[00:30] ok https://code.launchpad.net/~nskaggs/autopilot/page-object-docs/+merge/246504 is ready now as well [00:30] * veebers looks [00:32] balloons: Can you top approve this if you like (you're the 2nd reviewer): https://code.launchpad.net/~veebers/autopilot/adding-run-instructions-readme/+merge/247088 [00:34] veebers, I'd love too, but I have comments to change it . . [00:35] balloons: heh, even better :-) [00:36] done [00:36] thanks, will address soon [00:37] oops, I realized I lied veebers about docs/tutorial/running_ap.rst.. ignore that comment [00:39] balloons: hey, I get an error when building the docs of yours, also, not sure where to find the link to the content in the generated docs [00:39] veebers, ohh which docs error? and I'll find the link one sec [00:40] balloons: this is the error I get: ../page-object-docs/docs/tutorial/advanced_autopilot.rst:427: WARNING: Inline literal start-string without end-string. [00:42] balloons: nvm I found the link to the page, my bad [00:42] ok ;-) [00:42] gets me wondering if it should be more apparent. /me poinders [00:42] ponders even [00:44] veebers, weird I get no errors running my branch [00:46] balloons: hmm odd, this is what's I'm using to build: python3 setup.py build_sphinx -v -b html [00:46] I can see the error in the doc, not sure where it's come from though [00:46] veebers, ahh I get it now on a fresh checkout.. interesting [00:47] balloons: that's odd as it's not something you changed :-\ [00:47] * balloons is just running sphinx-build -b html docs htmldocs [00:47] indeed [00:48] I do see the error though.. simple fix [00:48] shall I include it? [00:49] pushed 525 with the fix [00:50] sweet cheers [00:54] veebers, "Text conflict in docs/tutorial/advanced_autopilot.rst" [00:54] wtf? [00:55] perhaps I should re-merge trunk' [00:55] balloons: ah right, thomi mentioned this morning that he's made changes to trunk that will need merging (whitespace changes) [00:55] also, have reviewed branch, NF [00:56] ty [00:59] balloons: hey it seems I've missed something completely (in your review) you say that link will die soon, where will it live after that? I thought there was just discussion as to the best way to get the docs in the right place [01:00] veebers, basically the static html as it is cannot live on @ developer.ubuntu.com. It will be integrated, but not hosted as-is [01:00] balloons: also, as this is in the README it's not generated into html (or anything else) so there has to be a link to something there (as it remains just text) [01:00] veebers, lol, long day.. right [01:00] you could just say refer to the docs then.. bad idea to put a link like that into a readme which will break and we'll never update [01:01] balloons: right, that's what you and thomis were talking about this morning. The documentation will still be available on dev.u.com right? /me feels he needs to re-read the backlog [01:02] veebers, yes it will be there.. Ideally I'd like it to be there as-is, but as it stands that's not in the plan. The plan is to import it as json and redisplay it [01:03] balloons: ugh right. So it sucks that we can't put in the projects own readme the url to where the docs live. I can't just say "Check the docs" in the readme, how would a dev find that if it's not there? (I'm ranting a little, I know) [01:05] veebers, yea I know. You could offer them the file based location [01:07] balloons: right I could but that feels a little dumb. I want them to see the good docs with it's links etc. I could always suggest building the docs and reading from there, but again it seems silly to do that if they exist on the web somewhere [01:09] balloons: so one thing I don't understand is, how come the link to the docs (on dev.u.com) will die or be unknown? [01:11] veebers, on the self.wait_select_single, what would you like to see? you are confused as to what 'object' is? [01:12] balloons: well, object doesn't contain the method wait_select_wingle, nor do I see it defined in that snippet (or implied) that it's implemented there [01:13] veebers, the url is currently on a server I have no control over and can't/won't be updated. It will eventually migrate into d.u.c API pages.. that's the plan at least, and it's how AP must go in if it's to live on d.u.c [01:14] I think it's ok as it's an advanced tutorial, and the source is given.. however, i'm happy to tweak it, I simply don't know how [01:14] balloons: but surely the url will be known right? Or are you stating that the current url and docs will be deleted before the migration? [01:14] balloons: I think a simple '. . .' above that code would be fine (considering it's an advanced tut.) much like line . . . [01:14] veebers, that link won't be deleted until it's done.. I may leave it there for some time after even; who knows what we'll do with eventually. But it's a horrible link to want anyway, imho [01:15] I'd rather see d.u.c/autopilot [01:15] or something easy and sensible [01:15] diffline 162 [01:16] balloons: right that makes sense. So we can update the readme once the migration happens right? It means up to that point the docs are available (and you can learn the location from the readme) and then when the url changes we'll update the readme [01:16] veebers, ok, hopefully everything is as you wanted now in that mp [01:16] * veebers looks [01:17] * veebers builds docs, looking good so far [01:17] veebers, perhaps a custom url shortner? [01:17] I'm ok with you doing it, it's your project so far be it from me to stop you, but I wanted you to be aware [01:19] balloons: sweet, bottom approved that MP of yours [01:20] balloons: I appreciate the heads up, it's just a little annoying that it seems hard to have the docs hosted somewhere (that should be an easy problem, right?) [01:20] ok, so one more for tomorrow [01:20] I think at this stage we'll use the existing url, then when the migration is finished and the autopilot docs are available there we'll update the url then (as it won't have broken inbetween) [01:20] veebers, I agree it should be easy which is why I was pushing / asking if you wanted to do it yourselves somewhere.. imho, I'd rather see autopilot have it's own site, with docs, intro, etc [01:21] balloons: while that might be nice, it'll also introduce further maintenance issues than what we have. I'm not saying it will never happen, it's just not top priority at the moment [01:21] yea, I know.. given the headaches though.. I'd rather own the silly thing [01:22] Your 4 layers removed, but I'm still 3 [01:22] *you are [01:23] ok veebers, that's it for me and mp's.. will you try to actually land them today? [01:24] I have the final big one I need to land and having the rest in makes it easier :-) [01:24] * balloons notes he is rather unlikely to be ready tomorrow anyway so :-) [01:24] balloons: thomi tells me that there are issues at the moment with landing. But yes the intention it to get them merged into trunk. [01:24] kk.. enjoy and thanks! [01:24] balloons: I'm pretty sure the bottom approved stuff is/will be top approved so should get merged [01:25] nw, and thanks to you too ^_^ [01:25] veebers, we will be in touch on the docs thing.. please bug me on it, so I can do the same to those making site decisions :-) [01:25] balloons: don't worry, I will ;-) [11:34] I'm seeing a failure in u-s-s AP tests on mako where a tap is interpreted as a long-tap [11:35] … I think. A list element is highlighted then, after a couple of seconds, de-highlighted. It's very intermittent. [11:36] Test is ubuntu_system_settings.tests.test_datetime.TimeDateTestCase.test_manual_tz_selection, and we're using scroll_to_and_click === barry` is now known as barry [17:59] balloons, hey :-) [17:59] howdy rpadovani [17:59] balloons, do you remember the problem I talked to you about? I'm not able to debug it, so I setup a branch for you, if you want :-) [18:00] balloons, https://code.launchpad.net/~rpadovani/ubuntu-calculator-app/bigNumber150122 [18:00] balloons, every time you run autopilot, there are random failures [18:00] rpadovani, I remember you telling me to ignore it. I suppose you no longer want me to ignore it :-) [18:00] thanks for the branch [18:00] balloons, that is :D [18:00] rpadovani, so is there a specific test that fails or ? [18:01] balloons, that's the problem, it changes at every run [18:01] balloons, 2 are the same, but are about precisions, [18:01] balloons, I suppose (but it's only a supposition) that it takes more time to load the library when it uses bignumber, so it doesn't have time to load it [18:04] rpadovani, if there's a good test to run over and over, that's what I was after [18:05] balloons, I didn't find it, every time I try to run only the test failed in the previous run, it works well [18:05] balloons, the last one failed on my pc is ubuntu_calculator_app.tests.test_main.MainTestCase.test_small_numbers [18:15] rpadovani, I'll run it a few times and see what happens. Might be a bit before I can have a look, otp [18:15] balloons, thank you! [18:15] yw.. I do love a good mystery [18:19] rpadovani, first glance it definitely looks like keypresses aren't being recieved sometimes [18:19] the fix for that likely didn't work [18:20] balloons, do you have in the log [18:20] Error: math.jsTypeError: Cannot call method 'eval' of undefined engine formula:0.000000001+1 [18:20] ? [18:23] rpadovani, yes it's gathering the logs as it runs [18:24] balloons, so I think it's the problem, the var that manages the engine isn't set [18:24] ahh.. if you think it's a timing thing, add a sleep in the setup step and see what happens [18:38] rpadovani, does that cause the autopilot.exceptions.StateNotFoundError: Object not found with name 'QQuickMouseArea' and properties {'id': 41} error? [18:38] balloons, ehm , don't know? [18:38] seems not [18:52] http://paste.ubuntu.com/9823350/