[00:07] http://paste.ubuntu.com/688727/ [00:07] it's working [00:08] I just need to add the tests for the multiple-sprint bit, and the fallback, with some mocking [00:10] but first I must eat [00:31] james_w: nigelb: I've got a fix for periods in url lookups, complete with testcase [00:31] yay [00:31] \o/ [00:31] where are we currently converting . to - ? [00:32] On that note, I should sleep. [00:45] james_w: nigelb: https://code.launchpad.net/~mhall119/summit/fix-period-in-url/+merge/75284 [00:47] mhall119, nice, thanks. [00:47] Approved [01:24] james_w: nigelb: cjohnston: http://ec2-50-16-76-22.compute-1.amazonaws.com/uds-p/ has been updated to the latest trunk [02:10] deployed the cache fixes to production (via the production branch) [02:10] so that we don't get hit by any dataloss from that [02:28] james_w: can you make sure that goes into /stable and trunk too please? [02:28] it went to stable [02:28] ok, I'll merge stable to trunk [02:28] you would like me to merge stable to trunk too? [02:28] ok, thanks [02:30] done [02:32] lpupdate is only pulling in 2 linaro sessions, is that all there is? [02:33] meeting linaro-summits-server [02:33] meeting linaro-summits-stb [02:37] that's all there is so far [02:37] we're blocked on the LP side currently, being able to approve the blueprints for UDS === newz2000` is now known as newz2000 [14:26] ping newz2000, are you there? [14:27] jawud: yes [14:33] On the following lijnk is an idea for ubuntu donations. It;s a Dutch forum but the proposal is English. Can you give me some feedback about the proposal and tell me your opinion? [14:33] http://forum.ubuntu-nl.org/algemeen-42/voorstel-voor-ubuntu-donaties/ [14:34] no hurry, I understand when you're bussy [14:52] nigelb: https://code.launchpad.net/~mhall119/summit/777171-percentage-problem-tests/+merge/75360 [14:53] nigelb: your branch was so out of date I was having trouble getting these tests into it, which is why it's proposed to trunk instead [14:57] jawud: I would suggest proposing this to Jorge Castro, aka jcastro. Do you know of him? [14:58] I've heard of him. Thanks, I will contact him. [14:58] what do you personally think of it? [15:00] jawud: I think it's interesting. I fully support ways to help non-coders make a meaningful contribution. [15:02] oke thanks! [15:25] mhall119: Yeah, I did it just after UDS :-) [15:26] nigelb: I know, not complaining, just telling you why I targetting trunk [15:26] mhall119: NP :) [15:26] I'm officially sick and been asked to rest for 2 days. [15:28] but you have a laptop right [15:28] j/k, take the time to get better [15:29] Heh. I think I want to sleep and get up on time for achacne [15:29] *change [19:26] Hey guys [19:27] is there a simple way to get the ubuntu website theme make of it one for kubuntu (graphics) and put text and stuff kubuntu related? [19:30] does tarmac not run the tests for summit? [19:31] sheytan: sure, you can branch one of the light-*-theme branches and convert it to kubuntu [19:31] james_w: I don't think so, no [19:31] mhall119: where do i get them? :) [19:31] sheytan: https://launchpad.net/ubuntu-community-webthemes [19:33] james_w: did you want to get the linaro customizations working on my ec2 box for testing? [19:33] mhall119, I'm just working on generic stuff currently, but that would be a good idea [19:38] https://code.launchpad.net/~james-w/summit/fix-period-tests/+merge/75411 makes all the tests work on trunk again for me [19:43] james_w: that's already done in https://code.launchpad.net/~mhall119/summit/777171-percentage-problem-tests/+merge/75360 [19:43] ah, ok then [19:43] james_w: why not land that MP? [19:44] mhall119, I assumed that tarmac wouldn't as the tests would fail [19:44] ah, ok [19:44] also I like a clean testsuite [19:44] as far as I know tarmac isn't running tests [19:44] I like a clean testsuite too, but only when it means things aren't broken [19:44] but with those two changes tied together we won't have one either way :-) [19:44] Thanks, [19:44] oops [19:44] yeah, but you are always going to have bugs [19:45] and they don't show up as test failures without writing a test [19:45] true, but we know there are bugs in trunk right now [19:45] these test cases just highlight that [19:45] flicked the switch so that it will land now [19:46] well, I could have told you there were bugs in trunk :-) [19:46] I'll try and update nigel's branch to where it A) passes these tests and B) can be merged into trunk [19:46] I can do that now [19:46] james_w: if you know of other bugs, write a test case for them [19:46] I've just finished https://code.launchpad.net/~james-w/summit/multiple-sprints/+merge/75413 [19:46] if you have time, that'd be great [19:47] the problem with nigel's code is that he's calling escape_strings() on things that don't need escaping [19:47] ah, that one [19:47] I have a fairly good idea of what needs doing on that, if you'd rather wait until I can get to id [19:48] I'll take a look, but may defer [19:48] having tests is useful for this :-) [19:56] yes [19:57] up to 26 tests now [19:57] much better than the 0 we had last cycle [20:04] hahaha [20:04] this is broken [20:04] I'm not sure where the url escaping should go [20:05] I don't think that django supports % in urls [20:05] well, it might be able to serve those urls [20:05] but it appears you can't put a link to one in a page [20:09] why not? [20:12] well, I'm a bit hampered in debugging as the exception when serving the page in the test is resulting in traceback page served with a 200 response code! [20:13] ah, no [20:13] something catches tracebacks when rendering a meeting div? [20:13] it's probably because of settings.DEBUG=True [20:14] are you talking about the % escaping tests? [20:14] I'm looking at test_percent_in_meeting_name currently [20:30] why does it need to do that [20:30] why not do that substitution when the %(style)s is added to the html? [20:33] because it puts the content of html in cache, but evidently the styles may change before the cache becomes invalid [20:34] so this lets it pull the html out of cache and then apply the style [20:36] ah [20:37] these tests seem non-deterministic [20:38] I guess non-specified fields get filled with random data by model_mommy? [20:39] yes [20:40] there is one non-deterministic test, but the others should be [20:40] one that is fixed is non-deterministically failing due to not all of them being fixed [20:41] which? [20:41] the test methods should all be isolated from eachother [20:45] it's because of the errors [20:46] so you fix one, and depending on whether or not a % appears somewhere else the test passes or fails [20:46] depending on whether the page gets a traceback or the div in ti [20:46] it [20:49] it sounds like there's some scenario my tests don't cover then [20:49] which also happens to not be escaped properly [20:54] yep, I'm adding some more tests now [20:56] I gotta have more test case [20:59] is there a reason these tests aren't calling meeting_div directly? [21:02] mostly because I wanted to check the entire page's output [21:03] also, to call meeting_div you need to build the Schedule object and initialized it properly [21:03] it was just easier to call the URL and check the final results [21:16] https://code.launchpad.net/~james-w/summit/777171-percentage-problem/+merge/75429 [21:21] thanks james_w, I'll check it out this evening