=== duflu_ is now known as duflu [07:37] mzanetti: Saviq: hey, did you submit any talk for FOSDEM? I don't see them now that the proposal period is over [07:46] didrocks, yeah, I did, and bregma, too [07:52] Saviq: what's the title of your conference? [07:52] I see bregma's one [07:53] didrocks, "Ubuntu on phones and beyond" [07:53] Saviq: ah, typo in your name :) [07:53] Saviq: ok, all good, it's listed ;) [07:53] * Saviq was worried did not fill the forms right, they're rather convoluted [07:55] Saviq: I won't even mention the backend-voting-side :p [07:55] (we are back to using some google spreadsheet for easyness, due to that) [07:58] ;) [07:58] spreadshits FTW! [08:08] heh [08:34] tsdgeos, hey, could you work today on making the apps scope non-unfavouritable? [08:35] yep [08:35] i was talking with xavigarcia about some card stuff [08:35] i will do that now [08:36] and then maybe fix the race in qtmir we indentified with gerry yesterday [08:36] that = non-unfavoritable [08:37] bign landing btw \o/ [08:37] -n [08:37] now if Cimi would review the moreAsyncDash ;) [08:38] Saviq: do you know if ci is back ? [08:39] tsdgeos, yeah, it should mostly be back [08:39] tsdgeos, if you're touching cards, maybe you can have a look at bug #1393008 [08:39] bug 1393008 in unity8 (Ubuntu) "Horizontal cards don't get backgrounds" [Critical,Triaged] https://launchpad.net/bugs/1393008 [08:39] yeah [08:39] that was what we were talking about [08:39] oh good [08:39] that's not a bug according to spec [08:39] : [08:39] :D [08:40] orly? [08:40] yeah [08:40] well if you read the spec in a certain way [08:40] of course [08:40] ah ok ;) [08:40] thostr_, says it was working before ;) [08:40] never [08:40] the code says [08:40] hasBackground = !isHorizontal && stuff; [08:41] don't think that changed lately [08:41] heh [08:41] Saviq: can you log into http://s-jenkins.ubuntu-ci:8080/job/unity8-ci/5023/rebuild ? [08:41] well log [08:41] does it even show up? [08:43] tsdgeos, do you have the new VPN set up? [08:43] probably not [08:43] tsdgeos, kicked that rebuild [08:44] tsdgeos, ok and as far as the spec goes, where did we read it that horizontal cards can't have backgrounds? [08:44] i'm writing it down [08:45] thanks [08:45] done [08:48] Saviq: imho our bug should be back to incomplete and add ubuntuux there for them to decide if what i understand is what they wanted to write [08:48] tsdgeos, sure [08:48] and bring back the designer that wrote that ^_^ [08:49] tsdgeos, yeah, you're right [08:49] tsdgeos, it clearly says it can't be shaped, so no background [08:51] well that's how we coded it yes [08:59] pstolowski, I'll start prepping the unity8 counterpart for dash bottom edge [09:00] pstolowski, but we'll have to wait for apps-special-after-all [09:04] Saviq, yeah, i'm going to prepare MP for that [09:05] pstolowski, you want to ensure that plugin side? [09:06] in case our 1 line of code fails :) [09:07] tsdgeos, truth is we might need to have multiple special scopes for OEMs, or at least replacing the apps one [09:07] tsdgeos, but then we don't need to implement that right now [09:08] i'll go back [09:08] to what we had [09:08] i.e. lame [09:08] showStar: model.scopeId != "clickscope" && (root.isFavoritesFeed || root.isOtherFeed) [09:08] damn, that also makes it unmovable [09:09] and we want it movable, right? [09:10] hmmm [09:11] and we need to make it unfavoritable also from das [09:11] h [09:11] Saviq, yes, I had it before in the plugin as a safeguard, i just need to find and revert that commit [09:11] is that something we do [09:11] or that the plugin does [09:11] * tsdgeos checks [09:11] Saviq, but you also need to special case it in the shell (not to show star) [09:12] something we do [09:12] pstolowski, yup [09:12] it's again something the scope should tell us [09:12] if we're going to make this generic for the FUTURE [09:12] trueth [09:13] or at least the registry [09:14] btw i just found a bug [09:14] not sure if it's ui or backend [09:14] but if you open manage dash, go to edit mode, reorder, leave edit mode and now click some scope, it will bring you to the "before reordering" position [09:15] pstolowski: if i had to say something i'd say that's on your plate ↑ [09:16] but maybe not :D [09:16] tsdgeos, worked fine here [09:16] tsdgeos, is "some scope" favourited or not? [09:16] otp [09:16] Saviq: yes, otherwise that's the point of the reordering breaking it [09:17] and now it works [09:17] oh there's one more thing [09:17] :S [09:17] the scope gets refreshed every time I click in Manage Dash [09:17] sounds wasteful [09:17] ah made it fail again [09:17] it may be on our side [09:18] will debug more [09:19] meanwhile https://code.launchpad.net/~aacid/unity8/apps-special-after-all/+merge/244548 [09:20] Saviq: ok my way to reproduce, have 3 fav scopes, move 2 to 1, click on 3rd [09:20] end in first [09:21] tsdgeos, confirmed [09:21] and it's probably our [09:21] s [09:21] looking at it now [09:22] tsdgeos, could we have a test for apps-special-after-all? [09:22] you want to test that !== works? :D [09:24] tsdgeos, no, I want to test that there is no star next to the click scope [09:24] or in the click scope's header [09:24] Saviq: which is basically testing !==, it will still break when backend decides to rename the click scope to "mooscope" and we're still checking for !== "clickscope" [09:25] tsdgeos, great, we'll have a failure that we might even be able to catch in proposed thanks to that test being executed on autopkgtest [09:25] soon [09:26] Saviq: no, we won't have a failure, why would we? [09:26] tsdgeos, because if they changed, the star would be there when it shouldn't be? [09:26] it's not like we use real scopes in tests [09:26] tsdgeos, oh well, right [09:26] so as i said, we're testing that !== works [09:26] meh [09:26] but if it makes you happy [09:26] it's 10 minutes of my time [09:26] so doing [09:27] tsdgeos, I think we should have that test indeed [09:27] tsdgeos, and while right now it might not make all kinds of sense [09:28] it hopefully will once we have a property on scopes that decide whether they can be unfav'ed or not [09:28] and nobody will remember to update the tests :D [09:30] Saviq, tsdgeos i guess I can add new role for such property in OverviewResultsModel [09:31] pstolowski, not enough [09:31] pstolowski, we need to know it for the scope as well, to show the star in the header or not [09:31] Saviq, tsdgeos and I'd special case apps in shell plugin for now (no configurability), but at least it won't bubble up [09:31] ah right [09:33] Saviq, so, new property in Scope interface? again special cased for now in shell plugin [09:33] ? [09:33] i'd leave it as it is now tbh [09:33] pstolowski, ↑ [09:33] pstolowski, once we know we do need to replace / add special scopes [09:34] is when we'll deal with that [09:35] Saviq, tsdgeos ack [09:47] Hola! [09:52] Saviq: tests added [09:52] tsdgeos, thanks [09:57] tsdgeos, did you push? [09:57] nope :D [09:57] and even then, there's a bug [09:57] k [09:59] now :) [09:59] Saviq: ↑ [10:01] tx [10:10] i hate listview [10:31] tsdgeos, that's re: went to the wrong scope? [10:31] ywah, i know what's wrong [10:31] my workaround for a listview "bug" breaks it [10:31] i'll add a few tests [10:32] which we have none [10:32] to at least make sure those two things work [10:32] in a separate MR [10:55] * Saviq starts being afraid of cherry-picking... === _salem is now known as salem_ [11:27] UUGH [11:27] yeah [11:27] Saviq: i appreciate you doing the cherry-picking, but it's most probably something the guy that did the original branch should do [11:28] Saviq: https://code.launchpad.net/~aacid/unity8/bottom_list_drag/+merge/244563 has the fix for the issue i found with dragging + click with a test for it and some other stuff [11:28] tsdgeos, it's not even that [11:28] tsdgeos, the real problem is the ordering [11:28] ah [11:28] tsdgeos, like now we're landing the bottom edge [11:28] tsdgeos, but there were things that changed for the performance fixes [11:28] right [11:29] tsdgeos, so it's conflicting (a bit) now, and will conflict again later, not to mention that I might actually break things [11:29] tsdgeos, and then... there's unity-api, which has a launcher change that we don't (yet) want in rtm [11:29] tsdgeos, so I need to cherry-pick that and do something weird with unity-api's version so that we don't break that... [11:30] so that we can actually reconcile in the futuer [11:30] but maybe it's gotta be what it's gotta be... [11:32] and then there's you not being able to make your mind on how the test should work and changing it back'n'forth ;P [11:32] but well, the test passed, so it must be right, right? [11:36] totally [11:36] :D [11:36] Saviq: which test? [11:37] tsdgeos, test_manage_dash_search_temp_scope [11:37] ah don't worry about that one, we don't support search yet anyway [11:37] so the skip() at the beginning is the important part [11:37] tsdgeos, that's searching in *temp* scope [11:37] tsdgeos, not in manage dash? [11:38] no [11:38] is searching in manage dash [11:38] ok [11:38] that's why it's called manage_dash_search [11:38] tsdgeos, anyway, there was no tryCompareFunction(), then you added it when doing "less delegates" and then removed it again in bottom swipe :P [11:38] he he [11:38] tsdgeos, it's fine, I'm just trying to keep the diff sane [11:40] mzanetti, can you please whip up a quick vivid silo with tsdgeos's fixes for the bottom edge? [11:40] kk [11:40] * Saviq tries not to land in rtm before vivid [11:40] mzanetti, https://code.launchpad.net/~aacid/unity8/apps-special-after-all/+merge/244548 and https://code.launchpad.net/~aacid/unity8/bottom_list_drag/+merge/244563 [11:40] mzanetti, if you could review the latter one, too, would be obliged [11:41] ah and https://code.launchpad.net/~aacid/unity8/scopeListPageHeaderScopeStyle/+merge/244293 too [11:41] tsdgeos, that's it, right ↑ ? [11:43] greyback__, just discovered hash -d in znc, better'n'wd ;) [11:45] mzanetti: Saviq: yeah that one too, though effectively it does nothing at the moment it's "better code" [11:45] Saviq: Sorry, I've no idea what that sentence means. znc the IRC bouncer? [11:45] greyback__, zsh, of course ;) [11:46] greyback__, hash -d alias=/some/dir/foo [11:46] greyback__, cd ~alias [11:46] Saviq: there's a "jump" plugin I like a lot. [11:47] ets you bookmark directories with a name, and then jump from one to another. Is very similar [11:48] greyback__, sounds the same :), but IIRC jump lets you do more [11:48] and have it aliased to "j" so I can just "j qtmir" to land in my qtmir root [11:51] yup [11:51] tsdgeos, https://code.launchpad.net/~unity-team/unity-api/rtm-14.09-staging/+merge/244567 please [11:54] pstolowski, is there a citrain row for bottom dash rtm yet? [11:54] Saviq: looks good, want me to approve? [11:55] tsdgeos, yup [11:55] Saviq, no, but i have another silo where i'm currently building two other fixes (including department jumping fix) [11:55] pstolowski, that's for vivid though? [11:55] Saviq, and it has shell plugin [11:55] Saviq, yes vivid [11:55] pstolowski, I'm talking rtm [11:56] this silo should only include bottom edge things IMO [11:56] Saviq, agree. but we need todays fix in vivid first [11:57] pstolowski, do you have rtm branches yet? [11:58] pstolowski, btw, you know we're limited to "3 fixes" per rtm silo? could we not land bottom edge separately? [11:59] Saviq, i only have rtm MP for plugin and it will need updating with todays fix [12:00] tsdgeos: I don't really understand what the issue is this one fixes: https://code.launchpad.net/~aacid/unity8/bottom_list_drag/+merge/244563 [12:00] pstolowski, sure, I don't have complete MPs yet, either [12:01] pstolowski, but want to start early [12:01] Saviq, yeah, I know of limits. actually, is it 3 MPs or 3 bugfixes? [12:01] :) [12:01] pstolowski, neither, it's 3 "fixes" [12:01] pstolowski, it can be more or less MPs [12:01] Saviq, yeah... i already prepared target rtm-14.09 branch for unity-api (which wasn't branched yet) [12:02] mzanetti: without the patch do this, get to manage dash, move second scope to first, click third scope, see how you're not in the third scope as you should be [12:02] pstolowski, yeah, got an MP for that up already [12:02] cool [12:02] tsdgeos: thanks [12:04] pstolowski, https://code.launchpad.net/~stolowski/unity-scopes-shell/manage-dash-rtm/+merge/244117 is the one? [12:04] mzanetti: note there may be a race in the test, i got it failing 1 every 20 or so if i loop it [12:04] tsdgeos: you fixing that? [12:04] having a look yes [12:04] Saviq, also, I've rtm silo 3 (rows #23) that needs qa sign off, or will block other landings of shell plugin [12:04] Saviq, yes, that's it [12:05] Saviq, it may also need dependency update [12:05] pstolowski, you'll need to chase product folk to accept https://launchpad.net/bugs/1394155 [12:05] Launchpad bug 1394155 in unity-scopes-shell (Ubuntu) "Crash of unity8-dash possible through Today scope" [Critical,In progress] [12:06] pstolowski, we'll need -api for the new silo too, right? [12:08] Saviq, unity-scopes-api? yes [12:09] tsdgeos: this one fails here all the time: test_manage_dash_move_current_click_other [12:09] mzanetti: hmmm [12:09] with the patch? [12:10] this branch: https://code.launchpad.net/~aacid/unity8/bottom_list_drag/+merge/244563 [12:10] you didn't revert the change in qml/Dash/Dash.qml [12:10] right? [12:10] can you add console.log("setCurrentScope", scopeId, scopeIndex); to Dash.qml::setCurrentScope after the for loop? [12:11] no, didn't revert that [12:11] trying the debug print [12:13] tsdgeos: qml: setCurrentScope MockScope1 1 [12:13] right [12:13] that's wrong [12:13] should be MockScope5 2 [12:14] you're probably hitting my race more often :D [12:14] so far 100% [12:14] i have a tentative fix, let me push it and see if it fixes it for you [12:14] ack [12:14] pull! [12:17] tsdgeos: pass [12:18] good [12:18] it's weird how something that was just a one off here was there all the time [12:18] yeah, I ran it as testDash [12:18] no xvfb [12:19] anyone knows where that "Using blocking call!!!" message is coming from? [12:19] I see that in my apps too [12:20] ok. passes in xbfb too. fixes the issue [12:21] mzanetti, never saw that [12:21] but judging by the 3 exclamation marks... Wellark ↑? ;P [12:21] lol [12:22] actually it's just one... the double ll! gave me that impression [12:22] rofl [12:22] *that* funny? [12:24] Saviq: wdyt, enough? Or should I implement a signal to be emitted to unity8, so unity does the sigstop: https://code.launchpad.net/~gerboland/qtmir/emitSigstopAtCorrectTime/+merge/244570 [12:25] greyback__, good enough for now [12:25] yep, thought was enough for patch [12:25] greyback__, later I want to be more granular [12:25] sure [12:25] greyback__, but that might as well wait for systemd [12:25] that'll fix everything [12:26] yeah, we'll be done then [12:28] tsdgeos: https://code.launchpad.net/~aacid/unity8/scopeListPageHeaderScopeStyle/+merge/244293/comments/603140 [12:28] mzanetti: honestly i think it's testing again that qml bindings work [12:28] but i can make it happen [12:29] tsdgeos: I'm rather thinking of the chain of qml bindings... if it's just easy please add a small one. if you need to start with creating a mock framework, it's probably not worth it === greyback__ is now known as greyback [12:30] tsdgeos: a small glitch I just noticed: go to manage dash, click on a temp scope, see how the arrow label at the bottom slides in and fades out... probably shouldn't do that [12:30] not related to the above branches I think [12:30] should be fine [12:30] yeah i saw that the other day [12:31] it's a thing between setting the scope and bla bla [12:31] could be fixed [12:31] mzanetti: would you mind creating a bug for me? [12:31] sure [12:31] tx [12:32] tsdgeos, did you notice how the scope header icons go transparent when you start dragging from the bottom? [12:33] tsdgeos: https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1401869 [12:33] Launchpad bug 1401869 in unity8 (Ubuntu) "bottom edge arrow label comes in when it shouldn't" [Undecided,New] [12:33] Saviq: i'd say is because of the disabling [12:33] Saviq: gotta go for an hour. I've tested this branch, works fine: https://code.launchpad.net/~aacid/unity8/scopeListPageHeaderScopeStyle/+merge/244293 [12:33] Saviq: however, we agreed to add a mock and test for it [12:33] Saviq: so once that's there, feel free to approve [12:33] Saviq: not sure how that would be fixed [12:34] the other branch is approved already [12:34] tsdgeos, right, makes sense [12:34] mzanetti, yeah, thanks, I'll take over now [12:34] kk. see you in a bit [12:38] Cimi, you there? [12:41] mzanetti: Saviq: wops, actually yep, i think it's wrong ^_^ [12:41] Saviq, ubuntu-ui-toolkit with my dialog changes have landed, btw [12:41] dandrader, coolz [12:42] Saviq, bumped unity8 dependency accordingly [12:42] dandrader, I don't think you should have [12:42] dandrader, as you said, it's not actually a requirement [12:42] dandrader, like things will work without it [12:42] dandrader, and then improve when you get new UITK [12:43] dandrader, what I mean by that is dependencies generally represent hard requirements [12:44] Saviq, right... but does it do any harm? "make tryDialogs" look proper at least [12:44] (with the new uitk) [12:44] when you rotate things [12:44] dandrader, do the test depend on the new UITK? [12:45] dandrader, the "harm" is unnecessary packaging changes, and suggesting that things won't work otherwise [12:45] Saviq, no strictly. but it will *look* broken without it [12:45] well, I will revert it then [12:46] Saviq, done [12:47] dandrader, thanks [12:47] Saviq, mzanetti: what!? [12:47] :) [12:48] where is that message? [12:48] Wellark, log output from apps sometime [12:48] greyback, I'll pop qtmir into the silo as well [12:49] ah no, you have one already [12:49] Saviq: well, I could write something like that, sure [12:49] greyback, did you ping folks for publishing silo 14? [12:49] but I usually augment my shit with a __PRETTY_FUNCTION__ or something :) === dandrader is now known as dandrader|afk [12:57] anyone have an idea why this renderer wouldn't show up with a background? http://paste.ubuntu.com/9488814/ [13:04] lol [13:04] cwayne_, bug #1393008 [13:04] bug 1393008 in unity8 (Ubuntu) "Horizontal cards don't get backgrounds" [Undecided,Incomplete] https://launchpad.net/bugs/1393008 [13:04] cwayne_, horizontal == no shape == no background, per spec [13:04] that is [13:04] (horizontal - summary) == no shape [13:05] mzanetti: Saviq: ok, make this work https://code.launchpad.net/~aacid/unity8/scopeListPageHeaderScopeStyle/+merge/244578 good call for testing ;) [13:06] now food! [13:14] Saviq: I just set as tested, I didn't ping [13:15] greyback, yeah, it helps to ping, gets them off their a$$es quicker [13:15] greyback, but I think we're low on landing team atm [13:15] * Saviq hates to be blocked by that [13:15] Saviq: ok, wasn't aware, but wasn't in rush either [13:17] Saviq: I removed one qtmir branch you added, as it's caused visual errors. Forgot to set it unapproved, done now [13:18] greyback, ok thanks [13:19] Saviq, oh, is that new? this renderer used to work (even if it shouldn't have :P) [13:20] cwayne_, no, the spec is old, we might just not have enforced it until recently (due to some refactoring) [13:20] right, that's what i meant [13:20] i know the specs haven't changed recently :) [13:21] but ok, so if i add even a blank summary it should work then? [13:21] cwayne_, yeah I think so [13:21] * Saviq checks [13:21] didn't seem to work for me, hm [13:22] cwayne_, on that note, yeah [13:22] cwayne_, so there's the bug [13:22] cwayne_, it should work where there is summary [13:23] hm, damnit [13:23] ok [13:28] greyback, should we drop the ::started() altogether then? [13:28] greyback, or is there no default impl? [13:32] Hmm, seems I'm not getting any scopes in the scope window on the new Unity 8 desktop mode. [13:33] Sits there with the circle turning. [13:35] Saviq: empty implementation in keeping with the other empty listeners in the file. Could probably remove the listener class entirely, but that for later MR IMO [13:36] Saviq: and anyway yeah, no default impl [13:36] greyback, kk [13:37] After closing the window and letting unity8-dash restart, it's now to a window that has no scopes. This is the unity8-dash log after restarting: http://pastebin.ubuntu.com/9489239/ === dandrader|afk is now known as dandrader [13:37] Saviq: Any ideas? ^^^^ [13:38] ChrisTownsend, is scope-registry running? [13:39] Saviq: Is "scope-registry" the name of the process? If so, then no. [13:40] ChrisTownsend, it's an upstart job [13:40] ChrisTownsend, process is, you guessed "scoperegistry"... [13:41] Saviq: Yes, "scoperegistry" is running. [13:41] ChrisTownsend, 'gsettings get com.canonical.Unity.Dash favorite-scopes'? [13:42] Saviq: Output: [13:42] $ gsettings get com.canonical.Unity.Dash favorite-scopes [13:42] @as [] [13:42] ChrisTownsend, wonder how you got there [13:42] ChrisTownsend, bottom swipe in the dash and add scopes to favorites [13:43] So it seems it's empty. How did that happen. I only did a dist-upgrade and this is what happened. It was fine yesterday in the pre-window mode. [13:43] ChrisTownsend, you can `gsettings reset` it, too [13:44] Saviq: I'll try that. [13:44] ChrisTownsend, I just dist-upgraded my -next vm and that didn't happen, not sure how you got to that state :/ [13:45] * ChrisTownsend shrugs [13:46] greyback, btw, we could use a test for that sigstop... any idea how to attack that? [13:47] * greyback thought u8 had one [13:47] autopilot/unity8/shell/tests/test_upstart.py [13:49] Saviq: fyi, the reset worked, so we'll chalk it up to an anomaly for now. Thanks for your help. [13:49] greyback, yeah, but that doesn't test the timing of it [13:49] ChrisTownsend, sure, thanks [13:50] greyback, we'd need to somehow add there a check that appmanager is ready when sigstop is emitted [13:52] Saviq: hmm, would need way for AP to be notified when AppMan is created. Am not sure how to do that with AP [13:52] greyback, as long as we get a signal/property, we can wait for that [13:53] greyback, on the AppManager object, that is [13:53] greyback, just tell me what to wait for and I'll hack the test up [13:56] Saviq: there is no signal emitted by AppMan on construction. Only idea I have is figuring out what its parent object will be, and listening for children being added to it to see if one is AppMan [13:57] greyback, it doesn't have Component.onCompleted does it [13:58]