=== thomi_ is now known as thomi [01:23] I'm getting a hang at the Google screen after enabling qt compositor... Trying a reflash now, but anyone else seeing this? [01:47] also getting an error when I do dist-upgrade === _salem is now known as salem_ === salem_ is now known as _salem [07:39] Saviq: i am in temp scope, and from dash overview "All" open a temp scope, that happens to be the same scope, i just go "back" instead of putting it on top, no? [07:40] tsdgeos, on one hand, yes... on the other there's going to be the scope stack (multiple temp scopes open on top of one another) [07:40] tsdgeos, so that'll need clarification,too [07:42] tsdgeos, so do the simplest thing you can do (open one on top anyway?) [07:42] i guess that's the simplest yes [07:43] and we'll add to the list [07:55] mzanetti, RFC https://bugs.launchpad.net/unity8/+bug/1343760/comments/1 [07:55] Ubuntu bug 1343760 in Unity 8 "[qtcomp] last focused app will rotate in spread" [Medium,New] [07:59] Saviq: done [07:59] tx [08:09] Saviq: having two temp scopes is veeeery confusing [08:09] hope they see it once it's running on their hands [08:10] tsdgeos, yeah well... that's what the stacked scopes will lead to anyway [08:10] but not with overview in the middle [08:10] overview in the middle is confusing [08:14] tsdgeos, well, you get an overview in the middle for different temp scopes as well (and for dash, for that matter) [08:14] not sure i understand what you mean [08:14] but let's finish this [08:14] and then see what people think [08:15] tsdgeos, I mean it's as confusing as when you have dash | overview | scope1 [08:15] and dash | scope2 | overview | scope1 [08:15] doesn't matter if scope2 is scope1 much [08:15] ah sure [08:18] Saviq: can you tell me what's wrong here? https://ci-train.ubuntu.com/job/landing-006-1-build/122/console [08:18] dh: Unknown sequence get-packaged-orig-source [08:19] mzanetti, that's not the problem [08:19] https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-006/+sourcepub/4301552/+listing-archive-extra is [08:19] mzanetti, prompt sessions [08:19] mzanetti, https://launchpadlibrarian.net/180198115/buildlog_ubuntu-utopic-amd64.qtmir_0.4.0%2B14.10.20140718-0ubuntu1_FAILEDTOBUILD.txt.gz [08:20] ah ok... that makes more sense now [08:20] I guess I need gerry for that [08:20] gonna do some taxes crap in the meantime [08:22] mzanetti, FWIW https://code.launchpad.net/~nick-dedekind/unity-mir/unity-mir.trusted_sessions/+merge/208324 this is what got merged in unity-mir to get trusted prompts going [08:40] Saviq: looks like qtmir silo is having issues since mir 0.5 was released. [08:41] dednick, yup, qtmir needs adaptation to prompt sessions [08:42] Saviq: hm. well if doesn't "need" adaptation. it should still be able to work without. but can't seem to upgrade to the packages in silo anymore. [08:42] dednick, it did FTBFS https://launchpadlibrarian.net/180198115/buildlog_ubuntu-utopic-amd64.qtmir_0.4.0%2B14.10.20140718-0ubuntu1_FAILEDTOBUILD.txt.gz [08:43] Saviq: ah, ned change to mir. er, i can fix that. [08:43] dednick, SessionAuthorized does not implement prompt_session_is_allowed [08:43] s/ned/new [08:43] dednick, sure, guys will be happy [08:44] dednick, btw, something *weird* happened with your trusted MP to unity-mir https://bugs.launchpad.net/unity-mir/+bug/1343639/comments/3 [08:44] Ubuntu bug 1343639 in unity-mir "can't swipe away keyboard or use extended keys on top row" [Critical,Fix released] [08:46] huh. that's weird. must have been an dodgey merge during the CI. [08:53] hi pete-woods, I was discussing with pstolowski what to recommend to scope developers to use as a C++ library to do networking. The tutorial code that we've got uses Qt, and I'm told that's probably not the best option. It seems you're writing the youtube scope with netcpp? Is that what you think we should recommend? [08:54] dpm: I'd say so, yes [08:54] Saviq: https://code.launchpad.net/~nick-dedekind/qtmir/mir-0.5-compatibility/+merge/227290 [08:54] dpm: otherwise you have to deal with integrating with the Qt event loop [08:55] dpm: I'd say the vimeo scope might be a more friendly example to point developers at than the youtube one, though [08:55] the youtube API is somewhat complex [08:55] mzanetti, dednick fixed qtmir ↑↑, I'll add to the silo [08:55] pete-woods, we don't have any example c++ scope using the latest version of the api. Do you have the vimeo scope code somewhere in LP? [08:55] Saviq: dednick: awesomes, thanks guys [08:55] dpm: https://launchpad.net/unity-scope-vimeo/ [08:56] it just uses a simple synchronous API [08:56] as opposed to the youtube scope, which is peppered with futures [08:57] pete-woods, awesome, thanks. Another thing that some scope developers might want to do is to parse JSON returned by the remote api they're connecting to. Do you have any suggestions for c++ libraries to do that? [08:57] dpm: again, see the vimeo scope - I'm using jsoncpp [08:57] excellent, thanks [08:58] dpm: see the youtube scope for a more difficult example, though (https://launchpad.net/unity-scope-youtube/) [08:58] as that's really what a more involved scope should look like [08:58] doing parallelised http requests (to some degree), gzip compression, etc [08:59] dpm: there are also (in my opinion) good examples of tests, using python's tornado as a fake HTTP server [09:01] pete-woods, great, thanks [09:02] pete-woods, for now, I think the vimeo scope should suffice. I'd like to get the scope template in the SDK to add some basic networking code to do a simple API call [09:02] sounds like a good idea to me [09:03] vimeo also has the tests, though, seriously think it's not a bad idea to include a simplified version of those in the template [09:03] reacting to just one API call, as you say [09:16] Saviq: just pushed the show dash overview from temp scope + all interactions [09:19] tsdgeos, ktx [09:19] tsdgeos, looks like qmluitests are back in business, btw [09:19] \o/ [09:19] So now i have a very small todo with a popup that doesn't want to die [09:19] and will try to cover all the interations in a qmluitest [09:19] not sure how well this can be tested [09:19] but will try anyway [09:20] Saviq: there's a magic item that opens a temp scope in tryDash [09:20] Title 2.2 in the red scope [09:38] tsdgeos, oh ok [09:53] tsdgeos, didn't design want that disabled? [09:53] mhr3: no [09:53] they want dash overview disabled from dash overview temp scopes [09:53] but not from general temp scopes [09:53] don't ask me why [09:54] pete-woods, i'm trying to gelp dpm with getting scope tutorial compiled with the latest api, and getting errors that come from the guts of libproperties-cpp-dev - http://pastebin.ubuntu.com/7813060/ [09:54] s/gelp/help/ [09:54] pete-woods, ideas? [09:55] dpm, also, here is a little bit that at least removes errors related to scopes api calls: http://pastebin.ubuntu.com/7813572/ [09:55] tsdgeos, i have to think too hard just to parse that sentence :P [09:56] pstolowski, slot is a qt keyword [09:56] pstolowski, thanks! pete-woods, for more context, that tutorial code uses Qt for networking. I realise we should now probably used netcpp based on the conversation earlier, but I think at this point the easiest thing is to fix the tutorial code to work with the latest API and update the networking part earlier -> http://bazaar.launchpad.net/~dpm/ubuntu-sdk-tutorials/api-updates/revision/119 [09:56] pstolowski, there was a fix for properties-cpp ages ago [09:56] err *later, not earlier [09:57] pstolowski, and it's still there :( https://code.launchpad.net/~marcustomlinson/properties-cpp/fix_qt_keyword_clash/+merge/223056 [09:57] mhr3, uh oh [09:57] dpm, ^ [09:57] dpm, easy fix though, -DQT_NO_KEYWORDS [09:58] awesome [09:58] ok, let me give this a go, thanks guys [10:02] yeah, beat me to it [10:15] MacSlow, one for you https://code.launchpad.net/~saviq/unity8/fix-test-notifications/+merge/227239 [10:15] oh well, mterry ack'ed it already, but no top-ack, so might as well [10:21] tsdgeos, hmm, something's broken with dash overview [10:21] E_MORE_DETAILS_NEEDED [10:21] somehow i can't activate any of the all scopes [10:21] tsdgeos, do you expect it to emit openScope()? [10:22] no [10:22] i expect it to have a scopeId [10:22] mhr3: how uptodate are you? [10:22] tsdgeos, pulled like 15minutes ago [10:22] ok [10:23] let me push the plugin [10:24] tsdgeos, lp:~unity-team/unity-scopes-shell/overview [10:25] mhr3: make tryDash works for you, right? [10:26] yep [10:28] ok, let me compile the plugin [10:28] see what i can help with [10:33] tsdgeos, also, searching and then reseting the search seems to break quite a lot [10:34] works for me in tryDash :D [10:35] still building [10:35] but i do unspeakable things to the models :P [10:45] mhr3: why is your all so weird? [10:45] i get holes :S [10:45] and scroll doesn't work :S [10:45] tsdgeos, guess a scope without art [10:45] yea, what's up with that [10:46] i don't know [10:46] it works fine in tryDAsh :d [10:46] should i be settings collapsed-rows: 0? [10:46] setting* [10:48] no unless something weird changed [10:49] mhr3: scopes.getScope(com.canonical.scopes.ebay) returns null [10:49] oh you call that? [10:49] yes [10:49] k, fixing [10:50] tsdgeos, you'd call that even when doing searches? [10:50] mhr3: what do you mean "evn when doing searches" [10:50] tsdgeos, searches can return specific canned queries [10:50] i don't know what's that [10:50] i open scopes [10:51] you give me a scopeId [10:51] i use it [10:51] tsdgeos, for example search for "u2", will return amazon result telling it to search for "u2" in the "music & dvd" department [10:52] if you just do getScope() all that is lost [10:52] well it's not what i was told to do [10:52] i was told, you search, have a scope, open it [10:53] that was cause you didn't want to wait for openScope [10:53] which it might and might not emit [10:53] iirc [10:54] what¿ [10:55] oh is openScope going to be any different anyway? [10:55] i can understand why you don't want to wait for a signal [10:56] oh, oh, we have the performQuery on the scope object now [10:56] could you just call that after getScope() with the result uri? [10:58] i can do whatever you want [10:58] i have just done what i've been told to do [10:59] and in the 99.05% i have been told anything i just improvised [11:01] tsdgeos, are you ok with having ownership of the temp scope you get via getScope() ? [11:02] i don't know what it means [11:02] so no [11:03] you'll own the instance [11:03] you're responsible for destroying it [11:03] that's not specified in getScope docu [11:03] is it? [11:04] no [11:04] because i'm pretty sure we do getScope in other places [11:04] and never destroy it [11:04] why should we do it in this case? [11:04] cause it used to give you only scopes that are in the model [11:04] but now you want temp ones [11:04] Holas [11:05] tsdgeos, or we introduce a new call for that [11:10] mhr3: if you're going to make me use a new function, may as well use openScope [11:11] mhr3: can you please detail exactly how i'm supose to interact with you [11:11] i'm giving you options [11:11] if openScope is fine with you, let's do that [11:12] anything is fine with me, you just need to tell me what you're going to do [11:12] obviously it's not going to work now [11:12] so don't complain it doesn't work [11:12] i did it so that it works for me [11:12] but it doesn't work for you, ok, tell me what you want [11:12] relax [11:13] i'm relaxed [11:13] i'm just trying to figure out what's the difference [11:13] and we identified that [11:13] so yey [11:13] and i'm going to give you a hint, don't tell relax to people that aren't relaxed [11:13] doesn't help === MacSlow is now known as MacSlow|lunch [11:15] i'm investigating the height thing [11:18] tsdgeos, so from my pov if you just call scope.activate(result) i'd be happy, that will emit either gotoScope (for favourites) or openScope(for the rest), or neither if it's a wikipedia result or whatever [11:18] that would have been useful before i started doing anything [11:18] mhr3: in search? in all? [11:19] everywhere [11:19] i'm not going to call that in favorites [11:19] fine [11:19] that means we could even get rid of gotoScope perhaps.. yey [11:20] although i don't know what's supposed to happen if you tap music from all [11:20] open is as a temp scope? [11:23] no [11:23] i go back to dash [11:23] i handle that [11:24] i check if it's on the favorites model or not [11:24] and then go to dash or not [11:24] booooo for using functions in bindings [11:24] they don't update when the variables used in the function but not in the function call change [11:28] tsdgeos, so you'll check scopeId, and if !favourite, call .activate() which you'll expect to emit openScope? [11:29] yes [11:29] mhr3: from the scope that was returned by getScope("scopesOverview") [11:29] yes? [11:29] yep [11:29] mhr3: and i have to call closeScope when i'm done [11:29] correct [11:30] and of course the opened scope can again emit openScope [11:35] not going to work [11:35] sorry [11:36] that doesn't work right now either in the dash [11:36] so i'm not going to fix it for this case [11:36] height of the "All" should be fixed now [11:36] it looks like it's wrongly calculated here at the moment since there's scopes without images that make it look empty [11:38] i wonder why is that [11:39] Error downloading https://dash.ubuntu.com/imgs/art/openlibrary-scope.jpg - server replied: Forbidden [11:39] You don't have permission to access /imgs/art/openlibrary-scope.jpg on this server. [11:39] is that firefox says [11:39] someone should fix it [11:40] facundobatista, ^ [11:40] mhr3: so getScope is going to return null for any non favorite scope, right? [11:40] mhr3: can we document that? [11:42] so that i make my fake plugin do the same [11:42] biab, interview [11:42] mhr3, tsdgeos, checking... [11:43] mhr3: tbh it is kind of weird, since we have the visible flag, so i would expect your "Scopes" to have all the Scope objects and set the visible flag accordingly [11:44] and then getScope would work [11:44] but ok, getScope() only works for visible scopes [11:45] tsdgeos, tbh i'd rather get rid of the visible flag, it's not like it was useful for anything, you just filter the invisible ones anyway [11:46] tsdgeos, while at it, can we make the overviewScope something really special? [11:46] mhr3: it's useful in my backend where i have all the scope objects created and then getScope works just fine for all scopes [11:46] as in Scopes.getOverviewScope() [11:46] mhr3: if you prefer, sure [11:47] mhr3: you're adding it to unity::shell::scopes::ScopesInterface then, no? [11:47] tsdgeos, doesn't work that great when you have dozens of objects and they're not that cheap [11:47] mhr3: actually it should be kind of a property [11:47] not a getOverviewScope? [11:47] or you prefer the getter? [11:47] are you creating it when i call that? [11:48] i mean it will always be non null when i call getOverviewScope? [11:48] tsdgeos, property is fine [11:48] yea, it could be null if things aren't initialized [11:48] i don't know, what do you prefer? [11:48] well if it can be null [11:48] i am going to need a property [11:49] can you please define the interface so i can copy it to my fake plugin? [11:49] k [11:50] going for lunch [11:53] Saviq: hey out of laziness I bumped the version number of unity8 with QtComp support to 8.0 - mzanetti not so sure about that. WDYT? === MacSlow|lunch is now known as MacSlow === jibel_ is now known as jibel [12:15] tsdgeos, lp:~unity-team/unity-api/scopes-v3 === _salem is now known as salem_ [12:51] mhr3: waht's overlaycolor? [12:57] tsdgeos, unrelated [12:57] ok [12:58] tsdgeos, btw the going back from search to surfacing was my bug [13:06] greyback, should we pass a QT_NODEBUG or something when building qtmir packages to silence its log output? [13:25] mhr3: pushed the activate+openScope thing [13:25] mhr3: should your part be ready or not yet? [13:26] tsdgeos, nope, need to override activate [13:26] ok, tell me when its done [13:26] +' [13:31] \o/ Dell technician's here [13:40] tsdgeos, btw anything you can do about the large result? it's white text on light background atm [13:40] tsdgeos, when you for example search for wiki in the overview [13:41] mhr3: not sure what you mean [13:42] mhr3: screenshot? [13:43] Saviq, Which MR is the one that splits out the dash? [13:43] wth.. my zeitgeist doesn't work [13:43] tsdgeos, http://imgur.com/EdrRvZ3 [13:44] i guess i need to set some background so it's not white [13:44] i'll see [13:44] mhr3: is that happening if i update all branches and compile? [13:44] yes [13:45] ok [13:46] tsdgeos, although maybe we need to update the scope, it's specifying the background color itself [13:47] dandrader: I think it's a good idea to keep the verbose logs printing, it'll help us tracking down bugs [13:47] still white on white... eek [13:48] confused [13:49] the designs no longer show the card being white... otoh there's no real visual design, just wireframes [13:49] well [13:50] the card is using scopeStyle.foreground as color for something i idenfity as background [13:50] Saviq: you there? [13:51] tsdgeos, wassup [13:51] Saviq: look at summaryColorWithBackground in CardCreator [13:51] Saviq: why is it using foreground? [13:51] should it be background? [13:51] paulliu, is it the design that the shutdown dialog stays behind the indicators panel and launcher? [13:52] tsdgeos, hmm? that's the colour of text? [13:52] Saviq: the color of the text is background? [13:52] or you mean summaryColorWithBackground is the color of the text? [13:53] ah right summaryColorWithBackground is the color of the text [13:53] yes it is [13:53] then why is that thing white [13:53] tsdgeos, maybe because it takes card background into account [13:53] or well, should at least [13:54] greyback, re: version... we didn want 8.0 to be when it becomes unity [13:54] greyback, i.e. we don't want to keep it being unity8 [13:54] greyback, but OTOH does one preclude the other... [13:54] Saviq: ah right [13:54] hmmmm [13:54] so how do we fix that? [13:55] Saviq: that = http://imgur.com/EdrRvZ3 [13:55] Saviq: I'll just do as you prefer on this matter. I don't have a strong opinion really [13:55] tsdgeos, ah that [13:56] tsdgeos, StyleTool has light/dark colours [13:56] tsdgeos, we need to take luminance of the background color [13:57] tsdgeos, and based on styleTool.threshold you go for dark or white [13:57] s/white/light/ [13:58] Saviq: in the cardcreator code [13:58] tsdgeos, it has cardStyle [13:59] scopeStyle [13:59] yes [13:59] tsdgeos, so summaryColorWithBackground needs to check the background color luminance (that should be ~happening already) [13:59] ok, let's see if i can make that happen [13:59] tsdgeos, and depending on scopeStyle.threshold use background or foreground [13:59] or well [13:59] light or dark [14:00] but yeah, that's something I didn't do yet [14:00] mzanetti, do you have a branch with split dash for tedg to use? [14:03] Saviq: no [14:03] Saviq: doesn't work for that yet as ApplicationManager doesn't pick it up yet [14:11] mzanetti, I'm not sure how to interpret that. Simple or complex? [14:11] tedg: well, what exactly do you want to do with it? [14:11] mzanetti, I want to overlay a trusted session prompt on it. [14:11] mzanetti, So I need the pid of it to pass to Mir. [14:12] tedg: hmm, I see. no, slightly to early for that... [14:12] tedg: when QtComp has landed (Monday or tuestday hopefully) we'll work full steam on that [14:13] tedg: will ping you as soon as I have a branch for you to play with [14:13] mzanetti, So, again, I have no understanding of what the difficulty is. So is that a couple days after thing or a week or two after thing? [14:14] tedg: ah, no, shouldn't be too hard I hope, greyback estimated 1 day to get it working [14:14] Okay, cool. [14:14] tedg: so definitely next week unless QtComp plays bad on us (not expected any more) [14:15] * greyback touches wood [14:15] heh :D [14:31] seb128: hey, just getting our qtcomp ducks in a row....would you mind approving https://code.launchpad.net/~unity-team/ubuntu-system-settings/use-qtComp/+merge/225540 [14:41] tube every day in 30 degree heat. yay! [14:42] Saviq: so something like http://paste.ubuntu.com/7814714/ ? [14:45] mhr3: so it's fixed with ↑ [14:45] mhr3: you need to run make again to pick up these changes === dandrader is now known as dandrader|afk [15:01] \o/ /me got my umph back === gatox is now known as gatox_lunch [15:07] dandrader|afk: any idea why my EdgeDragArea oesn't work in tests? [15:07] or anyone [15:09] could someone test-ping me here please? [15:10] MacSlow: ping [15:10] tsdgeos, thx [15:12] tsdgeos, once more please... [15:12] MacSlow: looooolo [15:13] tsdgeos, great...thx [15:15] tsdgeos, are you using touchDrag? [15:16] Saviq: touchFlick [15:16] can't find anything like touchDrag [15:17] tsdgeos, yeah, that's probably the same === dandrader|afk is now known as dandrader [15:17] s/same/one/ [15:18] it''s moving the scope below the drag area :S [15:18] dandrader: https://launchpadlibrarian.net/158819734/poweroff-buttons.png [15:18] dandrader: should be in-front-of those. [15:18] ok now [15:18] don't know what i did but works now :D [15:18] \o/ [15:20] tsdgeos, using mouse events instead of touch events? [15:20] dandrader: that was at the begnning, they i moved to touch and still didn't work, but i may have had the coords wrong [15:23] paulliu, so it's a bug, right? [15:23] dandrader: yes. [15:24] paulliu, does it happens with stock image? I was trying with qtcomp modifications on top [15:25] dandrader: I'm not sure. How did you make it behind them? [15:25] dandrader: If just pressing the power button, it should pop-up in the front of everything. [15:25] paulliu, well, just long pressed the power button and then slid down the indicators panel from the top edge [15:26] dandrader: Let me try. I didn't do that. Should be on stock image too. [15:26] paulliu, ie, while the dialog is being shown I have full access to the top bar [15:26] paulliu, and can also do a left-edge drag to slide in the launcher [15:27] dandrader: yeah. It is a bug. [15:27] paulliu, ok, will report it then [15:31] paulliu, https://bugs.launchpad.net/unity8/+bug/1344082 (don't know your launchpad nick) [15:31] Ubuntu bug 1344082 in Unity 8 "launcher and indicators panel can be accessed while shutdown dialog is shown" [Undecided,New] [15:35] tsdgeos, awesome [15:36] dandrader: ok. I'll assign myself. === gatox_lunch is now known as gatox [16:25] dednick, fyi: just started reviewing https://code.launchpad.net/~nick-dedekind/unity8/mirCompositor-appMocks/+merge/227317 [16:26] dandrader: cool. thanks [16:47] pete-woods, I was looking at the vimeo scope and noticed the PageHeader.Logo setting in the .ini file. However, when I try to run my scope using that setting via unity-scope-tool, I'm getting: [16:47] warning: ignoring unknown key PageHeader.Logo in group Appearance in file /home/dpm/dev/build-8tracks-scope-Desktop-Default/src/com.ubuntu.developer.dplanella.8tracks-scope_8tracks.ini [16:47] dpm: yes, the error is nonsense [16:47] dpm: ^ [16:47] but the thing is that the logo does not load [16:48] dpm_: I think you need to provide an absolute path to the logo [16:48] but your stuff is click-packaged, right? [16:50] * pete-woods is asking what is supposed to happen here [16:51] pete-woods, it will be click-packaged, it's just that it's not in a working state to be packaged yet. Right now I'm just getting the scope to run at all, using Qt Creator and unity-scope-tool [16:52] so I set up the .ini file and I was puzzled by the error. Let me try to hardcode the path for testing purposes... [16:52] dpm_: I'm just checking whether you're allowed to pass a relative path (as that's what you need if you're a click auther) [16:52] *author [16:52] dpm_: yes, the error message is annoying and misleading [16:52] I'm just using PageHeader.Logo = @SCOPE_INSTALL_DIR@/8tracks-blue.png atm [16:53] dpm_: what does that resolve to in your binary directory? [16:53] it might be looking in the real /usr/lib/ARCH/unity-scopes/SCOPE_NAME/… [16:53] (and maybe there's nothing there) [16:53] pete-woods, it's a relative path, right now using a non-click target, it's /8tracks-blue.png [16:54]