08:00:06 #startmeeting horizondrivers 08:00:07 Meeting started Wed May 25 08:00:06 2016 UTC and is due to finish in 60 minutes. The chair is robcresswell. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:00:08 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 08:00:12 The meeting name has been set to 'horizondrivers' 08:00:25 o/ 08:00:31 o/ 08:00:36 a quiet morning meeting 08:00:50 Usually is :) 08:00:55 a quiet evening meeting ;-) 08:01:01 [I'm also cooking dinner] 08:01:09 I do have a BP I'd like to run past you though 08:01:33 Sure. Just want to highlight the midcycle for anyone checking minutes 08:02:05 #info Remember to register for the Horizon midcycle! Everyone is welcome to attend 08:02:12 #link https://wiki.openstack.org/wiki/Sprints/HorizonNewtonSprint 08:02:21 * tsufiev still getting travel approval 08:02:28 #topic Blueprint Review 08:02:36 Fire away r1chardj0n3s 08:02:44 so it turns out I don't have a BP for you because it's already been approved ;-) 08:02:51 \o/ 08:02:52 I didn't do my pre-meeting homework, sorry 08:02:55 Which one was it? 08:03:01 https://blueprints.launchpad.net/horizon/+spec/admin-views-filter-first 08:03:08 got some folks working away on it 08:03:22 Yep, I instantly approved those due to previous discussion 08:03:26 didnt need to revisit it. 08:03:28 david's been working with them on it 08:04:12 Thats great. I'd like to continue refining the Django views while we move bits and pieces to angular. 08:04:23 I have one item on my BP agenda 08:04:31 Sure, go tsufiev 08:04:42 https://blueprints.launchpad.net/horizon/+spec/horizon-glance-large-image-upload was approved some time ago, but now it seems that its scope has changed 08:04:52 I have both Django and Angular implementation 08:05:33 and I would like to get Django implementation merged as well, because it could be backported to mitaka release, contrary to Angular implementation 08:05:52 so a Django implementation ... the opening line "bypassing Django" isn't really accurate any more? :-) 08:06:06 Can't backport a feature. Unless you meant locally for MOS etc 08:06:10 well... it is bypassing part of Django 08:06:29 robcresswell, yeah, I meant downstream backports 08:06:29 right 08:07:25 well, downstream backporting is not something that upstream would be really interested in, anyways it _might_ be interested in having this feature for Django - as long as Django is still supported 08:07:48 I don't have a huge issue with using the same blueprint. Its not overloaded (5/6 patches?) and it provides a good "story" for how the code moved forward. 08:08:00 Might make sense to update the wording in it though. 08:08:27 sounds good to me too 08:08:27 robcresswell, okay, I'll update the description section there 08:08:32 We're not strict on specs like the other projects. As long as the feature concept it sensible and the patches are tracked, we don't need to spend hours refining blueprints IMO. 08:08:40 concept is* 08:08:54 Thanks tsufiev 08:09:09 btw, r1chardj0n3s, I found out how to use the same approach for Swift upload 08:09:17 I mean, CORS feature 08:09:36 tsufiev: cool 08:09:38 I have one bp to mention, its more of a discussion point than a solid feature proposal. 08:09:46 so we can route around Django's FD issue ;-) 08:10:03 #link https://blueprints.launchpad.net/horizon/+spec/data-architecture 08:10:03 r1chardj0n3s, yep, will try to refresh my patch today 08:11:49 Its a vague title, but a discussion worth having with the angular rewrites 08:11:59 URL patterns and code structure seem to be hopping around a bit 08:12:09 yeah, interesting 08:12:20 doesn't mention s/instance/server though, so REJECT 08:12:27 08:12:31 :) 08:12:47 (I'm still not entirely sure how that conversation got derailed at the summit) 08:12:48 I like the idea, Admin dashboard quite often looks redundant 08:12:55 I mean certain panels within it 08:13:07 but the admin panels are basically fronting over admin access to services, right? 08:13:10 Anyway if you have any thoughts on it, please note them down. Wanted to get some early discussion on the BP before mentioning at midcycle. 08:13:14 so there's different permissions going on there 08:13:47 r1chardj0n3s: Right, but a lot of it is very minor. The admin networking views for example, are nearly all direct inheritance 08:13:53 right 08:14:06 Which adds a lot of boilerplate and duplication. 08:14:13 and all the addition 'admin' stuff could be hidden with policy.check() 08:14:15 So there are gains to be made in that regard 08:14:25 Precisely 08:14:47 Davids point is good though, you wouldnt want someone accidentally deleting someone elses resources, for example. 08:14:55 + we decided to stop using roles.admin permissions, in favor of checking policies 08:15:06 also it avoids the issue of us splitting out the angular implementation details of views because they're shared between the two sets of panels 08:15:17 that's a level of obfuscation I'm not happy about 08:15:37 Yep. So we would just have /images/ for example 08:15:51 * tsufiev just read David's remark 08:16:33 It ties in to the angular too, which atm is using /details// 08:16:37 yeah, David's point clarifies what I was thinking about there 08:16:45 oops, forgot a '/project/' in there 08:16:51 that would be a really good topic for UX research... 08:17:09 Yeah, I'll see if Piet has any thoughts 08:18:05 robcresswell, do you know how the process of discussing the changes is organized now in UX community? 08:18:18 I have fallen out from its workflow recently :/ 08:18:50 I do not. They were organising research, but I don't recall the process 08:18:54 the next meeting is this friday 08:19:34 okay, should come there and see what's going on... 08:20:46 I've dropped that bp into the agenda so I can raise it there, see if there is any interest 08:22:28 Cool, thats all from me and the agenda is empty. 08:22:59 quiet, quick meeting :-) 08:23:07 * tsufiev in the process of updating glance-image-upload bp, will finish 10 minutes later 08:23:11 So if you have any thoughts on that bp, please leave comments; its just intended to generate early discussion so that its not a new idea at the midcycle (since time is limited) 08:23:24 yep 08:23:39 Thanks all 08:23:40 sure 08:23:46 #endmeeting