yea, it's nasty, but should successfully workaround it [20:13] tedg, any last comments? [20:13] Naw, it's fine. We should figure out a longer term solution. [20:13] Port libupstart to gdbus seems like the best to me :-) [20:13] tedg, yup, but I have a gut feeling that really means looking into upstart timeouts and such [20:13] tedg, +1 :) [20:13] * tvoss looks at tedg [20:13] slangasek, done [20:14] tedg: ah, so the app-launch lib is also running within unity? [20:14] slangasek, Yes [20:14] ok [20:15] so indeed, we will need to deal with that before this makes any difference [20:15] Saviq, I don't quite understand that lifecycle mp tbh [20:16] tedg: however, I think the change to the plugin is per se correct - no sense in keeping a ref open from a plugin that's going to use it twice over the length of the server [20:16] tedg: where does the code for the app launcher live? [20:16] slangasek, Well effectively that plugin was keeping it open for everyone. The library just effectively refs and unrefs the singleton if it exists already. [20:16] slangasek, lp:upstart-app-launch [20:17] and if it doesn't already exist? [20:17] It creates a new one [20:17] * slangasek nods [20:17] Which isn't the end of the world, but eh, it's nice to share. [20:17] :-) [20:17] right [20:17] tvoss, that's not really related to lifecycle [20:18] tvoss, but it does fix the difference between qtubuntu's appmanager and unity-mir's one [20:18] tvoss, so the tests pass [20:18] tvoss, but we need to make sure nothing else breaks [20:18] tvoss, so I'll do that tomorrow [20:18] slangasek, I think that the start/stop observers would be a long running connection as well. But, that's gdbus, so there's a thread on the Unity side watching that connection. [20:18] tedg: so my patch to the plugin was predicated on the belief that this socket was only being used twice, and that it was wasteful to keep it open in between (since by "fixing" the code to register with the main event loop, we would then have to cope with a stream of data from upstart that we know we don't care about). But if upstart-app-launch also uses it... how often do we expect it to open/close the connection? Once per launched app [20:18] Saviq, ack, I will give it a spin first thing tomorrow morning. Just drop me a mail if you want something special to be checked [20:19] slangasek, Yes, on a start or stop (rare, but fallback to ask Upstart to stop) [20:20] We ask nicely once, then call the executioner. [20:21] tedg: ok. so I think it's still preferable to just reopen the connection when needed [20:21] however, if libupstart-app-launch already does the correct nih_unref(), I wonder why jodh said the indicator plugin change was insufficient [20:22] Well, I could be doing it wrong... :-) [20:22] I think that's the most libnih code that I've written. [20:22] yes, but that may make two of us. :) [20:22] because your code looks an awful lot like mine [20:25] tedg: upstart_app_launch_stop_application() doesn't nih_unref() the proxy, that looks like a bug to me [20:27] slangasek, Yup, sure does. [20:28] would that explain what's happening here, though? [20:28] since people are saying they reproduce this from a pristine phone, seems unlikely to me that stop() would have been called [20:28] * tedg guesses no [20:29] * slangasek looks for more bugs, then [20:31] slangasek, https://code.launchpad.net/~ted/upstart-app-launch/stop-proxy-unref/+merge/190479 === bschaefer_ is now known as bschaefer [20:51] mterry, hey [20:51] mterry, is that known that setting an image from /usr/share/background for the greeter doesn't work? [20:52] mterry, I played a bit with that today, I get no background when I do that, it works when I set one in the phablet userdir though [20:52] seb128, no it is not known. I wonder why that would be. I usually tested from a file in that dir [20:53] seb128, this is via AS? [20:57] mterry, right, using 'gdbus call --system -d org.freedesktop.Accounts -o /org/freedesktop/Accounts/User32011 -m org.freedesktop.Accounts.User.SetBackgroundFile '/usr/share/backgrounds/163_by_e4v' [20:57] seb128, I assume AS keeps it, but greeter doesn't pick it up correctly? [20:57] mterry, that leads to an empty background (the image is there, I've install ubuntu-wallpapers-saucy) [20:57] mterry, yes [20:58] mterry, it works with /home/phablet/Images/something though [20:58] mterry, which means it works with the content picker since the picker does a file copy in the userdir [20:58] mterry, e.g not such of an issue, I just wondered if I should report it [20:58] or if it was known [20:58] seb128, yeah seems odd. I'd guess some weird file permission thing, but seems unlikely in /usr/share/ [20:59] seb128, oh wait [20:59] seb128, does your /home/phablet example have a file suffix? [20:59] seb128, qt is a bit picker than Gtk about that [20:59] seb128, like with warty-final-ubuntu.png, it won't load, because it's really a jpg [21:01] mterry, the /home/phablet one works ... what do you mean "suffix"? it was called "ubuntu.jpg" [21:01] mterry, the system one was '/usr/share/backgrounds/163_by_e4v' iirc [21:01] seb128, but the /usr/share one didn't have a suffix? I'm guessing qt didn't load it because it lacked one [21:01] seb128, try one with a .jpg ending [21:01] .jpg [21:02] mterry, I wonder if I screwed up my copy paste or something [21:02] oh weird, Ok [21:02] :) [21:02] mterry, my device is busy without other things atm, I'm going to try again tomorrow [21:02] mterry, probably a pbcak [21:02] pebkac [21:03] mterry, thanks ;-) [21:04] seb128, I suspect it's real. Especially if it only happens on Mir. :) That thing brings out the bugs in everything [21:04] true speaking there [21:04] mterry, I'm going to try again tomorrow and let you know how it goes [21:05] mterry, blame robert_ancell for the Mir issues (well, until he comes back to desktop, then blame kgunn) [21:05] seb128, that wouldn't be wise on my part :) [21:05] seb128, how rude! [21:06] just try it mterry [21:06] lol [21:06] robert_ancell, don't take offense, I'm just being French :p [21:16] you can tell because he's twirling his moustache === jhodapp is now known as jhodapp|bbiab [22:01] mzanetti, Hey, can you give me access to the presentation you sent via mail? === jhodapp|bbiab is now known as jhodapp === salem_ is now known as _salem === jhodapp is now known as jhodapp|afk