/srv/irclogs.ubuntu.com/2015/05/27/#ubuntu-app-devel.txt

=== aaron is now known as ahoneybun
=== _salem is now known as salem_
=== salem_ is now known as _salem
=== aaron is now known as Guest26610
=== Guest26610 is now known as ahoneybun
shank_Hi all03:41
shank_Im trying to develop a game with SDL03:41
shank_on ubuntu03:41
shank_is SDL the right lib to use for this?03:41
dholbachgood morning07:00
=== sturmflut-work is now known as sturmflut2
davidcalleGood morning all07:21
=== chihchun_afk is now known as chihchun
=== avi is now known as Guest51819
=== chriadam is now known as chriadam|away
=== tsdgeos_ is now known as tsdgeos
=== ondra_ is now known as ondra
nerochiarooSoMoN: quick question: why is there a need to check if actualUrl.toString() when dealing with AddressBar.internal.simplifyUrl ?11:38
oSoMoNnerochiaro, where exactly? this is done in two places, which one do you refer to?11:47
nerochiaroAddressBar.qml in onActiveFocusChanged (around line 325 in my branch)11:48
nerochiarooSoMoN: ^ the reason I ask is because I need to refactor the code there and I am not sure I understand the intent11:49
nerochiarooSoMoN: my current understandind is that we want to simplify the text only when the addressBar has active focus. and if the page is loading we want to use the requestedUrl, if it has loaded we want to use the actualUrl11:50
oSoMoNnerochiaro, I guess the reason is that there is no point in trying to simplify an empty URL. Try removing the check and run the unit tests. If they pass it should be safe to remove.11:50
nerochiarooSoMoN: can you confirm that my understanding of how the simplification logic should work is correct, as written above ?11:53
=== _salem is now known as salem_
oSoMoNnerochiaro, « we want to simplify the text only when the addressBar has active focus » : that’s precisely the opposite12:26
oSoMoNnerochiaro, if the page is loading, we’re not simplifying12:27
oSoMoNotherwise we use the actualUrl12:27
nerochiarooSoMoN: so when do we use the requestedUrl ?12:28
nerochiaro    onRequestedUrlChanged: {12:28
nerochiaro        if (!activeFocus) {12:28
nerochiaro            text = internal.simplifyUrl(requestedUrl)12:28
nerochiaro        }12:28
nerochiaro    }12:28
nerochiarooSoMoN: also, unrelated, is there anything special that I need to do when running unit tests with a modified oxide branch ? when running webbrowser-app manually with the modified branch I set the env variables you suggested some time ago to point to the oxide build, and everything works. but if i run ctest from the same shell, a number of QML tests fail. how do you do it ?12:34
=== salem_ is now known as _salem
=== _salem is now known as salem_
oSoMoNnerochiaro, nothing special should be needed, what errors are you getting?13:19
nerochiarooSoMoN: no errors, just tests failing that don't fail when i run them from a shell that does not have any oxide-specific env vars set13:20
oSoMoNnerochiaro, well if there are failures can you paste them somewhere so I can investigate? run the tests with ctest -V13:21
nerochiarooSoMoN: failures are here: http://termbin.com/0pnp and I prepare the shell where I run ctest -V by going to oxide/build and running 'source' on this file: http://termbin.com/huin13:24
nerochiarooSoMoN: hold on, something wrong with the failures file i uploaded13:25
=== mzanetti is now known as mzanetti|otp
nerochiarooSoMoN: they seem to fail only when i launch them manually (using tst_QmlTests) but not with ctest -V . I will investigate further and update you. sorry for the noise13:27
balloonsdholbach, so we should talk about https://code.launchpad.net/~nskaggs/help-app/functional-test-template/+merge/260206. I threw that together quickly for someone who wanted to help out with writing some tests.13:42
dholbachballoons, I think it's great13:46
dholbachI'm just not sure about adding another ppa :-/13:46
balloonsit's not permanent, but yea, it kind of needs a ppa. If you make it part of the build, that complicates things13:47
balloonsif you leave it as -is until it can be a proper build depends, we can still modify the jenkins job to add the ppa13:48
balloonsthe other thing I wanted from you is a list of user stories to automate; aka, what functional tests need to be written specifically. Steps and expectations would help those looking to write the tests13:49
balloonsalso, I found a bug and I want to make sure it's not a dupe. Internal linking seems broken. Clicking on the 'scopes' internal link puts you in a browser window with no theming, menu, or controls13:50
kalikianat1mp: loicm: zsombi I'll leave it as-is now https://developer.ubuntu.com/en/blog/2015/05/26/keeping-track-evolving-qml-components/ let me know if you like it, and I can publish it later today13:54
dholbachDS-McGuire, what do you think about https://daniel.holba.ch/blog/2015/05/rfc-help-app-design/#comment-2048027884?13:54
=== mzanetti|otp is now known as mzanetti
oSoMoNnerochiaro, the failure is most likely because tst_QmlTests needs an -import parameter to locate oxide, see tests/unittests/qml/CMakeLists.txt for how the test is run13:57
nerochiarooSoMoN: makes sense. thanks13:57
kalikianazsombi: https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/noDepNoteForComponents/+merge/26030213:58
kalikianait's not catching a 100% of the warnings but very very close13:58
dholbachballoons, for now I think it'd be good to just add the foundations for more functional tests13:58
zsombikalikiana: wow, nais!!!!13:58
dholbachballoons, I'm not quite sure which ones we need13:59
dholbachballoons, maybe just navigate to a page and back again...13:59
kalikianazsombi: crossing one out on the list of things I s/always/NEVER EVER/ wanted to achieve in life. get deep into the qml type system....13:59
balloonsdholbach, I suggested nagivating to every page to start14:05
dholbachcool, sure - why not?14:05
balloonsI listed a few others I thought of on the bug. But yes, I'd like the mp to go as-is.. no tests, just foundation14:06
loicmkalikiana: good stuff in there, it's fine for me14:06
loicmkalikiana: the construction of that sentence could be better IMO "It's designed to be used a way to serialize the public QML API in a way that is human readable as well as easy to process in a pogrammatic fashion."14:10
kalikianaloicm: how about "It's designed to serialize the public QML API in a way that is human readable as well as easy to process in a pogrammatic fashion."14:12
loicmkalikiana: much better :)14:12
sturmflut2sverzegnassi: Great job on the new Document Viewer!14:18
sverzegnassisturmflut2, thank you! Hope to add SD card  write support soon... :)14:21
sturmflut2sverzegnassi: Since when does it render PDFs so fast? I just put a "Linux Voice" issue on the SD card of my bq and it rendered the first page within five seconds. This document is very complex, even my high-powered desktop sleeps for two seconds before anything is displayed.14:27
popeyheh, i use Linux Voice as my tests too14:30
sverzegnassisturmflut2, the rendering happens in multiple threads (at the moment the number of threads is hardcoded at 4). Each page is rendered in a single thread, in order to use all the power of the SoC.14:30
sturmflut2sverzegnassi: Very nice!14:31
popeyit's _really_ hard to get all those cores to wake up14:33
kalikianapublished https://developer.ubuntu.com/en/blog/2015/05/26/keeping-track-evolving-qml-components/14:45
nerochiarooSoMoN: i think that https://code.launchpad.net/~uriboni/webbrowser-app/find-in-page/+merge/258225 should be good now15:12
nik90popey: ping15:25
popeynik90: hello!15:26
nik90popey: Hi :)15:26
popeynik90: am on a hangout, but type at me :)15:26
nik90popey: sure, it looks like https://code.launchpad.net/~brendan-donegan/ubuntu-clock-app/disable_location_prompt/+merge/242682 is going to be merged at which point clock-app should be free of release blockers. Can you get QA to test it and get out a release when possible?15:26
nik90popey: as always i have kept the clock-app changelog updated at http://pad.ubuntu.com/clock-app-changelog15:27
popeynik90: you bet!15:27
nik90popey: once this release goes through, I am going to migrate clock-app to the new 15.04 framework in preparation for OTA-415:27
nik90popey: thnx15:28
nik90rpadovani: hey would you have time to review https://code.launchpad.net/~nik90/ubuntu-clock-app/fix-empty-state-warnings/+merge/260173 today? It is a harmless MP15:34
rpadovaninik90, lucky guy, I've just 5 minutes, on it :-)15:35
nik90rpadovani: ;)15:35
STEiN__Really silly question, I'm trying to write an app for Ubuntu touch- but I can't seem to find a way to select phone for build (It all says Desktop Desktop) I read the Convergence page online, does that mean I can just install my app to my phone even though it's not written for phone?15:47
nik90STEiN__: hi there...what kind of project is it (qmlproject, cmake, qmake)?15:48
STEiN__I just hit new HTML5 project, because I couldn't find a project that said that it was compatible with phones15:49
nik90STEiN__: either way, click on Projects in the sidebar on the left and then choose the phone kit. That should build it for the phone.15:49
STEiN__I normally write apps for Android/Mac (Odd combo but eh..)15:49
nik90STEiN__: I presume, you have run through the initial ubuntu-sdk configuration like creating chroots, kits etc?15:50
STEiN__Nope, I just followed the Wiki to install the Ubuntu SDK haha15:50
STEiN__It jumps right from Installing the SDK --> Create an App Project15:51
nik90STEiN__: I would recommend reading through https://developer.ubuntu.com/en/start/ubuntu-sdk/ to get you acquainted.15:51
nik90STEiN__: and after the create an app project, it links to running apps from the sdk15:51
nik90which shows how to run your app on the desktop, phone and any other device for that matter15:51
STEiN__Cool, thanks :)15:52
nik90yw15:52
nik90balloons: hey, I'd like some help with debugging the jenkins issue in https://code.launchpad.net/~nik90/ubuntu-clock-app/fixed-bottomedge-status/+merge/25399815:53
nik90balloons: manually testing it on the desktop and phone, I can't hit the issue that AP seems to face when run by jenkins15:54
nik90and code has been approved by zsombi, so its bound to be correct.15:54
balloonsnik90, I like that theory.. bound to be correct15:56
balloonssure, give me a few mins I'll have a look15:56
nik90;)15:57
nik90well I trust zsombi's judgment when it comes to alarm code15:57
=== chihchun is now known as chihchun_afk
=== rickspencer3_ is now known as rickspencer3
=== chihchun_afk is now known as chihchun
balloonsnik90, the location fixes landed16:23
nik90balloons: yup..I have requested popey to ask QA to test clock-app and push out a release16:24
balloonsI kicked off a re-run of https://code.launchpad.net/~nik90/ubuntu-clock-app/fixed-bottomedge-status/+merge/25399816:25
nik90balloons: quick question, which ubuntu version is jenkins running on?16:25
balloonsand I'll have a look in a moment16:25
nik90balloons: I just proposed https://code.launchpad.net/~nik90/ubuntu-clock-app/migrate-to-15.04-framework/+merge/260336 where I use Ubuntu.Components 1.2 which I know for sure has landed in vivid, but it fails complaining that it cannot be found.16:26
nik90more specifically the unit tests are failing because QtQuick 2.4 and Ubuntu.Components 1.2 are not found16:29
=== chihchun is now known as chihchun_afk
sverzegnassimzanetti: ping16:41
mzanettisverzegnassi, hi16:41
sverzegnassimzanetti, o/ do you have some spare time for talking about tweakgeek and uttt?16:44
mzanettisverzegnassi, sure16:47
balloonsnik90, indeed, it's on utopic17:15
balloonslet's get that fixed17:16
STEiN__hey nik90, are you still here?17:29
mcphailHow do I force landscape orientation for an app?17:30
kalikianat1mp: still around? wanna have a look at a small review? https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/unitTestApiCheck/+merge/260348 nothing added, just a little test case for what we have17:34
mcphailIs it even possible to force orientation? I thought it was, but can't find a reference in the API. I note cuttherope uses rotation:90 in the WebView, but I don't want to use a webview17:40
ogra_iirc you can only forst the change of orientation to off ...17:41
ogra_but if your portrait app gets started while the phone is in landscape it will come up in landscape17:41
mcphailthat's not great news17:41
ogra_shell rotation support will likely fix that ... if that ever lands17:42
mcphailFor normal apps, fixed rotation is evil. But for games...17:42
ogra_dont tell me :)17:43
mcphailyeah17:45
zsombikalikiana: one comment on https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/noDepNoteForComponents/+merge/26030217:46
kalikianazsombi: hmm that is a good point. but if I hard-code lib that makes it impossible to run out of the builder folder...17:51
zsombikalikiana: perhaps just detecting that Ubuntu/Components is in the path would be enough17:51
* mcphail is feeling grumpy. Getting the infrastructure in place for a games platform is rather difficult17:51
kalikianamcphail: like bacon2d?17:52
mcphailkalikiana: bacon2d seems nice enough, but rather limiting. Would be nice to get something like the quake engine running. Being tied to Qt/QML or webviews for all games is not great17:53
kalikianamcphail: depends on what you want of course. you can also use quake in js17:57
kalikianain that sense 'tied to' is relative17:58
mcphailkalikiana: being stuck with emscripten is not great. SDL2 _works_, but needs shell support for things like rotation and fullscreen17:58
kalikianathere are javascript APIs for rotation and fullscreen, if anything you have an advantage because you don't need to implement it from scratch18:00
kalikiananote I'm not trying to convince you of anything - I was mostly curious about what you were after18:00
mcphailkalikiana: the platform is 90% capable of running most open-source (hence able to be compiled for armhf) games with little modification. But I really need that other 10%18:02
mcphailkalikiana: simply having the equivalent of the Android manifest, where you can force landscape mode, would be great18:03
balloonsnik90, in the interim, here's a test run on vivid: http://91.189.93.70:8080/job/generic-mediumtests-vivid/1121/18:03
nik90balloons: which MP is this run for?18:04
kalikianamcphail: that's already in the works, using the .desktop file18:05
kalikianazsombi: updated https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/noDepNoteForComponents/+merge/26030218:06
mcphailkalikiana: That's great news! Cheers me up, 'cause I'm getting a bit grumpy18:06
zsombikalikiana: happroved18:07
zsombikalikiana: I still cannot find th ereason why the AP fails on tests...18:08
zsombikalikiana: if I run the QML files, and try to invoke the funcs, they seem to show up... no idea why the AP doesn't bring them up...18:09
kalikianazsombi: hrm18:11
zsombikalikiana: though test_caret_hidden_if_empty seems not to delete the text, and if I run test_textinput.header.qml, and if I delete the text from the header's input, the caret is still visible...18:11
kalikianazsombi: that's because you're typing then18:12
kalikianaer wait18:13
zsombikalikiana: wow, if the test tyes to delete the inpu tcontent with the cross button, that doesn't work18:13
kalikianahmmmmm18:13
kalikianazsombi: which one is that exactly? lemme run it here18:14
zsombikalikiana: ubuntuuitoolkit.tests.components.test_textinput.CaretTextInputTestCase.test_caret_hidden_if_empty18:14
=== oSoMoN_ is now known as oSoMoN
zsombikalikiana: weird... the test assumes the text input is empty... but the test shows it is not...18:26
kalikianazsombi: interesting, fails in the header one for me as well18:26
zsombikalikiana: yes, the assert is for the carret not being visible, but it is, as th einput has text... how the heck was this working before?!18:28
zsombikalikiana: if I remove the text from the test, it passes....!18:29
kalikianazsombi: odd indeed, the others don't have text, and the test would basically make no sense if there was any....18:30
zsombikalikiana: yes... so just wonder how th eheck was this working before? it wasn't my MR which got this test, and I didn't change the test either18:31
kalikianaI'm quite sure it passed at one time, otherwise it wouldn't be in staging18:31
zsombikalikiana: but that's wrong anyway, as the caret should be shown if there's text...18:32
kalikianazsombi: hm?18:32
kalikianathis case is about an empty field where there is no text and thus no handler18:33
zsombikalikiana: well, the test HAS text :)18:33
zsombikalikiana: and noone deletes that text18:33
kalikianazsombi: it does. but that would never pass. my only explanation is that a merge went wrong and it changed18:34
zsombikalikiana: so if I remove the text from the test file, it passes18:34
kalikianayes18:34
zsombikalikiana: lemme check staging... but the prob is that my MR doesn't show any change there18:35
zsombikalikiana: ok, staging has the same, everything is the same18:39
zsombikalikiana: and fails on staging as well!!!!!18:42
* zsombi eod18:42
SvenBo90Hello, can somebody help me? When I try to create an emulator on qtcreator I go to Devices and click on the +-button. But nothing happens. When I have a look on the console it says QQmlComponent: Component is not ready PopupUtils.open(): Failed to create the popup object.21:02
popeyjdstrand: seen https://myapps.developer.ubuntu.com/dev/click-apps/2744 ?21:48
jdstrandpopey: I haven't, but we can't allow that22:06
=== salem_ is now known as _salem

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!