13:01:03 #startmeeting Weekly Cockpit meeting 2016-07-25 13:01:03 Meeting started Mon Jul 25 13:01:03 2016 UTC. The chair is andreasn1. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:01:03 Useful Commands: #action #agreed #halp #info #idea #link #topic. 13:01:03 The meeting name has been set to 'weekly_cockpit_meeting_2016-07-25' 13:01:08 #topic agenda 13:01:30 .hello harishanand 13:01:31 harish: harishanand 'Harish Anand' 13:01:37 .hello andreasn 13:01:38 andreasn1: andreasn 'Andreas Nilsson' 13:01:44 * timers 13:02:10 * test lib / phantom functions 13:02:35 * atomic scan 13:04:26 all rigt. 13:04:31 #topic timers 13:04:52 dperpeet andreasn i have corrected the issues u have pointed out 13:04:58 https://github.com/cockpit-project/cockpit/pull/4645 13:05:01 nice, I haven't had time to look yet 13:05:06 basically simplified repeat options in timer code 13:05:18 avoided switches 13:05:23 it's shaping up really well 13:05:29 thanks dperpeet for pointing it out 13:05:38 i missed that 13:05:42 mistakes 13:06:07 words like 1st, 2nd, 3rd et cetera for dropdowns would they need to be given in _() ? 13:06:15 I could have those under _() because its rendered from mustache 13:07:21 I think andreas commented that it would be better to just have the numerals 13:07:25 i.e. 1,2,3,... 13:07:29 I think that's widely accepted 13:07:43 oh thats was our decision 13:07:50 makes translations a lot easier 13:08:00 okay then we can have it like that 13:08:13 what about Monday tuesday et cetera 13:08:14 I wasn't super firm in my decision, so I said that if 1,2,3 is easier, I'm good with having it like that 13:08:53 so lets go with that 13:09:36 day of the week needs to be translated 13:09:43 so in html with the translatable attribute 13:09:46 and _() in js 13:09:56 Monday tuesday are part of dropdowns, they too are inside mustache. I could have those under _().? 13:10:12 I think if the mustache script is part of the html, then translatable 13:10:50 dperpeet do i need to give translatable=yes inside mustache script 13:11:17 if it's in the html file, then yes 13:11:39 okay. 13:11:42 if you want to make sure, then populate via js 13:11:45 and use _() 13:12:06 try to be consistent with other mustache code on that page 13:12:09 and you'll be fine 13:12:09 i thought of doing that for days of the week 13:12:21 I have added transalable=yes for other english terms inside services.html. 13:12:58 dperpeet I haven't written the blog yet, i thought of writing it after completing tests for timer. 13:13:03 sorry about that. 13:13:16 I think you have enough to write a blog post now, right? 13:13:21 tests probably deserve their own 13:13:40 yes they are probably related to code changes 13:14:14 I will have it done this week, also the tests. 13:14:40 also andreasn suggested on changing "last trigger" to "last run" in timers page. 13:14:46 should that be part of this pr? 13:15:14 nah, that can be separate 13:15:19 different part of the UI 13:15:26 just complicates things 13:15:34 I agree with andreasn1 13:15:36 okay! i thought they all relate to same timers 13:15:53 rest i think is good 13:16:39 all right, that's all for timers? 13:16:50 yes andreasn1 13:17:04 #topic test lib / phantom functions 13:18:14 I was looking at some test stuff 13:18:36 and sometimes we use :contains in our css selectors in the integration tests 13:19:00 e.g. to wait for a "Register" button, we use "button.btn-primary:contains(Register)" 13:19:08 this needs jquery to work 13:19:24 we never noticed since pretty much all of our pages used jquery 13:19:30 until we started porting to react 13:19:55 after discussing possible workarounds with andreas, I decided to patch some of the js code we inject during testing 13:20:12 namely https://github.com/cockpit-project/cockpit/blob/master/test/common/phantom-lib.js#L15 13:20:40 I think it's worth supporting a simple version of :contains here 13:20:52 where :contains() is the last part of the selector and for plain text only 13:21:05 I feel I'm not the best judge on these kinds of things, so others probably have stronger opinions on this than me :) 13:21:17 so nothing like "div:contains(foo) a" 13:21:56 the problem is that if we use our wait_present and wait_in_text checks separately in the tests, we can get race conditions 13:22:08 this way we have a javascript selector that checks both conditions at the same time 13:22:49 as a side benefit, the code will throw a readable error if :contains is used in a way we don't support 13:23:28 and it only affects the code injected during testing, not the tests or the actual codebase 13:23:47 so most devs won't have to worry about this at all 13:24:11 should I post this a standalone pr? 13:24:20 for better discussion 13:24:26 yeah, might be best 13:24:29 ok 13:24:31 will do! 13:24:33 end of topic 13:25:53 #topic atomic scan 13:26:00 achakrab: ^ 13:26:22 okay 13:26:30 so i've continued work on the POC for now 13:26:41 so I asked around for user cases, and the feedback was a bit sparse, but I think I have enought to put together personas now 13:26:45 and workflows 13:26:47 so I'll work on that next 13:26:50 okay sure 13:27:07 once you get that through, we can have a handoff then? 13:27:30 yeah, if you and dan finds those all right I'll do mockups 13:27:30 who should i speak with about the design? 13:27:35 that would be me 13:27:41 okay cool 13:28:03 how are things on the coding side of things? 13:28:07 so you'll come up with cases and the design and once you're done, then we can probably schedule a meeting with me, you and dwalsh? 13:28:43 sure, sounds good 13:29:57 how are things going with the prototype? 13:31:55 achakrab: ? 13:35:24 open floor next I gues 13:35:26 s 13:35:30 #topic Open Floor 13:38:59 not much 13:39:06 sounds like a wrap. Thanks everyone! 13:39:25 #endmeeting