=== greyback__ is now known as greyback [07:19] morning [07:31] Saviq, wohoo https://jenkins.qa.ubuntu.com/job/unity-phablet-qmluitests-utopic/919/ [08:09] Saviq: hey, do you know where I can get a Ubuntu One staging account? [08:16] mzanetti, https://login.staging.ubuntu.com/ [08:17] Cimi, yeah, I know... worst thing is this was weird because of a duplicate name: in CardTool's test... [08:21] dandrader, thanks for your review and sorry for you having to look at this mess, it was a witch hunt initially 'cause of the dupe name: [08:24] Cimi, re: can we have a card without a title, for real, no, but the code was triggered when templates and components were empty (so we might hit it on startup) [08:26] Saviq, I figured you must have been too tired when you wrote it as it was quite late :) [08:27] dandrader, yeah, and I was hunting for the actual failure because obviously it wasn't reproducible locally (when it really was, only different test suite :P) [08:31] MacSlow, hey, what's the state of the fix for bug #1354406 ? [08:31] bug 1354406 in unity8 (Ubuntu) "PIN-locked phone becomes unresponsive after simultaneous alarms" [High,In progress] https://launchpad.net/bugs/1354406 [08:40] Saviq, which timezone are you working? :) [08:41] Cimi, ALL OF 'EM [08:41] sometimes [08:41] hah [08:42] Saviq, alt nav requires a test though [08:42] Cimi, yeah, just getting on that [08:53] greyback, should I report a bug on mir asking for the surface type to be passed to PlacementStrategy::place()? [08:56] dednick: hey. I'd need help on trusted sessions [08:56] have a minute? [08:56] might be an hour though :D [08:57] dandrader: sure [08:57] Saviq, https://code.launchpad.net/~saviq/unity8/preview-customizations/+merge/230736 [08:57] mzanetti: sure [08:57] Cimi, nope [08:57] dednick: ok.. there's this bug that promptsessions are removed on focus change [08:57] dednick: I tracked it down to platforms/mirserver/promptsessionlistener.cpp [08:57] Cimi, if (item.hasOwnProperty("scopeStyle")) item.scopeStyle = root.scopeStyle; [08:58] dednick: PromptSessionListener::prompt_provider_removed is called [08:58] dednick: but from here I'm not sure where to go [08:59] mzanetti: "bug" [08:59] ? [08:59] dednick: so what happens is that such a prompt session is opened [08:59] mzanetti: it's a "feature" actually [08:59] oh [08:59] Cimi, hmm or by now all of them have one indeed [08:59] Cimi, fixing [08:59] dednick: so opinions seem to differ on that: https://bugs.launchpad.net/qtmir/+bug/1355173 [09:00] mzanetti: prompt sessions are explicitly cancelled by qtmir when focus is switched [09:00] Ubuntu bug 1355173 in QtMir "Switching windows with a Trusted Prompt Session active looses the trusted prompt session" [Critical,New] [09:00] mzanetti: perhaps we don't want it anymore though [09:01] dednick: dunno... I got assigned this bug... I said I've no clue what's going on but I will look at it [09:01] greyback, I meant the surface state. but having also surface type might be good as well [09:01] dandrader: the more info the better IMO [09:01] yeah [09:02] dednick: hmm... so assuming this is the wanted behavor [09:02] dednick: should the thing that opens the trusted session be informed? [09:02] dednick: tedg says that the app still thinks the prompt is shown, its just visually not there any more [09:02] mzanetti: have you read the bug comments? :) [09:02] err, yes... [09:03] did I miss something? [09:03] * mzanetti rereads [09:03] mzanetti: well, it explains things there. not a solution though [09:03] oh I see [09:03] right... so this needs to be removed (at least for now) I guess [09:04] mzanetti: teds problem is that the payments is not complete enough to deal with the behaviour [09:04] dednick: well, rather online accounts is not complete enough [09:04] but yes [09:04] mzanetti: so yeah. let me get a codepoint for you [09:05] mzanetti: anything that calls: [09:05] :/ [09:05] fek [09:05] http://bazaar.launchpad.net/~mir-team/qtmir/trunk/view/head:/src/modules/Unity/Application/application.cpp#L439 [09:05] mzanetti: ^ [09:06] actually, probably only the application::setState [09:06] I see [09:06] still want to close on app destruction [09:06] ok... I know my way around in applicationmanager... thank... I tought that call would come from mir [09:07] mzanetti: goes to mir [09:07] oh ok... explains things [09:08] mzanetti: generally unity8 shouldn't be worrying about stopping prompt sessions, but this was a requested feature for the time being [09:09] so I guess the "solution" for now is just to comment away that soptPromptSession call and add a FIXME on top to reenable it when UOA is ready for trusted sessions [09:09] * mzanetti tries [09:09] mzanetti: yep. think so. although not sure how this is going to affect other things. [09:09] greyback, https://bugs.launchpad.net/mir/+bug/1357259 [09:09] Ubuntu bug 1357259 in Mir "PlacementStrategy::place() should be given the surface type and state" [Undecided,New] [09:09] dednick: well, I think this is the only place that uses trusted session yet [09:10] or do you know about others? [09:10] dandrader: ta [09:10] mzanetti: it was more of a "guard" than a feature. since some prompt helpers may not support dealing with more than one prompt request at a time. [09:10] hmm... ok [09:10] mzanetti: actually, i think a number of things do. camera, maps, etc. [09:11] dednick: maps? [09:11] resource access requests. location-services to be exact. [09:11] we have maps? [09:11] osmtouch. [09:11] open street maps [09:11] mzanetti: browser google maps as well. [09:11] indeed... [09:11] the camera does [09:12] now that you say it, the dialog slides in from the bottom instead of just appearing [09:12] mzanetti: what is the camera requesting? [09:12] * mzanetti retries [09:12] hmm... its gone [09:13] I think it was location [09:13] yeah, i think it only asks once, then remembers [09:13] I declined it, won't come back [09:14] so that means no geolocation for me any more until I --wipe the phone? [09:14] mzanetti: hm. there's a file you can delete which makes it ask again, but can't remember now. [09:14] heh [09:14] ok... no worries.. I don't think my camera should access GPS anyways [09:15] mzanetti: yeah, it's per app i think [09:18] mzanetti: but a way to test if it still works after removing the stop, is to try get two prompts on the screen at the same time for different apps [09:18] like one for camera and one for osmtouch. [09:18] I c [09:18] ok [09:18] mzanetti: but to be fair, i don't think that's as important as getting payments to work :) [09:19] yeah... I agree [09:19] i like money [09:19] although not giving it away [09:19] :D [09:20] yeah... so far I definitely made more money with apps then spent on them :D [09:20] still not much :D [09:34] Cimi, pushed alt nav test [09:34] Saviq, thx [09:34] ah, wait, want to add another one [09:42] dednick: any chance you remember which file it is? can't find it [09:42] the one to reactivate the camera app's promptsession [09:43] mzanetti: uuu. give me a minute. i'll try find it. lost my note :( [09:46] mzanetti: rm -rf ~/.local/share/UbuntuLocationService [09:46] uh... interesting [09:46] mzanetti: think you need to restart the trust-store after that [09:48] mzanetti: presumably there will be a settings page where you can change these things at some point. [09:49] dednick: just figured there is a setting in the camera app [09:49] dednick: but once you declined that popup it breaks the setting [09:49] mzanetti: there might be one for camera as well. [09:49] not sure [09:49] i think it's per service rather than per type [09:50] mzanetti: per server per app i mean [09:51] mzanetti: it's a db lookup. servce->app = [yes/no] i think [09:51] so location->camera, or location->osmtouch [09:53] Cimi, done [09:59] dednick: sqlite3 on that file and "delete from requests" does the job without restarting anything [09:59] well, restarting the camera-app ofc [10:00] mzanetti: cool [10:09] mzanetti, bug #1357230 is without a pass / pin lock? [10:09] bug 1357230 in Unity 8 "After booting, indicators and launcher can't be pulled in over Greeter" [Critical,New] https://launchpad.net/bugs/1357230 [10:10] Saviq: yep [10:11] dednick: yeah... as suspected, there are issues with the use case you said [10:11] expected === shiznix_ is now known as shiznix [10:22] dednick: https://code.launchpad.net/~mzanetti/qtmir/workaround-1355173/+merge/230952 [10:24] dednick: thanks for your help [10:26] mzanetti: no worries. just want to talk to tvoss about it first. maybe something can be done about the multiple app thing at the same time. [10:26] dednick: yep === dandrader is now known as dandrader|afk [10:30] dednick: am reviewing your prompt-surface-model stuff, it wouldn't be in a silo by any chance? [10:33] greyback: not yet. if you code review, can get a silo i guess [10:41] paulliu, hey, I put a hopefully final review comment on your reboot branch [10:50] sodding wifi [10:57] greyback__: fair point... but I guess it might be not that easy to convert UOA to trust sessions [10:58] greyback__: but I don't agree that the "new" issue is as bad as the "old" one [10:58] greyback__, can get it in 17 for you [10:58] greyback__, assuming it can land soonish? [10:58] * greyback__ popular all of a sudden [10:59] greyback__: chances are like 0 that you open two apps with the location question without confirming/declining it first [10:59] Saviq: no, I was just being lazy and hoping packages existed [10:59] greyback__: while chances that you click on install on an app before having set up your UOA account are much bigger [10:59] greyback__, https://code.launchpad.net/~nick-dedekind/unity8/prompt-surface-model/+merge/230813/comments/561274 [10:59] http://jenkins.qa.ubuntu.com/job/unity8-utopic-armhf-ci/1010/artifact/work/output/*zip*/output.zip [11:00] Saviq: yes but for unity8 whic also changes [11:00] mzanetti: I don't like introducing one bug to fix another. [11:00] there must be a better option [11:01] greyback__, does not compute, but also for qtmir you'd need to request they collect the packages, not sure why it doesn't [11:01] Saviq: oh sorry, I misread the link. Thought you were pointing me to qtmir packages which do build [11:01] @unity, need volunteers for https://code.launchpad.net/~unity-team/unity8/seeLessStuckBottom/+merge/230948 https://code.launchpad.net/~unity-team/unity8/tablePreviewWidget/+merge/230787 [11:06] mardy: hey, what's your opinion on https://bugs.launchpad.net/ubuntu-system-settings-online-accounts/+bug/1355173 ? [11:06] Ubuntu bug 1355173 in QtMir "Switching windows with a Trusted Prompt Session active looses the trusted prompt session" [Critical,New] [11:06] mardy: is there already some work ongoing to convert the UOA setup to a trusted prompt session? [11:07] mzanetti: it's in silo 13, we are testing it [11:08] oh! [11:08] that's good news [11:08] mzanetti: thanks for pointing me to that bug report, I was not aware of that [11:08] mardy: well, we just figured an hour ago that this is actually the issue [11:08] :) [11:08] at least I did [11:08] dednick: ^^ Hold off t [11:09] dednick: so seems the workaround is not required after all === dandrader|afk is now known as dandrader [11:09] mzanetti: ah. ok. awesome! [11:10] mzanetti: wait a second, are we talking of one trusted session, or of two? [11:10] ? [11:10] mardy: err... one I guess [11:10] mzanetti: i got to go out for a bit. continue later. your solution has unfortunately been rejected. :) [11:10] mzanetti: OK, so the parties involved are only the client application and OA, right? [11:11] mardy: hmm... no [11:11] mardy: dash opens payment ui [11:11] mardy: payment ui opens UOA [11:11] but I don't think that should affect you... [11:11] Saviq, I'll take https://code.launchpad.net/~unity-team/unity8/seeLessStuckBottom/+merge/230948 [11:12] dandrader, thanks [11:12] mzanetti: does dash use a trusted session to open the payment UI? [11:12] mardy: yes [11:12] mzanetti: mmm... then we have two trusted sessions, because OA will also create its own [11:13] mardy: yeah... but unity can deal with that... not sure if UOA even knows about it [11:13] mzanetti: OA will create a trusted session having the payment UI as initiator, and the OA UI as dialog [11:13] mzanetti: ah OK, that's fine. No, OA won't know about it [11:13] Saviq, But I will need your help. What's this "See less" thing? [11:13] dandrader, expansion button below categories [11:13] dandrader, let me find a scope for you :) [11:14] dandrader, start unity8-dash BINARY=$PWD/builddir/src/Dash/unity8-dash ARGS=-mousetouch [11:14] Saviq, yeah, I was looking for such buttons yesterday evening but couldn't find any in the UI [11:15] dandrader, bottom-swipe to overview, go to Yahoo [11:15] Saviq, right. see them there [11:16] mzanetti, can I bother you for https://code.launchpad.net/~unity-team/unity8/tablePreviewWidget/+merge/230787 [11:17] Saviq: yeah... didn't do my daily review today yet [11:20] mzanetti, pushed a small fix there [11:20] ok [11:20] mzanetti, speaking of which, https://code.launchpad.net/~dandrader/unity8/lifecycle/+merge/230090 has been updated as per your feedback :) [11:20] ok [11:21] dandrader, you too, just pushed a s/"grey"/palette.baseText/ to seeLessStuckBottom [11:21] Saviq: is the right-edge-ppa still in use? [11:22] mzanetti, no, it's all yours [11:22] cheers [11:22] Saviq: btw... regarding the hotfix for the user testing... I guess we should properly fix that behavor? [11:23] mzanetti, yeah, please file a bug with ubuntu-ux and assign Vesa [11:23] mzanetti, there's probably a few more cases where we need to have a closer look at this [11:23] mhm [11:23] ok [11:23] /somefood [11:26] vesar: https://bugs.launchpad.net/ubuntu-ux/+bug/1357333 [11:26] Ubuntu bug 1357333 in Unity 8 "Launcher shouldn't hide on long left edge swipes in some circumstances" [Undecided,New] [11:26] I guess we should talk about this in our next weekly [11:36] greyback_, https://code.launchpad.net/~dandrader/qtmir/lifecycle/+merge/230092 is updated. needs your review [11:36] dandrader: ack [11:37] Holas [11:39] Saviq, http://paste.ubuntu.com/8053056/ [11:39] Saviq, I think it looks fine, tested with all grey/all red and normal [11:39] Saviq, we can tweak shade values a little when we will have real scopes to test [11:41] Cimi, we have real scopes to test [11:42] Cimi, just go to the overview, All [11:42] Cimi, "#B2FFFFFF" [11:43] Cimi, k, I'm applying that to my branch then [11:44] Saviq, that was opacity of white line? [11:44] Cimi, yes [11:44] Saviq, 0.7 alpha? [11:45] Cimi, yeah [11:45] ok [11:45] Saviq, ok so in the patch, you can try Qt.rgba(1.0, 1.0, 1.0, 1.0) instead 0.7 [11:45] or maybe something little bit more, 0.8 [11:46] which them multiplies for 0.6 global opacity of the rectangle [11:46] I think 0.7 was too much on dark colours [11:46] Cimi, it doesn't work with a solid background unfortunately, z-ordering is bad between header and content, need to talk with Albert about htis [11:46] Cimi, the rectangle is full opacity [11:46] ah there [11:46] no it's not ok [11:47] Saviq, the opacity of the divider is full because we have full color here, no? === MacSlow is now known as MacSlow|lunch [11:47] Cimi, it's full because we want it to be full [11:48] Saviq, anyway you got the point of color manipulating [11:48] we can always tweak them later [11:48] Cimi, it's not like "I didn't get the point" ;) [11:48] yeah whatever :) [11:48] Cimi, it's "I GOT A DESIGN THAT DIDN'T HAVE IT" :P [11:49] and even explicitly asked about it [11:50] Cimi, if I"m honest, this doesn't work too well with any coloured divider, let me get you some shots [11:52] Saviq: if we release the new scopes (and shell plugin) to RTM, will that upset the version of unity8 in RTM? (i.e. does it have the v4 shell API there) [11:52] pete-woods, it does, yes, and scopes shell, too [11:52] pete-woods, I landed that in rtm [11:53] Saviq: awesome, thanks! [11:53] Cimi, check out the "Zrzut ekranu..." ones in https://drive.google.com/drive/#folders/0B32jwBcbaPloR1IyRF91ZWs0bFE [11:53] amazon looks ~ok [11:53] I have those [11:54] Cimi, k wait for it [11:54] Saviq, what's wrong with those? [11:55] Cimi, please tweak these to your liking http://paste.ubuntu.com/8053144/ [11:55] Saviq, you want more contrast? [11:55] Cimi, I can't see the inset at the top TBH === alan_g is now known as alan_g|lunch [12:02] mzanetti, with all your SIM cards, can you please have a look at bug #1357321 [12:02] bug 1357321 in Unity 8 "scope images do not load in HSDPA or 3G" [Undecided,Confirmed] https://launchpad.net/bugs/1357321 [12:02] ok [12:05] Saviq, I bumped to http://paste.ubuntu.com/8053224/ [12:06] Cimi, my diff was different, please rebase [12:06] Cimi, no need to put it in the bottom container [12:06] Cimi, http://paste.ubuntu.com/8053144/ [12:06] Saviq, I prefered it as before for the normal headers, but this should work lil better with other combinations [12:07] Saviq, just change 1.08 to 1.1 maybe? [12:07] try it [12:07] it's tastes... [12:07] Cimi, that feels negligible [12:07] Saviq, I see the difference :) [12:08] Saviq, better than 1.08 [12:08] Cimi, yeah, indeed [12:08] Saviq, we can always tweak them next week [12:08] Cimi, pushed [12:09] I'm coming too if you are there [12:09] uh oh [12:10] ah, building helps [12:12] ugh no it doesn't [12:16] Cimi, it's actually slightly more complex in the nav branch, but I'll manage [12:17] facundobatista, let's do it here [12:17] facundobatista, somehow I'm getting a art:mascot mapping for recommendations [12:18] facundobatista, and yes, no art == breakage [12:18] so no title either [12:19] Saviq, the client scope scope may be changing it, in the log is clear that the server is returning art=art [12:20] facundobatista, yeah, sure, as if I change the mapping to art:art in the scope tool, it works [12:20] facundobatista, so unity-scope-scopes? [12:20] Saviq, the "related search" case is different, though... the server is returning mascot=mascot in the category, and a mascot in the result, but empty :/ [12:20] this is a server issue! [12:20] I knew it! ;) [12:21] it's ok if in that case I return a mascot in the result (with a proper image) and mascot=mascot in the category component mapping? [12:21] "art": { [12:21] "field": "mascot", [12:21] facundobatista, ↑ breakage [12:21] Saviq, yes, that breakage is in the recommendations [12:22] facundobatista, it's like so: mascot is something the shell looks at, and the plugin maps from result based on the mapping [12:22] Saviq, once dash loads a scope, it will never unload it? [12:22] dandrader, the *dash* will not [12:23] dandrader, overview will, as it destroys the temporary pages [12:23] facundobatista, so whatever you set the mapping as, the plugin will try and map from the result to show the shell [12:23] Saviq, what's overview? that thing from where you can access all scopes? [12:23] dandrader, yes ;) [12:23] facundobatista, so if you want the mascot in the shell to display what you send in the "mascot" field in the result [12:24] facundobatista, you want mapping to be mascot:mascot indeed [12:24] right [12:24] thanks [12:29] facundobatista, are you sending overlay color at all or are we fine with the 15% black we have everywhere? [12:30] Saviq, I think we're ok [12:30] facundobatista, I'll drop the mapping then [12:30] Saviq, btw, don't bother about the URL I requested, I see oopses from what you surely tried (scopes scopes previews from PL) [12:30] facundobatista, you got me ;) [12:32] facundobatista, re: previews for results, I was told that there should be no preview for anything other than scopes, we'd have to clarify that [12:33] facundobatista, the other thing I just read... "Note - the search results page uses a dark theme but any subsequent screens (including previews) are in the standard theme." [12:33] that's not possible (for preview to be in a different theme than search page) [12:33] so let's ignore that for now ;) [12:35] ok [12:37] facundobatista, https://code.launchpad.net/~saviq/unity-scope-scopes/fix-overview-search-mapping/+merge/230966 [12:42] tedg: just an FYI, I've an MR up with basic LTTng tracepoints added. It'll take me more time however to actually see what might be slow [12:43] https://code.launchpad.net/~gerboland/qtmir/lttng/+merge/230967 [12:44] here is output from one test run: http://pastebin.ubuntu.com/8053449/ === alan_g|lunch is now known as alan_g [12:56] Cimi, hmm is the bottom highlight supposed to be over the divider still or below it, over the content? [12:57] inside [12:57] aaand remerge [12:58] oh crap really? [12:58] Notification center? [12:58] mzanetti, how's that going to be in german ↑? [12:59] seb128, french ↑? [13:00] * Saviq foresees ellipses [13:00] Saviq, Centre de notifications [13:00] *maybe* [13:00] I bet german is longer ;-) [13:01] weeell must be someone thought it through, right? [13:01] MacSlow|lunch, poke, for when you get back, about notifications and placement on screen [13:17] Saviq, I think should be over the content [13:17] Cimi, then dots are not centered vertically [13:17] Saviq, because is the lightening of the edge of the content [13:17] Cimi, but anyway, because background is on top now anyway [13:17] Cimi, let's leave above, when we fix z-ordering [13:18] Saviq, dots are centered vertically now [13:18] Cimi, we'll fix that too (offset dots by... .5dp) ;) [13:18] Cimi, yes, because both darkening and lighting is *inside* the divider [13:18] Saviq, is like a cast shadow on the divider, no? [13:18] Saviq, lighting should be outside [13:18] Cimi, yes, but can't be right now due to z-ordering [13:19] Saviq, so you have to do fully opaque line [13:19] Cimi, needs fixing in LVWPH [13:19] under the lighting === MacSlow|lunch is now known as MacSlow [13:19] mterry, what about notification positioning? [13:20] put a rectangle line of opacity 1.0 using bg color [13:20] Cimi, TBH I don't get why it should be non-opaque at all [13:20] then on top the highlight [13:20] Saviq, it should what? [13:20] Cimi, don't get why it's 0.6 at all [13:21] Cimi, unless it's to let the non-solid (image) background through [13:21] Saviq, that one [13:21] Cimi, anyway, can we fix after? I have to remerge 5 branches every time I change that one :P [13:21] Saviq, sure [13:22] Cimi, we'll sort out z and put it in the right place [13:22] Saviq, was working before [13:22] Cimi, no it wasn't, not with a solid background [13:22] Saviq, on my patch, the lighting outside [13:22] Cimi, no it didn't, not with a solid background [13:23] Cimi, the red scope [13:23] Cimi, in tryDash [13:23] I'll merge and try your then [13:24] Cimi, you and your ideas... now I need to try and find out the top color of the navigation background [13:24] Cimi, which is impossible in the case when it's an image of course [13:24] mzanetti: what was the "other use case" you'd break with the qtmir workaround for uoa using trusted prompts ? [13:24] Saviq, we can use shader for that [13:25] Cimi, sounds expensive for a stupid effect like that [13:25] Saviq, it is :D [13:25] Cimi, when 99.99999% of the time it's going to be solid [13:25] I think we acn just approximate that [13:27] Saviq, the highlight at the bottom of the divider is a punch in the eye [13:28] Saviq, can we remove it for now? [13:28] what's wrong with it? [13:28] Saviq, try with open library [13:28] 18 GU [13:28] or yahoo finance [13:29] ok yeah, not too great [13:29] Cimi, ok, putting it outside, we don't have solid backgrounds anywhere anyway [13:29] yeah [13:30] Saviq, add TODO [13:33] Cimi, check now [13:33] before I start remerging again [13:33] Saviq, did u push? [13:33] Cimi, now [13:34] Saviq, ok that looks nice [13:34] apart weather channel [13:34] but that's their fault [13:35] the divider is supposed to be darker than the page [13:35] because is a bottom layer [13:36] Cimi, ok, anything more to fix in that branch? because I'm merging :P [13:36] Saviq, text in overlay card is not aligned horizontally [13:37] was it due to my card branch? [13:37] it's even on your TODO [13:37] my TODO? [13:37] :D [13:37] Cimi, yes, you put it on your TODO [13:37]