/srv/irclogs.ubuntu.com/2013/12/17/#ubuntu-app-devel.txt

ahayzenballoons, will do00:00
ahayzenballoons, will probably add them tomorrow if Victor doesn't :)00:00
balloonssounds great. you guys have been doing great work00:00
balloonsand the test writing has been fine00:00
balloonsgood stuff!00:00
balloons*fine = fun00:00
balloonslo00:00
ahayzenhaha00:01
ahayzenballoons, should this be able to land now, if we approve it again? Jenkins was playing up before https://code.launchpad.net/~vthompson/music-app/fixes-1259962/+merge/19860100:02
balloonsahayzen, yes most likely00:02
ahayzenballoons, haha Victor just did it literally then lol00:02
ahayzenballoons, thanks for ur help, we'll try and stay green :)00:03
balloonsahayzen, you are most welcome00:04
ahayzenballoons, oh one very last thing the bug for infographics should i add the needs-autopilot-test or just report as bug first before we decide if to do it or not?00:05
ahayzenballoons, FYI the bug with autopilot testcase https://bugs.launchpad.net/music-app/+bug/126158700:13
ubot2`Launchpad bug 1261587 in Ubuntu Music App " Autopilot Testcase Needed: When the library has an empty state" [Undecided,New]00:13
balloonsahayzen, you can still put it in there, but don't mark the status as active00:15
ahayzenballoons, ok and tht MP didn't land :/00:15
=== chriadam|away is now known as chriadam
dholbachgood morning08:00
=== mpt_ is now known as mpt
=== dholbach_ is now known as dholbach
JamesTaitGood morning all; happy Maple Syrup Day! :-D09:14
=== chriadam is now known as chriadam|away
owlhi09:33
oSoMoNnerochiaro, hey, would you mind approving https://code.launchpad.net/~osomon/gallery-app/fix-ftbfs-cmake-moc/+merge/199198 ?09:42
=== zequence_ is now known as zequence
nerochiarooSoMoN: looking10:05
nerochiarooSoMoN: ok, makes sense. approved10:06
nerochiarooSoMoN: weird that it built fine on desktop10:06
oSoMoNnerochiaro, yeah, it shouldn’t have built in the first place10:08
oSoMoNnerochiaro, thanks for approving, the next MR up for review is https://code.launchpad.net/~osomon/gallery-app/wait-for-confirm-dialogue/+merge/19907610:08
oSoMoNnerochiaro, I’ve added a bunch of mostly minor comments to your MR10:37
nerochiarooSoMoN: i'm looking into them10:38
=== xnox_ is now known as xnox
nerochiarooSoMoN: regarding the comment about my TODO for the transitions going in and out of open pictures: I had a second look and I think the current behavior is acceptable. My next work item after this MR is merged is making sure the list comes back to the previous position, so I can have another look at the transitions when I do that. But I think for now I'll just remove the TODO from the code, address the rest of your10:56
nerochiarocomments so we can go ahead and merge this MR. Sounds like a plan ?10:56
oSoMoNnerochiaro, sounds like a plan, I’ll do another quick round of functional testing after you address all my comments, and we should be good to go10:57
nerochiarooSoMoN: and I think I'll actually change the name of lastOpenedPicture to currentMediaInViewer. much clearer11:02
oSoMoNnerochiaro, sounds good11:02
oSoMoNnerochiaro, I’ve triggered a new CI run on your MR, now that my fix for the build on trusty has been merged11:21
nerochiarooSoMoN: thanks11:21
nerochiarogreyback: do you remember if there is already a bug report somewhere about providing from unity8 a standard way for AP tests to request closing an app "cleanly" (i.e. as the user would do using the shell) ?11:31
greybacknerochiaro: no bug report I'm aware of11:31
nerochiarogreyback: if I file one, to which package should it be addressed to ?11:32
greybacknerochiaro: unity8 I think. Didn't we think it best if unity8 had a support package for AP to add common utils like this?11:33
nerochiarogreyback: yes11:35
nerochiarogreyback: there's already a package called unity8-autopilot though11:36
greybacknerochiaro: true, that includes AP tests for u811:36
greybackthere are probably handy things in that which other projects would appreciate if shared11:36
nerochiarogreyback: so should i report a bug on unity8-autopilot to add this king of emulator ?11:39
greybacknerochiaro: yeah11:40
nerochiarogreyback: https://bugs.launchpad.net/unity8/+bug/126172011:59
ubot2`Launchpad bug 1261720 in Unity 8 "add to unity8-autopilot emulator for closing apps" [Undecided,New]11:59
nerochiarogreyback: unity8-autopilot is built out of unity8 it seems11:59
greybacknerochiaro: yes it is12:00
nerochiarooSoMoN: pushed all changes you requested, I think12:00
oSoMoNnerochiaro, ok, will test in a bit12:04
=== gatox is now known as gatox_dr
oSoMoNnerochiaro, it seems you didn’t remove the TODOs from your code13:07
nerochiarooSoMoN: oh, ok, doing it now13:10
nerochiarooSoMoN: pushed13:11
nerochiarooSoMoN: heading out for a bit of food13:11
oSoMoNsame here13:12
zsombitmoenicke: ping14:04
tmoenickezsombi: pong14:06
timpzsombi: can you add the proposed new functions to https://docs.google.com/a/canonical.com/document/d/1Ng3fJxg-LaUgQU3ITNIg0cOB0Ii00XuX3a8PAs7lpns/edit# ?14:10
zsombitmoenicke: do you know the height of the OSK by heart?14:11
zsombitimp: yes, I will14:12
timpokay :) thanks14:12
zsombitimp: uhm... how you thought? where ? :D14:13
zsombitmoenicke: any hunch? 30 GU or less?14:14
timpzsombi: I don't understand the question14:14
timpzsombi: in proposal 2? (or proposal 2A). I think we can mark proposal 3 as not chosen14:14
zsombitimp: there's no "real" API proposal there, just examples thrown here'n'there...14:14
zsombitimp: ahh... and the P2 is not up to date at all...14:15
tmoenickezsombi: iirc 468 w/out predictive text14:16
timpzsombi: indeed. iconSource, pagePreloaded are nonsense now14:16
tmoenickezsombi: but you should be able to read it from the rectangle14:16
timpzsombi: I'll edit that part14:16
zsombitmoenicke: yes, I should, just was trying to test it without OSK support...14:17
zsombitimp: ok14:17
zsombitmoenicke: so 468 pixels, right?14:17
zsombitmoenicke: what's the pixel/GU ration of the Nexus?14:18
timpzsombi: I'm done14:19
zsombitimp: me2 :)14:20
zsombitimp: do you happen 2 know the GRID_UNIT_PX of Nexus?14:20
timpzsombi: I removed the old questions and the gallery section which is also ages old14:20
timpzsombi: galaxy nexus?14:20
timpzsombi: can I look it up on the device?14:20
tmoenickezsombi: iirc the GN has 18px/gu14:21
zsombiah, it's 1814:21
timpphablet@ubuntu-phablet:~$ export | grep GRID14:21
timpdeclare -x GRID_UNIT_PX="18"14:21
zsombitmoenicke: got it also, thx14:21
nerochiarooSoMoN: well spotted, there was a missing dependency. i fixed it now, pushed all changes you requested14:21
zsombitmoenicke: at some point next year, we should agree on the DatePicker to be added to the OSK layout14:22
oSoMoNnerochiaro, thanks, I’m testing on my device now14:22
oSoMoNnerochiaro, in the meantime, could you please review https://code.launchpad.net/~osomon/gallery-app/wait-for-confirm-dialogue/+merge/199076 ?14:23
tmoenickecool14:23
timpzsombi: it should be possible to use the ConditionalLayouts for our UITK gallery, right?14:23
nerochiarooSoMoN: do you mind if i take care of it after standup ? need to check something before it's too late in the day and sdk people go home14:23
timpzsombi: currently it is a bit of a hack. not a bad hack, but still14:24
oSoMoNnerochiaro, sure, as long as it goes in today14:24
nerochiarooSoMoN: ok14:24
zsombitimp: yes, it should be14:24
zsombitimp: the last one of teh DatePickers: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/pickerpanel/+merge/19925014:26
timpzsombi: jenkins still doesn't like the previous one14:28
timpzsombi: I added the new MR to my list14:29
zsombitimp: the datepicker fails again on docs... but where? it does pass on my local machine...14:29
zsombitimp: the error is not really descriptive14:30
timplast time it seemed to fail on generating the docs :s14:30
zsombitimp: that's all I see: SRC=documentation ./documentation/docs.sh /tmp/buildd/ubuntu-ui-toolkit-0.1.46+14.04.20131129/documentation; make[1]: *** [docs] Error 114:33
zsombitimp: and only on armhf14:34
=== gatox_dr is now known as gatox
timpkalikiana_: ^ didn't we see something like that before? and did you have a look at the cause then? (I vaguely remember something like that)14:37
dholbachdpm, mhall119: do you know where http://developer.ubuntu.com/resources/tutorials/getting-started/creating-click-packages-with-cpp-extensions/ went?14:46
dholbachhum, nevermind, it was linked from the test site14:47
dholbachhrm14:47
mhall119dholbach: from the developer.u.c test site?14:53
dholbachyeah14:54
dholbachI found the blog post to the app showdown from last time which I referred somebody to14:54
popeybeuno: is there a problem with the store? I clicked an app in the dash and get a spinner... http://popey.com/~alan/phablet/device-2013-12-17-145531.png14:56
popeyhmm, seems okay now14:57
beunopopey, network glitch?15:00
popeypossibly15:05
dpmhi kenvandine, if we'd like to invoke the camera app from the reminders app to take a picture to attach it to a note and upload it to Evernote, is this something we can do with the content picker? I.e. passing images around. How would it actually work?15:14
kalikiana_timp: I think it may be a race condition, that's why the previous fix didn't actually fix it… I have a possible fix in one branch; I'll isolate it and show you; but since it's a race hard to say if that fixes it… you see the problem15:14
kenvandinedpm, i don't think that's a use case we have a way of dealing with now15:17
timpkalikiana_: argh.15:17
=== oSoMoN_ is now known as oSoMoN
timpkalikiana_: yep, that can be a big hassle. :(15:18
kenvandinedpm, i think what we need for this is a way to open camera app and tell it to store the picture in the reminders app15:18
mzanettikenvandine: that still wouldn't go back to the reminders after a picture has been taken15:20
kenvandineit would need to15:20
mzanettikenvandine: I think we'd need to convert the camera app to a "picker"15:20
kenvandinemaybe15:20
mzanettidpm: so yeah... given that it's just about writing Camer {} in qml I'd say we do that ourselves for now15:21
kenvandineso the camera app being special hurts us here15:21
timpkalikiana_: anyway, thanks for looking into these annoying failures :)15:21
timpkalikiana_: you've been working quite a lot on those15:21
kalikiana_timp: somebody has to, right? :-D15:24
kalikiana_here you go https://code.launchpad.net/~kalikiana/ubuntu-ui-toolkit/splitBuildInstall2/+merge/19930615:24
timpkalikiana_: yes, *somebody* has to. That's why I am especially happy that you did it ;)15:26
dpmmzanetti, that sounds good to me. kenvandine, I'm not sure I can follow what you mean by the camera app being special, could you clarify?15:27
kenvandinewell camera-app stores the images in the gallery-app15:28
kenvandineit doesn't own it's content15:28
kenvandinewhich is "weird" in our design15:28
kenvandineall content is owned by a single app15:28
dpmaha, gotcha15:28
timpkalikiana_: thanks. I add it to my review list, I need to read up on debian packaging to fully understand what's happening.15:28
kalikiana_timp: I still don't know why it's only armhf… my theory is it might be parallelizing so stricter order would help15:31
dpmthanks kenvandine15:31
mzanettikenvandine: one more question. should I be able already to use the gallery content picker somehow?15:36
kenvandineyup15:37
kenvandineusing the content-hub api15:37
* mzanetti googles for docs15:37
kenvandinehttps://code.launchpad.net/~ken-vandine/+junk/hub-importer15:37
kenvandinehas an example15:38
mzanetticool. thanks.15:39
kenvandinehttp://developer.ubuntu.com/api/qml/sdk-1.0/Ubuntu.Content/15:40
kenvandinemzanetti, ^^15:40
mzanettiyep. found that15:40
=== Ursinha-afk is now known as Ursinha
=== teknico_ is now known as teknico
=== gatox is now known as gatox_lunch
hakermaniaHow can GNOME 3 and Unity has so similar layout and GNOME 3 be so much faster ?17:10
hakermaniaEven with the Unity blur and the logging of the applications off, GNOME 3 remains 7 to 10 times faster.17:11
EDYhi developers17:20
=== EDY is now known as Guest52113
Guest52113any developer girl here?17:21
popeyGuest52113: we tend not to distinguish between male and female developers round these parts17:28
popeywhy?17:28
=== gatox_lunch is now known as gatox
balloonsnik90, ping19:51
nik90balloons: pong21:03
balloonsnik90, :-) So I wanted to chat about alarms and your merge proposal21:04
balloonsnik90, so first things first. What's up with alarms? Who were you speaking with about getting the visualization to show up?21:05
nik90balloons: I sent some messages over to charles kerr who was incharge of the notifications. But I did not get any reply.21:06
nik90And I haven't pursued it again recently21:07
nik90so technically no one is on the alarms issue21:07
balloonsnik90, ok, so I started to chase charles a bit but wasn't sure if I had the right person21:08
balloonsnik90, however, afaik does charles need to make a change on the platform side or do you simply need information?21:09
nik90balloons: he is actually responsible for indicator-datetime service which runs in the background all the time. It is supposed to trigger the notifications when an alarm is triggerd21:10
charlesnik90, balloons, there's an email discussion today started by dpm about alarms that's discussing the current status, I can forward some of that along to you21:10
nik90charles: yeah sure. I want to get the alarms notifications works asap. It would help identify if there are any issues on the clock app side21:11
balloonscharles, thank you that would be helpful. In a nutshell is there work for you to do yet?21:13
charlesballoons: yes, there's indicator-datetime and alarm api work for this scheduled for right after the Christmas break21:16
charlesit's discussed a little in the third mail I just sent to you21:16
charlesthere's also a link to the draft API there, too21:16
balloonscharles, excellent, I got the mails ;-) Thank you much21:17
balloonsI'll update the bug with those details so we have the status recorded..21:17
balloonsnow nik90 onto your merge proposal ;-)21:18
balloonscharles, do keep us up to date on the bug report if you would, I know you will21:18
nik90charles: I thought that the indicator-datetime part was complete since it is marked as Fix commited on the bug report21:18
nik90Or is this not yet landed on the phone yet?21:18
nik90balloons: sure :)21:19
nik90balloons: I read your comment on the MP21:19
charlesnik90: well, there is some code in indicator-datetime right now that connects directly to EDS and should theoretically pull up snap decisions21:20
balloonsnik90, I'm running again on the latest build and not able to reproduce it21:20
nik90balloons: but here is the thing. The error message seems to indicate test_timer.py file which was not modified in the MP. Could this be a MIR issue or something?21:20
charlesI'm not sure how much confidence to put into the existing datetime code; one thing that will happen with the API discussed in those emails is a better testing framework to do unit tests against the EDS code21:21
balloonsnik90, it's possible I ran something else.. anything is possible. I'm not seeing any issues now21:21
nik90charles: okay. Yeah I kept an active look on the datetime trunk and thought same21:21
nik90balloons: :)21:21
nik90balloons: so merge?21:21
charles:)21:21
balloonsa couple more runs and yes, I'll approve21:21
=== Noskcaj10 is now known as Noskcaj
nik90charles: I manually got the latest trunk build and tested alarms on phone but it didnt trigger any snap notifications21:22
balloonswell maybe I have questions to, heh..21:22
nik90charles: so I am assuming that it needs some debugging21:22
nik90balloons: hehe. let me know and I will answer them :-D21:22
balloonswhy do you have a timeout loop on self.main_view.get_world_cities_list().count != 0? you are using a wait_select so it shouldn't be needed anymore.21:22
nik90balloons: line number pls?21:23
balloons?21:23
nik90found it21:23
nik90balloons: let me look at the code and let you know. 1 second21:24
nik901 minute21:24
balloonsin test_create_lap, can we confirm the stopwatch moves (aka, things increment) even if you want to sleep21:24
balloonsnik90, I'll just keep posting in here so we iterate faster21:24
balloonsand finally, you should abstract out the code to create a lap, and call it instead of calling self.test_create_lap() inside test_delete_lap21:25
balloonsthe tests look like the pass just fine now.. So I would say we're good, just need info and tweaks21:25
nik90balloons: what do you mean by abstract out the code to create a lap?21:27
nik90do you mean it shouldn't contain tests and only focus on creating a lap?21:27
balloonsnik90, so instead of calling the actual test function, call a function that can be used by both tests to create a lao21:27
balloonsyes.. not a good idea to call a test as part of a test..21:28
nik90okay that can be done. I will define a new function in test_stopwatch.py file itself21:28
balloonsyep. that should work fine21:30
nik90balloons: okay I now remember why a timeout loop is required.21:34
nik90this test is to specifically retrieve the results returned from the oonline API21:34
nik90however if you open the clock app and press add city, you will notice that the list is already populated with local results21:34
nik90hence it is necessary to wait few seconds after performing a search to allow for the online results to populate the list21:35
nik90otherwise it just grabs the first local city accidentally instead of the actual online one21:35
balloonsnik90, do you expect the list to increase in size?21:35
nik90giving a false positive21:35
balloonsalso, we need to work on not retrieving the online result but :-)21:36
nik90balloons: as in more cities being added to the list?21:36
balloonsyes.. atm, the code would continue no matter what21:36
nik90balloons: yes, the local city list is temporary. We need to get the final result from the design team21:36
nik90but it is a low priority one since online search is available21:36
balloonsohh.. while it's !=0.. weird.. it's just wait 10 seconds21:36
balloonsSo if it was 0 it would fail... that's basically just a 10 second sleep21:37
nik90actually yeah :P21:37
balloonsso if you want online results, grab the original count, then wait for it to increment21:38
nik90no the original count won't increment21:38
nik90the original list would be completely replaced with the online results21:39
balloonsohh?21:39
nik90so the original count doesnt matter21:39
nik90the listview can only have one source21:39
nik90so when an online search is made, it replaces the local list with the online one21:39
nik90and hence you see only online search results made by the user21:39
balloonswell, I'll defer to you since I'm not failure with the objects21:39
balloonsthe point is, it's not doing what you think it is :-)21:40
nik90yeah I am strongly suspecting so21:41
nik90I think I need to think about this more21:42
nik90okay I will see if this can be tested any other way tonight.21:43
nik90Regarding the second question you had about the stopwatch lap21:43
nik90I am using sleep to allow it to sleep 2 seconds and then press the lap button21:44
nik90to ensure that the lap consists of some time21:44
balloonsnik90, right I'm ok with that. I'd also like to see us check that some time elapses.. or am I missing something21:46
balloonsi see you check that a lap is created.. perhaps it's enough and I'm offbase21:46
nik90well I could add a check to see if the stopwatch text changes after it has been started21:46
balloonsohh right.. the point is, even with a 2 second sleep you can't be sure the stopwatch has moved21:47
balloonsbingo :-)21:47
nik90but that is already done in the the test_start_stop_reset_stopwatch() test21:47
nik90so why test it again in another test concerning laps?21:47
balloonscan you have a 0 second lap?21:47
balloonsin theory it could happen21:47
balloonsunder the current test if 2 seconds isn't enough time for the clicks to process etc21:48
nik90true21:48
nik90alrite I will add a assert statement there21:48
balloonsit's a possible point of failure due to timing..21:48
nik90to check if the stopwatch label changes21:48
balloonsnot probable I suppose, but possible21:48
balloonsyes, that's it21:48
balloonsgotta watch out for those things.. ok that's all my commentary :-)21:48
nik90regarding your online server mockup21:49
nik90I do not know how to mock that21:49
nik90but here is the thing21:49
nik90the interaction between choosingn a local city and choosing a online city is the same21:49
nik90the only difference between one is fetched online21:49
balloonsright. I am OK with having the test in there for now. If it flakes out in the dashboard we'll revisit.21:50
nik90okay21:50
balloonsPerhaps the others would disagree with me :-)21:50
nik90:-)21:50
balloonswe'll solicit feedback before merging if needed.. it;s easy to disable it21:51
nik90true21:51
nik90I will also talk to thomi about this and keep u updated21:51
nik90meanwhile I will fix the other 2 points u mentioned21:51
balloonsI put everything in the mp as a summary21:52
balloonsthanks nik90 !21:52
nik90np21:54
nik90balloons: u still there?22:27
nik90ballons: I dont think the online search can be tested. As I mentioned earlier the listview containing local results is cleared and overridden with online results. So if you look at the model.count it goes from 91(local) -> 0(after clearing) -> 5 or whatever number of online results.22:29
nik90balloons: But the thing is it is not possible to distinguish between the 91 (local) and 5 (online) results since they populate the same listmodel22:29
nik90balloons: so I am not really sure what I should be doing here to fix it :)22:30

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