=== salem_ is now known as _salem [00:57] hi! [00:58] I'm trying to write my first web-app. For this I need to modify the user agent. but --user-agent-string="test" does not seem to work. Any suggestions? [01:00] Plus I would like to test it on my Ubuntu Phone (the bq e4.5 thing). But every time I connect it to my pc I get: Checking installed emulator package. ii ubuntu-emulator 0.16-0~136~ubuntu14.04.1 Search configured emulator instances. Detecting device..; And then it vanishes again [01:03] Anyone online? [01:11] I'm trying to write my first web-app. For this I need to modify the user agent. but --user-agent-string="test" does not seem to work. Any suggestions? [01:12] hey newbie-13442 [01:12] newbie-13442: There are people online; I am one of them though this is the first time for me writing something here [01:13] Okay, thanks for the feedback :) [01:13] If you're not getting help immediately, I'd recommend you to get yourself an IRC bouncer so you can read things that are written here while you're away, so you can get answers without actually being online [01:13] newbie-13442: if you have a device you don't need the emulator really [01:13] * ahoneybun needs one [01:13] Though I guess you'll have to have your own virtual server for that [01:14] Hm, why would one need the emulator at all, you can build apps for x86 / x86_64 ... [01:15] In my Ubuntu VM, the 'Computer' build target or whatever it's called was there from the beginning [01:15] I've never had it built for the desktop [01:16] why not? [01:16] jplatte: I use the the computer icon [01:16] hard to explain [01:16] deployed it local [01:16] but never built it [01:17] Uhm, wat [01:17] It has to be built before you can 'deploy' it. It's just done automatically [01:17] deploying does not really build [01:17] you can't run it outside of the sdk anyway [01:17] I'm pretty sure you can [01:18] Though I wouldn't want to, anyway [01:20] Locally it runs. Still I would like to test it on my phone. But i could not find any instructions online how to configure the use of a mobile [01:27] newbie-13443: same person [01:27] ? [01:32] yes, just switched browser [01:37] newbie-13443: I'm looking for the docs === ahoneycutt is now known as ahoneybun [01:53] editing the /usr/lib/x86_64-linux-gnu/qt5/qml/Ubuntu/Components/Extras/Browser/ua-overrides-desktop.js file and adding rules for the websites worked. Locally. For now. Not a nice solution though. [01:56] editing the /usr/lib/x86_64-linux-gnu/qt5/qml/Ubuntu/Components/Extras/Browser/ua-overrides-desktop.js file and adding rules for the websites worked. Locally. For now. Not a nice solution though. [01:57] @ahoneycutt [01:57] remains the problem, that I don't get the app onto my device === ahoneycutt is now known as ahoneybun [01:57] newbie-13443: what side are you making a web app for? [01:58] *website === chihchun_afk is now known as chihchun [07:57] Mirv: hi! I forgot to follow-up: indeed copying the qmake-cross + mkspecs works, I could build my app :-) [08:12] mardy: wow, not that I didn't believe it'd work, but I'm positively surprised anyway ;) [08:13] mardy: my main concern was that the qmake could say that it wants qt 5.3.2 instead of rtm's 5.3.0, but luckily that wasn't the case then === chriadam is now known as chriadam|away [08:22] Morning all o/ [08:26] good morning [08:47] morning :) [09:54] against what package do i file sdk bugs ? [09:54] (do we have some generic product ?) [10:08] * ogra_ looks on developer.u.c to see if there is any bug guidance for sdk issues [10:10] doesnt seem like :( [10:11] ogra_: qtcreator-plugin-ubuntu if IDE and ubuntu-ui-toolkit if UITK [10:12] no generic product otherwise === JMulholland_ is now known as JMulholland [10:32] Mirv, thanks ! [10:32] the docs should really provide some links for that [10:32] davidcalle, ^^^ [10:33] ogra_, +1 [10:39] balloons, it'd be great if you could review/test https://code.launchpad.net/~dholbach/help-app/1434415/+merge/253844 [10:39] * ogra_ files bug 1435778 [10:39] bug 1435778 in ubuntu-ui-toolkit (Ubuntu) "can not set applicationName property from variable" [Undecided,New] https://launchpad.net/bugs/1435778 [10:47] kalikiana, ^^^ this is the issue we talked about yesterday btw ... [10:48] (i tried an alternative method too, that fails the same way) [10:59] ogra_: kalikiana: from the logs I do see the app name being set to foo-bar.ogra [10:59] ogra_: kalikiana: however the organization name is still QtProject... [11:00] right [11:00] something uses the empty "applicationName" before using the imports [11:00] and seems to fall back to QtProject [11:01] dpm, balloons: there's also https://code.launchpad.net/~dholbach/help-app/1435783/+merge/253952 if you have a bit of time [11:01] ogra_: actually the applicationName when changed also alters the organizationName, so seems that is not set at all... [11:02] "~/.cache/QtProject/foo-bar.ogra/Cache" is clearly using the correct name in some manner [11:02] but it's odd that the other values don't match [11:03] it might be whatever uses this Cache folder is using APP_ID [11:05] kalikiana: shouldn't the QtProject be missing as we set the org name to an invalid string? [11:09] zsombi: that's why I said it might be something using APP_ID directly [11:09] kalikiana: ah, yea, sorry [11:10] dholbach, what does make -s do? [11:10] re: your merge proposal [11:10] kalikiana: afaik we don't have our own launcher, do we? [11:10] kalikiana: meaning we still use qmlscene [11:11] dpm, "quiet" [11:11] it's irrelevant [11:11] or "silent" rather [11:11] ok, thanks [11:11] zsombi: outside of testing, yes [11:17] dholbach: I was using your excellent tutorial on producing .debs for a PPA. You use bzr-specific commands. I prefer git and note there are debian-package git extensions. Have you used them? Do they work as well as bzr? === ara is now known as Guest75262 [11:31] zsombi, this app definitely runs "qmlscene $@ Main.qml" [11:32] the code without the global var stuff (and still quite messy, its just a prototype yet) is at http://bazaar.launchpad.net/~ogra/junk/alternate-webapp-container/files/head:/ [11:55] can I get the entire document from a delegate of a ListView that has a U1db.Database as its model? I can get to docId and contents individually, but I'd like to pass on the entire Document to a method === _salem is now known as salem_ [12:11] popey, dholbach, mhall119, I just noticed this: http://ubports.com/ [12:11] davidcalle, balloons ^ [12:12] it seems the guy who is working on the oneplus one port created it [12:12] they've got a nice feature where one can request for a port, and see which ones are the most popular [12:13] perhaps we could work with them to keep the list of devices? [12:19] zsombi: ping [12:20] nik90: meeting, will ping U later [12:22] dpm, that's nice. Maybe I could make a list of recently enabled devices and open a thread on their forum? As a first step to build a common list and maintain it together? [12:24] zsombi: ok [12:24] davidcalle, could be a good idea [12:37] dpm, nice :) [12:44] and there's https://code.launchpad.net/~dholbach/help-app/1435772/+merge/253954 too [12:44] ^ balloons [12:52] ack dholbach [12:53] thanks :) [13:02] renatu, ping [13:04] balloons, did you see the mail I sent? I'm not quite sure what we can do to fix the jenkins builds... [13:04] charles, ping [13:04] both pyflakes/pyflakes3 and the missing PPA looks like stuff not in the control of lp:help-app [13:05] dholbach, I looked yesterday at the build issues and thought it was on our end, but we can fix if not [13:05] it should be using the coreapps ppa for pelican [13:05] for pyflakes, we can change it to pyflakes3, it's not running pyflakes3 [13:06] ok... so we need to move away from https://code.launchpad.net/~ubuntu-touch-coreapps-drivers/+archive/ubuntu/collection? [13:06] where is pelican? [13:06] it uses https://launchpad.net/~ubuntu-touch-coreapps-drivers/+archive/ubuntu/daily [13:06] ok, I'll upload pelican to that one then [13:08] balloons, shall I also direct the daily build over to ppa:ubuntutouch-coreapps-drivers/daily? [13:09] or shall I stop the daily build altogether? [13:09] pelican is now building in the right ppa: https://code.launchpad.net/~ubuntu-touch-coreapps-drivers/+archive/ubuntu/daily/+build/7105050 [13:09] dholbach, we can swap the ppa ci uses if you wish, but the other core apps use that ppa, so it makes more sense to swap to it I think [13:10] so yes, add builds to it [13:10] no, I'll move everything [13:10] cool [13:12] https://code.launchpad.net/~dholbach/help-app/1434415/+merge/253844 will need to be landed to make the build pass on utopic [13:12] (separate from the pelican requirement) [13:15] now that I merged from trunk the diff should be smaller as well [13:15] all in all we also have more tests in the pipe now :) [13:32] working on getting pyflakes changed to pyflakes3 [13:34] where is that defined? [13:38] in the jenkins job. Should be done in just a sec and we can re-run [13:42] is that in a branch somewhere? [13:43] or in a jenkins ui somewhere? [13:43] the latter [13:43] ok [13:43] well.. actually it might be in the branch config [13:45] let me know when you found it - I'd be interested to learn more about that bit as well :) [13:46] well I've simply asked CI to swap it. At one point there jobs read from a config we could make an mp against in bzr, but they still have the power [13:46] meaning, they could just change the value in jenkins anyway [13:47] ok [13:47] davidcalle, popey, mhall119, dpm: who's up for the Q&A today? [13:48] dholbach, popey, would you be up for it? Mike and I have a call at that time [13:48] ok [13:49] dpm, is it in 1h10m or 2h10m? [13:49] dholbach, in 2h10, despite what our calendar says [13:50] :-) DST <3 [13:50] cool [13:52] dpm: I'm up for it [13:52] great, we have a plan, then [13:52] #blamemhall119 [13:52] its at the right time in the UOA calendar [13:56] dholbach, why do you suspect we need pyflakes3 though? [13:57] * balloons is curious now and wondering as pyflakes2 seems to work fine locally [13:57] mh? [13:57] we use python3 [13:57] sure.. the weird part is, so do all the other core apps and we've not had an issue [13:58] daniel@daydream:~/dev/apps/help$ pyflakes internals/*.py internals/tests/*.py internals/translations/*.py [13:58] internals/translations/build.py:314: undefined name 'NotADirectoryError' [13:58] daniel@daydream:~/dev/apps/help$ pyflakes3 internals/*.py internals/tests/*.py internals/translations/*.py [13:58] daniel@daydream:~/dev/apps/help$ [13:59] NotADirectoryError is builtin in python3 [14:01] maybe the other core apps only use features and classes which exist in both (and are imported from the same modules) [14:01] some use only py3 code and modules actually [14:02] but anyways, we'll use py3 [14:02] right [14:03] I was just saying that they might be using code where the written code imports modules and classes which are the same on python2 and python3 [14:07] hello! I can't get the webview to work, I've copy pasted the example in the documentation [14:22] nik90: I'm back, sorry, it took a bit more than expected [14:24] zsombi: hey np...I figured out the source of the clock app crasher and while trying to fix it I ran into some stuff that I wasn't sure about [14:24] zsombi: so it turns out that clock app is crashign due to the bottom edge which uses the get() function to find out the next closest active alarm [14:25] zsombi: and it checks every minute (which sucks) [14:25] nik90: so it crashes when it calls the get()? [14:25] zsombi: yes [14:25] hmm!!! [14:26] zsombi: and that's why the clock app pretty much crashes randomly at different places...it doesnt matter if the user is in the settings page, or the add world city page..it still crashes because of the bottom edge [14:27] zsombi: so I figured that I would use the signals provided by alarm model which include dataChanged, countChanged and modelReset to update the bottom edge [14:27] nik90: if teh crash comes due to the get(), then we are in trouble... [14:27] nik90: yes, you coudl use that, however this get is a crazy dude [14:29] zsombi: the thing is I need to use the get() to iterate through the alarm list and figure out the most recent active alarm [14:29] I can't use any other method [14:29] zsombi: and guess what, to do that I am using the same method that caused the crash in the edit alarm page [14:29] where I passed the alarm mode [14:29] s/mode/model [14:30] zsombi: so I figured I would better solve this crash by using the signals to retrieve the recent alarm and pass the *alarm date* alone which can then be used to display "next alarm in 3h 40m" etc.. [14:31] nik90: then we are still having issues in the model... [14:31] zsombi: there by no more alarm model passing (which is the cause of the crash) === om26er_ is now known as om26er [14:31] nik90: get() should never crash [14:32] zsombi: could get() perhaps be crashing when it is called before the alarm model is fully initiated? [14:32] zsombi: so in essence it is trying to read a null state or something? [14:33] nik90: nope, it's a different thing then... last time I saw some QHash hickup, it does not run in a separate thread, so there shouldn't be any fancy problems like that [14:33] pmcgowan: ^ [14:34] nik90: do you have the bug for the app for this? [14:34] nik90: let's assign toolkit for it as well [14:35] zsombi: yeah, 1 sec.. there are 2 bugs reported [14:35] nik90: one is enough :D [14:35] zsombi: bug 1431579 and bug 1434504 [14:35] bug 1431579 in Ubuntu Clock App "Clock App crashes when clicking on the settings icon" [Critical,Confirmed] https://launchpad.net/bugs/1431579 [14:35] bug 1434504 in Ubuntu Clock App "Clock app crashes when trying to set a new alarm" [Critical,Confirmed] https://launchpad.net/bugs/1434504 [14:36] zsombi, ok will start tracking that as a blocker [14:36] pmcgowan: right [14:39] nik90: does the New Alarm bug also call AlarmModel.get()? [14:39] nik90: or is the bottom-egde swipe which tips in there as well? [14:40] zsombi: well the code bottomEdgeTitle: alarmUtils.set_bottom_edge_title(alarmModel, currentTime) is called every minute which in turns calls the alarm model get function [14:41] nik90: ok, so if you comment that code out, does the New Alarm crash still? [14:41] http://bazaar.launchpad.net/~ubuntu-clock-dev/ubuntu-clock-app/utopic-3.0/view/head:/app/alarm/AlarmUtils.qml#L137 [14:41] zsombi: nope it doesn't crash anymore [14:41] nik90: excellent, so the get() is our monster [14:42] yeah definitely [14:43] nik90: so smells like all teh crashes are having teh same root cause [14:44] nik90: we have 3 bugs for this it seems - shall we mark them dupes and add a task for uitk? [14:44] bug 1422693 bug 1429273 bug 1434504 [14:44] bug 1422693 in Ubuntu Clock App vivid-15.04 "[Vivid][Regression]Cannot change alarm switch status (enable/disable alarms)" [High,Fix released] https://launchpad.net/bugs/1422693 [14:44] bug 1429273 in Ubuntu Clock App "Clock App crashes when editing on alarm on vivid" [Critical,Fix committed] https://launchpad.net/bugs/1429273 [14:44] bug 1434504 in Ubuntu Clock App "Clock app crashes when trying to set a new alarm" [Critical,Confirmed] https://launchpad.net/bugs/1434504 [14:44] nik90: I just need a good minimalistic QML file to repro this [14:45] zsombi: I will try to create a minimalistic project for that [14:45] popey: bug 1422693 is fixed and no longer relevant, no? [14:46] nik90: no project, a QML file... :) [14:46] bug 1422693 in Ubuntu Clock App vivid-15.04 "[Vivid][Regression]Cannot change alarm switch status (enable/disable alarms)" [High,Fix released] https://launchpad.net/bugs/1422693 [14:46] ah right [14:46] the other two though. [14:46] zsombi: well since it happens only with pagestack and passing alarm model around, I need multiple qml files ;P [14:46] * popey dupes [14:46] popey: yes, they can be marked dupes [14:47] nik90: perhaps a component loaded dynamically would also do teh job [14:47] zsombi: ack [14:47] ok, lets use bug 1429273 - task added pmcgowan [14:48] bug 1429273 in ubuntu-ui-toolkit (Ubuntu) "Clock App crashes when editing on alarm on vivid" [Critical,In progress] https://launchpad.net/bugs/1429273 [14:48] popey, thanks [14:49] popey: haha, I added one to https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/1431579 [14:49] Ubuntu bug 1431579 in Ubuntu Clock App "Clock App crashes when clicking on the settings icon" [Critical,Confirmed] [14:50] popey: that is also a dupe of teh same toolkit issue [14:50] also marked dupe [15:00] nik90: interesting! seems get() sometimes returns null, so if you dereference a null pointer then crash can happen [15:06] nik90: so beware of the null result get() may return [15:13] zsombi: oh..so if I do, var currentAlarm = alarmModel.get(i).. .I need to check the validity of currentAlarm by if(currentAlarm) ? === chihchun is now known as chihchun_afk [15:31] nik90: this is crazy! I made a simple test with a model, initTestCase generates 10 alarms and 10 test alarms, when I start to remove the test alarms in clearTestCase, I get null atth e3rd test alarm removal [15:32] nik90: the alarmList[0] sais the Alarm is a valid pointer, and that pointer is passed across the whole chain, fromadapter to model, and by the time arrives in QML is null... wth? [15:32] hmmm [15:35] nik90: QDEBUG : components::AlarmModelStressTest::cleanupTestCase() (ucalarmmodel.cpp:217, UCAlarmModel::get) - get 4 UCAlarm(0x1a37350) "testAlarm_4" [15:35] QDEBUG : components::AlarmModelStressTest::cleanupTestCase() (file:///home/zsombor/CloudStation/projects/alarms/alarmmodel-get-crash/tests/unit_x11/tst_components/tst_stress_alarmmodel.qml:73, clearTestAlarms) - null [15:35] nik90: the first line is the trace from teh AlarmModel::get() just before returns the alarm [15:35] nik90: the second trace is from teh QML right after the get() returns [15:35] nik90: and that oen is null [15:40] zsombi: it fails while accessing from the qml side..have you noticed this before (in any other projects) ? [15:41] nik90: not really... I don't recall any such weirdness [15:43] nik90: and the thing is that it is not teh first one which fails... [15:43] nik90: it si random... [15:43] is [15:46] yeah the crash happens at different times as well for me.. sometimes straight on app startupp .. sometimes while navigating to another page [15:46] and surprise surprise..it doesn't crash when there are 0 alarms [15:54] nik90: hmm... I just added a mutex into the AlarmList.... and seem sdelete is the guilty one [15:54] seems [15:56] zsombi: but I don't call delete though during my crashes [15:56] nik90: right.... and you still get NULL pointer [15:56] which is weird, you shouldn't [15:57] nik90: or you may, if the given index is out of model.count [15:57] nik90: but even then you;d get a crash... [15:57] zsombi: well I do check (!alarmModel || !alarmModel.count) [15:57] before trying to access the alarm model [15:57] nik90: but the thing is you most likely call ... yes L) [15:59] nik90: ok, mutexes don't help at all, as I thought.... [15:59] nik90: the only thing I can foresee is that the alarm is deleted... for some reason [16:01] zsombi: I can't get the sample app to crash.. I used a Loader to load the alarmModel, passed it around to the function and accessed it pretty much the same way as the clock app does [16:02] I added debug statements after pretty much every line of the function that causes the crash and it promptly crashes at the "var currentAlarm = alarmModel.get(i)" line [16:07] nik90: so you cannot get it repro with a sample code? [16:08] yes [16:08] nik90: damn.... [16:08] nik90: I can clearly see it returning null... [16:11] nik90: I hav etraces in Alarm destructor, the alarm that is supposed to be returned is not deleted :/ [16:14] thsi stupid backend is killing me... [16:21] zsombi: let me using the alarm model signals to update the bottom edge title and avoid passing alarm model as an argument to other functions [16:22] zsombi: since that was the cause of the crash in the edit alarm page that we fixed [16:22] nik90: ok, but beside that, calling a get() which has a valid return value in C++ and in QML a null result is utmost brainkiller [16:28] nik90: DAMN IT! [16:29] nik90: if I create a temporary Alarm object and not the stock one, it works [16:38] nik90: ok, so seems we have a solution [16:38] zsombi: ? [16:39] nik90: you make sure you don't pass any model in between the pages [16:39] nik90: and I have to create temporary Alarm objects each time get() is called [16:40] nik90: I'll have an MR from toolkit [16:41] zsombi: well I instantiate the alarm model in MainView and have to pass it to the alarm list page. If not I need to declare a AlarmModel twice [16:41] zsombi: and if I don't instantiate the alarm model in MainView then the bottom edge wont work [16:41] nik90: ah... hmm [16:42] zsombi: the crash appears when I pass the alarm model and use the get() function.. Fortunately I don't do this in the alarm list page since I use a listview to get the alarm objects and show them [16:42] that combination is what I previously mentioned that I will try to avoid [16:53] nik90: bzoltan_: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/alarmmodel-get-crash/+merge/253981 [16:54] nik90: get this MR before ytou do anything with your app [16:54] zsombi: ack. Testing it now [16:55] nik90: thx, lemme know what happened, I most likelly will see it tiomorrow morning [16:55] zsombi: ok. I need to grab dinner soon but I will definitely have the results by tonight [16:55] nik90: the test I have it with teh MR reproduced the get() crap [16:56] nik90: thx!!! [16:56] * zsombi eod [16:56] davidcalle, what can we do about bugs 1313764, 1434887, 1435473, 1435471, 1317065? [16:57] bug 1435473 in Ubuntu App Developer site "API page empty" [Undecided,New] https://launchpad.net/bugs/1435473 [16:57] bug 1313764 in Ubuntu App Developer site "Unity Web API does not have a 14.04 framework doc" [High,Confirmed] https://launchpad.net/bugs/1313764 [16:57] bug 1317065 in Unity Website "Move old indicator docs to unity.ubuntu.com" [Undecided,Confirmed] https://launchpad.net/bugs/1317065 [16:57] bug 1434887 in Ubuntu App Developer site "links in unity webapps dev portal have broken (404) links" [Undecided,New] https://launchpad.net/bugs/1434887 [16:57] bug 1435471 in Ubuntu App Developer site "cookbook page empty" [Undecided,New] https://launchpad.net/bugs/1435471 [17:02] Hello anyone, anyone I can ask about QTCreator not detecting the Aquaris Ubuntu? [17:02] dholbach, for web api and cookbook, we could probably redirect to the guide page. [17:03] I'm wondering what kind of content is missing now [17:03] Does anyone have any idea why a Aquaris Ubuntu Edition doesn't appear in the Ubuntu Device section of the Ubuntu-SDK? [17:03] is this stuff that should be on unity.u.c or am I conflating two different issues? [17:04] davidcalle, ^ [17:05] dholbach, it's not the same thing (afaik), the web api was docs put together by the old webapp team, standalone html files, with how to integrate website into the launcher (count), indicators (messaging, music control), notifications [17:05] websites* [17:06] dholbach, it's still somewhere on lp, and is deprecated (13.10 only) [17:08] dholbach, that's why I think we should redirect to the guide. [17:08] AlanBell, which docs are you missing there?^ [17:08] kalikiana, t1mp, zsombi, perhaps one of you guys can help Risikoruk? ^^ [17:15] dpm: davidcalle: dholbach: if you see Risikoruk back would you proxy him that he needs to enable development mode in the Settings → About this device → Developer mode. [17:16] bzoltan_: I don't know Risikoruk :/ [17:16] and need to run now [17:17] bzoltan_, I'm out in a few minutes, but sure, if I see him [17:18] same here [17:19] i think he found help in -touch === balloons is now known as iwishiwasascoola === iwishiwasascoola is now known as balloons [17:24] zsombi: hmm that seems to have done the trick. The app did crash *once* at startup. I will keep testing throughout the night to see if I can reproduce it. [17:24] zsombi: without your MP it crashes pretty much on every run which I don't see now [17:53] nik90: faced another issue http://paste.ubuntu.com/10671409/ [18:06] daker: I have seen those errors as well, but it still installs the necessary packages required by the sdk [18:07] daker: just ignore them and proceed to create the desktop file and open the app using lxc [18:09] ... qtcreator is still compiling... [18:31] bzoltan_: kalikiana: what else is needed to merge https://code.launchpad.net/~kalikiana/u1db-qt/qmlDocs/+merge/208818 ? [19:05] Bit stuck using the Ubuntu SDK. I just got my Ubuntu phone this morning & I'm trying to copy an application I've written to it. Can't seem to work out what to do. Is there a tutorial or an explanation of what to do? [19:20] keef: have you enabled developer mode on the phone? [19:23] mhall119: I need to tweak the docs as per marco's comments, I'll do that now, then it can be passed to ci [19:23] thanks kalikiana [19:24] mhall119: thank you for checking, sometimes I just have too many tabs open… [19:25] * mhall119 looks at his tab bar [19:25] I understand completely :) [19:28] whenever i see "tab bar" all i can think is that someone is dyslexic [19:28] or it's time they should pay it and go home [19:29] lol [19:32] dobey: I've got a bit further, I'm using a different cable to connect the device & it's detected it, but the Ubuntu>Install application on device... comes up with a progress bar, but it doesn't appear on the device [19:34] keef: ah ok. i'm not quite sure how all that works exactly, and i've not used it. maybe someone else can help you though [19:35] dobey: Thanks anyway :) I'll keep bashing my head against it [19:35] keef: if that progress bar comes up, then it means that the SDK hasn't detected your device [19:35] keef: Can you run "adb devices" in the terminal and see if your device is listed [19:36] rpadovani: ping [19:36] keef: make sure your screen is unlocked when you do the "install on device' perhaps ^^ [19:37] based on what nik90 said [19:43] Mirv: still around? mind passing this on to ci? https://code.launchpad.net/~kalikiana/u1db-qt/qmlDocs/+merge/208818 [19:53] akiva-thinkpad: what was that name of the help app again on launchpad , === greyback__ is now known as greyback [19:54] justCarakas, hmmm? [19:54] help app? I think that is ahoney's project [19:54] or are you talking about my autopilot plugin? [19:55] justCarakas, https://code.launchpad.net/help-app? [19:56] yes balloons that one thx :) [20:01] seb128: I might finally have a fix for your bug 1381432 [20:01] bug 1381432 in Ubuntu Clock App "Display "no active alarm" while checking" [Medium,In progress] https://launchpad.net/bugs/1381432 [20:13] nik90, pong [20:13] rpadovani: hey, how is it going? [20:14] nik90, busy as hell, but all it's more than good :-) What about you? [20:14] rpadovani: Good. I saw your g+ post about getting your glasses removed..I was thinking of doing the same later this year using a lasik surgery [20:16] rpadovani: hey, when you find time, could you review https://code.launchpad.net/~nik90/ubuntu-clock-app/improve-setting-listitem/+merge/252930 [20:16] nik90, you totally have to do it: no pain at all. I had 10 dioctries for each eye, so the surgery was a bit long, but I had no problem. Two days of blind to a life without glasses [20:16] nik90, sure, I do it tonight :-) [20:17] rpadovani: yeah I was planning to do it in feb what I went to india, but didnt have time to get it done [20:18] s/what/when [20:18] rpadovani: well enjoy a glass free experience :-) [20:18] \o/ ty [20:23] Who's feeling knowledgeable about QML? Specifically, I have a checkbox which, when changed, gets its value (true or false) written to the database. However, that checkbox's initial value depends on that read *from* the database. So, if you create a variable isEnabled, set CheckBox { checked: isEnabled; onCheckedChanged: updateDB(checked); } and updateDB() saves the value and then updates the isEnabled var... [20:23] you get an infinite loop. [20:23] How is one *meant* to do this? [20:23] it'd be easier using u1db rather than sql, but I don't think that a scope can read a u1db database. === ahoneycutt is now known as ahoneybun [20:29] aquarius: hey [20:29] aquarius: why don't you instead use the onClicked signal instead [20:29] * nik90 has run into this paradox before :P [20:30] nik90, because there's also a Label next to the checkbox, and clicking the label should also toggle the checkbox [20:31] hmm ok [20:32] the Label contains a MouseArea which calls checkbox.clicked(), but I am exceedingly mistrustful of the idea that programmatically invoking .clicked on a thing will properly fire its onClicked handler :) [20:32] aquarius: let me try it on my end, 1 min' [20:36] aquarius: then when you click the label, change the checkbox's checked value and also update the database [20:37] ya, but then if the database write *fails*, I have to switch the checkbox back agan, from inside the db update function, which is a massive layer violation :) [20:38] aquarius: I tried setting the label's mouse area to call checkbox.clicked() and verified that it does call the onClicked signal correctly [20:39] I mean you are firing a signal, so I don't see why its handler wouldn't be called by qml [20:39] huh. Maybe what I do is *cancel* the clicked handler [20:39] and *only* set the checkbox value from the db function [20:40] nope; you can't cancel a clicked handler [20:40] well, not by returning false, anyway [20:41] is that possible? [20:41] aquarius: well since you set the checkbox value to the database's value, if the db write fails, wouldn't the checkbox also return to its original value? [20:41] I am not sure how to cancel the clicked handler [20:41] yep. but that's what I *want* it to do, because the db value didn't change [20:43] aquarius: paste.ubuntu.com/10672572 [20:44] I am not sure how to simulate db write fails thouhg [20:44] s/thouhg/though [20:45] yeah, that's what I'm basically planning to do, I think [20:45] write a mock for the DB API which fails [20:46] returns an error, or whatever [20:49] yeah. I think I was just overcomplicating things and I didn't need to. It works with onClicked [20:50] sweet. So, app basically works [20:50] now all I need is the scope to go with it :) [20:51] hmm, is there a way to make a qml app fullscreen without having to use a window ? [20:51] for which cwayne has said he'll write some docs, if he gets a chance. [20:51] MainView doesnt seem to support something like toggling fullscren [20:51] ogra_, um, how does it not have a window? :) [20:52] well, none i have defind [20:52] *defined [20:52] i mean the qml Window element :) [20:52] oh! [20:52] that makes more sense ;-) [20:54] I don't *think* so. MainView imports Qt.Window, but it is not itself an ApplicationWindow :( [20:54] i have a webview inside a page inside a mainview ... if the webview goes fullscreen, it only does that inside the mainview ... so the panel stays visible and the app behaves like maximized [20:54] because your app isn't creating the window; qmlscene is [20:54] right [20:55] qmlscene --fullscreen will run it fullscreen, wooo, but that's not toggleable from code inside that window, I don't think [20:55] lol [20:55] yeah, not what i want [20:56] well, if it bothersme enough i might try to use a Window ... [20:56] ogra_: I heard that's planned in the SDK to add Mainview support for fullscreen apps [20:56] how does the browser do that i wonder [20:56] ogra_: the gallery apps does this using c++ code afaik [20:56]