[00:08] perrito666, you could use vivid, 15.04 [00:25] balloons: you mean as a kit? [00:26] http://pastebin.ubuntu.com/9534675/ <-- the error === _salem is now known as salem_ === salem_ is now known as _salem [06:55] good morning === oSoMoN_ is now known as oSoMoN [10:05] What is the recommended framework to put on the manifest.json nowaday? :) (I'm having "'ubuntu-sdk-14.10-dev2' is obsolete. Please use a newer framework") [10:08] ybon, here's a non-human friendly version to that answer: https://myapps.developer.ubuntu.com/dev/api/click-framework/ [10:08] perfect, thanks :) [10:08] ubuntu-sdk-14.10 is probably the answer you're looking for [10:08] but, give a man a fish... [10:09] :) [10:09] Good morning all; happy Chocolate Covered Anything Day! :-D === john-mca` is now known as john-mcaleely === _salem is now known as salem_ === rmescandon is now known as rmescandon|brb [12:07] morning all, Ill re-ask just in case, I am using ubuntu 14.10 and just created a new go+qml project in ubuntu sdk, the basic qml created yells the following error [12:07] http://pastebin.ubuntu.com/9534675/ [12:10] h === rmescandon|brb is now known as rmescandon === ybon1 is now known as ybon [12:55] Anyone know why my breakpoints aren't being triggered in my scope? I'm setting the breakpoint in QtCreator and press the "Start Debugging" button, but the breakpoints are never triggered. [13:06] zbenjamin, perhaps you can help TommyBrunn with his question about breakpoints? ^ [13:26] It would be most appreciated. I've never written C++ before, and not being able to use the debugger makes it a tad harder. :) === rmescandon is now known as rmescandon|lunch === chihchun_afk is now known as chihchun [13:58] dpm: benjamin is on holidays === chihchun is now known as chihchun_afk === davmor2_ is now known as davmor2 === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === rmescandon|lunch is now known as rmescandon === chihchun_afk is now known as chihchun === boiko_ is now known as boiko === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [17:39] ahayzen, are you on trusty? [17:43] balloons, i'm on utopic :) [17:49] ahayzen, did you ever run on trusty with the sdk ppa? [17:50] balloons, yup i did all the time :) [17:50] ahayzen, did you encounter issues with component versions? [17:50] balloons, i don't think so [17:51] balloons, the bigger issue i had is that the music-app stopped running on trusty when we moved to a later version of mediascanner :/ [17:51] balloons, but it seemed to handle 1.0 1.1 etc ok if thats what you mean? [17:51] ahayzen, right.. It's weird it's not for someone I'm trying to help [17:52] thanks for the sanity check [17:52] balloons, weird whats the error? [17:52] it doesn't find the 1.1 version of components, the 2.3 version of qtquick, etc. It's all symptoms of the version being too old [17:53] can you even get 2.3 on trusty? [17:53] I think it's a dist-upgrade issue.. ahayzen that was partly my question :-) [17:53] 2.3 = Qt 5.3 which isn't in trusty right? [17:53] ahayzen, that's my understanding [17:54] same [18:02] balloons, an italian user had the same problem yesterday. QtQuick in trusty repository and in sdk ppa for trusty is still at 2.2, and I didn't find any trusty PPA that is trustable enough. You can download Qt5.3 from the website [18:02] 'trustable' hah [18:02] rpadovani, interesting [18:07] bzoltan_, refresh my poor memory again. What's the thoughts of having newer versions of qt on the lts? [18:08] see the above conversation bzoltan_ ^^ [18:09] balloons: I am not sure of I capture the issue ... [18:10] balloons: we stopped backporting the new Qt releases and the Ubuntu UI components to the LTS [18:10] bzoltan_, if I'm running 14.04 LTS and I want to develop an application I can use the sdk ppa to get the newer frameworks. However, trusty shipped with qt5.1? I think, so if I require say QtQuick 2.3, the app won't run on my desktop [18:11] balloons: that is correct. Sadly the LTS desktops are not targets for app development with the SDK. The reason is very simple.. it i snot possible to co-install the 5.1 Qt with the 5.3 Qt. [18:12] balloons: and we did want 5.3 Qt because it has signifucant improvements and bugfixes [18:15] balloons: I mean we do not stop anybody from the community to apt-get source the utopic/vivid Qt packages and backport them to Trusty in a PPA. After that it would be trivial to backport the UI Toolkit source packages to LTS too. So it is not impossible to make a full desktop backport. [18:16] bzoltan_, sure. So what does the sdk ppa really get you on 14.04? [18:16] balloons: but it can be only a community initiative because I can not promise that such backporting would be bulletproof. We do not test the Qt stack on LTS anymore [18:16] the idea is you get newer components, but must use / test / run them inside an emulator? [18:16] or device I guess [18:16] balloons: QtCreator with our plugins, a click what can create you 14.10 and 15.04 chroots for building projects and emulators [18:19] bzoltan_, thanks, that does all make sense. I just need to update my docs for folks wanting to write tests; it's really utopic or better to do it. 14.04 really won't cut it [18:19] based on the above [18:21] balloons: Utopic is not much better than Trusty from that aspect. The UITK is newer, yes and Qt is newer.. but once we move on to Qt5.4 the situation will be the same. Actually for writing tests it should be all the same... [18:22] balloons: but true, that even I struggle with writing tests for non relesed UITK components...vivid is new new enough when I need to write tests for a component what exists only in the staging branch :) [18:22] bzoltan_, I recommend people writing tests to be running the devel version, but that's not always possible. At least if you can run the last release you can get by [18:23] ahh yes, I've been there [18:24] balloons: it would be cool to be able to run AP tests on the emulator [18:25] bzoltan_, it technically works; depending on the test some run better than others. Actually I'm curious now about setting up an lxc container and running apps and tests inside of it [18:25] I think nik90_ did this long ago? [18:26] balloons: I have heard about 4-5 people mentioning the LXC as the finding of the golden pot ... all of them gave up :) It looks like a viable path, but it is not [18:26] ..does audio work in the emulator yet? this was the limitation stopping us trying music's autopilot in the emulator last time [18:27] balloons: how do you mean that AP test work on emulator? How the input events made? Can you use touch and keyboard events? [18:27] I feel like generally I want a fast environment that lets me iterate on tests quickly during development, so that's the desktop as I get all my tools as well. For confirmation, I would simply run on a device. For that reason, I'm not sure the emulator makes a ton of sense as a target, as I would not want to run tests against it during test development (too slow) [18:28] bzoltan_, you can run AP tests as-is on the emulator unmodified. The issues I found last time I tried was apps talking too long (timeout issues) and memory issues (emulator kills AP thread) [18:29] balloons: yeah I use lxc to run the tests [18:29] I just created a new vivid lxc [18:30] http://nik90.com/fiddling-around-with-lxc-containers/ [18:31] balloons: wow, that is new to me. Good to hear. Last time I tried (several months ago) I could not even install the AP packages [18:31] nik90_, hmm. So you are able to develop and test happily on 14.04 then? [18:33] bzoltan_, nik90_ so we could / should ship an lxc container package that would set this up for you transparently and run the sdk from it yes? [18:40] nik90_: balloons: I would not ever suggest that to anybody. You are free to do it, but I do not promise any sort of support for the SDK or for the UITK when it is used in LXC or other virtual environment. The sole reason is that we do not have resources to test the whole SDK in LXC. [18:42] nik90_: can you create click chroots or run an emulator in the LXC? [18:44] balloons: yes, I ain't leaving 14.04 :D [18:44] bzoltan_: nope [18:45] bzoltan_: but that's got to do with how I create the chroot.. [18:45] bzoltan_: which is why I use 14.04 with emulator , and then use the lxc container only for testing (ap + qml tests) [18:45] since AP and QML Tests require tests to be run natively [18:46] nik90_: I hardly ever run any test natively, but i am privileged to have a real device. [18:48] nik90_: I think that ubuntu people were too long spoiled with the comfort that the development - build -runtime environment were the same ... well, not really anymore :) But I see developers are still emotionally bonded to the old practices. [18:48] bzoltan_: I try to run the tests natively before every MP since I used to use my real device for daily dogfooding with rtm [18:49] nik90_: so instead of fooling ourself with the fake comfort of running tests on desktop :) i would rather fix the QA process and make a real device able to to test MPs [18:50] bzoltan_: + [18:50] bzoltan_: its just that, running things natively can be done within 1-2 seconds which is so fast...compared to running on emulator or on a device... [18:50] but that said, I have stuck to 14.04 with emulator for development [18:50] so I am doing what you recommend [18:51] nik90_: and that is the desk of balloons :) The ultimate solution would be click package all the AP stuff with the app you want to test and do not taint the device [18:51] nik90_: I am happy that you use 14.04 :) because we do not have much dogfooders on LTS [18:53] interesting interesting [18:53] so you can't create a build target inside your lxc sdk nik90_ ? [18:56] nik90_, and yes the sdk ppa gives you enough to develop with the newer frameworks, even though you are on 14.04 and can't run the app natively that you are developing [19:05] balloons: we have functions in the SDK to run AP tests for an app. I would love to extend it to "run the AP tests of the app on the target device". Sadly it is not possible on a stock image... because bunch of packages are not installed [19:06] bzoltan_, you need to use adt. Have you ever used it? [19:09] basically autopkgtest should setup the entire environment on the device and run the tests. You should be able to easily script into that and return the results to the user [19:09] it's what CI is transitioning to [19:17] balloons: adt? Does not ring a bell [19:18] bzoltan_, autopkgtest? [19:19] balloons: that one I know... but how can I use it on a readonly stock device? [19:19] bzoltan_, I talk about it in this post: http://www.theorangenotebook.com/2014/09/autopilot-test-runners.html [19:21] balloons: wow...`adt-run --click com.ubuntu.music --- ssh -s adb` is that all??? [19:21] bzoltan_, yes. [19:22] bzoltan_, for the sdk, I would imagine you should pass the local folder too so it runs the local tests [19:22] balloons: dude :) you made me two wows in an hour. that is a new record this week [19:22] balloons: thank you a bunch [19:23] bzoltan_, since developer mode landed, support is added for that as well, you just need to pass your password [19:25] bzoltan_, this might also be useful: https://people.debian.org/~mpitt/autopkgtest/README.running-tests.html [19:26] technically adt-run will also run and setup inside an lxc environment as well.. that might be an easier suggestion per our previous conversation about setting up an lxc container just for tests [19:28] balloons: no I cannot create a chroot inside the lxc [19:28] balloons: I get some permission errors [19:29] mm, that partially makes sense nik90_ [19:29] I talked to strgraber and he said that was an issue [19:30] but that's fine....I dont use the lxc for anything else [20:16] ahayzen, is this still valid? https://bugs.launchpad.net/music-app/+bug/1261587 [20:16] Launchpad bug 1261587 in Ubuntu Music App " Autopilot Testcase Needed: When the library has an empty state" [Low,Triaged] [20:16] does music need anymore AP tests ahayzen ? [20:21] nik90_: balloons: Yeps, that was the showstopper for me last time when I tried to use LXC for testing the SDK [20:23] balloons, we always need more ap tests :) [20:23] balloons, yeah that one is still valid...and we've found that there is a bug in trunk todo with empty library :/ [20:23] balloons, so that test would have been useful :P [20:24] balloons, and i need to rebase those url-dispatcher tests to the remix and try and land them again [20:25] ahayzen, I just sent a mail to the list making sure all the needs are filed as bugs, and tagged with needs-autopilot-test [20:25] balloons, ok i'll have a scan through the bugs this evening :) .. i've just finished my last exam so i'm 'free' again \o/ [20:26] ahayzen, awesome.. I don't want Letozaf_ to get bored ;-) [20:29] balloons, hehe [20:29] balloons, i've been playing about with QmlUnitTests while writing a Bacon2d game... i guess we should be using them more in the music-app? [20:30] balloons, is there a guide somewhere with more info about *where* to use them? like that acceptance test guide for AP? [20:33] ahayzen, are you writing QmlUnitTests for your bacon2d game? [20:33] :-D [20:33] kenvandine, i have a basic one that loads one of my objects and checks it collides with another ;) [20:33] awesome [20:33] kenvandine, i also have autopilot tests to check that the menus/pause/resume functions [20:33] mzanetti, hi [20:34] ahayzen, that's great, i've started on some qmltests, mostly to detect API breakages [20:34] kenvandine, i'm nearly ready to release a 'beta' to the store :) just a few minor bugs to resolve [20:34] but we need both [20:34] ahayzen, that's awesome [20:34] can't wait to see it [20:34] kenvandine, but i think my bacon2d version is massively out of date, it is the one from the sprint? [20:35] kenvandine, yeah i just finished exams today so i can get back to work on it :) [20:35] yeah, we've broken the API since then :) [20:35] i think... [20:35] maybe it was before [20:35] kenvandine, what did you break? .. just of interest [20:36] yeah, that landed on nov 3 [20:36] it was the box2d stuff [20:36] ahayzen, on qmltests, yes. I would point you at the following: http://doc.qt.io/qt-5/qtquick-qtquicktest.html http://doc.qt.io/qt-5/qml-qttest-testcase.html and http://summit.ubuntu.com/uos-1411/meeting/22396/writing-unit-tests-for-qml-applications/ [20:36] upstream qml-box2d refactored their code to make it easier for us to integrate it [20:36] balloons, thanks :) [20:36] so in integrating their changes, we had to change our API a bit [20:36] ah right [20:36] but this is much better and it'll be easier to maintain [20:36] one of the things we wanted to do before releasing 1.0 [20:37] i also noticed that i can use the sensors (TiltSensor) with no apparmor policy... but looks like jdstrand is away? [20:37] i have one more planned API break before 1.0, which is pretty small [20:37] would be awesome to have a 1.0 :) [20:37] oh? [20:37] yeah it says "policy pending due to LP: #1227116" [20:37] Launchpad bug 1227116 in apparmor-easyprof-ubuntu (Ubuntu Trusty) "qtsensors tilt and rotation is non-functional" [Medium,Fix released] https://launchpad.net/bugs/1227116 [20:37] on the app-confinement docs next to the 'sensors' policy [20:38] but ^^ bug is fixed and the sensors now work...so i suspect that policy *should* be enforced? [20:38] yeah, should be [20:38] i'll ping him next time i see him online :) [20:39] ahayzen, https://github.com/rpadovani/100balls/pull/6 [20:39] if you want to see the changes i needed to make to 100balls to support the API changes [20:39] ahayzen, the other API break is going to involve added a state property to Game, which is an enum [20:39] and making running on the scene read-only [20:40] so you have to pause the game by changing the property on game [20:40] and it'll handle things like pausing the game when the app suspends, etc [20:40] i pause it by changing the running property at the moment? [20:40] right [20:40] but to do this right, we need more than a boolean [20:41] so i'm making it an enum on Game [20:41] ah cool :) [20:41] which the controls the running property on the currentScene [20:41] it'll be much better :) [20:41] :) [20:41] not quite ready to land it [20:41] i need to figure out how to wire it up to the spirte animations [20:41] so they stop too [20:42] heh [20:46] kenvandine, is there a way of telling if something has come to a 'rest' in a certain direction... eg tell when an object hits the ground and stays? [20:47] kenvandine, i tried just listening to the velocity is zero but that doesn't help as then it counts it as zero when it is at the top of the 'jump' [20:47] Letozaf_: hi [20:48] unless i just listen to when it collides [20:48] ahayzen, i think there is an sleeping property [20:48] but that might be noisy [20:48] kenvandine, note it can move left/right as well [20:48] you could just listen to collisions [20:48] mzanetti, rpadovani told me some very big branches are landing in reminders, I was wondering If It's better to wait for writing autopilot tests, or may I carry on all the same ? [20:49] Letozaf_: hmm... good question [20:49] mzanetti, :-) [20:49] ahayzen, in pathwind i listen for collisions then change the sprite based on what it collided with [20:49] Letozaf_: this is the branch: https://code.launchpad.net/~mzanetti/reminders-app/offline-mode/+merge/244713 [20:49] so hits the ground it changes to walking [20:50] yeah i think collisions maybe best i'll have a play [20:50] Letozaf_: the ui didn't change much [20:50] Letozaf_: but it supports offline mode now. [20:50] Letozaf_: I'd be happy to get some feedback of that [20:50] basically my character can 'jump' but you can only jump when your touching the ground...at the moment if you happen to hit jump at the 'top' of the jump then you can double jump :) [20:50] ah [20:50] cool [20:51] mzanetti, ok I will take a look and let you know [20:51] so i think having something that just knows when it has collided with the ground and then allows you to jump again may be enough [20:51] ahayzen, i think you can also setup a contact and see if you are touching something [20:51] i've never done that [20:51] ooo that would be cool [20:51] how would i do that though lol [20:51] i've never tried that :) [20:52] mzanetti, I see that it needs fixing for CI, shall I try it all the same ? [20:53] ahayzen, Box2DContact has a isTouching function [20:53] kenvandine, is that in the QML interface? ...i've just been using this as my reference http://bacon2d.com/docs :) [20:55] ahayzen, ok, i think you can just use onBeginContact and onEndContact [20:55] between fixtures [20:56] so when you begin contact with a specific fixture set a property [20:56] and onEndContact you toggle it? [20:56] maybe [20:56] kenvandine, ah yeah that should work :) [20:56] there is a contacts example in the source [20:56] uses those signals [20:57] so onBeginContact check what the other fixture is, and if it's the one you care about set your proerty [20:57] property [20:58] i already use onBeginContact for my 'ball' :) [20:58] i usually do that by adding an objectName property to my entity [20:58] great [20:58] so i can use other.target.objectName [20:58] although target there is part of the new API [20:58] but you can do other.parent [20:58] i think [20:59] yeah other.parent is what i've used [20:59] the latest stuff adds a target property to the body [20:59] which is set to the entity [20:59] sweet :) [21:00] thanks for the pointers this should stop my double jumping issue hehe :) [21:01] Letozaf_: pushed a fix [21:02] mzanetti, ok thanks [21:02] ahayzen, great! === salem_` is now known as _salem === chihchun is now known as chihchun_afk