/srv/irclogs.ubuntu.com/2013/06/18/#ubuntu-app-devel.txt

=== chriadam|away is now known as chriadam
=== _salem is now known as salem_
=== salem_ is now known as _salem
dholbachgood morning06:57
mzanettitmoenicke: hey... how's the progress with the keyboard testing?08:20
mzanettitmoenicke: we would like to start integrating the shell tests with some OSK usage08:23
mzanettitmoenicke: do you have some emulators ready we could reause?08:23
mzanettireuse08:23
tmoenickemzanetti: lets talk to bill about this in the afternoon08:24
JamesTaitGood morning all, happy Autistic Pride Day! :-D08:31
seb128hum08:36
seb128is there a "ubuntuorange" color in the toolkit? ;-)08:36
dpmhi seb128, the sdk guys can probably give a better answer, but I haven't found it, I just specify "#DD4814" :)08:46
seb128dpm, ok, that's cheating :p08:46
seb128dpm, hey btw, and thanks ;-)08:47
dpm:)08:47
seb128I was going to do that but I figured out that it would be nice to have a named color for that one08:47
seb128I guess quite some apps want to use to give an ubuntu look08:47
timpseb128: no, there is no UbuntuOrange yet https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/109820908:50
ubot2Ubuntu bug 1098209 in Ubuntu UI Toolkit "define the ubuntu orange color in the SDK" [Wishlist,Confirmed]08:50
seb128timp, thanks ;-)08:57
guschom26er: can you review this? https://code.launchpad.net/~schwann/gallery-app/gallery-video-autopilot/+merge/17001509:08
om26ergusch, looking09:08
guschom26er: thx09:08
dholbachdpm, seems like the developer.ubuntu.com/packaging link is broken09:14
dholbachdpm, do you know what happened there?09:14
dholbachdpm, FWIW http://developer.ubuntu.com/resources/tools/packaging/ is broken too09:23
dpmdholbach, looking at it now09:35
dholbachdpm, thanks09:37
zsombiguys, do we have a way to detect the running QML application name? (i.e. desktop file name, main QML file)? Arguments SDK component aint seem to provide that info...09:39
timpgusch: I'm a bit stuck with autopilot (I don't have much experience with it).10:02
timpgusch: when I return the "cancel icon", I know that I need tools.back from the toolbar. But is there a way with autopilot to do that besides adding objectNames to the back/cancel button everywhere?10:02
guschtimp: you can use any property10:03
timpgusch: an easy way to do it is by looking for the button with text "Cancel" or "Back", but that will break on systems that use different languages10:03
guschtimp: correct10:03
timpgusch: #toolbar_items = toolbar.select_single("ToolbarItems") #return toolbar_items.back10:03
timpwithout the #10:03
timpthat doesn't work, but you get the intention. I want to get the back-property of toolbar_items10:03
guschuse select_many instead of select_single10:04
guschtimp: ^ that might help10:04
timpwhy many? there is only one10:04
guschfor sure?10:04
timpthe problem is AttributeError: Class 'ToolbarItems' has no attribute 'back'.10:05
timpin qml I have this in ToolbarItems: property Item back10:05
timpI just don't know how to access it from python10:06
gusch"back" is a QML property?10:06
timpyes10:06
guschwhich type?10:07
timpItem10:07
timpwell, I noticed that now I have several subclasses of ActionItem (including Button, ToolbarButton), and if select all ActionItems it doesn't seem to include Buttons10:08
guschthen return toolbar_items.select_many("QQuickItem")[0]10:08
guschthen "return toolbar_items.select_many("QQuickItem")[0]"10:08
guschor whatever index the back item has (that's a least not worse than the current test ;)10:08
timpohh I need to use the cpp classname, not the qml name?10:08
guschyep10:08
guschyou can run "gallery-app -testability" - and "autopilot vis"10:09
guschthe select the gallery, and you can browse the QML tree nicely10:09
timpok, thanks10:10
timpmaybe I can do it all via the objectName10:15
timpgusch: it is a bit confusing to me. I have ActionItem, and subclasses of that called ToolbarButton and Button (in qml)10:23
timpgusch: autpilot vis shows me a classname "Button" for the classic buttons, and "ActionItem" for the ToolbarButtons10:23
guschActionItem for ToolButtons ?!?10:24
guschthat sounds weired10:24
timpgusch: and it seems that the classname for autopilot must match exactly, so I cannot say "ActionItem" and select all instances of (subclasses of) ActionItem10:24
guschbut if you can set a unique objectName, that would be the best for testing for course10:24
guschtimp: yes, the names have to match exactly10:25
timpgusch: yes, I was trying that. but it doesn't work for me because of the classname10:25
timpsome times it is ActionItem (for back buttons), some times Button (for cancel buttons)10:25
timpI guess it'll work if I introduce an additional get_toolbar_back_icon function in addition to the get_toolbar_cancel_icon10:26
timpjust weird that for ToolbarButton I need to search for ActionItem instead.10:27
timpthe full implementation of ToolbarButton is this: https://pastebin.canonical.com/93000/10:27
timpperhaps it doesn't change the className because I didn't add properties to it10:27
timpthat would mean the tests that I am planning will break if I ever add a property to ToolbarButton10:28
guschtimp: damn - but might be10:31
om26ergusch, Hi! gallery-app crashes on 'Auto Enhance' btw10:32
timpgusch: thanks for the help. I'll have a break now and then I give it another try.10:32
guschom26er: yep https://bugs.launchpad.net/gallery-app/+bug/118034510:32
ubot2Ubuntu bug 1180345 in gallery-app "Auto enhance crashes gallery app and deletes photo" [Critical,Confirmed]10:32
guschom26er: you could do a quick fix in C++ if you want ...10:33
guschom26er: in src/photo/photo.cpp comment line 805 (QApplication::processEvents();)10:34
om26ergusch, cool, that works10:36
guschom26er: well - it blocks the UI - so I need another (proper) fix10:39
=== oSoMoN is now known as oSoMoN|afk
om26ergusch, should I keep your workaround with my test branch ?11:24
guschom26er: good question11:25
guschom26er: I'd say yes - and update the bugreport11:25
om26erok11:27
nerochiaroom26er: just to recap: all the autopilot tests that are in the branch at the momnent for the browser plugin are working, right ? i just tried and they all seem to pass successfully11:29
nerochiaroom26er: didn't you say there were some failing ?11:29
om26ernerochiaro, yes, I have disabled a few so you would not see them failing11:29
om26erlook in test_history.py11:30
nerochiaroom26er: ok, thanks11:30
nerochiaroom26er: oh, i thought you had disabled them with the decorator, not with comments11:36
nerochiaroom26er: so it's basically only test_history and test_chromeless, right ?11:37
om26ernerochiaro, yep11:37
nerochiarooSoMoN|afk: when you're back i have one question for you on a traslations-related issue11:38
nerochiarooSoMoN|afk: please pingme11:38
=== MacSlow is now known as MacSlow|lunch
=== greyback is now known as greyback|lunch
om26ergusch, I added tests for auto-enhance and 'delete last photo' .. https://code.launchpad.net/~om26er/gallery-app/add_test_for_auto_enhance/+merge/17005412:06
om26erfor the latter case the bug would need to be fixed before the branch goes in.12:06
guschom26er: cool - looking12:06
=== oSoMoN|afk is now known as oSoMoN
oSoMoNnerochiaro: ping12:15
guschom26er: looks good - I guess now it's my turn ;)12:18
om26ergusch, yep :D12:18
nerochiarooSoMoN: hi, i've been trying to figure out why my translation files were being installed under /usr/local/share instead of /usr/share and I'm not really finding any difference with the other packages that use translations. Wonder if you could have a quick look at lp:~amanzi-team/ubuntu-ui-extras/ubuntu-ui-extras-initial and confirm that debuild does the same for you and give any clue12:20
oSoMoNnerochiaro: sure, I’ll look into it shortly and get back to you12:22
=== greyback|lunch is now known as greyback
nerochiarooSoMoN: ok, i'll go for some food in the meantime12:28
=== _salem is now known as salem_
=== chriadam is now known as chriadam|away
guschnerochiaro: can you review the camera? https://code.launchpad.net/~schwann/camera-app/camera-video-orientation/+merge/17005312:43
dholbachdpm, could you find out anything?12:46
oSoMoNnerochiaro: independently from the issue you asked me to look into, building a package with debuild fails with the following error:12:49
oSoMoNdh_install: qtdeclarative5-ubuntu-ui-extras-plugin-autopilot missing files (usr/lib/python*/dist-packages/ubuntu-ui-extras/*), aborting12:49
timpzsombi: could you review https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/toolbarItems/+merge/170070 ?12:49
timpzsombi: don't happrove yet, when you think the MR is good, I'll first send an e-mail to ubuntu-touch mailing list12:50
dpmdholbach, back online at http://developer.ubuntu.com/resources/tools/packaging/ - it was my fault, I had reshuffled a lot of the content on the resources page and had made the packaging page private by mistake12:50
gemaboiko: ping12:52
boikogema: pong12:53
gemaboiko: I'd like to have a quick chat with you as to where you are functionality - wise with the phone-app12:53
gemaboiko: we are going to start testing it in the lab and I'd like to know what to expect12:54
gemaand how to prioritise the testsing12:54
boikogema: yep, nice!12:54
dholbachdpm, thanks12:55
gemaboiko: so I had an interesting morning today not being able to do some things with the app, it'd fail to hang up, or to send the conference id to a conference call12:55
gemaboiko: it'd also not be able to call to the provider on a service number12:55
gemaboiko: are these things on your TODO or do you think they are there12:56
boikogema: so conf call is not supported yet12:56
gemaok12:56
boikogema: failing to hang up is something new to me12:56
boikosalem_: ^12:56
gemaboiko: I was calling the canonical conference system12:57
gemaI had tried to enter the conf id12:57
gemathen tried to hang up12:57
gemadidn't happen12:57
boikogema: oh, you mean DTMF then ok12:57
salem_gema, ah, I have seen this before. I think it only happens if you try to hangup from the dtmf screen.12:57
gemasalem_: waht is the dtmf screen?12:58
salem_gema, the screen to type numbers during an ongoing call.12:58
gemasalem_: ah, ok12:58
gemasalem_: which screen should I use to hang up?12:58
salem_gema, well, you should be able to hangup from that screen, but as it seems we have a bug there. I would suggest you to tap back, and then hangup from the main screen12:59
gemasalem_: ack, do you want me to raise a bug?13:00
boikogema: please do13:00
gemaack13:00
salem_gema, that would be great. thanks13:00
boikosalem_: I'll try dtmf in today's image to see if it works13:00
salem_boiko, ok. I think this bugs is not 100% reproducible, but I have seen it before13:01
boikosalem_: ok, let's see what I can find here13:01
gemaboiko, salem_ : https://bugs.launchpad.net/phone-app/+bug/119215813:02
ubot2Ubuntu bug 1192158 in phone-app "Unable to hang up from dtmf screen" [Undecided,New]13:02
gemaI used your nomenclature, feel free to change it :)13:02
gemaboiko: on the bright side I was able to gather some energy consumption data with a call that lasted 28 mins13:03
gema(until credit ran out)13:03
boikogema: oh, and how did it go?13:03
gemaboiko: I still need to do the math, I will let you know soon13:03
gemahow long I expect the battery to last with an engaged call13:04
oSoMoNnerochiaro: that said, regarding your issue, I’ve added some debugging messages to src/Ubuntu/Components/Extras/Browser/po/CMakeLists.txt, and the value of CMAKE_INSTALL_LOCALEDIR when building a package is share/locale, which looks correct13:04
boikogema: not sure how long, cause battery usage does not depend exclusively on the phone-app13:06
gemaboiko: I know, that's why we are measuring the whole system :)13:07
gemaboiko: but I need the phone-app to behave if we are going to measure this kind of thing13:07
boikogema: yep13:07
gemai.e. being able to make calls, keep them up, hang up13:07
gemaboiko: do you have an ETA for the conferencing?13:08
=== MacSlow|lunch is now known as MacSlow
nerochiarooSoMoN: yes, it looks correct, but then if you look under obj-i686-linux-gnu/  you will see that /usr/local/ is prepended to that path13:12
boikogema: that's DTMF, conferencing is something different (creating a conf call on the phone)13:13
boikogema: I'll test that, and I bet salem_ will also give it a try too13:13
oSoMoNnerochiaro: I don’t see anything there, probably because the package fails to build13:13
oSoMoNnerochiaro: have you tried building from a clean copy of your branch?13:14
boikogema: actually there is a simple test for that: call some other phone, and press the keypad buttons to see if any sound is heard on the other side13:14
=== dholbach_ is now known as dholbach
nerochiarooSoMoN: i'm always doing that from a clean branch now. i'm using debuild -uc -us -b and the obj-i686-linux-gnu/ dir is there even if the build fails (which i confirm it does with the issue you mentioned, i'm fixing it)13:15
gemaboiko: it is not at the moment13:16
boikogema: so there is indeed a but there, the phone has just finished flashing, let me try it here13:17
nerochiarooSoMoN: ok, fixed the issue you were noticing before, pushed13:17
nerochiarooSoMoN: but there are other problems, i'm working on them13:17
gemaboiko: ack13:17
boikogema: just tested using today's image, and DTMF seems to be working here, I'll do some more tests13:19
oSoMoNnerochiaro: next error when building package:13:19
oSoMoNdh_install: tmp/ubuntu-ui-extras-initial/obj-x86_64-linux-gnu/tests/unittests/Ubuntu/Components/Extras/libubuntu-ui-extras-plugin.so exists in debian/tmp but is not installed to anywhere13:19
oSoMoNnerochiaro: does the package actually build successfully on your machine?13:19
nerochiarooSoMoN: no, it does not, i was trying to approach one issue at a time. and the translation one is the one i could not explain13:19
=== francisco is now known as Guest7324
nerochiarooSoMoN: this one you just pointed out to me I know where it comes from13:20
oSoMoNnerochiaro: well how can you tell there’s an issue with the package if the package doesn’t build at all?13:20
nerochiarooSoMoN: because i see the translation files installed in the wrong dir13:20
boikogema: are you using today's image?13:20
oSoMoNnerochiaro: fix the package build first, and then see if there are issues, not the other way around13:21
nerochiarooSoMoN: fair enough13:22
guschnerochiaro: can you review the camera? https://code.launchpad.net/~schwann/camera-app/camera-video-orientation/+merge/17005313:33
guschom26er: here the delete fix https://code.launchpad.net/~schwann/gallery-app/gallery-delete-last-photo/+merge/17007713:33
guschboiko: can you review? https://code.launchpad.net/~schwann/gallery-app/gallery-previewmanager-mediacollection/+merge/16999813:33
boikogusch: probably after the standup only, is that fine?13:34
guschboiko: ok - no problem13:34
guschboiko: that branch sits and waits for about a week already ;)13:35
boikogusch: hehe, ok :)13:35
timpgusch: are all gallery autopilot tests supposed to pass on my local machine?13:39
nerochiarogusch: i'm stuck with some build issues to solve ASAP, but i can try to take care of that later13:39
guschnerochiaro: ok13:39
timpgusch: I fixed all the button tests in my branch, but I have two (seemingly unrelated) fails:13:40
guschtimp: yes13:40
timpFAIL: gallery_app.tests.test_photo_viewer.TestPhotoEditor.test_photo_editor_crop(with mouse)13:40
timpFAIL: gallery_app.tests.test_photo_viewer.TestPhotoViewer.test_double_click_zoom(with mouse)13:40
guschom26er: ^13:40
timpin the crop test, I don't see the frame for cropping when the test is run.13:40
guschtimp: but they all run fine locally and even on Jenkins13:40
guschtimp: whaaaat?!?13:41
timpin the double_click_zoom test, the mouse cursor moves to the center of the photo, but I don't see zooming13:41
timpgusch, om26er oh. doubleclick to zoom, and cropping also don't work for me when I run gallery myself13:41
timpah..13:41
timpprobably I broke that.13:42
timpbecause I commented out some HUD stuff.13:42
timp^so that I can run it without having a HUD. But the code is broken because of my chagnes13:42
timp*changes13:42
guschtimp: you shouldn't do that ;)13:42
gemaboiko: no, I am using the first saucy image, I will be upgrading soon13:43
boikogema: ok13:43
timpgusch: I don't have Ubuntu.HUD13:43
gemaboiko: for some reason I cannot turn it on right now, need to look into that13:43
timpgusch: last time I installed shell stuff it broke my whole system. (that was months ago)13:43
gemaboiko: we are looking at 7.3 hours worth of battery13:43
gemawith an active call13:44
gemapmcgowan: ^13:44
guschtimp: well, then uncooment it, and push hope the best, and jenkins will test it13:44
guschtimp: well, then uncomment it, and push hope the best, and jenkins will test it13:44
pmcgowangema, good13:44
timpgusch: yes, I will do that.13:45
seb128hum13:47
seb128did anyone look adding qtsystems (https://qt.gitorious.org/qt/qtsystems) to the touch image?13:47
seb128that might be a question for #ubuntu-touch13:47
om26ergusch,  timp, ah sorry, unity notifications are not working, looking now13:53
guschom26er: should be sorted timp broke the tests himself ;)13:54
=== ckpringle_ is now known as ckpringle
seb128Kaleo, hey14:08
seb128Kaleo, I've a quick question for you about keyColor if you are around14:08
=== oSoMoN_ is now known as oSoMoN
Kaleoseb128: sure14:17
seb128Kaleo, is there any way to use that in a ListItems.Standard?14:18
seb128Kaleo, I was talking with tiheum, the icon he has atm are using a #cccccc grey rather than #88888814:18
seb128he recommended using keyColor to change the grey14:19
seb128but I'm not sure that works with listitems?14:19
Kaleoseb128: where did you get keyColor from?14:22
Kaleoseb128: you know it's part of the Icon class right? http://developer.ubuntu.com/api/ubuntu-12.10/qml/mobile/qml-ubuntu-components0-icon.html14:22
seb128Kaleo, right ... I guess my question is "can we set the property for an icon: embedded in a ListItem.Standard"?14:23
Kaleoseb128: http://developer.ubuntu.com/api/ubuntu-12.10/qml/mobile/qml-ubuntu-components-listitems0-standard.html#icon-prop14:23
seb128or do I need new icons with the proper grey?14:23
Kaleoseb128: you are probably confused by that bad API & documentation for ListItem.Standard14:23
pmcgowangema, kind of interesting comparison here http://www.anandtech.com/show/6440/google-nexus-4-review/214:23
Kaleoseb128: ultimately the API should be:14:24
Kaleoseb128: ListItem.iconSource and ListItem.iconName14:24
Kaleoseb128: the source would be any url to any image file14:25
Kaleoseb128: iconName would be the name of the icon in the theme14:25
Kaleoseb128: right now the icon property can take 2 different values14:25
Kaleoseb128: either a URL to an image file14:25
gemapmcgowan: yep, we are pretty close to those figures :)14:25
Kaleoseb128: or any Item derived class14:25
pmcgowangema, yep, and I like the way they describe the testing14:26
seb128Kaleo, if I set an iconSource, does it use an Icon {} element to build the widget, and can I set keyColor: for that Icon{} in some way?14:26
gemapmcgowan: I will definitely be reading that in detail14:26
Kaleoseb128: so no it does not use an icon14:26
timpKaleo: fyi, I'd propose to add iconName to ActionItem (all ListItems, Buttons, ... inherit from that now) and property var icon can be deprecated14:27
Kaleoseb128: since the API does not accept an icon name14:27
Kaleoseb128: it would not work14:27
seb128Kaleo, ok, I guess that settle it, I will ask for new icons14:27
seb128Kaleo, thanks14:27
Kaleoseb128: so all you can do right now14:27
Kaleoseb128: is to create an Icon class yourself14:27
Kaleoseb128: ListItem.icon: Icon {}14:27
gemapmcgowan: although I am expecting us on the nexus 4 to be pretty close to their figures (since it is mainly hw usage all those tests)14:27
Kaleoseb128: (and there is no "iconSource" property)14:27
pmcgowangema, right14:28
gemapmcgowan: but it is comforting to know we are :)14:28
timpKaleo: there is, from its parent ActionItem14:28
pmcgowanexactly14:28
timpKaleo: but it is still ignored in the list items14:28
Kaleotimp: don't make it more confusing for seb12814:28
Kaleo:)14:28
Kaleotimp: so yes, please make sure API includes iconName & iconSource14:29
Kaleotimp: and wherever we are using Item icon, we probably need to keep it for now since people rely on it14:29
Kaleotimp: but we can make it so that it does not accept a URL anymore14:29
seb128Kaleo, I've the feeling I will run into https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1190751 if I try your workaround ;-)14:29
ubot2Ubuntu bug 1190751 in Ubuntu UI Toolkit "ListItem.Standard icon's item geometry is incorrect " [Undecided,New]14:29
Kaleotimp: (since iconSource will support that)14:29
Kaleoseb128: yes, that's a dupe btw14:30
Kaleoseb128: we can fix that14:30
seb128Kaleo, sorry for the dup, I looked for reports bug didn't find it14:30
Kaleoseb128: no worries14:31
timpKaleo: Item icon was added because iconSource was not flexible enough, but I don't remember the specific use-case14:32
timpbrb food14:32
seb128Kaleo, timp: thanks for the help!14:33
timpKaleo: hmm. instead of iconName: "name" it is now possible to say icon: Icon { name: "name" }14:34
Kaleotimp: yes that's what I told seb12814:36
seb128timp, Kaleo: I tried that with an icon and ran into the scaling issue ... I will keep the color issue instead until that bug is fixed :p14:39
seb128"pick your bug" ;-)14:39
Kaleoseb128: k14:40
guschom26er: are you looking for my review?14:51
om26ergusch, hmm, missing context. what review?14:52
guschom26er: here the delete fix https://code.launchpad.net/~schwann/gallery-app/gallery-delete-last-photo/+merge/17007714:52
om26ergusch, I am going to do the functional review to make sure if it fixes14:55
guschom26er: fine14:55
=== ckpringle_ is now known as ckpringle
=== salem_ is now known as _salem
boikogusch: MR approved16:19
guschboiko: \o/16:19
guschom26er: please merge trunk and push again for https://code.launchpad.net/~om26er/gallery-app/add_test_for_auto_enhance/+merge/17005416:47
om26ergusch, on it16:47
gusch:)16:47
=== _salem is now known as salem_
renato_om26er, ping18:03
om26errenato_, pong18:03
renato_om26er, I am getting this error on my project when I try to run lcov: https://pastebin.canonical.com/93039/18:05
renato_do you have any idea?18:05
om26errenato_, no, no clue :/18:08
=== salem_ is now known as _salem
=== _salem is now known as salem_
om26errenato_, there might be some compatibility problem? https://bugs.launchpad.net/ubuntu/+source/lcov/+bug/116375818:40
ubot2Ubuntu bug 1163758 in lcov (Debian) "geninfo: Argument "=====" isn't numeric in numeric gt (>) at /usr/bin/geninfo line 1126." [Unknown,Confirmed]18:40
renato_om26er, but the real problem is: geninfo: ERROR: /home/renato/Projects/phablet/contact-service/fix-dbus-lost/build/src/CMakeFiles/address-book-service-lib.dir/qindividual.cpp.gcno: reached unexpected end of file19:10
renato_this crash gcov compilation, and I am not able to see any resuts19:10
om26errenato_, tried split_crc ? http://ltp.sourceforge.net/coverage/lcov/geninfo.1.php19:15
om26errenato_, that error is specified there19:15
renato_om26er, let me try19:18
=== francisco is now known as Guest75516
renato_om26er, I am not using geninfo command19:26
renato_only lcov19:26
=== boiko_ is now known as boiko
=== marlinc is now known as marlinc|away

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