=== chihchun_afk is now known as chihchun | ||
=== JoseeAntonioR is now known as jose | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
=== chihchun is now known as chihchun_afk | ||
=== chihchun_afk is now known as chihchun | ||
elopio | ping brendand: you have flake8 errors on your config branch. | 14:55 |
---|---|---|
elopio | I like it a lot that you moved the name to an argument. | 14:55 |
brendand | elopio, odd. let me check | 14:58 |
elopio | brendand: now that you are adding commnad line arguments, I would like to see a test for that. Maybe if you move the parsing to a different function it will be easier. | 15:02 |
elopio | alesage: quick review here please, to follow the rule of two: https://code.launchpad.net/~canonical-platform-qa/ubuntu-sanity-tests/flash_in_readme/+merge/247935 | 15:08 |
alesage | elopio k | 15:40 |
paulliu | alesage: https://code.launchpad.net/~paulliu/unity8/indicator-power-autopilot-test-20150121 | 15:47 |
paulliu | alesage: I can get the Notificatiion if I use create_notification.py | 15:48 |
paulliu | alesage: But I get no Notificaiton show when I set the battery low. | 15:48 |
brendand | elopio, see my reply on https://code.launchpad.net/~brendan-donegan/ubuntu-sanity-tests/fix_wizard_test/+merge/247995 | 15:48 |
brendand | elopio, the continue action is only used for a passphrase | 15:48 |
elopio | brendand: ack. Sounds good. | 15:49 |
paulliu | alesage: I think, for http://people.canonical.com/~thomir/ueqa-code-proposals/proposals/indicator-power/index.html UI Helpers part is almost done. | 15:49 |
elopio | I'll flash and try to run it. | 15:49 |
brendand | elopio, you're sure that a test for the command line args is necessary in https://code.launchpad.net/~canonical-platform-qa/ubuntu-sanity-tests/specific_config/+merge/247841 ? it's a very simple bit of code... i could add something if you really feel strongly about it | 15:53 |
alesage | paulliu, I'll test this morning thanks | 15:58 |
paulliu | alesage: When I run test111, from the screen of the phone I saw the power-icon changed. But the Notification dialog doesn't show up. | 15:59 |
alesage | paulliu, I'll have time to have a look--when running charles' tests I def saw the notification when the battery gets low | 16:02 |
brendand | elopio, btw can you offer some feedback on https://bugs.launchpad.net/ubuntu-sanity-tests/+bug/1415877 | 16:02 |
ubot5 | Error: launchpad bug 1415877 not found | 16:02 |
elopio | brendand: yes, I feel strongly about it. I want 100% coverage. | 16:02 |
elopio | brendand: wouldn't it be solved by my optional options branch? | 16:04 |
elopio | now all options are requested, but they can be left blank. If there is a default, it will be used. | 16:04 |
brendand | elopio, if it does, then great | 16:05 |
elopio | brendand: I'm not sure. You still can use passcode and assign letter to the passphrase. | 16:05 |
elopio | avoiding that would require some validation. | 16:05 |
brendand | elopio, but the problem is that what the provided passcode/phrase is considered to be is purely dictated by that setting | 16:08 |
brendand | elopio, e.g. 1234 can be either a passcode or passphrase | 16:08 |
brendand | elopio, so you can't really validate it | 16:08 |
brendand | elopio, i think i have a plan for the command line test | 16:12 |
=== chihchun is now known as chihchun_afk | ||
elopio | brendand: we would need the validation to happen after all the config has been read. | 16:13 |
brendand | elopio, what would you validate then? that the password matches with the security method given? | 16:14 |
elopio | brendand: cool about the tests. You can look at autopilot for ideas too. | 16:14 |
elopio | brendand: that if security is passcode, password has to be four digits integeeer. | 16:15 |
brendand | elopio, that would be fine with passcode, you can check that it's a 4 digit string | 16:15 |
brendand | elopio, but for passphrase you can't | 16:15 |
elopio | brendand: for passphrase, any value is accepted, what would you validate there? | 16:15 |
brendand | elopio, right you can't validate anything | 16:16 |
elopio | brendand: I think that the welcome wizard card also involved checking it against the test docs in moztrap. Can you please check if we are following all the same steps? | 16:32 |
brendand | elopio, actually the moztrap test specifies only to use a passcode | 16:32 |
brendand | elopio, oh you mean in general | 16:32 |
brendand | elopio, actually there is one thing | 16:33 |
elopio | brendand: yes, to make sure we are doing the same thing in each page. | 16:33 |
brendand | elopio, we don't go through the wifi wizard properly | 16:33 |
brendand | elopio, since the connection already exists the wpa key is never asked for | 16:33 |
brendand | elopio, so we need to delete the connection before running the test | 16:33 |
brendand | elopio, we can simply rm the connection file | 16:34 |
elopio | brendand: ok. Should we do it as part of this card, or make another one for that? | 16:34 |
brendand | elopio, i'll make a bug for that - i was focused on fixing the error | 16:34 |
brendand | elopio, another card, yes i think so | 16:35 |
alesage | elopio I'm reviewing optional_options fyi | 16:35 |
elopio | brendand: ok, please make the card and put an estimate in ideal hours to it. | 16:35 |
elopio | alesage: thanks. | 16:36 |
brendand | elopio, added in TODO. size 5 | 16:42 |
elopio | brendand: thanks. | 16:50 |
elopio | brendand: ah, and there's one more thing to annotate about that card. Since we are in developer mode, the security screen is actually doing nothing. | 17:01 |
brendand | elopio, in that it's not actually changing the password/phrase. yes | 17:01 |
brendand | elopio, i doubt there's anything we can do about that though | 17:02 |
elopio | brendand: yes. Maybe, make an annotation on the moztrap doc saying that it is automated but the security still needs to be manually checked. | 17:02 |
elopio | ? | 17:03 |
brendand | elopio, well that's not in the sanity tests ;) | 17:04 |
elopio | You are right there. It just checks that the screens can be filled and accepted. | 17:04 |
elopio | so, we are actually automating it all :) | 17:05 |
brendand | elopio, i'm going offline til the standup | 17:05 |
brendand | elopio, see you then | 17:05 |
elopio | bye | 17:05 |
elopio | enjoy life. | 17:05 |
Letozaf_ | balloons, hi | 20:15 |
balloons | Letozaf_, hello | 20:18 |
Letozaf_ | balloons, I had a chat with Roman on that shorts app flaky test, I am trying something to see if I can fix it | 20:19 |
balloons | Letozaf_, excellent | 20:20 |
Letozaf_ | balloons, wait... it will be excellent if it works :-P | 20:20 |
balloons | :p | 20:20 |
Letozaf_ | balloons, looks like it passed, this time, do you want to see if you can break it ? | 20:24 |
dobey_ | is it possible to use the phone emulator as the virt-server for adt-run? balloons ? | 20:25 |
balloons | Letozaf_, I will | 20:25 |
Letozaf_ | balloons, :) thanks | 20:26 |
balloons | dobey_, I do the opposite. Start the emulator, and use adb | 20:26 |
dobey_ | balloons: ah, was hoping one could do adt-run --- xvfb-run ubuntu-emulator devel-proposed or something | 20:27 |
balloons | Letozaf_, I queued up 4 runs | 20:32 |
balloons | also running locally on my desktop | 20:33 |
Letozaf_ | balloons, awesome! thanks. | 20:33 |
Letozaf_ | balloons, I am also running the tests several times in sandbox on my desktop | 20:33 |
balloons | first glance, I think it's fine ;-) | 20:34 |
balloons | Letozaf_, lol, jenkins failed it | 20:35 |
balloons | Letozaf_, I too am getting an occasional failure perhaps on test_edit_topic | 20:36 |
Letozaf_ | balloons, :'( | 20:36 |
balloons | I guess it might be worth actually looking closer at what is going on | 20:36 |
balloons | it's significantly better. I was getting several errors every time, now it usually passes | 20:37 |
Letozaf_ | balloons, yeah, let me take a look :) | 20:38 |
balloons | Letozaf_, so it seems it's failing in the assert, which is probably an easy fix. We should see how we actually want to check the edit topic worked or not | 20:39 |
Letozaf_ | balloons, if you look at this video:http://91.189.93.70:8080/job/generic-mediumtests-vivid/873/artifact/shorts_app.tests.test_rssreader.TestMainWindow.test_edit_topic.ogv | 20:40 |
Letozaf_ | balloons, you will se that the topic is not changed, it does not click in the right place | 20:40 |
Letozaf_ | balloons, looks like it's unable to get "LabelVisual", text="Ubuntu" | 20:40 |
Letozaf_ | balloons, in the valueselector... here def change_feeds_topic(self, name): in __init__.py | 20:41 |
balloons | I assume there's a sdk helper we could be using | 20:41 |
Letozaf_ | balloons, hah it passed now :P | 20:42 |
Letozaf_ | balloons, well at least, some times it passes :P | 20:42 |
* balloons still playing around locally | 20:43 | |
balloons | Letozaf_, yes it's very very close | 20:45 |
balloons | Letozaf_, I made some simplification tweaks; lp:~nskaggs/ubuntu-rssreader-app/fixing-flaky-test | 21:03 |
balloons | still running locally to play with it | 21:03 |
* Letozaf_ is looking at balloons' tweaks | 21:04 | |
balloons | trying to get a failure with some more debugging output | 21:04 |
balloons | Letozaf_, so basically sometimes change_feeds_topic fails | 21:14 |
Letozaf_ | balloons, yeah looks like soemtimes it fails to get the LabelVisual object to click | 21:14 |
Letozaf_ | balloons, it sometimes clicks on the valueselector object instead of the labelvisual one, in the video it looks like that | 21:15 |
balloons | well Letozaf_ I think it's simply because the location is wrong.. autopilot has the unexpanded location and clicks it instead | 21:16 |
balloons | Letozaf_, care to merge my changes? | 21:19 |
Letozaf_ | balloons, ok I will | 21:19 |
Letozaf_ | balloons, done | 21:21 |
balloons | I have an idea for the other timing issue I hope | 21:21 |
balloons | going to try and add a wait_for on newTopic inside of change_feeds_topic | 21:22 |
Letozaf_ | balloons, ok | 21:22 |
balloons | http://paste.ubuntu.com/9943397/ | 21:22 |
balloons | trying to break it now | 21:23 |
Letozaf_ | balloons, :) | 21:23 |
Letozaf_ | balloons, hope it works | 21:24 |
balloons | no luck breaking it.. fingers crossed | 21:24 |
* Letozaf_ crossed fingers | 21:24 | |
balloons | I say ship it ;-) | 21:24 |
Letozaf_ | balloons, :) | 21:24 |
balloons | Letozaf_, so what is happening is a a timing issue | 21:25 |
Letozaf_ | balloons, yeah looks like | 21:25 |
balloons | the list can be expanded but listitem isn't yet visible. As such, autopilot has the coordinates for the parent object, which is the label | 21:25 |
balloons | make sense? | 21:25 |
Letozaf_ | balloons, oooh!!! yes makes sense | 21:26 |
balloons | 10 runs all passed | 21:26 |
Letozaf_ | balloons, yay! | 21:27 |
balloons | file that one into your brain if you see it again. If autopilot is clicking in the wrong place, it's probably clicking the parent object because the object you are trying to click isn't visible or it's coordinates are changing | 21:27 |
Letozaf_ | balloons, yes thanks I definitely will | 21:27 |
balloons | you got most of it, adding the wait_for on expanded | 21:29 |
balloons | I wouldn't have thought to have added the visible.wait_for also | 21:29 |
Letozaf_ | balloons, yes but you fixed the last bit :) thanks for your help | 21:29 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!