inside a "twelve-col" div is the same size as a "six-col" div inside an "eight-col" div. [12:57] davidcalle, I tried never to touch any of those items [12:57] dholbach, you should, it's "fun" ;-) [12:57] no :) [12:57] I have enough fun in the area I'm working on :) [13:00] * dpm reads that twice [13:03] davidcalle, good to know, thanks! I generally try to have one single twelve-col parent, in which page would you use cols inside an eight-col parent? [13:03] dpm, all the text (as requested by you) :p [13:04] dpm, jk, images in text. I'm aligning them to the grid [13:04] davidcalle, yeah :) but I mean, in which page have you used columns inside that? [13:05] do you have an example? [13:05] dpm, https://developer.ubuntu.com/en/apps/html-5/tutorials/cordova-camera-app-tutorial [13:06] dpm, the page is not finished (still waiting on the middle section), but images are four-col [13:07] dpm, of course, you can also switch back to twelve for images [13:07] davidcalle, ok, gotcha. But do images not float to the left in any case? [13:08] man, btw, the tutorials and the guides are looking fantastic with the new site [13:10] davidcalle, another thing we could do for long images is to put them on the empty 4-col at the right. This is just a thought, I'm not asking to change it :) [13:11] dpm, yes, but with cols, all image sizes are consistent, you can align them, side by side, etc. Which brings back some flexibility to that wonderful editor we will learn to love one day. [13:11] :-) [13:11] ok, makes sense [13:12] dpm, indeed, but the android dev site does it and it's not that great in terms of readability (but it works well to introduce a new section or a new theme, though) [13:12] ok [13:12] ( a bit like a hero img) [13:30] timp: what up? [13:32] mzanetti: ahayzen was having some questions about running apps fullscreen on the phone (is that possible from qml, or do you need cpp for that?). I didn't really know the answer so I was looking for a unity8-dev to help out [13:34] ah. I'm not sure tbh... I know the gallery app uses it so it must be possible somehow [13:34] might well be that it just does QQuickView->showFullscreen() === _salem is now known as salem_ === salem_ is now known as _salem [13:36] timp: ahayzen: http://bazaar.launchpad.net/~phablet-team/gallery-app/trunk/view/head:/src/gallery-application.cpp#L242 [13:36] timp: I guess this means MainView lacks [13:40] mzanetti: thanks [13:41] mzanetti: MainView is written in qml only and inherits from some other qml components [13:41] mzanetti: so it may be a bit tricky to add a "fullscreen" property there [13:41] should be possible though.. Not sure if we want that. Do many apps need a fullscreen mode? [13:42] dholbach: yes, thjat can be removed [13:43] dholbach: is that something you are looking at or shall I? [13:46] dpm, do you know of any popular apps which use URLDispatcher (that is: they register themselves for handling a "special" URL) other than the clock ("alarm:"), the music app ("album:"), and the messaging app ("message:") ? === rmescandon is now known as rmescandon|lunch [13:49] timp, mzanetti, thanks :) i've nearly managed to hack window around mainview to work [13:50] jdstrand, ping [13:50] ahayzen: hey [13:50] jdstrand, Hey, i've been writing a game that uses a TiltSensor and have noticed I do not need any apparmor policies? I have checked here http://developer.ubuntu.com/apps/platform/guides/app-confinement/ and it says by the 'sensor' policy 'policy pending due to LP: #1227116', but that bug is now resolved. So should this policy exist and be enforced? [13:50] Launchpad bug 1227116 in apparmor-easyprof-ubuntu (Ubuntu Trusty) "qtsensors tilt and rotation is non-functional" [Medium,Fix released] https://launchpad.net/bugs/1227116 [13:52] ahayzen: what policy version are you using? [13:53] jdstrand, 1.2 [13:54] ahayzen: you are supposed to get haptic for free, and then get fill access to usensorsd via 'sensors' [13:55] s/fill/full/ [13:55] jdstrand, i seem to be able to run with no policies and can access the TiltSensor readings [13:55] oh, and there there is: 'apps may always use the accelerometer and orientation sensor' [13:55] /etc/xdg/QtProject/Sensors.conf r, [13:56] ah [13:56] ahayzen: please let me know how you get it to work in qml only [13:56] I'd need that for a bunch of apps too (didn't really investigate) [13:57] timp: yes, I'd say many need that. especially games [13:58] mzanetti, basically i've done this https://code.launchpad.net/~andrew-hayzen/volleyball2d/fullscreen/+merge/245196 [13:58] mzanetti, just put the MainView inside the Window as timp said [13:59] ahayzen: nice. thanks [13:59] mzanetti, i then just had to play about with my detection of when it is in portrait [13:59] mzanetti: Should games use MainView? That's there mainly to add the app header, and (currently) automatic orientation [13:59] hmm let me check ahayzen's code :) [14:00] timp, my game uses Page {} for the settings etc [14:00] timp, but yeah most just directly put the Game {} directly instead of MainView IIRC [14:00] ahayzen: oh the change is not that big :) nice! [14:00] timp: how does that work with applicationId stuff then if you don't use a MainView? [14:01] ^^ was my other thought [14:01] hmm [14:01] mzanetti: that's the applicationName property? [14:02] yep [14:02] ahayzen: love that game btw [14:02] have been playing an hour at least yesterday :D [14:02] mzanetti, awesome thanks :) still some stuff to clean up but the basis is there [14:02] but man... it's hard! [14:02] mzanetti: all that MainView does with that is this: http://pastebin.ubuntu.com/9569973/ [14:03] didn't win a single time [14:03] yeah there is no AI difficulty levels yet [14:03] so that can easily be repoduced without the MainView [14:03] timp: ah ok, yep, that seems ok [14:03] but ahayzen's solution with a MainView inside a fullscreen Window looks like a good solution anyway :) [14:03] its just on 'hardest' ... i know a way to *always* win though ;) although i'll patch that soon with some new AI logic hehe [14:04] I managed to lose 5:4 at best [14:04] hah [14:08] jdstrand, will do [14:09] dholbach: thanks! [14:13] done :) [14:15] cool [14:15] aquarius, reminder does, but I'm not sure that is in trunk yet [14:28] jdstrand: hey, when developing apps it's really annoying that the system constantly prints some "Error opening shm" into the app's debug messages. I've looked into the code that prints it and I have a feeling it's denied by apparmor to access that. [14:29] what would you say would be way to go to fix that? I guess we won't allow it to access that... but then I don't really know if lttng is even useful without it [14:33] dpm, ooh, really? who do I ask about details of that? [14:35] aquarius: mzanetti ☻ [14:36] http://bazaar.launchpad.net/~reminders-app-dev/reminders-app/trunk/view/head:/src/app/main.cpp#L60 [14:36] http://bazaar.launchpad.net/~reminders-app-dev/reminders-app/trunk/view/head:/src/app/qml/reminders.qml#L56 [14:37] http://bazaar.launchpad.net/~reminders-app-dev/reminders-app/trunk/view/head:/manifest.json.in#L11 [14:37] and http://bazaar.launchpad.net/~reminders-app-dev/reminders-app/trunk/view/head:/reminders.url-dispatcher [14:37] aquarius: ^ [14:37] aha, evernote: URLs. [14:38] * aquarius rtfs :) [14:38] ah, darn, I can't provide the *text* for a note via uridispatcher. [14:39] nope... I guess we could add that [14:39] does the reminders app run on 14.04? [14:39] yes [14:39] (desktop) [14:39] it's in apt [14:39] although I'm not sure if url-dispatcher does [14:39] might add it myself. I'll think about it. It's all part of my Current Master Plan :) [14:40] its quite straight forward... you just create that json file describing you url, register that in the manifest file [14:41] then, if the app is running you'll get calls from UriHandler.onOpened [14:41] *nod* yep [14:41] if not, the app will be started and the uri passed as cmdline arg [14:43] hrm. Trying to build it in Ubuntu SDK, I get "error: [CMakeFiles/com.ubuntu.reminders.desktop] Error 127" [14:43] useful, Ubuntu SDK. Not. [14:44] aha! there is a COmpile Output tab with actual errors in. GOod :) [14:45] * aquarius installs intltool === ian is now known as Guest41228 [14:48] mzanetti, how do I run it? I can build it now, but clicking the big Run triangle in Ubuntu SDK says "no executable specified" [14:48] ah, I need to select "reminders2", do I? [14:49] hm. "module "QtQuick" version 2.3 is not installed " [14:52] mzanetti, are you sure this works on 14.04? http://packages.ubuntu.com/search?keywords=reminders-app only has it packaged for 14.10 and vivid [14:54] aquarius: hmm... trunk might not. however you should be able to just change the imports to QtQuick 2.2 and it should work [14:54] oh, you said 14.04 [14:54] then you might need QtQuick 2.1 :D [14:56] aquarius: but unless you install it with click, I don't think url-dispatcher works on the desktop [14:56] changing to 2.2 seems to work [14:56] ok [14:56] I'm fine with urldispatcher not working - I can just run the app with the url on the command line [14:56] and edit the tests [14:56] but! having changed to 2.2, I still get an error when running [14:56] qml/components/RtfButton.qml:26:32: Invalid alias location [14:56] property alias iconSource: icon.source [14:57] don't even know what that means [14:57] * aquarius googles :) === rmescandon|lunch is now known as rmescandon [15:01] hmm [15:01] aquarius: I guess you can work around that by changing it to: property alias iconSource: icon.iconSource [15:01] I have temporarily worked around it by making that a string property, which means that I won't get any icons, but I can live with that :) [15:01] let me dig out the old docs [15:01] ok [15:02] wtf? Now it's claiming that AccountServiceModel doesn't have an applicationId property, which indeed it does not according to the docs, but then how did it ever work for anyone? [15:02] applicationId: "com.ubuntu.reminders_reminders" [15:02] qml/reminders.qml:248:9: Cannot assign to non-existent property "applicationId" [15:03] aquarius: well, it probably does have that in the newer version, no? [15:03] but you're not *using* a newer version; you're using Ubuntu.Components 1.1. [15:03] http://developer.ubuntu.com/api/qml/sdk-14.10/Ubuntu.OnlineAccounts.AccountServiceModel/ [15:03] unless the SDK makes breaking changes while keeping the same version number? [15:03] fair point [15:04] mardy: ^ [15:04] what? that's a possibility? I was joking! [15:04] * aquarius does the angry look. [15:04] this is what version numbers are *for*! [15:04] aquarius: well, it's in the import of Ubuntu.OnlineAccounts [15:04] mardy, ping, then :) [15:04] and yeah, according to the docs it's still at 0.1 [15:05] ya, if you were using a newer version than I have then I'd have failed on the import [15:05] which means that new properties got added to Ubuntu.OnlineAccounts 0.1 rather than a new version number [15:06] yes [15:06] and that means it won't run on 14.04 but it's impossible to tell why :( [15:06] * aquarius looks sad. [15:06] well, not impossible to tell, but yeah, you do have a point [15:07] aquarius: in any case, you can just comment that line away and get going [15:07] it won't break much iirc [15:07] yay! now it starts up at least [15:08] sooo. now the real online accounts fun :D [15:08] you have to add a evernote account in systemsettings (the desktop one) [15:08] you can install online-accounts-plugin-evernote for that [15:08] *nod* [15:08] then restart the app and *it should* work [15:08] aquarius, mzanetti: the applicationId was indeed a later addition, let me check when it happened... [15:08] am just filing an aggravated bug for mardy to read ;) [15:10] mardy, aha, you're here. I don't mind that new properties were added, that's fine -- I mind that the version number stayed the same, so I can't tell whether the version I have will work or not. [15:10] aquarius: so, it happened in May, and indeed I forgot to bump the version number [15:10] mardy, http://developer.ubuntu.com/api/qml/sdk-1.0/Ubuntu.OnlineAccounts.AccountServiceModel/ and http://developer.ubuntu.com/api/qml/sdk-14.10/Ubuntu.OnlineAccounts.AccountServiceModel/ have different API and the same version number :( [15:11] is it worth me filing a bug about it, or is it now too late? [15:11] aquarius: right, please file a bug, I guess we can bump the version number, then push the new lib to 14.10 and re-update the docs [15:13] https://bugs.launchpad.net/accounts-qml-module/+bug/1404272 filed :) [15:13] Launchpad bug 1404272 in accounts-qml-module "QML Accounts API changed without version number change" [Undecided,New] [15:14] mzanetti, do I use my existing evernote account, or do I create a new one for the sandbox? [15:16] aquarius: however you want. if you want sandbox, you need to install account-plugin-evernote-sandbox and start the app with --sandbox [15:16] mzanetti, 14.04 only has the sandbox plugin for online accounts, afaict [15:16] hmm, really [15:17] yeah... could be [15:17] $ apt-cache search evernote [15:17] account-plugin-evernote - Allow Ubuntu to access the Evernote developer sandbox [15:17] right.. then we created the production one only afterwards [15:17] ah, using my real account doesn't work, so I'll create a new one :) [15:17] I'd like to recommend you to use the offline-mode branch which would allow you to use the app without evernote account [15:17] however, I really don't know how that behaves on 14.04 [15:18] I would, but you only handle evernote: URLs, right? [15:18] we handle them offline too :) [15:18] if you want to try: https://code.launchpad.net/~mzanetti/reminders-app/offline-mode/+merge/244713 [15:18] :-) [15:19] I thought before I should probably change it to reminders:// now that we don't have a hard dep on evernote any more [15:19] indeed :) [15:21] aquarius: there are rumors you saw an ubuntu phone :) [15:21] exciting times, eh? [15:21] yes! [15:21] I got a nice little briefing from cparrino and rupinder [15:22] I really like the pitch -- not locking your stuff up behind a grid of icons. [15:23] hm. We now start up OK, but I get the spinner for ever... and a ton of warnings [15:23] Error fetching username: "Not connected." [15:23] that seems relevant... [15:24] dpm, what do we do about http://developer.ubuntu.com/en/apps/qml/cookbook/? [15:25] aquarius: that's with trunk or the offline-mode branch? [15:25] trunk [15:25] http://paste.ubuntu.com/9570579/ [15:26] hmm... it should get the account from OA now and then connect [15:26] line 15 is where userstore.cpp throws the error about not being connected [15:26] ah no, it did that [15:27] * mzanetti looks up the code [15:28] ah right... mismatch in the plugin and the app (as you have the old OA plugin) [15:28] aquarius: edit reminders.qml:262 [15:29] change it to EvernoteConnection.hostname = "http://sandbox.evernote.com" [15:29] or actually https://sandbox.evernote.com [15:30] ok, progress... no more spinner. Now the connection actually fails and the app catches it [15:30] hmpf [15:30] have a meeting now [15:31] I don't think that's the right URL. I'll look into it [15:31] getting Thrift: Fri Dec 19 15:31:06 2014 TSocket::open() getaddrinfo() Name or service not known [15:31] for both http and https [15:32] ahaha! [15:32] it's a hostname, not a url :) [15:32] and now I am connected! [15:32] ah :) [15:56] mzanetti, have filed https://bugs.launchpad.net/reminders-app/+bug/1404289 whining about needlessly requiring qtquick 2.3 ;-) [15:56] Launchpad bug 1404289 in Ubuntu Reminders app "Imports QtQuick 2.3 when it doesn't need it and works with 2.2" [Undecided,New] [15:57] mzanetti, also, http://paste.ubuntu.com/9570863/ improves the URL parser a bit, allowing arbitrary query strings, and implements one example: you can specify the title of a newly created note. [15:58] mzanetti, it's not a proper merge request because I also had to make about a million comment-this-out-so-it-works-on-14.04 changes which should not be merged :) [15:58] mzanetti, I'd have allowed specifying the body of the new note in the querystring as well, but I don't know how to do that; NotesStore.createNote does not return the note it's just created! [16:00] aquarius: thanks for the feedback :) [16:00] no worries about the merge proposal, I'll put it in here somewhere. [16:00] and yes, currently it's all async, however, now with the offline mode support that could (or probably should) be changed [16:01] you can probably think of a zillion other parameters that could be passed in, but I'll leave that to you :) [16:01] yeah... we just added the ones that would be useful to call from the (upcoming) scope [16:02] being async is the right thing to do, certainly! NotesStore.createNote(notetitle, null, function(newnote) { console.log("this is the new note:", newnote); }); [16:02] that is, provide a callback [16:03] better still, return a Promise, but that's probably a bit advanced for QML's slightly noddy JS parser ;) [16:03] well, the callback is there NotesStore::noteAdded() [16:03] oh! [16:03] but I guess what I'm going to do is to return the note straight away [16:04] I should have done that then :) [16:04] and then when I get the evernote sync reply I change it's guid [16:04] current trunk is quite bad with error handling though [16:04] so if it goes wrong you won't get anything atm [16:05] which again is a non-issue with offline mode because it'll just try to sync later and keep the note offline for the time being [16:05] fair enough. I'm not sure if I'd use this (I'm still thinking about how it would fit into my current app plan) but it seemed an easy enough contribution to make :) [16:08] mzanetti: I guess you are referring to bug #1260491 [16:08] bug 1260491 in apparmor-easyprof-ubuntu (Ubuntu) "please explicitly deny access to /run/shm/lttng-ust-*" [Undecided,Fix released] https://launchpad.net/bugs/1260491 [16:08] mzanetti: "Talking with tedg, these shouldn't be available to confined apps and we can safely explicitly deny the access to silence the errors." [16:09] mzanetti: so, apparmor is denying it, but silently. lttng should probably suppress the log message [16:10] jdstrand: ah, yeah. but the code is not ours and it doesn't have a possibility to silence things at compile time or so (already checked that). [16:10] maybe that could be added [16:11] but yes, that seems to be it [16:12] thanks [16:19] anyone got examples of what they'd consider a really nicely visually designed SDK app? [16:24] aquarius: I guess the internal ones, like messaging, dialer etc. Some of the core apps look quite good too, like the new music app. And imo nik90_ did a good job with that tv series app (forgot the name) [16:24] mzanetti, nik90_ 's 'flashback'? [16:24] yep, that's the one [16:24] aah, yeah, I did like nik90_'s app, indeed [16:25] new music app \o/ :) [16:26] the new music app does look good, but it looks good because it has loads of cool art to choose from, which is not the case for my app :) [16:26] very good point [16:26] yeah, I'm always struggling with that too [16:26] nik90_, who did the visual design for Flashback? [16:27] hah yeah .. we don't have bottom edge yet though :/ [16:28] ah! there was that little group of people, wasn't there? who were offering designer services [16:28] torico or something [16:28] * aquarius googles. [16:28] yep! torico. https://plus.google.com/107423272812169770223/posts/HiVAkp1Yn1W [16:28] wonder if they're still up for it? [16:29] huh. seems perhaps not; hardly anyone in the channel :) [16:29] ahayzen: just opened the music app: http://i.imgur.com/csLySx2.png [16:29] hah oh dear [16:30] mzanetti, we are still waiting for empty state designs from the designers as well :/ thats just some text we put there but we should probably wrap that lol [16:30] yep, wrapping would help :) [16:30] mzanetti, the 'tap to shuffle music' is now hidden if you don't have any music as well :) [16:30] seems reasonable too :) [16:30] 'The form [from torico] "Application Ideas " is no longer accepting responses.' [16:30] oh well :) [16:31] ahayzen: actually you probably should just pop the complete stack [16:31] mzanetti, the toolbar is a rectangle over the top of everything? [16:31] ahayzen: some image that shows a PC and Phone connected by cable and some text would be nice I guess [16:31] ..its custom [16:31] mzanetti, yeah thats what we wanted [16:32] mzanetti, but we need design to design it...we even have a walkthrough prototype somewhere [16:39] ahayzen: hah. I just won! [16:39] mzanetti, well done :) [16:39] ahayzen: while you're waiting on design, you can add multiplayer over bluetooth to it [16:39] mzanetti, yeah i've already thought about that... bluetooth vs wlan [16:39] mzanetti, as it runs really well on the desktop as well [16:41] mzanetti, note the click package is arm only you'll have to branch and run to do it on the desktop a the moment [18:44] timp: Kaleo: https://plus.google.com/u/0/111980561516715514914/posts/FUrmN5F9uWi what would cause this? [18:52] mhall119, different toolkit maybe? [18:54] mhall119: my quick answer is that the phone is vivid and the desktop is not. [18:55] bzoltan_: assuming they use the same UITK version in their imports, like Ubuntu.Components 1.0, shouldn't they look the same? [18:55] or will they still get 1.1 versions of the components if they're available? [18:55] or 1.2 or whatever is on the phone [18:56] mhall119: we had 1.1 release both on V and on U [18:57] mhall119: anyhow, the first question I would ask is about the image number and the desktop series [18:57] ok, ignore the specific version numbers, the question is whether it'll use a newer version that what was specified if a newer version exists on the device [19:00] mhall119, or maybe is only a bugfix that is landed on vivid, and not on Utopic yet - so the experience is different atm, but at the end Components 1.1 will have the same behavior both on desktop and on phone [19:01] mhall119, I think this is a consequence of this bugfix: #1386208 [19:02] see comment nr 8 [19:02] bug 1386208 [19:02] bug 1386208 in ubuntu-ui-toolkit (Ubuntu) "Purple overflow panel in SuruGradient theme" [Undecided,New] https://launchpad.net/bugs/1386208 [19:04] thanks rpadovani === sephiap_ is now known as sephiap === ken_ is now known as kenvandine [19:49] mzanetti, hey [19:55] Letozaf_: hi [19:57] mzanetti, I am having some issues running reminders autopilot tests: https://dl.dropboxusercontent.com/u/10083981/Ubuntu%20errors/Evernote-1.png and http://paste.ubuntu.com/9572868/ [19:57] mzanetti, I am running them on vivid with autopilot3 [19:58] Letozaf_: need to try. gimme a bit [19:58] mzanetti, no problem I can wait :-) === _salem is now known as salem_ === salem_ is now known as _salem [20:11] Letozaf_: hmm... I don't get this error here [20:12] mzanetti, you are on vivid like me ? [20:12] yes [20:13] mzanetti, I used to write autopilot tests for reminders time ago, could it be I have some directory with old configurations to cancel ? [20:13] rpadovani: no UITK changes will land on Utopic as Utopic is closed. [20:14] rpadovani: mhall119: it is important to note that convergence is between form factors and not between Ubuntu series [20:16] Letozaf_: actually now I can reproduce it [20:16] Letozaf_: still need to figure how to fix it [20:16] mzanetti, ok [20:16] mzanetti, you can let me know by email if you are busy now... I am not in a hurry [20:17] bzoltan_, no, wait, so we could have different behavior on different releases with the same component version? I expect the behavior of , I dunno, Text, it's the same both on vivid and utopic if I imported the same version of Ubuntu.Components (e.g 1.1) [20:21] Letozaf_: there seems to be a bug in the autopilot tests [20:22] Letozaf_: each time I run them it creates a new account [20:22] and then fails to use the newly created account [20:23] it still works on jenkins for some reason [20:24] * Letozaf_ is looking at the tests [20:45] mzanetti, I am unable to delete the Evernote accounts created by autopilot tests in online accounts I have Permission Denied error [20:45] Letozaf_: yeah, I had to delete the accounts.db too [20:45] ~./config/libaccounts-glib/accounts.db [20:45] warning: it'll drop ALL the accounts [20:45] mzanetti, ok thanks, I was wondering were it was :) [20:45] mzanetti, no matter [21:03] * Letozaf_ needs to reboot