/srv/irclogs.ubuntu.com/2014/07/17/#ubuntu-unity.txt

=== _salem is now known as salem_
=== salem_ is now known as _salem
achiangtrying to use sqlite3 in my qml/qt app (build host is trusty). should i be using the system's libsqlite3-dev package, or is there some Qt version of that package to use?04:47
achiangreason i'm asking is because i encounter this during cmake: http://pastebin.ubuntu.com/7807167/04:47
achiangbut also encountering this during runtime: http://blog.forwardbias.in/2011/08/when-sqlite-queries-fail-for-no-reason.html04:48
achiangso wondering if ubuntu suffers that issue or not04:48
achianghm, nevermind. dumb question05:08
tsdgeosSaviq: did we lose qmluitests altogether in CI?07:09
tsdgeosconfiguring fails :/07:10
tsdgeosfginther: ↑ ?07:10
tsdgeosoh come on07:21
tsdgeoswhy does unity8.shell.tests.test_emulators.GenericScopeViewEmulatorTestCase.test_open_preview fail in remote and not for me :/07:21
tsdgeoswe don't record videos of autopilot anymore?07:36
Saviqtsdgeos, we do07:39
tsdgeoscouldn't find it07:39
SaviqFYI, I'm incapacitated07:40
Saviqlaptop died07:40
tsdgeosfor https://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-mako/2329/07:40
Saviqdoesn't get through POST07:40
tsdgeosused to be in build artifacts, no?07:40
tsdgeosSaviq: oh :/07:40
Saviqtsdgeos, yeah, gotta call Dell, hope they'll make it happen yesterday07:40
Saviqtomorrow I mean07:40
Saviqthough yesterday'd be better07:41
Saviqtsdgeos, indeed no video there07:41
Saviq:|07:41
Saviqbrb07:42
Saviqhuh, apparently you can hit Dell Tech Support via Twitter these days ;)07:59
tsdgeosyeah well08:01
tsdgeostwitter is about the only way i get companies to do what i need08:02
tsdgeosget infinite calls of company X and ask them to stop calling -> nope08:02
tsdgeoscomplain on twitter -> yes sure, of course we're sorry (have to see if they still call again though)08:02
Saviqnever tried yet08:04
Saviqand TBH love Dell support for not bullsh$tting, I told them it doesn't go past POST, response was "OK, we need to replace the motherboard, will be there tomorrow"08:05
Saviqthis is probably the single best feature of Dell's notebooks (or well, you could say they shouldn't break at all...)08:06
tsdgeos:D08:06
tsdgeosSaviq: you paid some extra support, no?08:06
Saviqtsdgeos, yeah, NBD, but it's money well spent08:07
Saviq4 years, too08:07
tsdgeosi'll think about it with my refresh laptop08:07
Saviqnot sure of course what the actual cost was08:07
seb128Saviq, jibel: bug #1336675 ... isn't the sim unlock dialog part of unity8?08:07
Saviqbut something in the order of €300-350 probably08:07
ubot5bug 1336675 in ubuntu-system-settings (Ubuntu) "SIM unlock - UI hangs when user enters wrong PIN" [Undecided,New] https://launchpad.net/bugs/133667508:07
Saviqseb128, it is08:07
* seb128 reassign08:08
seb128Saviq, thanks08:08
seb128jibel seemed to insist it's a setting issue for some reason ;-)08:08
Saviqseb128, it's probably an indicator-network one actually08:08
Saviqyeah, as pat assigned08:08
seb128Saviq, that's where Pat reassigning it, but jibel disagreed and put it back on settings08:08
Saviqseb128, I'll actually verify in a mo, flashing now and got a locked SIM here08:09
seb128Saviq, thanks08:09
jibelseb128, there is a system-settigns crash that's why I reassigned. I can retry but I don't like the idea of locking  my sim card :)08:11
seb128jibel, having some sort of bt would be nice, settings are not even supposed to be running at this time08:11
Saviqjibel, can't confirm your bug, where were you putting the wrong PIN? after "Unlock SIM" in the indicator?08:24
Saviqjibel, put wrong PIN twice and then the correct one, all worked08:25
Saviqexcept for the "you have foo tries" label, which got displayed in a random place08:25
jibelSaviq, on the keypad, after you select 'Unlock SIM' in the indicator, you enter the wrong pin and validate08:25
Saviqjibel, yeah, worked here08:25
Saviqput the wrong one twice08:26
jibelSaviq, on second try the UI froze completely08:26
jibelSaviq, I'll try again08:26
Saviqktx08:26
jibelSaviq, seems to be okay now. I'm closing the reporT.08:31
Saviqtsdgeos, the qmluitests fail looks like a network issue08:46
Saviqhopefully temporary08:46
tsdgeosSaviq: you think so? seems like a package is conflicting or something08:47
Saviqtsdgeos, ah the last one, I've seen that before, it looks like cgmanager does not configure due to a post-install script fail08:48
Saviqtsdgeos, https://launchpad.net/ubuntu/+source/cgmanager/0.27-0ubuntu708:49
Saviqtsdgeos, but apparently stuck in proposed08:49
tsdgeosgrrr08:49
tsdgeosunstuck!08:49
Saviqcaused a regression in lxc autpkgtests08:49
Saviqhttp://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html08:49
Saviqalthough that looks like a misattribution...08:50
Saviqtsdgeos, the ap test failed for me here on dash_overview branch08:52
tsdgeosSaviq: interesting08:52
Saviqtsdgeos, although differently08:52
tsdgeosSaviq: log?08:53
Saviqand it might be a result of small screen08:53
Saviqtsdgeos, http://pastebin.ubuntu.com/7807918/08:53
* Saviq tries on external screen only08:53
tsdgeosyeah i think it's failing to swipe08:54
Saviqyeah it selected some text in my terminal08:54
Saviqyeah passed on 1080p08:55
greybackSaviq: how does it feel to be an orange-sunglasses-wearing Apple hipster?09:16
Saviqgreyback, not too great, I'm not dealing well with the keyboard (mostly Fn, alt, cmd and F? keys get me)09:17
SaviqI've been using this from time to time, but working on this isn't that great09:18
greybackSaviq: you'll get used to it :)09:18
Saviqgreyback, no I won't, getting my shiny Dell back tomorrow ;P09:18
tsdgeosSaviq: we are in dash overview, search, select a scope that is a favorite, so we go back to regular dash (not a temp scope), when showing dash overview again, is it still in search?09:23
tsdgeosi vote no09:24
Saviqdednick, terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<boost::program_options::unknown_option> >'09:24
Saviq  what():  unrecognised option 'prompt-file'09:24
Saviqtsdgeos, no no09:24
dednickSaviq: erm. huh?09:24
dednickwhat's that from~?09:24
Saviqdednick, unity8 startup after the env change in unity8.conf09:24
dednickhm.09:24
tsdgeosSaviq: no no means yes? or means you agree with me doubly?09:25
dednickdamn. maybe alan_g's change hasn't been released in image09:25
dednickdidn't think it'd matter though09:26
Saviqtsdgeos, when you close the overview, you reset it09:26
dednickalan_g: ^ ?09:26
greybacktsdgeos: "no no" != yes09:26
greyback:D09:26
tsdgeosgreyback: depends on the programming langauge ;)09:26
alan_gdednick: which change? what is "image" built from?09:26
tsdgeosSaviq: okidoki09:26
greybackno means no, and no no also means no09:26
Saviqtsdgeos, if you asked just one question, you'd know which one the answer is for ;)09:26
Saviqaaanyway09:27
Saviqyeah, reset it09:27
tsdgeosSaviq: i actually only asked one question, even if it was long :D09:27
Saviqindeed ;)09:27
greybackno no no no no no no no no no no no there's no limit09:27
anpoksometimes one no does not cut it09:27
Saviqgreyback, you're in a good mood today ;P09:27
Saviqalan_g, I'm getting abort after setting the MIR_SERVER_PROMPT_FILE env var09:28
dednickalan_g: the prompt-file work in mir.09:28
Saviqdoes Mir interpret all MIR_SERVER_* vars, trying to match them to command-line options?09:28
dednickSaviq, alan_g: sounds like a bug in mir though. setting env vars shouldnt abort if they aren't being used...09:28
anpokSaviq: afaik yes09:28
Saviqyikez09:28
Saviqdednick, ok, I'll drop it from our silo for now, then09:29
alan_gSaviq: all the variables starting MIR_SERVER_ must be valid options09:29
dednickSaviq: hm. which silo is it?09:29
Saviqdednick, 3, unity8 updates09:29
alan_gIt is sensible to reject misspellings09:29
Saviqdednick, I'll add to silo 18 then09:30
dednickSaviq: i think 18 has been marked for release already09:30
* alan_g still wants to remove the hack that ignores misspellings on the command line09:31
Saviqdednick, ok hmm09:31
Saviqdednick, nope, it's been blocked by the clean exit09:31
Saviqcamako, can I add a unity8 MP to silo 18? https://code.launchpad.net/~nick-dedekind/unity8/trusted-socket.prompt-file/+merge/227051 this one09:32
dednickSaviq: ok. might as well add it in there as well. since it's related09:32
Saviqyeah owners are not around anyway09:32
camakoSaviq, I don't know how that'd affect packages already built... It shouldn't invalidate anything right?09:33
Saviqcamako, nope09:33
Saviqcamako, isn't it 4am where you at?09:33
camakoSaviq, nope, I'm in Turkey :-)09:33
Saviqcamako, not what the directory says! ;)09:34
camakosaviq, though we'd need to do another round of testing09:34
camakoSaviq, just for a few weeks for the summer... My base is still TX09:34
Saviqcamako, we can release it later separately if you'd rather, dednick?09:35
Saviqwhat's the impact when the var's not there?09:35
dednickSaviq: trust sessions wont work09:35
camakosaviq, it's okay.. go for it09:36
Saviqso why isn't it in silo 18 in the first place :D09:36
dednickSaviq: because yeterday it was marked for release, but then i think we found the issue with clean exit09:36
Saviqcamako, what's the ETA on silo 18? I've a unity8 landing in flight, will have to rebuild anyway after that lands, so might not add it there just yet09:36
dednicklater09:36
Saviqdednick, 'stood09:37
SaviqETA on my silo landing would be about an hour09:37
camakosaviq, silo 18 is being tested by myself and tvoss, as we speak. The clean exit fix...09:37
Saviqcamako, ok, land it then, we'll land the env var separately09:38
camakosaviq, sure.. I think it'll land faster that way09:38
Saviqyeah agreed09:38
Saviqtsdgeos, you wanna check your changes with autopilot3... https://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-mako/2341/console09:54
Saviqbrb09:55
tsdgeosyeah...09:56
tsdgeosdamnit :D09:56
tsdgeosthat much wait for nothing09:56
* Saviq fixed the keyboard (except the damn Fn is still on the far left :P)10:10
Saviqcome on, muscle mem, kick in!10:10
tsdgeosSaviq: ok, search done10:42
tsdgeosdid we get any answer on what needs to happen for the corner cases we asked about?10:42
Saviqtsdgeos, no10:42
Saviqgtg file tax reports, biab11:02
dednicktsdgeos: hey. having a problem with not being able to read a qt property from qml. "QMetaProperty::read: Unable to handle unregistered datatype 'QList<TestObj*>"11:07
dednickwhich one do i have to register? seem to have tried everything, but doesnt work...11:08
facundobatistaHolas11:08
mhr3dednick, qRegisterMetaType<QList<...>>("QList<...>") ?11:09
tsdgeosdednick: i think by just making it be a QList<QObject *> it should work11:09
tsdgeosor you do what mhr3 says11:10
dednicktsdgeos: TestObj is a qobject11:10
tsdgeosdednick: sure, just change the declaration to be QObject *11:10
mhr3then indeed what tsdgeos said :)11:10
dednicktsdgeos: it's class TestObj : public QObject11:10
tsdgeosdednick: declaration of the property11:10
dednicktsdgeos: i c11:11
tsdgeosit can be solved registering the list or the list and the pointer too11:11
tsdgeosbut if you're not too attached to the list being a QList<TestObj*> instead of a 'QList<QObject*> the second is just an easier fix11:11
mhr3tsdgeos, but when doing what i suggested it would be just an opaque type in qml, instead of an array, right?11:12
tsdgeoshmmmmm11:13
tsdgeosmaybe11:13
tsdgeosnot sure tbh11:13
=== MacSlow is now known as MacSlow|lunch
mzanettiSaviq: about silo 6. afaics we still can't just dist-upgrade to that one.11:17
mzanettiSaviq: is that a problem or something that won't be a problem when landed?11:17
dednicktsdgeos: thanks. it was namespaces that was mucking me about...11:23
tsdgeosah11:23
tsdgeosor that11:23
dednickwould be nice if qmllistproperty worked with repeaters...11:24
tsdgeosqml scoping is weird11:28
tsdgeosaka bad :D11:28
tsdgeosmost of the times11:28
dednickhm. i thought repeaters worked with lists11:33
dednicktsdgeos: ^ ?11:33
tsdgeoshmm11:33
tsdgeosshould work wiht an object list yes11:34
tsdgeosis it not working?11:34
dednicktsdgeos: not with a QList<TestObj*>11:36
dednickwhich is a qobject and registered11:36
dednickthink it only works with QQmlListReference11:36
tsdgeosdednick: can you see if it works with a QList<QObject*>11:37
tsdgeosjust in case11:37
tsdgeoshmmm11:38
tsdgeosit should work wth TetsObj* too11:38
tsdgeossee http://qt-project.org/doc/qt-5/qtquick-modelviewsdata-cppmodels.html#qobjectlist-based-model11:38
dednicktsdgeos: ah, well the property probably has to be a qlist<qobject*>11:39
tsdgeosdednick: docu says "will be accessible as named roles when a QList<DataObject*> is exposed to QML"11:40
dednicktsdgeos: although the code does "QList<QObject*> dataList;"11:41
tsdgeosdoes it?11:42
tsdgeosah11:42
tsdgeosdidn't see it11:42
Saviqmzanetti, there's no ubuntu-touch in there11:44
Saviqmzanetti, and there can't be, not until qtmir lands in distro11:44
mzanettiSaviq: ack. so that means we're good. Not having some issue that would block landing, right?11:45
Saviqmzanetti, no, just ordering11:45
tsdgeosmzanetti: ping11:47
mzanettitsdgeos: pong11:47
tsdgeosmzanetti: https://code.launchpad.net/~aacid/unity8/fixPageHeader/+merge/22718011:48
mzanettiack11:49
tsdgeoswait11:49
tsdgeosactually11:49
tsdgeosi think it wants to be pageHeader...11:49
tsdgeosmzanetti: read the commetn i just made11:50
mzanettitsdgeos: hmm... need to read more of the code11:51
tsdgeosthink that comes from the pageheader reordering you did11:51
mzanettiyeah, probably11:52
tsdgeoslunch11:53
dednicktsdgeos: yep. qlist<qobject*> working for me.11:54
=== _salem is now known as salem_
=== MacSlow|lunch is now known as MacSlow
=== dandrader_ is now known as dandrader
Saviqpaulliu, nice work on the shutdown dialog!12:50
Saviqpaulliu, we might need to make sure unity8 logs out when shutting down, so that the spinner is displayed, but later12:51
paulliuSaviq: ok. But jenkins failed a lot.12:51
Saviqpaulliu, yeah, jenkins is not happy today12:52
tsdgeosdednick: interesting12:59
tsdgeosshouldn't be that hard make it work for a qobject* derived class i'd say13:00
dednicktsdgeos: wouldn't have thought so...13:01
tsdgeosdednick: maybe file a bug, don't think it'll ever get done but once we hire that mithical guy that's going to do Qt development for us it might work as a first-easy thing to do13:01
tsdgeosmzanetti: pushed the correct fix i think13:10
mzanettitsdgeos: ack, thanks13:14
dandraderSaviq, do you have a device with latest(or recent) image flashed (non-qtcomp)?13:18
Saviqdandrader, yes, mako13:18
Saviqdandrader, can flush13:18
Saviqsilo 3 that I have there13:18
Saviqif needed13:18
dandraderSaviq, that's N4 right? when you play a video and press the volume keys, does the actual sound volume changes?13:19
Saviqbut it's landing already, so probably same same13:19
kgunndandrader: i thot i saw this same thing about 10 images ago....i think its inconsistent13:19
kgunndandrader: if you change vol in indicator panel....i bet the video app suddenly allows vol changes13:20
kgunnvirgin image13:20
Saviqkgunn, video app doesn't really do anything with volume, if that's what happens, bug's with us13:21
kgunnyep13:21
kgunni was thinking it might be down actually13:21
SaviqI think I lost my media player ?¿13:21
kgunnearly morning run of qtcomp still looking fast and fluid greyback13:22
* Saviq gotta flash, media player dead13:23
Saviqdandrader, in any case, do you see the vol icon/slider in indicator going up/down?13:23
dandraderSaviq, yes13:24
Saviqdandrader, hmm!13:24
dandraderSaviq, if that moves then it's all fine right?13:24
Saviqdandrader, it should be, yes13:24
dandraderSaviq, as in, it's not our fault :)13:24
Saviqdandrader, but if you can't hear volume changing... kinda means it bypasses pulseaudio or sth13:24
Saviqdandrader, but yeah13:25
Saviq(bypassing PA actually makes sense - stagefright - but volume should still be applied)13:25
camakoSaviq, I'm seeing a failure in unity8 AP test : Is this a known problem?13:26
camakoFAIL: unity8.shell.tests.test_lock_screen.TestLockscreen.test_can_unlock_passphrase_screen(Native Device)13:26
Saviqcamako, reproducible?13:26
Saviqcamako, you can run just that one test13:27
camakoSaviq, I only ran it once13:27
Saviqdaaamn ^W13:27
camako:-)13:27
Saviqcamako, pass the whole unity8.... part before (13:27
Saviqcamako, to phablet-test-run or however you're runningit13:28
camakolike so?: phablet-test-run -n -p unity8-autopilot unity8.shell.tests.test_lock_screen.TestLockscreen.test_can_unlock_passphrase_screen13:28
Saviqcamako, you only need -p unity8-autopilot once per flash13:28
Saviqcamako, but yeah, that's it13:28
camakoah ok13:29
camakolemme try13:29
camakoSaviq, ran 3 times and passed... Perhaps just low timeout value?13:34
Saviqcamako, yeah, it's fine13:34
camakoSaviq, ack thanks13:34
Saviqcamako, got the failure message about still?13:34
Saviqyikes, 18k people laid off...13:35
Saviqthat's a number13:35
camakoSaviq, Yeah will paste13:35
camakohttp://pastebin.ubuntu.com/7809058/13:36
camakoSaviq ^^13:36
Saviqcamako, some input got lost probably13:37
Saviqcamako, due to high load or so13:37
camakoSaviq, that's what I figured13:37
=== dandrader is now known as dandrader|afk
kgunnSaviq: is that the msft # ?13:53
Saviqkgunn, yup13:53
kgunnyeah thats big13:53
Saviqhave to be quite big to be able to lay off 18k people... without hiring most of them first just to fire them...13:55
cwaynejesus, thats like, 30 canonicals worth of people13:56
Saviqyup13:58
=== dandrader|afk is now known as dandrader
kgunncwayne: thanks for that....i love thinking about stuff that way14:10
kgunnits like 1/2 my city14:11
Saviqheh... yay for -j12 on a 2-core CPU...14:13
Saviqload 40... ;)14:13
kgunnpaulliu: thanks for landing reboot!14:27
=== alan_g is now known as alan_g|tea
kgunnSaviq: camako ...seems unity8 & mir both migrated to archive now (@mir dashboard is tad behind jenkins actual)14:37
kgunni'm gonna get a silo for unity8....that ok ?14:37
Saviqkgunn, +114:38
camakokgunn, yeah14:38
kgunncool, now let's see if i can copy/paste the right mp :P14:39
MacSlowSaviq, the odd thing was I didn't even see the others lips-icon turn red14:39
SaviqMacSlow, sounds like mumble didn't like you14:40
camakokgunn, not sure how soon mir 0.5 will appear in the archive. Does it not take a couple of hours?14:40
MacSlowSaviq, I'm always suprised how google hangout always manages to work and mumble fails every now and then14:40
MacSlowhas to be the NSA-support ;)14:40
kgunncamako: yeah, i just spoke to ogra, they're gonna spin an image immediately when its out of the proposed pocket14:42
dednicktsdgeos: can you finish reviewing https://code.launchpad.net/~nick-dedekind/unity8/indicator.dismissal/+merge/225338 please?14:43
dednickpeople want it!14:44
dednick @_14:44
tsdgeosdednick: oh i was there only complaining about merging, wasn't planning on doing any actual review :D14:44
dednickahha14:44
dednickslack14:44
kgunndamn it...i missed stand up14:44
dednickSaviq: would you mind taking a look? thostr_ wants it to land. https://code.launchpad.net/~nick-dedekind/unity8/indicator.dismissal/+merge/22533814:45
tsdgeosSaviq: can you do above dednick's review? ↑14:45
=== greyback is now known as greyback|lunch
Saviqtsdgeos, yeah will do14:45
tsdgeosi can see how me commenting as fails to merge may make others think someone else is already reviewing and they go to another MR14:45
tsdgeosSaviq: do you think i should keep doing all those merge checking comments?14:46
Saviqtsdgeos, if you want, yeah, but maybe with no vote?14:46
Saviqtsdgeos, unless you Abstain them later ;)14:46
tsdgeosor that14:46
tsdgeostx14:46
tsdgeoswill do14:46
* tsdgeos goes back to the fake unity scopes plugin coding14:46
mhr3tsdgeos, Saviq, will you be adding the dconf schema to u8 for the favourites?14:51
Saviqmhr3, no, you will14:51
tsdgeosmhr3: which favourites?14:51
tsdgeosah14:51
Saviqmhr3, we don't even know where they're stored, you know14:51
Saviqmhr3, because you'll be abstracting it for us nicely :)14:52
mhr3Saviq, i don't want dconf schema for a qml plugin14:52
mhr3weird14:52
Saviqmhr3, it's not for the qml plugin, it's for the scopes system14:54
Saviqmhr3, otherwise what's the purpose of the Scopes model14:54
mhr3Saviq, why should scopes system care that some scopes are special for the user?14:55
Saviqmhr3, because he marked them as favourite14:56
mhr3for us all scopes are equal14:56
mhr3Saviq, then you own the schema :)14:56
Saviqmhr3, I know you want to stop using the scopes scope in the overview14:57
Saviqmhr3, but as long as Scopes is a model and the Favourites in overview are a ResultsModel, I really don't think I want to tell you what the list of favourites is14:58
mhr3Saviq, scopes have visibility attribute, scopes scope isn't visible actually14:58
mhr3Saviq, i'd like if Scopes had scopeIds prop which you'd just bind to the dconf key14:59
mhr3would be awesome14:59
Saviqmhr3, would be useless14:59
Saviqmhr3, why does Scopes need to know it at all14:59
Saviqmhr3, I can just request the favourites one by one15:00
Saviqmhr3, in that case15:00
mhr3cause you want to tell it which scopes to present in the model15:00
SaviqI don't want the model from it15:00
Saviqif I have the list myself15:00
Saviqit's backwards this way15:00
Saviqmhr3, IMO a cross-scope notion of favourites needs to happen anyway15:00
Saviqmhr3, it's only a question of time when we add the stars in apps, music albums, whatnot15:01
Saviqmhr3, and then reorder them15:01
Saviqmhr3, if I have to maintain all that myself in the shell, what use are the scopes then15:01
mhr3and what do scopes have to do with apps and music albums and what not15:02
Saviqthey're just content, same thing15:02
Saviqscopes are results in the scopes scope15:02
Saviqaand there they go15:02
mhr3isn't that a reason to keep it somewhere where it's not domain specific?15:02
Saviqmhr3, but how is it not domain specific, store it in Google Drive for all I care15:03
Saviqsay there's another shell built with the scopes backend15:04
Saviqwhy would that shell reimplement the favourites concept?15:04
mhr3Saviq, i can ask the same thing from the other end15:05
mhr3Saviq, if there are multiple shell implementation why should all have the same favourite scopes15:06
Saviqmhr3, because the user marked the *scope* favourite, not that scope in that frontend15:06
mhr3i disagree, without the frontend the scope is just a bunch of meaningless data15:07
Saviqmhr3, sounds like a next-week conversation15:09
Saviqmhr3, but really, if you don't want to expose "collections" (be it favourites or all of them) of scopes for us, we might as well get rid of the Scopes model, there's no point to have it15:09
Saviqand the scopes scope goes away15:10
Saviqin surface mode at least15:10
Saviqbut I would completely not understand why15:10
SaviqI told you already I would gladly just take the scopes scope's Favourites category to drive the dash itself15:10
mhr3Saviq, imo the Scopes object is there just to allow you to get proxies to the individual scopes15:11
Saviqmhr3, well it's not, right now, it's providing me with a model of favourite ones15:11
mhr3right15:11
Saviqmhr3, if it doesn't need to do that (because I have the list of favourites from somewhere else already)15:11
Saviqit can just stop being a model15:11
Saviqno point in me passing you a list of favourites for you to give it back15:12
mhr3sounds fine to me, it's not necessary for it to be a model15:12
mhr3then again, we have schedule to fit into15:13
Saviqmhr3, yeah, nothing like that's gonna happen before RTM15:13
mhr3and making it not-a-model sounds expensive15:13
Saviqmhr3, it wouldn't be that expensive actually, but let's come up with a plan before we start doing anything15:16
Saviqmhr3, can we agree to, at least for now, do it where it happens currently (where UNITY_SCOPES is interpreted)?15:17
Saviqmhr3, or do you not want to deal with (un)favouriting?15:18
Saviqmhr3, where would the overview get the list from then?15:18
=== gatox is now known as gatox_lunch
mhr3Saviq, the overview is already misusing it, it's doing getScope("scopesOverview") ffs15:23
Saviqmhr3, not thatlist15:24
Saviqmhr3, the actual list of favourites15:24
mhr3it's just an array of ids where each is expected to be a scope15:25
Saviqmhr3, so what you're saying is that there won't be a favourites category that the scopes scope supplies15:25
Saviqmhr3, so all you're going towards is dropping the scopes scope altogether, and making unity8 deal with all this internally15:26
mhr3i'm saying it's the exact same thing that currently the Scopes model exposes15:26
mhr3so it's just bound together15:26
Saviqmhr3, except the overview expects a ResultsModel *from* scopesOverview scope15:27
Saviqmhr3, that supplies the list as scopes15:28
Saviqs/scopes/results/15:28
Saviqfor both favourites and All15:28
Saviqin two surfacing categories from the scopes/overview scope15:28
Saviqmhr3, really, to us data for the overview is just as for any other scope, it's categories, with results, with a card template15:30
Saviqmhr3, the only thing different is how we deal with activation15:30
Saviqoh yay, and now my mako is dying15:30
mhr3Saviq, to me the overview thing isn't really a scope15:30
mhr3yet i understand why you want it to have the same data model15:31
Saviqmhr3, why is that?15:31
Saviqmhr3, it's just a scope of scopes, how is it different than the scopes scope we currently have15:31
Saviqit just looks differently, data-wise it's exactly the same15:32
mhr3because it's hard linked to all the other scopes15:32
mhr3it's on top of everything15:32
Saviqit's just a master aggregator scope15:32
Saviqso what15:32
mhr3it's a "master" scope for everything15:32
mhr3root of the scope tree if you will15:32
Saviqeverything else in that tree is a scope, why can't that be a scope, too15:33
Saviqit's even designed to behave like a scope15:35
Saviqwith a few exceptions, exceptions that we plan to have all over the place for "our" scopes anywy15:35
mhr3Saviq, even design themselves consider apps and scopes scope super-special15:37
mhr3and they indeed are, there's a reason those two shouldn't be unfavouritable15:37
Saviqmhr3, but they're only super-special in the sense that they can do more15:38
mhr3at some point scopes will be just tiny qml programs :)15:38
Saviqmhr3, not that they are not scopes15:38
Saviqtiny, right15:39
mhr3Saviq, you could say the same about all our apps too :)15:39
mhr3ok, sudoku as a scope... maybe not15:39
Saviqmhr3, while are we building this whole scope system at all then?15:39
Saviqs/while/why?15:39
Saviqaargh15:40
Saviqbad day15:40
mhr3cause apps are slow to launch? :)15:40
Saviqmhr3, and scopes are not15:40
mhr3and they're so 2010 :)15:40
Saviqok... no more run_on_device.sh :|15:56
Saviqnot enough space in the image...15:56
josharensonAnyone know why I'm getting this traceback? http://pastebin.ubuntu.com/7809679/   What is supposed to be setting this bogus lib path, and why do other tests work? (running AP test on tablet manually via autopilot run <package.test.method>)16:00
Saviqjosharenson, the tests themselves set those paths16:02
josharensonI've literally copied an existing test, and when I call self.launch_unity() it crashes16:03
josharensonsaviq, but the other tests work fine16:03
josharensonill set some break points and see, just wondering if this was an easy fix16:03
=== gatox_lunch is now known as gatox
Saviqjosharenson, you doing this on device?16:04
josharensonyes16:04
Saviqjosharenson, ok, do you have unity8-fake-env installed?16:04
Saviqjosharenson, that path should be there16:04
josharensonsaviq, haven't manually installed it...16:04
Saviqjosharenson, except LD_LIBRARY_PATH should not point at it never16:04
mterrydednick, thanks for review of dialer-above!16:05
dednickmterry: no worries16:05
Saviqjosharenson, maybe the test you copied was decorated to not use the lightdm mocks?16:05
josharensonsaviq, I removed all the decorators16:06
Saviqjosharenson, yeah, that's your problem most probably16:06
josharensonsaviq, ah, thought I was simplifying things...16:06
Saviqjosharenson, the decorators deal with the lightdm mocks per-test16:07
josharensonack16:07
Saviqjosharenson, granted, no decorators should work, too, so there's something wrong16:07
josharensonsaviq, it seems to be getting farther (before crashing) with the fake env package installed16:08
Saviqjosharenson, you should probably just install unity8-autopilot16:08
josharensonok16:08
Saviqjosharenson, and then just use PYTHONPATH to point at your own suite16:08
josharensonsaviq, and run from host I assume16:09
Saviqjosharenson, no, not necessarily16:10
Saviqjosharenson, well, depends on what you mean "from the host"16:10
josharensonsaviq, phablet test runner16:10
Saviqjosharenson, doesn't really matter16:10
josharensonok16:10
Saviqjosharenson, or well, it does, won't let you set PYTHONPATH16:11
josharensonah true16:11
Saviqkgunn, I'll add https://code.launchpad.net/~nick-dedekind/unity8/indicator.dismissal/+merge/225338 to your silo, ok?16:31
Saviqthostr_, will be happy16:31
thostr_indeed16:31
thostr_like xmas and easter the same day16:32
Saviqkgunn, oh btw, we got two test plans... https://wiki.ubuntu.com/Process/Merges/TestPlan/unity8 https://wiki.ubuntu.com/Process/Merges/TestPlans/Unity816:32
=== greyback|lunch is now known as greyback
kgunnSaviq: ack16:35
Saviqkgunn, kicked a rebuild, should be good in ~36 mins16:35
Saviqman it takes too long :/16:36
Saviqadd to that building the image and autopkgtests that ~will happen in CI airline...16:36
kgunnSaviq: how come a rebuild ?16:36
Saviqkgunn, added the MP?16:37
kgunnoh sorry nvmd16:37
kgunnmisssed the scrollback16:37
Saviqkgunn, ah... not enough ACKs...16:37
kgunn;)16:37
Saviqkgunn, I'll be available to run the test plan if you need me16:38
kgunnthanks16:38
dandradermzanetti, greyback, on my device if I press the power key I see it getting repeated KeyPress events for ever16:43
dandraderwhich effectively makes the device screen blank until you "restart lightdm"16:44
greybackdandrader: news to me.16:44
greybacktry restarting?16:44
dandradergreyback, you mean rebooting?16:44
greybackyes16:45
mzanettigreyback: dandrader: seen that too16:46
mzanettionly if you manually restart lightdm16:46
mzanettidoes not happen when booting on its own or with run_on_device16:46
dandraderahhahaha, unbelievable16:46
dandradermzanetti, I'm fiddling with focus as well in the key handling work16:48
dandradermzanetti, so better not having us both working on it.16:49
mzanettidandrader: I just fixed that one bug report we had16:49
mzanettidandrader: not touching it any more16:49
kgunnSaviq: for landing qtmir have you been talking to any core devs ?16:54
kgunni was asking slangesek to review for packaging...16:54
kgunncurious if we were doubling down16:55
=== alan_g is now known as alan_g|EOD
Saviqkgunn, slangasek asked me about it around last weekend17:00
Saviqkgunn, and said he just needs a GO signal when we're done with code reviews17:01
kgunnSaviq: awesome!...just sync'd w the guys...sounds like Mon afternoon euro time will be it...17:02
kgunntry to land it on Tues17:02
Saviqk17:02
kgunnolli_: ^17:02
kgunnqtcomp17:02
dednick\o/17:06
dandradermzanetti, greyback: pushed key handling + focus refactoring17:14
dandraderhopefully I didn't break anything17:14
greybackdandrader: ok, will test. Did you try on the desktop?17:14
dandradergreyback, no. never did try it on the desktop17:14
greybackdandrader: well now's a great time to try :) I recommend using your non-primary machine17:15
dandraderoh, the poor g+ laptop17:15
dandraderafter lunch17:16
=== dandrader is now known as dandrader|lunch
mterrySaviq, did you still have concerns about the locking-hash branch?17:47
kgunnSaviq: hmmm, can you try silo 12 ? ....i updated and made sure i rm'd welcome-wiz-had-run....17:51
kgunnand it never booted17:51
kgunnso just rebooted, but i just get endless spinner17:51
kgunnhoping its me17:52
kgunnah shit....it is me17:52
kgunnnvmd17:52
kgunnidiot17:52
kgunnforgot to update mir17:52
Saviqmterry, right, sorry, should we add a --pam option, though? so that even with ./run.sh we can force usage of PAM?18:02
Saviqmterry, your call, doesn't even have to be in the same MP18:02
Saviqmterry, ah and for later, is there any way we could test the PAM interactions?18:04
mterrySaviq, that would be nice, yeah.  I usually just test on the phone but yeah18:04
Saviqmterry, re: testing PAM interaction?18:12
mterrySaviq, sorry got distracted18:12
mterrySaviq, you mean manually or via tests?18:12
Saviqmterry, tests18:13
mterrySaviq, well that's certainly possible as long as our tests can futz with setting passwords and the like18:14
Saviqmterry, when you have a moment, review fail on my part, didn't update the mock: https://code.launchpad.net/~saviq/unity8/fix-test-notifications/+merge/22723918:14
mterrySaviq, like, suitable for running AP tests on the device, but not so much on your own machine18:14
Saviqmterry, well, yeah, so food for thought, then18:15
Saviqmterry, source of the above issue: https://code.launchpad.net/~larsu/qmenumodel/add-nameowner-property/+merge/22642818:16
mterrySaviq, we had same problem with split greeter18:16
Saviqmterry, ACK18:17
Saviqmterry, thanks to you I learned something about QFutures ;)18:18
mterrySaviq, we could LD_PRELOAD low-level methods that PAM needs.  So fake which users are available.  That way we'd at least hit the PAM stack18:18
mterrySaviq, yeah that was a journey for me too :)18:18
Saviqvery wow ;)18:19
mterrySaviq, looking at that branch btw, just have to at least build to answer the checklist  :)18:21
Saviqdoes not compute18:21
Saviqmterry, so what do we still need to happen for password support to land? settings?18:23
mterrySaviq, ugh18:23
mterrySaviq, so settings needs seb signoff, dialer-above branch needs security signoff.  Then we transition into the Foundations world, where there's quite a bit needed (mostly written, but working on tiny details about some password tools hardcoding /etc in some places)18:24
Saviqmterry, right, so we're still going for the out-of-/etc passwd?18:24
mterrySaviq, (because the current plan is to have password databases in /var/lib/extrausers)18:25
mterrySaviq, yeah, in-etc just isn't feasible18:25
Saviqk18:25
kgunnlunch & running errand while unity8 ap test runs18:43
kgunnall other manual tests look good for silo1218:43
=== dandrader|lunch is now known as dandrader
dandradermzanetti, greyback, gonna merge trunk in unity8/mirCompositor19:17
greybackdandrader: ok19:17
dandraderwow, still online!19:18
dandradergreyback, btw, found any new issues?19:18
=== salem_ is now known as _salem
=== Guest88980 is now known as cyphase
=== _salem is now known as salem_
=== salem_ is now known as _salem

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