12:01:08 #startmeeting Horizon 12:01:09 Meeting started Wed Aug 5 12:01:08 2015 UTC and is due to finish in 60 minutes. The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:01:10 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 12:01:12 The meeting name has been set to 'horizon' 12:01:35 anyone here to talk Horizon? 12:01:50 mostly to listen ) 12:02:07 Hi 12:02:17 hey o/ 12:02:42 I was going to talk about the domain stuff but I picked your brain about it earlier :-) 12:03:05 o/ 12:04:31 General items first and then chaos 12:04:40 as no one added agenda items 12:04:53 L-2 was tagged last week 12:05:10 https://launchpad.net/horizon/liberty/liberty-2 12:06:01 our pace on blueprints is down from Kilo, not sure the reason why 12:06:25 so much refactoring 12:06:35 maybe thicker blueprints 12:06:44 I think we've finally finished the eslint twiddling tho, so there's that 12:06:45 and refactoring 12:06:53 \o/ 12:07:09 just a few of those still to be merged, but I believe the substantive work is all done 12:07:18 Now we can have some content? :p 12:07:20 (if robcresswell was looking for something to review... ;) 12:07:26 On it. 12:07:33 haha, I'm sure we can find some more files to rename first! 12:07:42 I have a feeling L-3 is going to be frenzied 12:07:54 * tsufiev votes for topology stuff 12:08:06 because I'm an idiot, I'm focusing on getting our damned selenium suite fixed :/ 12:08:49 tsufiev: curvature is getting some final tweaks before merging, I think we're close though 12:09:28 #topic Review Liberty Priorities 12:09:43 #link https://etherpad.openstack.org/p/YVR-horizon-liberty-priorities 12:10:28 so a lot of the tech debt items are far along 12:11:46 for plugins, work continues. I have a patch up to make horizon load content only via enabled files 12:12:03 the mixed model made testing difficult 12:12:19 I saw that. Interesting patch. 12:12:19 lhcheng is working on moving trove 12:13:08 and plugin docs is still a WIP on my end 12:13:51 session cleanup has a patch proposed that needs review 12:14:01 have any link for plugin docs? 12:14:14 and the eslint work is close to completion, finally 12:14:31 masco: only local copy will push soon for more eyes 12:14:48 david-lyle, ok no problem. 12:15:08 Django 1.8 support is turning into an onion 12:15:16 yes 12:15:21 stinky? 12:15:28 and lhcheng is peeling away layers 12:15:41 building on work started by mrunge 12:16:12 we need to land that in time to roll g-r to include 1.8 12:16:15 hope it doesn't make him cry ) 12:16:38 so we can't let that one go to the very end as g-r freezes about 2 weeks before L-3 12:16:44 david-lyle, it's one of my highest prios right now 12:16:52 excellent 12:17:23 I will also give a plug for trove and sahara reviews 12:17:42 #link https://etherpad.openstack.org/p/sahara-reviews-in-horizon 12:18:09 #link https://etherpad.openstack.org/p/trove-reviews-in-horizon 12:18:49 for both trove and sahara, I consider +1s from the service team and 1 +2 sufficient to approve 12:19:18 so horizon cores, if there are positive reviews from the team and looks good to you, go ahead and approve 12:19:42 got you 12:19:47 Got it 12:19:49 they have a lot of work there that's struggling to get into Horizon 12:20:15 eventually that will be part of the documentation :) 12:20:37 Last general update 12:20:50 Domain scoped tokens in the sessions are very close 12:21:17 one last bit of logic encapsulation and the d-o-a patch will be ready 12:21:30 great!! 12:21:33 and then there's a fairly good sized Horizon patch to consider 12:21:39 superb 12:23:08 there has been active ongoing discussion which you can track here: 12:23:14 #llink https://etherpad.openstack.org/p/domains_planning 12:23:20 #link https://etherpad.openstack.org/p/domains_planning 12:23:25 less l this time 12:23:47 that's the priorities wrap up 12:23:56 I think we're actually pretty close on most 12:24:19 oh and I will be releasing a new release for d-o-a in kilo soon 12:24:42 and d-o-a for liberty as soon as domain support goes in 12:25:05 would like to have 1.8 support in that release as well, but I'll need to monitor the timing 12:25:14 #topic Open Discussion 12:25:42 david-lyle, what do you think about the chances to backport WEBROOT fix into Kilo? 12:26:02 tsufiev, do you have a link? 12:26:10 mrunge, I've started it here: https://review.openstack.org/#/c/209108/ 12:26:24 still WIP, I hope to make a final version in few days 12:26:32 I'm worried with all the moving files it won't be clean, but I haven't looked at your patch yet 12:26:44 tsufiev, it fixes broken behavior. 12:26:58 but as david mentioned, lot's of files moving around 12:27:02 mrunge, okay, I take it as chances are good enough ) 12:27:31 that files moved just makes it more difficult, but I think it makes sense to backport 12:27:40 yes, same here 12:27:59 launch instance is broken in kilo, when not using / as webroot 12:28:17 oh I have a patch to devstack to move it by default 12:28:19 glance metadata is broken in kilo, too 12:28:28 * david-lyle shamelessly self-promoting 12:28:40 link? or it didn't happen... :D 12:28:46 mrunge, yeah, we encountered all these issues in our distro Horizon 12:28:58 #link https://review.openstack.org/#/c/206263/ 12:29:22 things keep moving under me and I need to fix up the offline compression 12:29:26 tsufiev, I believe you. that's the reason we're still with old launch instance 12:29:49 once I have that working again I think it will be ready 12:30:03 great, thank you david-lyle 12:30:27 but that will help us prevent future breakages 12:30:38 yes! 12:30:41 or at least get quicker feedback 12:30:51 and maybe we should change location from time to time? 12:30:58 it also fixes the fact that keystone now stomps on our users page URL 12:31:21 I mean, move from /dashboard to /horizon to / to .... 12:31:22 hurgleburgler has a patch to make that easier 12:31:55 #link https://review.openstack.org/#/c/206592/ 12:32:09 david-lyle, btw, I suspect that the problem with OFFLINE_COMPRESS on devstack may be caused by CUSTOM_THEME_PATH, so Diana's patch may fix it 12:32:16 I've been fighting other first so I need to review it 12:32:38 tsufiev: it was working before the patch to the themes changed 12:33:33 but yes it should help a great deal 12:34:13 oh yeah, forgot to mention, robcresswell excised the router dashboard from the horizon tree 12:34:31 and moved it to its own repo 12:34:51 seems like that should've been a blueprint 12:38:57 anyone have anything else? 12:39:25 david-lyle: It was a blueprint 12:39:39 david-lyle: Just a partial one; still want to remove the rest of the Cisco code 12:39:42 uhm, do we need a more regular blueprint review process? 12:40:07 mrunge: most likely :) 12:40:19 I mean, reviewing blueprints is a question coming up from contributors 12:40:55 current status is a bit undefined, leaving ALL blueprints in a semi-accepted state 12:41:50 So that's one thing I've actually spent very little time on since being core, admittedly. Mainly done bug cleanup and reviews. 12:42:00 How much discussion/ group acceptance does a blueprint need? 12:42:33 Can a core approve it? Or do they all need to be brought up on mailer/ meeting, which is what I tend to advise people who ask. 12:43:04 so for Kilo I tried to get more reviews from across the horizon team and horizon drivers, that was short lived 12:43:16 robcresswell: any core can approve a BP 12:43:36 back in the old days(R), you proposed a blueprint, and submitted code for it. once code was approved, bp was approved 12:43:37 some need more discussion than others, but that's a judgement call 12:44:34 there is a whiteboard field in blueprints. just write down your thoughts on that bp 12:44:55 At the moment, aside from the bigger BPs, most seem to be left and just used as tracking for bigger features. 12:45:12 Yeah, will put some more time into BP review 12:45:45 a couple of things from my point of view 12:46:05 I don't like a heavy handed process around adding new feature support 12:46:21 I do think architectural changes make sense to review in BP 12:46:48 but we don't have a stable API to maintain, and that makes our life easier, I think 12:47:14 and if you don't have code for your proposal, it's not worth a whole lot yet 12:47:54 but I understand that with other services without approval, you shouldn't even start, so that makes it confusing 12:48:02 so we have to start work on the BP before it get approve? 12:48:24 masco: no, but we have 400+ bps 12:48:40 many of those are approved and never moved anywhere 12:48:47 Sounds reasonable, I'll keep those points in mind 12:48:58 so getting the bp approved doesn't move us very far without the code 12:49:16 ok got it 12:49:31 the other option is we could schedule something to regularly review bps 12:49:39 and try to stay on top of it 12:49:49 makes sense 12:50:15 but I really dislike having a huge bundle of inactive approved bps 12:50:29 if it is in top it will get attention 12:50:51 we could try to clean up old blueprints 12:51:01 I think, there are still many duplicates 12:51:10 masco some times it just takes a ping to say, "no, I'm actually serious about this one" 12:51:10 by doing some activity the submitter can bring in to top 12:51:48 mrunge: I cleaned some up 12:51:56 not near enough 12:52:03 I'd be up for that, scheduling something to run through them 12:52:11 problem is I start going through them and think, we really should do this 12:52:17 :) 12:52:43 you have think that before approve ;) 12:52:54 same here david-lyle , still there are too many. and most of them are really good 12:53:00 well some are from 2013 12:53:06 and we still haven't done them 12:53:15 but we should 12:53:45 * david-lyle mumbles about linting and bit twiddling 12:54:54 limited sample, but who's in favor of regularly reviewing in a meeting setting 12:55:13 sounds good to me. timezone permitting ;) 12:55:15 #startvote review bps in regular meeting ? yes, no 12:55:16 Begin voting on: review bps in regular meeting ? Valid vote options are yes, no. 12:55:17 Vote using '#vote OPTION'. Only your last vote counts. 12:55:24 #vote yes 12:55:25 #vote yes 12:55:31 #vote yes 12:55:32 #vote yes 12:55:34 #vote yes 12:55:56 #vote yes 12:56:00 :D 12:56:05 #endvote 12:56:06 Voted on "review bps in regular meeting ?" Results are 12:56:07 yes (6): robcresswell, mrunge, masco, r1chardj0n3s, david-lyle, tsufiev 12:56:12 so no then 12:56:15 :P 12:56:22 (having the regular meetings really helped some kilo stuff along, I see no reason why it couldn't also help other things along) 12:56:43 ok, I will schedule a time 12:56:50 awesome 12:56:58 do we want alternating TZs I suppose? 12:57:14 really the only way any more 12:57:40 mmm 12:57:46 alternate like our weekly meeting, but the other way around? 12:58:06 like: today, 20:00 UTC a bp meeting? 12:58:14 sure, we could that 12:58:24 just a thought 12:58:47 not sure it is the brightest idea ever, since sliced bread 12:58:54 * robcresswell foresees confusion 12:59:14 * mrunge too, but you just need to remember: horizon meeting 12:59:17 review meeting in alternating tzs then 12:59:20 I like it 12:59:35 either bp review or weekly meeting 12:59:47 let's start next week, so I have time to send an email and get people informed 13:00:33 and we have to come with a list of BPs to review on that day either via wiki link or eitherpad 13:00:33 cool 13:00:43 I will send said email 13:00:53 and we're out of time 13:01:00 * r1chardj0n3s heads off to bed 13:01:02 o/ all 13:01:12 #action schedule/email about bp review meeting 13:01:16 good night r1chardj0n3s 13:01:17 #undo 13:01:18 Removing item from minutes: 13:01:34 #action david-lyle schedule/email about bp review meeting 13:01:43 Thanks everyone 13:01:49 #endmeeting