=== _salem is now known as salem_ === salem_ is now known as _salem === tvoss|eod is now known as tvoss [05:47] bregma: regarding the unity7 AP tests, just another reference point from two successive test runs, the '+':s of which prevented the automatic release this time http://pastebin.ubuntu.com/5979939/ [05:48] in case there'd be something that could be made more solidly succeeding [08:03] jamesh: hi! [08:03] jamesh: did you get my e-mail? [08:03] sil2100: hi. I saw your comment on the MP. Was there some other email? [08:04] Thanks for your work on lucene++ though: we're deciding what to do about the test failures now [08:05] Looks like we'll disable them for now, since they appear to be broken in trunk too and pass locally [08:08] jamesh: sounds good for me! NP, and once this gets in let's deal with releasing media-scanner to the world [08:08] sil2100: yep. [08:12] Saviq, ping [08:12] Cimi, pong [08:13] Saviq, ciao [08:13] Saviq, so can you test the last rev with autopilot? [08:13] Cimi, sure, will do [08:14] Cimi, and well, CI can, too ;) [08:14] there's just one minor thing I belive [08:14] background property of root, not sure it will have to be an alias or not [08:15] readonly alias might be good [08:21] Cimi: just ran the autopilot tests with the background branch... all passed [08:21] mzanetti, and you confirm is failing without? [08:21] mzanetti, a previous rev... [08:21] Cimi: no... I never saw it failing [08:22] Cimi: ah... hmm... there is one thing: [08:22] Cimi: if there is no image set because of a wrong url. then the greeter has some weird behavior. i.e. it doesn't fully slide away etc [08:23] Cimi: I could see how that might has broken the tests in environments where no background is set [08:23] mzanetti, rev 167 should fail [08:23] mzanetti, mmm the branch is supposed to fix those [08:24] mzanetti, it should revert to default bg in case [08:24] Cimi: yep... let me just try without my fix if I can reproduce the issue [08:24] mzanetti, set your background to an invalid one [08:24] mzanetti, or to an empty string [08:24] mzanetti, that's when they failed [08:24] mzanetti, 'cause greeter was never "ready" [08:24] Saviq: yeah... I've fixed that one now... but let me verify [08:28] mzanetti, thought was fixed in the branch... [08:28] Cimi: Saviq: yep. confirmed. with an empty background it failed with revision 174, but works with 175 [08:28] of this branch lp:~unity-team/unity/background-ugly [08:29] mzanetti, k cool [08:29] I'll clean up and push to unity8 :P [08:29] :D [08:41] larsu: good morning! [08:42] dednick: morning :) [08:42] larsu: would you mind taking a look? https://code.launchpad.net/~nick-dedekind/qmenumodel/remove.unitymenuaction/+merge/179800 [08:42] larsu: we're leaking :( [08:43] larsu: had to remove UnityMenuAction [08:44] dednick: hm, activate() does magic now... [08:44] dednick: fine by me if it's fine by you [08:45] and I guess it is, since you wrote it :) [08:46] dednick: when using https://code.launchpad.net/~nick-dedekind/unity8/indicator.positions/+merge/179505 clock is not on the right anymore, is this ok? [08:47] larsu: well, i wasnt really sure about the activate. could do activateWithData if you prefer [08:47] but then magic will just be shifted to unity8 [08:47] tsdgeos: er. let me check that out. doesnt sound right [08:48] dednick: on the device, on the pc is on the right [08:48] tsdgeos: pc? [08:49] dednick: Personal computer, laptop [08:49] tsdgeos: hm. i havent seen that. what indicator is that? [08:49] :D [08:49] ok [08:49] let me start again [08:50] dednick: fair enough. I'm approving. [08:50] dednick: if i run unity8 on my laptop, the clock is correctly on the right, but if run unity8 on the phone it is not on the right [08:50] dednick: clearer now? [08:50] tsdgeos: ahha. yes [08:51] tsdgeos, pushed fixes to app-preview-data [08:51] oka [08:51] tsdgeos, will talk with the online guys later to see if it should work already [08:51] tsdgeos, so don't top-approve yet [08:51] tsdgeos: it's possibly because the files are still being read from install folder. let me check it out [08:51] tsdgeos, thanks [08:54] dednick, looking at lp:~nick-dedekind/unity8/indicators-behaviour.tabs - MainView is kinda scary to use there, it's meant to *be* the main view... [08:55] Trevinho, andyrock, bregma: ping [08:55] sil2100, pong [08:55] Saviq: indeed it is scary. unfortunately it's the only thing that tabs work on at the moment. I spoke to SDK, but it's not really on their priority list. [08:55] dednick, mhm :/ [08:55] sil2100, what's up? [08:55] andyrock: hi! Poor you! You always respond to my pings ;) And you know what my pings are usually about! [08:55] andyrock: failing AP tests, oh noes! [08:56] Saviq: dednick: yeah i remember looking at that and then running away :D [08:56] Saviq: dednick: also i guess you have that "let's override this internal property" for that reason, right? [08:56] Saviq: PageStacks are the same. i have no idea why they seem relucant to do something about it. [08:56] tsdgeos: yep [08:56] dednick, mhm [08:56] sil2100, :D do you have a list of need-fixing tests right? [08:57] andyrock: so, I actually have two lists, let me bring those in [08:57] tabs seem like a fairly normal thing to have in a view... [08:58] +1 [08:59] larsu: any idea why the version format of qmenumodel has suddenly changed? [09:01] dednick: no. Do you mean the version of the package? [09:01] dednick: wait, i may not be running the correct code on the phone [09:02] larsu: yeah, it was 0.2.6daily13.06.07-0ubuntu1 , now 0.2.6+13.10.20130812-0ubuntu1 [09:02] dednick: no clue - maybe they changed the format of the daily builds and qmenumodel didn't have a release in a while... [09:03] larsu: yeah, i guess so [09:04] larsu: i think we should bump it up to 0.2.7 [09:05] dednick: sure, feel free to do so (I'm a bit preoccupied with i-messages right now -- I'll by on vacation from tomorrow) [09:05] larsu: ok :) i'll stop bugging you in that cas [09:06] dednick: heh, didn't mean it like that. Just wanted to let you know I don't have time for changing the version ;) [09:06] larsu: sure. i meant it in hurry up and finish messaging ;) [09:06] ha :) [09:09] Saviq: how do you actually go about bumping package version up? is all you need to to is add new section to debian/changelog? [09:09] dednick, yes [09:09] dednick, just add a *released* version entry [09:09] dednick, just the upstream version (i.e. 7.83.1 in the case of unity8) [09:10] dednick, daily release takes care of the rest [09:10] Saviq: ok, thanks [09:10] we do all the heavy work ;) [09:11] prefer still -0ubuntu1 though if you want to easily rebuild locally the package [09:11] (but that's just for your own ease) [09:17] Saviq: what's your opinion on https://code.launchpad.net/~aacid/unity8/less_repeated_stuff_cmake/+merge/179875 ? [09:17] ok or ko? [09:17] quite interesting that unity8 is version 7... ;) [09:18] dednick: I think its common to have 7.80 for 8-alpha, 7.90 for 8-beta etc [09:18] mzanetti: ahh i see. like v0.X [09:18] dednick, yup [09:19] tsdgeos, I'm torn, on the one hand I agree it's spurious, on the other I like the plugins self-contained, so that you can effectively take 'em out and the build should work ~without intervention [09:20] tsdgeos, but it's not really the case anyway, probably [09:20] Saviq: probably you'd need stuff from the toplevel cmakelists too [09:20] as the project() call and whatnot [09:21] tsdgeos, yeah, exactly [09:21] tsdgeos, so yeah - ok [09:21] okidoki [09:23] * mzanetti agrees [09:24] Saviq: this right? https://code.launchpad.net/~nick-dedekind/qmenumodel/bump-0.2.7/+merge/179879 [09:24] dednick, +1, maybe a slightly more descriptive changelog entry would be useful, though [09:24] :) yeah, i was just thinking that [09:25] dednick, commented on lp:~nick-dedekind/unity8/indicators-behaviour.tabs [09:26] * mzanetti loves this ^^ [09:26] mzanetti, +1 [09:27] looks like a simple change (design-wise) but improves usability in so many ways [09:27] mzanetti, we're actually supposed to get the same for the dash [09:27] mzanetti, to replace the bottom bar [09:28] hmm... not sure how that will look. but lets see. [09:29] Saviq: thanks [09:30] mzanetti, like the video you're getting in your email, more or less [09:30] tsdgeos, you got it, too, but don't worry about it for now [09:30] mzanetti, tsdgeos, dednick you might be interested in the fallout from IoM, though [09:30] https://docs.google.com/a/canonical.com/document/d/1ypEDFzqRcYXKNcEVzAEG5KgdCWfgwiUHlmP5NledYzo/edit [09:31] not huge, but good to keep in mind [09:31] i dont like the sound of "fallout" [09:31] hehe [09:31] mzanetti, I think background property should be readonly alias of backgroundImage.source [09:32] Cimi: we write it in line 448 [09:32] is a self approval of a version bump ok? :) [09:33] dednick, approved [09:33] mzanetti, that's bad [09:33] Saviq: thanks [09:33] Cimi: I don't see why [09:33] mzanetti, because it's not in sync with the shell [09:33] dednick, btw, it might make sense next time to explain why the release is needed (if there is such a reason ;)) [09:33] mzanetti, so my branch breaks it [09:33] mzanetti, I am using source independently [09:33] source of crossfadeimage btw [09:34] Cimi: have to go to a meeting now. I'll ping you in 20 mins for this [09:34] Saviq: "I like to give myself work" [09:35] and annoy people with useless MPs [09:35] dednick, ;) [09:35] mzanetti, what's background property? [09:35] who needs it? [09:38] how many edges do we have on this phone? i could swear we only have 4 to play with... [09:38] rotfl [09:38] dednick, it's *the* edge [09:38] :D [09:38] there can be only one [09:38] dednick, but there's 4 on the back, too! [09:39] there's at least 2 on the left side :D [09:39] dednick, and with the Edge's design, there's like 24 total! [09:39] we going to have underhand touch detect? stroke with right index finger opens the calculator. [10:14] mzanetti, you back [10:14] mzanetti, I think it's used by the lockscreen [10:14] Cimi: yes, now I am [10:14] mzanetti, the real shell background don't need it [10:14] Cimi: what exactly is the issue? [10:15] mzanetti, I was wondering if we could have dropped the background property of shekk [10:15] shell [10:16] Cimi: right... there seems to be some issue indeed [10:17] * Saviq biab [10:17] Cimi: I'd vote for making it an alias [10:17] mzanetti, why does the greeter change the property? [10:17] mzanetti, I believe only for the lockscreen [10:17] Cimi: the background changes if you select another user [10:17] actually I don't know what we should do for multi user [10:18] mzanetti, but what happens when it loads gsettings? [10:18] it shouldn't change the bg of the shell [10:18] I'm confused [10:18] do you have a tablet to test? [10:18] Cimi: you can test on the desktop [10:18] ok [10:19] but how about multiple users with multiple gsettings? [10:19] Cimi: good question :/ [10:19] the greeter background is stored in accountsservice, not gsettings [10:19] well accountsservice has a list of users with properties [10:19] keyboard layout, background image [10:19] Cimi: I think this issue will go away once the greeter is split out of the shell [10:19] the greeter read it from there [10:19] the desktop greeter [10:20] Cimi: as the greeter just updates its own image with the stuff from accountservice as seb128 said. [10:20] that's the plan for the new greeter as well afaik [10:20] Cimi: and once you log in, the shell can load the background from gsetting and that won't change through the greeter any more [10:22] Cimi: so I think we can just drop line 446-449 [10:23] andyrock: btw! I looked at raring right now, and I see you made a fix to compiz that disables the default show desktop shortcut [10:33] mzanetti, but that way it won't change when you change user? [10:35] andyrock: I'm not entirely sure this is SRUable though [10:36] andyrock: I know that the shortcut works only for the first time, but still it would probably be best to simply fix showdesktop [10:36] Cimi: yes, because in a real multi user environment, ever user runs his own instance of the shell [10:36] andyrock: or is it broken fundamentally beyond fixing ;) ? [10:37] mzanetti, so when you change user it changes gsettings? [10:37] Cimi: no... ever user runs his own instance of the shell [10:37] so if you change user, a second shell will be launched [10:37] mzanetti, I mean in the greeter [10:38] mzanetti, how can the greeter change bg when you scroll users? [10:38] Cimi: no... the greeter just reads the stuff from logind or whatever which in turn reads if from the gsettings for the selected user [10:39] Cimi: then when you log in with a user, a new shell gets started which just reads the setting from gsettings. [10:39] Cimi: makes sense? [10:40] mzanetti, the lock screen seems to read shell.background [10:40] Cimi: yeah, that's fine... [10:40] lockscreen != greeter? [10:40] Cimi: yep [10:40] ah ok [10:40] Cimi: the lockscreen's background will go away eventually once we can do blur [10:42] mzanetti, ok pushed [10:42] to unity8? :P [10:42] https://code.launchpad.net/~unity-team/unity8/unity8.background-gsettings-fix/+merge/179884 [10:45] small comment added [10:47] jamesh: hm, we probably need to create a project for media-scanner, right? [10:47] jamesh: since I doubt hollywood is the right name [10:53] mzanetti, done [10:53] mzanetti, I merged the two tests [10:54] Cimi: good [10:54] mzanetti, nope [10:54] ? [10:54] mzanetti, wrong code hold on :) [10:54] right :) [10:55] mzanetti, ok now done [10:55] mzanetti, had to check for expectedUrl not just url [10:55] Cimi: ok. giving it a last spin before approving. code looks good [10:56] Cimi: one note: In general its more safe to use removeAll() instead of removeOne() in such unregisterObject() methods [10:56] Cimi: won't do any harm here, but if that api is used by someone else you get nasty effects if someone calls register() twice and unregister() only once [10:56] ok [10:56] good to know [10:57] Cimi: removeAll() might be a bit slower tho... of in case you're doing performance critical stuff its again a tradeoff [11:00] sil2100, sorry was afk [11:00] let me check [11:01] sil2100, i proposed a fix for 0.9.9 [11:01] sil2100, because 0.9.9 is still saucy ;) [11:02] hmmm [11:02] Saviq, https://code.launchpad.net/ubuntu-settings-components has packaginf [11:02] Cimi, k === dpm_ is now known as dpm [11:03] andyrock: sadly, 0.9.9 is for raring ;) [11:03] andyrock: target_branch: lp:compiz/0.9.9 [11:03] tsdgeos: any news on the ListView positioning with snapping in the center? [11:03] sil2100, but it's saucy too right? [11:04] andyrock: well... for saucy we're using no branch right now, maybe it was based on 0.9.9 but we're not doing any releases [11:04] *for saucy [11:04] mzanetti: ah no, didn't do anything, i was about to start wokring on it last week but then i realized i wasn't sure what would be "good enough" for us [11:04] andyrock: we'll be using 0.9.10 for saucy [11:04] andyrock: so every commit now that goes into 0.9.9 is for raring only [11:04] tsdgeos: do you need anything from me in that regard? [11:04] andyrock: since that's the branch that's targetted for daily release for raring ;) [11:04] sil2100, cool... so just revert that rev [11:04] andyrock: sooo, I think we need to consult this with the SRU guys if it's ok or not [11:05] mzanetti: i think i do, that's why i stopped, but don't remember now, i'll try to have a second look again today and ask you something if needed [11:05] in that case I need to propose a branch for unity 7.0 too [11:05] unity raring branch [11:05] tsdgeos: thanks [11:08] sil2100, in that case we should backport http://bazaar.launchpad.net/~unity-team/unity/trunk/revision/3443 [11:09] we are not disabling the show_desktop [11:09] we just moved the option [11:09] Saviq, can we chat on the backend for welcome wizard? [11:09] sil2100, we have good reasons :D [11:09] andyrock: ah, so we moved it out of compiz to unity? [11:09] Cimi, sure, should have seb with us, though, and he doesn't seem around [11:10] lunch time in france [11:10] andyrock: if so, it would be nice to have the disable-shortcut-for-compiz and unity-add-showdesktop-to-unity in raring ;) [11:10] Cimi, probably, can you throw something in our calendars? === MacSlow is now known as MacSlow|lunch [11:10] sil2100, do you want me to propose a branch for unity 7.0? [11:10] Saviq, ok [11:10] andyrock: would be awesome [11:10] sil2100, can I do it after lunch? [11:11] andyrock: of course! I'll be going out for lunch pretty soon as well [11:13] Saviq: greyback: i am going to put some time on the category expansion animation, am i right that we agreed that we'd only animate the delegate height when no contentY position was needed? [11:23] tsdgeos: just thinking out loud: if the bottom of the contracted category is off-screen, there's no need to animate the height, just expand it immediately and animate contentY (if needed). [11:24] tsdgeos: if bottom of contacted category on screen, I think animating both height & contentY would be visually ok. [11:25] you see something similar on iOS, when viewing a folder of apps. [11:26] greyback, I think it's more or less simple: if you can see the animation - animate, otherwise - don't [11:26] tsdgeos, ↑ rather [11:26] :D [11:26] ok [11:43] Cimi: there is still another issue :/ [11:44] Cimi: dunno why, but if the gSetting has an invalid picture url at startup the background doesn't work :/ [11:44] Cimi: weird thing is, the Greeter works, it's just the background that doesn't [11:44] Cimi: I gave up now and added the dummy picture back to determine if the url is valid or not [11:51] Saviq: one flacky test: http://10.97.0.1:8080/job/autopilot-saucy-daily_release/960/ [11:52] didrocks, yeah, we saw that before, need to add some debugging there [11:52] Saviq: we'll maybe wait for next daily run then [11:53] (2h from now) [11:53] didrocks, it's just typing "password\n", no idea how it fails :/ [11:53] Saviq: the video doesn't help? [11:53] don't you think it's just timing out? [11:53] didrocks, yeah it does - it shows that the password was wrong [11:53] didrocks, which is *weird* [11:53] interesting [11:53] didrocks, that's why I'm saying we need more debugging in there [11:54] didrocks, but obviously it'll just work on the next run [11:54] ok, let's wait for now [11:56] sil2100: FYI ^ === MacSlow|lunch is now known as MacSlow [12:16] Looking! [12:19] sil2100, didrocks in autopilot-saucy-daily_release job 956 bamfdaemon crashed (that's why all those failures) === _salem is now known as salem_ [12:20] do you know how can I get the .crash file? [12:20] should be easy to debug the issue [12:20] and fix it [12:27] andyrock: I don't find it, but you think that we everytime have high level of failures because of bamf? [12:27] didrocks, sil2100 https://code.launchpad.net/~saviq/unity8/autopilot-debug-passphrase-entry/+merge/179915 [12:27] (for the previous runs) [12:27] andyrock: I would then look at https://errors.ubuntu.com/?release=Ubuntu%2013.10&package=bamf&period=month [12:27] seems there is a popular crash [12:27] Saviq: are .debug() showed by the default level execution during AP? [12:28] didrocks, are you not running with -v ? [12:28] Saviq: it's not me, it's autopilot [12:28] so not sure how it's running you :) [12:28] didrocks, are you not running autopilot with -v? [12:28] we do [12:28] didrocks, then that's it - it's autopilot debugging [12:29] didrocks, not "our" debugging [12:29] Saviq: sorry, I read too quickly, I thought the process part was in unity8 [12:29] ok, approving then [12:29] didrocks, I expect that enter is pressed (and authentication requested) before the key was processed properly [12:29] didrocks, so a simple Eventually(Equals()) should fix that [12:30] didrocks, but let's see what (if anything) we get from the debug logging [12:30] Saviq: probably yeah [12:30] right [12:33] didrocks: still, a pretty nice outcome - just one red light ;p [12:34] right ;) [12:34] * didrocks stare at Saviq! [12:34] almost 2 runs green in a row [12:34] * Saviq hides [12:34] :) === mdeslaur_ is now known as mdeslaur === jhodapp|afk is now known as jhodapp [13:04] tsdgeos, ok, the app preview seems to be fine, it's the backends that are still not up to par [13:04] tsdgeos, I think we can land it [13:06] tsdgeos, it will at least make the guys fix it then :) [13:06] sil2100: hi. About the name for the mediascanner, currently the "hollywood" name is exposed in the name of the grilo plugin [13:07] sil2100: the packaging in the branch uses ubuntutv-mediascanner for the source package name, and just "mediascanner" in the binary package names [13:07] Saviq: ok for me [13:07] * tsdgeos approves [13:14] mzanetti, I just read the MR [13:16] Cimi: ok... what do you think? [13:16] Cimi: I kinda hate the fact that we need that dummy Image... [13:17] Cimi: but I've found another issue and spend quite a lot of time trying to fix it... haven't been able to do so without the dummy image [13:18] mzanetti, first of all, why not gsettingscontroller? [13:18] ? [13:18] Cimi: ah, you mean the failed test? [13:18] the one you put the log [13:19] Cimi: didn't investigate why it happens, but tests fail if I run them locally [13:19] Cimi: most likely the mock plugin isn't installed, or it is installed, but the real one has preference or something like this [13:19] mzanetti, you might need the import? [13:20] Cimi: does it work for you if you do this: [13:20] cd builddir && make testShell [13:20] ? [13:20] yes [13:20] let me branch again [13:20] I get this error in that case [13:26] mzanetti, might be wrong import sequence [13:26] order [13:27] still compiling though === pete-woods is now known as pete-woods-lunch [13:27] mzanetti, works here on a branch new branch [13:28] *brand new [13:32] Cimi, mzanetti, tsdgeos, dednick_, nic-doffay standu [13:32] p [13:33] Saviq, mistake, I thought today was wednesday :P [13:33] i was there [13:33] what happend? [13:33] Saviq, can we have today the meeting for the backend? [13:33] Cimi, sure, it's you who put it in for tomorrow :) [13:33] Saviq, lol saw the wrong day :) [13:38] someone knows where web is? [13:38] seb [13:41] Cimi: travelling to Berlin [13:41] he will be back on IRC in ~1h I guess [13:41] didrocks, ah ok… you know when will he be online/available? [13:42] didrocks, and working? [13:42] yeah [13:42] didrocks, or holiday? [13:42] ok [13:42] if seb is on IRC, he's working ;) [13:42] Saviq, let's move the meeting later then ^ [13:50] eh. sorry. [13:50] didrocks, pretty much like sms pillaz [13:50] hah [14:12] Cimi: I fixed the comment [14:13] Cimi: where is the whitespace issue? the whitespace test passes here [14:18] mzanetti, new line [14:19] mzanetti, empty line in gsettings [14:20] Cimi: ok. fixed... I've no clue why the tests don't pass here... [14:20] who can test? [14:20] tsdgeos: can you? [14:21] mzanetti: what? [14:21] https://code.launchpad.net/~unity-team/unity8/unity8.background-gsettings-fix/+merge/179884 [14:21] tsdgeos: does "make testShell" work for you? [14:21] in general or with that branch? [14:21] this branch [14:22] * tsdgeos tests [14:23] jamesh: ping [14:24] sil2100: pong [14:24] jamesh: quick question - media-scanner will be named media-scanner, right? Not hollywood? [14:25] jamesh: since I saw unity-scope-hollywood [14:26] jamesh: also, do we intend to daily-release hollywood/media-scanner ? === alan_g is now known as alan_g|tea [14:27] mzanetti: fails [14:27] FAIL! : qmltestrunner::Shell::test_wallpaper(red) Uncaught exception: GSettingsController is not defined [14:27] sil2100: I don't really know about the naming. The "hollywood" naming is exposed in the Grilo plugin name, which does make up part of the public API [14:27]