/srv/irclogs.ubuntu.com/2014/06/26/#ubuntu-touch.txt

guyJust got Ubuntu Touch.01:24
=== guy is now known as Guest42159
=== chriadam|away is now known as chriadam
CyborgCygnusSo only just discovered the ubuntu tv section on the main website, forgive me if I'm annoying anyone but I seem to find it hard to find more thorough info on ubuntu os's for all devices. Ubuntu TV will that be installable on tv's like smart tv's?02:05
CyborgCygnusIgnore that^, wrong channel sorry.02:05
=== Guest30773 is now known as IanWizard-Cloud
=== japaric4 is now known as japaric
=== fginther is now known as fginther|away
=== chihchun_afk is now known as chihchun
=== dpm-afk is now known as dpm
lotuspsychjeogra_: morning06:00
lotuspsychjedid you see meizu ubuntu touch showed off in china?06:00
dholbachgood morning06:07
=== RAOF_ is now known as RAOF
=== oSoMoN_ is now known as oSoMoN
Kaleoogra_, popey, easy way to prevent the screen from locking on the phone?07:40
Kaleo+do you know an07:40
didrocksKaleo: hey! it's quite late for you, isn't it?07:45
Kaleodidrocks, I'm in Paris until this weekend :)07:48
didrocksah, I was afraid for unsane hours for you :)07:48
Kaleodidrocks, nah :) I gave up on that07:48
pittihello all07:53
pittiI have a question about click apps with compiled code07:53
pittie. g. ubuntu-filemanager-app has07:53
pitti    "architecture": "armhf",07:53
pittiin its manifest07:53
pittithat'll build com.ubuntu.filemanager_0.3.212_armhf.click on *any* platform, even if I build it on amd6407:54
pittiI can change that to ["armhf", "i386", "amd64"], and then it'll build com.ubuntu.filemanager_0.3.212_multi.click which I can install on x86 (tests run fine)07:54
pittibut that's still wrong, as in both cases the .click contains amd64 code (regardless of whether it's tagged _armhf or _multi)07:55
pittibut if I don't specify "architecture" at all, it just says _all.click, which is also wrong07:55
pittiis there an equivalent of "any" which will behave like dpkg, and add the arch you built it on?07:55
pitticjwatson: ^07:55
pittihttps://click.readthedocs.org/en/latest/file-format.html#control-area doesn't talk about the details of that07:55
ogra_Kaleo, unity8-autopilot ships a script ... /usr/share/unity8/unlock-device07:56
Kaleoogra_, nice!07:56
Kaleoso hmm07:56
ogra_Kaleo, http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/touch/view/head:/utils/host/reboot-and-unlock.sh07:56
KaleoI can just call it before running autopilot07:56
lotuspsychjeevery boot my brightness resets on halfway again on nexus7, you guys got that issue on other devices aswell?07:56
lotuspsychjewould be nice if it sticks 100% brightness07:57
mandelpopey, monring! yes, but we have to land the cmake job that generates them08:01
mandelpopey, Elleo_ created a branch, so it will be landing asap08:01
lotuspsychjenice fix on the 'save image' on webbrowser, ive been able to set matrix wallpaper on startscreen08:14
lotuspsychjejust the main device background stays white default08:14
lotuspsychjei hope you guys get your hands on the meizu image soon  to check it out :p08:15
pittididrocks, sil2100: I'm currently running the autopilot tests of various click apps with autopkgtest in a container08:22
didrockspitti: oh waow, nice! :)08:22
pittididrocks, sil2100: for the most part that works fine, but I stumbled over a failure in calendar-app, which fails with ImportError: No module named 'dateutil'08:22
pittididrocks, sil2100: indeed python[3]-dateutil isn't installed on the phone (or my container)08:22
pittididrocks, sil2100: so hence my question: when we run the autopilot tests, do we actually install the .click, or do we build a .deb and isntall that?08:23
pittithe calendar-app-autopilot package indeed does have the python-dateutil dep08:23
pitti(and not python3, argh!)08:23
didrockspitti: we install the .click08:23
pittididrocks: oh, but supposedly we also install the <app>-autopilot .deb?08:24
didrocksif it's a click apps installed on the phone, the whole debs world is ignored08:24
didrocksno08:24
didrockssergio is using a script to branch latest trunk08:24
didrocksand take the autopilot from it08:24
pittididrocks: then I don't understand how calendar-app's tests can succeed08:24
didrocksthat sounds weird as well08:24
didrockscalendar-app is for sure a click app on a vanilla system ?08:25
didrocksogra_: do you have your list handy? ^08:25
didrockspitti: sometimes, we had click apps, but they were installed as debs on the image08:25
didrocks(during the transition, this was all puzzling)08:25
pittididrocks: hm, not sure; I just picked some random things from https://launchpad.net/ubuntu-phone-coreapps and locally built .clicks from it08:25
ogra_http://people.canonical.com/~ubuntu-archive/click_packages/click_list08:25
pittiright, there's calendar08:25
pittiroot@ubuntu-phablet:~# dpkg -l *calend*08:25
pittidpkg-query: no packages found matching *calend*08:25
popeyphablet-click-test-setup --click com.ubuntu.calendar08:25
popey^^ thats what sets up autopilot08:26
didrockspitti: the thing is that autopilot and other things are installed first08:26
didrocksso maybe something is pulling it in first ?08:26
popeynot installed, copied to ~/home/phablet/autopilot08:26
ogra_https://wiki.ubuntu.com/Touch/Testing08:26
pittiright, autopkgtest is doing something similar now (if the filesystem is r/o, it installs the test depends into /tmp/... and sets all the $*PATH envs08:26
popeyrun that command and you'll see (for example) python_dateutil-2.0.egg-info in /home/phablet/autopilot/08:27
popeyi.e. we don't use debs08:27
pittipython_packages = (08:28
pitti    'python-mock', 'python-dateutil', 'python3-dateutil'08:28
pitti)08:28
pittiooh08:28
pittiyay hardcoding08:28
pittiso this should definitively become a test dependency in calendar-app's manifest then08:28
pittibut at least that explains how it can currently work in production :)08:28
pittipopey: thanks for pointing out08:28
popeynp08:28
didrocksso, installed from pypi08:30
pittididrocks: right, but the point is, it's currently an externally hardcoded test dependency; that's what I wanted to know08:31
pitticf. balloons's u-phone@ discussion, we want to move these test depends into the click's manifest08:31
pittisimilar to dep-808:31
didrocksyeah, and so, it's duplicated between the .deb and the python_packages list08:31
didrocksyeah, one place to rule them all, or DRY ;)08:32
pittididrocks: well, the -autopilot debs will go away (hopefully), right?08:32
didrocksnot only the -autopilot, but the whole deb as well08:32
didrocksthe day we can have .click built from source08:32
pittididrocks: how do you mean? I currently build them directly from the lp checkout08:38
pittiah well, build deps08:38
pittithey seem to be pretty much constant for arch:all clicks, but of course not for the compiled ones08:38
didrocksyeah08:38
didrocksI mean for compiled ones08:38
pitticlick building indeed feels fairly incomplete still08:39
pittiRuntimeError: Application Launch Failed: Application crashed.08:41
pittiw00t08:41
pittiI'll leave that to running on an actual phone then :)08:41
pittibut 3/4 success is quite nice for running locally with a very fast turnaround08:42
JamesTaitGood morning all; happy Chocolate Pudding Day! :-D09:16
svetaaha.09:20
=== sletta_ is now known as sletta
=== ycheng is now known as ycheng-afk
AskUbuntuUbuntu for Android on HP Slate 21? | http://askubuntu.com/q/48836809:58
lotuspsychje!devices | AskUbuntu10:05
ubot5AskUbuntu: You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices10:05
cjwatsonpitti: I suspect "click build" is deficient here and you just have to set the manifest right yourself.  bug would be welcome10:14
pitticjwatson: sure, filing one10:14
pitticjwatson: thanks; nice timing, was just about to re-ask on the ML :)10:14
pitticjwatson: done, bug 133459710:23
ubot5bug 1334597 in click (Ubuntu) "Please support "any" architecture when building instead of hardcoding an arch" [Undecided,New] https://launchpad.net/bugs/133459710:23
cjwatsonta10:26
=== ondra- is now known as ondra
=== chriadam is now known as chriadam|away
=== vrruiz_ is now known as rvr
sil2100oSoMoN: hello! We also noticed that in recent images the number of failing tests for webbrowser-app increased11:19
=== bschaefer_ is now known as bschaefer
Jazzcat14736251Hi folks :)11:25
Jazzcat14736251I was wondering how best to find out information about getting ssh working (if this is possible)11:26
Jazzcat14736251I've copied my id_rsa.pub to /home/phablet/.ssh/authorized_keys and adb shell tells me ssh service is running11:29
popeyJazzcat14736251: does the right user own .ssh and authorized_keys, and have the right permissions?11:30
Jazzcat14736251I believe the permissions are correct, perhaps not the user11:31
Jazzcat14736251-rw-------  1 root    root     405 Jun 19 00:35 authorized_keys11:32
=== chihchun is now known as chihchun_afk
popeynope, thats nto right11:33
popeyshould be owned by the user, phablet11:33
oSoMoNsil2100, let me take a look11:33
sil2100oSoMoN: brendand tried looking into what's wrong but didn't see anything obvious11:33
Jazzcat14736251cheers, will fix.11:33
sil2100But we're having a lot of failures recently11:33
oSoMoNsil2100, looks like a mix a different types of failures, some of them due to the toolbar not being open correctly (using the UITK emulator), and others where the OSK doesn’t show up11:36
oSoMoNI don’t think any of those failures are actually on the browser side11:36
Jazzcat14736251popey, thanks again, I chowned the file and folder, now it is working ^_^11:38
popeysweet11:38
=== alan_g is now known as alan_g|lunch
=== MacSlow is now known as MacSlow|lunch
=== fginther|away is now known as fginther
=== MacSlow|lunch is now known as MacSlow
=== alan_g|lunch is now known as alan_g
jdstranddoes the weather app work for anyone else on latest promoted? seems to be broken for a while now12:48
jdstrandI start it and am told it couldn't load weather data. then, I can't do anything12:49
jdstrandbottom toolbar is empty, nothing to click, etc12:49
pmcgowanjdstrand, I have 81 and it works here, and there are no app updates pending12:53
jdstrandI think I'll delete the data dirs12:53
jdstrandpmcgowan: thanks12:53
popeyjdstrand: works fine here.12:58
jdstrandI got it working12:59
jdstrandbut I have a feeling I am going to have a problem again12:59
jdstrandin the distant past I would set it to a nearby city since the city I live in wasn't listed12:59
jdstrandmore recently, the city I live in starting showing up13:00
jdstrandso I chose it13:00
jdstrandtoday when resetting it up, I looked for my city, but was told no data sources13:00
jdstrandso I used the nearby city13:00
jdstrandthen for giggle I tried adding my city again13:00
jdstrandand it worked13:00
jdstrandmy thinking is there may be an intermittent problem with my city13:01
popeyjdstrand: move13:13
popeyfix-committed13:13
jdstrandheh13:13
=== chihchun_afk is now known as chihchun
cwaynemardy: ping13:25
mardycwayne: hi!13:25
cwaynemardy: hiya!  so I saw account plugin click hooks are in the image now?!13:26
cwaynebut I also saw that my fitbit plugin is borked :( (even the deb-packaged one in universe)13:26
mardycwayne: yes, click hooks are there, but the thing is not really ready yet13:28
cwaynemardy: so i changed log level to 2 in /etc/signond.conf, and I still can't quite see what's wrong13:28
cwayneah13:28
cwaynewell, I'll still need to solve my deb-packaged one being broken anyhow :)13:28
mardycwayne: so, what's the symptom?13:29
cwaynemardy: so it launches the site to authenticate, and when you fill in your info then click accept, everything dies and the account isn't added13:29
cwaynei tried using a standard qml-plugin (like twitter's) instead of the one I made, but same issue13:30
mardycwayne: the logs should help, I think13:31
=== chihchun is now known as chihchun_afk
oSoMoNdidrocks, hey, I have a webbrowser-app MR with minor packaging changes, I’ll need a core-dev to ack them, would you mind doing that when you have a spare moment?13:36
oSoMoNthe MR is: https://code.launchpad.net/~osomon/webbrowser-app/contextual-selection/+merge/22376013:36
didrocksoSoMoN: sure, looking13:36
cwaynemardy: let me get some logs, 1 sec13:38
didrocksoSoMoN: done13:41
oSoMoNdidrocks, thanks! I’ll remove the Replaces stanza13:41
didrocksoSoMoN: perfect, consider the +1 then (but I can approve if you need that in your process :))13:42
oSoMoNdidrocks, it’s probably better for the sake of the process, although I don’t think it’s mandatory13:42
didrocksoSoMoN: just tell me once pushed then :)13:43
oSoMoNdidrocks, note that I used Replaces because the package installs the same files, but in a different location, but IIUC Replaces is only when there is a conflict in fully-qualified filenames, right?13:43
cwaynemardy: just syslog? or would anything else be useful?13:44
didrocksoSoMoN: yeah, it's only the full path that is taken into account13:44
mardycwayne: online-accounts-ui logs might help13:44
didrocksthere is no md5sum magic or whatever :)13:44
oSoMoNdidrocks, pushed13:44
didrocksoSoMoN: approved13:45
oSoMoNthanks!13:45
mardyexport OAU_LOGGING_LEVEL=213:46
mardyexport OAU_DAEMON_TIMEOUT=900013:46
mardyonline-accounts-ui13:46
mardyafter killing any running instance13:46
mardycwayne: ^13:46
didrocksyw ;)13:48
cwaynemardy: segfaulted,i tried ubuntu-app-launch online-accounts-ui and nothing showed up13:48
cwaynehuh13:48
cwaynenow it worked13:48
mardycwayne: oops! invoke it as "online-accounts-ui --desktop_file_hint=/usr/share/applications/online-accounts-ui.desktop"13:49
cwaynemardy: hah, but now that i got all the logging stuff setup, adding an account works13:50
mardycwayne: see? :-)13:50
cwaynelol13:50
cwayneseems weird13:50
cwaynemust be an intermittent issue13:50
cwaynemardy: segfaults for me13:52
tomhengrootstock-touch-install run for a while, it report device not found, what that mean ?13:54
cwaynemardy: so i found that if i hit enter after putting in my password, it works. but if i press the button on the page, it crashes.13:54
mardycwayne: you mean a button which is inside the webpage?13:55
cwaynemardy: yes13:57
mardycwayne: might be an oxide bug, but it would be nice to have a stacktrace to confirm13:57
cwaynemardy: I can't seem to launch online-accounts-ui from cli without it segfaulting :/13:58
mardycwayne: with the desktop_file_hint?13:58
mardycwayne: and are you killing any previously running instance?13:58
cwayneyes and yes13:58
cwaynelet me reboot13:58
cwaynemardy: still segfaults14:02
mardycwayne: backtrace?14:05
cwaynemardy: none, just does this: http://paste.ubuntu.com/7706033/14:06
mardycwayne: but does it starts if you launch it from the system settings UI?14:07
cwaynemardy: yep14:08
mardycwayne: and what's the command line used in that case?14:08
cwayneno idea tbh14:08
mardycwayne: maybe some environment variable is missing in the shell (though I wonder, because it's working here)14:08
mardycwayne: ps -fe | grep online14:08
cwaynemardy: nothin, http://paste.ubuntu.com/7706049/14:09
mardycwayne: that's just impossible, if online accounts is open there *must* be an online-accounts-ui process running14:11
mardycwayne: ah, you are logged in as root!14:11
mardy(though you still should see the running processes...)14:11
cwaynemardy: just where I was doing the ps -fe14:12
cwaynei did a sudo -iu phablet in another shell and am now running /usr/bin/online-accounts-ui --desktop_file_hint=/usr/share/applications/online-accounts-ui.desktop14:12
cwayneand it's not showing up..14:12
=== japaric3 is now known as japaric
cwaynemardy: hm, what about this: application-legacy-online-accounts-ui-1403791371914694.log:** (process:3741): WARNING **: Desktop file 'online-accounts-ui' is set to not display, not copying14:13
mardycwayne: so, you need to run that, and then open online accounts from the ui; then, you'll start see messages in the console14:14
cwayneoh i didnt know i needed to do it from UI too14:15
cwaynemardy: there we go : http://paste.ubuntu.com/7706101/14:16
cwaynelooks like it segfaults, gah14:16
mardycwayne: can you please report a but and attach those logs?14:18
mardycwayne: do you think you can run it under gdb?14:18
cwaynei can try14:18
mardygdb /usr/bin/online-accounts-ui and then "run --desktop_file_hint=/usr/share/applications/online-accounts-ui.desktop"14:18
cwaynewhat do i log it against, u-s-s-o-a?14:18
mardycwayne: yes14:18
stgraberogra_, barry: heads up, I've updated image-signing.tar.xz on system-image just now, let me know if you see any problem happening as a result of that.14:19
ogra_stgraber, ok14:19
barrystgraber: ok14:19
stgraberwe're not yet using the new signing key though, just shipping the new keyring as the old one would have expired on Saturday14:19
barryhopefully, all the key update logic is as well tested as i think it is ;)14:19
stgraberthe old key expires in a year from now so I'll have time to do some more checks before I switch over to it14:19
stgraberbarry: I sure hope so because we don't exactly have much breathing room if it doesn't ... :)14:20
barrystgraber: is there a way we can test this now?14:20
stgraberbarry: I think my phone is at least one image behind, so I'm testing it now14:20
barrystgraber: cool.  i'll try to reflash my device with an older image and do an upgrade14:21
stgraberbarry: so it only downloaded the blacklist, IIRC we're caching the tarballs unless they fail to validate right?14:21
stgraberbarry: so I'd have to set my phone's clock to say a week from now and retry14:22
barrystgraber: correct14:22
stgraberTue Aug 26 14:22:54 UTC 201414:22
stgraberlet's try that again now :)14:23
cwaynemardy: alright, so i got the crash after running from gdb, now what? paste the output?14:23
stgraberbarry: we've got a problem...14:23
stgraberbarry: http://paste.ubuntu.com/7706148/14:24
stgraberbarry: a re-run appears to fix it though, so maybe not entirely dramatic14:24
stgraberbarry: the second run, I get (as expected): http://paste.ubuntu.com/7706151/14:25
mardycwayne: type "bt"14:25
cwaynemardy: http://paste.ubuntu.com/7706163/14:26
barrystgraber: interesting.  i'll have to do some debugging to understand what's going on14:27
barrystgraber: i think it shouldn't error out the first time14:28
mardycwayne: not useful :-(14:28
mardycwayne: but please report a bug anyway14:28
barrystgraber: the trick will be devising a test for this condition14:28
stgraberbarry: well, it shouldn't error out ever, but yeah, I think it got confused because I did a first run with -n, then did the time change and then did it again14:28
cwaynemardy: done: https://bugs.launchpad.net/ubuntu-system-settings-online-accounts/+bug/133468814:28
ubot5Ubuntu bug 1334688 in Online Accounts setup for Ubuntu Touch "Clicking "allow" button from website crashes ussoa, but hitting 'Enter' on keyboard works and account is added" [Undecided,New]14:28
cwayneim sad it's not a bug in my account-plugin, would probably have been easier to fix :(14:29
barrystgraber: could you file a bug please?  at least then we'll have a good record of what happened14:29
stgraberbarry: sure14:29
=== alan_g is now known as alan_g|tea
=== alan_g|tea is now known as alan_g
alf__Elleo: Hi! Are the QML files that ubuntu keyboard uses embedded in the plugin .so file?14:57
Elleoalf__: no, the main files are installed to /usr/share/maliit/plugins/com/ubuntu/15:00
alf__Elleo: great, thanks15:01
Elleono problem15:01
Elleoalf__: I don't know if the keyboard position mappings *should* have worked with Mir 0.2, since it also got resized to 1222 in image #8815:02
Elleoalf__: I mean it *did* work, but I don't know if that was actually depending on something wrong in Mir 0.2 which was fixed in 0.3 for example15:02
alf__Elleo: That's what I am trying to figure out :)15:03
alf__Elleo: Is the related calculating happening in Keyboard.qml reportKeyboardVisibleRect ?15:03
alf__s/calculating/calculation/15:04
=== vying is now known as Guest601
Elleoalf__: yeah, it's basically assuming that it's being positioned on a fullscreen item so it can take its x/y values as being global screen cordinates15:04
Elleoalf__: which seems nasty, but as far as I could see QML doesn't provide any proper mechanism for translating between window cooridnates and screen coordinates15:05
cwayneElleo: hey, any update on that pinyin bug?15:05
Elleocwayne: not sure; we passed that along to the SDK team, will just find the bug report tracking it15:06
Elleocwayne: https://bugs.launchpad.net/bugs/1324955 <-- looks like they're targetting it for june, so presumably it'll be looked at soon; might want to check with zsombi15:08
ubot5Ubuntu bug 1324955 in Ubuntu UI Toolkit "Automatically commit text in AbstractButton" [High,New]15:08
cwaynezsombi: ping, re ^15:09
zsombicwayne: I'm affraid we cannot have that in June :( but on the next MS15:09
zsombikalikiana: dude, could you actually take care of this bug? ^^15:10
cwaynezsombi: it's a pretty high priority, critical for chinese images15:12
zsombicwayne: I understand that :)15:12
zsombicwayne: well, actually you coudl also contribute a fix for the text inputs :)15:13
cwaynehah15:13
cwayneso basically http://patcheswelco.me/ ?15:14
zsombicwayne: we discussed this with Elleo that the fix should be in teh text inputs (more precisely in InputHandler) and it seems to be a pretty small fix15:14
cwaynezsombi: so should the bug be also affecting network-manager then?15:15
zsombicwayne: the problem is that I have ~3 MRs on teh text inputs pending, and I cannot take it anymore till eo June15:15
cwayneor whatever draws that snap decision15:15
zsombicwayne: why would it?15:15
cwayneis that not where that text input is?15:15
clevasShould ubuntu work on a knock off tablet from china?15:15
cwayneer network-indicator sorry15:15
zsombicwayne: this one you mean? https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/132495515:15
ubot5Ubuntu bug 1324955 in Ubuntu UI Toolkit "Automatically commit text in AbstractButton" [High,New]15:15
cwayneyeah15:16
zsombicwayne: I don't see the connection between those...15:16
cwaynethat was why the bug was logged in the first place. inputting a wifi password with pinyin doesn't work because it's not auto-committed15:20
cwayneand I thought you were saying it needed to be fixed where the text input is15:20
kalikianazsombi: checking15:54
zsombikalikiana: it's a pretty straight FW stuff, in InputHandler you need to call Qt.inputMethod's commit when teh text input looses focus15:56
kalikianaI could be wrong but didn't somebody propose a patch for that some time ago?15:59
kalikianait feels familiar15:59
kalikianahrm I don't see a branch like that so I guess not16:01
alf__Elleo: Does maliit_geometry.visibleRect need to be in screen coordinates?16:04
alf__Elleo: i.e, what code expects it to be in screen coordinates?16:04
=== dandrader is now known as dandrader|afk
Elleoalf__: I believe those values get reported to Mir; they're what defines the actual area of the screen that receives touch events for the keyboard16:09
Elleothey go through the UbuntuApplicationApiWrapper class which opens a socket somewhere and sends them off16:10
ElleoI'm not especially familiar with that code though, as until this bug showed up I hadn't had any reason to go near it before16:10
alf__Elleo: I see in InputMethod::onVisibleRectChanged() that we are passing the coordinates to the inputMethodHost() (I don't know what that is)16:11
Elleothat'll be the maliit framework iirc16:13
Elleonot sure what it needs them for, below that you'll also see they're being reported via the application wrapper though: d->applicationApiWrapper->reportOSKVisible16:14
Elleowhich I believe is what actually sends it off to something responsible for shaping the input area (maliit-framework wouldn't do that)16:14
Elleoalthough it might actually16:16
Elleoyeah, I know nothing about this side of the keyboard really :P16:16
Elleoyeah, iirc the InputMethodHost ends up going to the maliit framework which then in turn interacts with the QInputContext, etc. it's taking the values so it can inform apps what areas to avoid displaying content in16:22
Elleopretty sure its the transmission through applicationApiWrapper which tells Mir(/or something else?) what area to take touch input from16:23
Elleobut don't take my word as gospel on that, like I say this is a part of the Keyboard I've never had to do anything with before16:24
alf__Elleo: fair enough, who should I ask about this part of the code?16:25
Elleoalf__: I'm not sure; the keyboard was somewhat abandoned when I joined the company, I just sort of inherited it16:26
Elleoalf__: bfiller might have an idea if there's anyone still at Canonical who'd be more familiar with that stuff16:27
Elleoalf__: otherwise bzr blame can at least tell you who touched it last :/16:27
Elleoalf__: from the logs it looks like it was mostly Thomas Moenicke and Guenter Schwann who were involved in that stuff, but I don't think either are with us anymore?16:31
Elleoalf__: so for now I guess I'm your best bet :/16:32
=== dandrader|afk is now known as dandrader
Elleoalf__: fyi there is a branch going to land working around this in the keyboard (it just looks at the discrepancy between the real screen geometry and the item size), so we have something to unblock the issue for now; I'd guess the main thing to figure out is if it's indicative of something wrong in Mir or not16:53
alf__Elleo: so short story, the problem is that the "fullScreenItem" is not really fullscreen anymore. The keyboard rectangle it reports is relative to "fullScreenItem". That would be fine for Mir since it expects surface-relative coords for the input region. But somehow absolute coordinates reach mir and this messes up input handling (we ignore the area of 58 at the top of the OSK and therefore can't swipe)16:53
alf__Elleo: from what I have seen up to now, it doesn't seem to be a problem in Mir. But I will look a bit more tomorrow.16:54
Elleoalf__: is it possible that Mir 0.2 was expecting absolute coordinates?16:55
* alf__ checks16:55
alf__Elleo: probably so16:58
Elleoalf__: I wonder what's translating between the relative coordinates in keyboard and the absolute coordinates then16:58
Elleopossibly something in the ui session stuff that UbuntuApplicationApiWrapper talks to?16:59
alf__Elleo: UbuntuApplicationApiWrapper is a bit mysterious... I only see a connection to which we sent data16:59
Elleoyeah, I have no idea what it's doing, just that it does something...17:00
alf__Elleo: anyway, I will investigate more tomorrow, thanks!17:01
=== alan_g is now known as alan_g|EOD
Elleoalf__: okay, thanks for helping out with this :)17:02
=== dpm is now known as dpm-afk
=== dandrader is now known as dandrader|lunch
=== rusty is now known as Guest81848
jgdxdoes the background image change depending on the device rotation?18:54
=== barry` is now known as barry
dobeyjgdx: doesn't change here19:02
dobeyjgdx: but on my phone(s) uniy8 doesn't rotate anyway19:02
dobeyor do new images do that now?19:02
jgdxdobey, which channel? Rotates on #89 here (mako)19:02
dobeymy mako has a totally dead battery right now19:03
dobeyon hammerhead i'm stuck on image $50 because regressions :(19:03
dobeydash isn't rotating on my mako on #9119:04
jgdxactually, uss will change the background if you rotate in-app. ha19:05
jgdxdobey, thanks, you led me to my answer19:05
dobeydoesn't here :)19:05
jgdxonly temporary19:06
dobeyeh i'm flashing it to 99 now19:06
dobeyTassadar: hi! did you look into the screen blank issue on hammerhead any further? i haven't been able to, myself19:07
TassadarI did, but I can't figure out how to fix it19:07
Tassadarthe drivers are shit :P19:07
Tassadar:/19:07
dobey:-/19:07
Tassadardo you know when did it start?19:08
dobeyaround image #7119:08
dobeywhen the split greeter was added iirc, but i don't know if that's related19:08
dobeytaking the split greeter back out obviously didn't fix it19:08
Kohelet_so I'm working on a port to xperia z ultra and while building I get19:11
Kohelet_make: *** No rule to make target `/home/kohelet/freshTogari/out/target/product/togari/obj/STATIC_LIBRARIES/libelf_intermediates/export_includes', needed by `/home/kohelet/freshTogari/out/target/product/togari/obj/RECOVERY_EXECUTABLES/extract_elf_ramdisk_intermediates/import_includes19:11
Kohelet_from what I understand this is because libelf isn't found, but libelf is on the system and I also have it in my manifest.xml19:12
jgdxis it possible to easily resize the window of an app in autopilot? I'm sure I've seen a example somwhere. /me investigates.19:15
=== boiko_ is now known as boiko
=== Kohelet_ is now known as Kohelet
Koheletwas able to figure it out...19:27
dobeyTassadar: any ideas how i can help figure it out?19:31
Tassadarif you can determine when it started, then we'll know which change broke it19:32
Tassadarbut tasemnice probably already deleted builds around 71 :/19:33
popeyTassadar: the n5 builds or our builds?19:39
Tassadarn519:39
Tassadarhe can't test it with official builds)19:40
popeyah19:40
popeyi keep an archive of our builds19:40
=== gnu is now known as Guest53248
dobeyTassadar: it was around that time when it broke :-/19:49
ahayzenSaviq, ping20:01
mhall119alex_abreu: when I create a new webapp project, it doesn't actually create a QtCreator project file, just a .desktop and .json file20:09
alex_abreumhall119, yes this is a bug, I have a branch that is almost done (but got postponed) that fixes this at the qtc level20:09
alex_abreumhall119, I'll try to make it land soonish20:10
Saviqahayzen, pong20:30
ahayzenSaviq, Are you the right person to ask about unity8 bugs?20:30
Saviqahayzen, sure, just leave the questions around, I'll reply when around20:31
ahayzenSaviq, cool :) ... When I have an application open if I do a full swipe from left to right, sometimes, you only see the background behind and not the scope until the swipe is complete?20:31
ahayzenSaviq, is this a bug/design? if is a bug it is known or do i need to report it?20:32
ahayzen*it is a bug is it20:32
Saviqahayzen, bug #133326120:39
ubot5bug 1333261 in Unity 8 "Left-edge long-swipe dash fade-in is not smooth" [Medium,Triaged] https://launchpad.net/bugs/133326120:39
ahayzenSaviq, ah cool thanks you guys are already on it :)20:40
Saviqahayzen, we probably won't spend time on it (it's caused by Qt 5.3), because we20:40
Saviq'll be changing this whole transition20:41
ahayzenah i see ... well i'll look forward to the new transition :)20:41
mhall119alex_abreu: my webapp isn't launching, won't even get to where it writes a log file in ~/.cache/upstart/20:50
mhall119https://code.launchpad.net/+branch/~mhall119/+junk/bbqpad-app is the code20:50
mhall119can you help me figure out what I'm doing wrong?20:50
mhall119I'm using the devel channel image20:51
mhall119r8720:51
alex_abreumhall119, remove the "," at the end of http://bazaar.launchpad.net/~mhall119/+junk/bbqpad-app/view/head:/app.json#L820:52
alex_abreumhall119, add a "-" between session and cookies in http://bazaar.launchpad.net/~mhall119/+junk/bbqpad-app/view/head:/app.desktop#L520:52
alex_abreumhall119, those are know issues w/ fixes landing very soon20:53
mhall119alex_abreu: those did the trick, thanks!20:56
alex_abreumhall119, np20:57
AlbertAogra_: can  I get top approval on this one: https://code.launchpad.net/~albaguirre/ubuntu-touch-session/remove-power-off-delay-arg/+merge/22446920:59
=== dandrader|lunch is now known as dandrader
WellarkSaviq: still around?21:48
Wellarkis that suru theme going to land?21:49
WellarkI have a MP against system-settings I would love to get landed but I don't want to collide with the theme landing silo21:49
compulsivehi all22:37
popeyhi22:41
compulsivehow's it goin22:41
popeygreat22:42
compulsivecool22:42
compulsiveis ubuntu touch even going to hppen22:43
popeyHell yeah!22:43
compulsiveI hope22:43
popey ☑ That was easy22:44
=== sveta is now known as svetlana

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