12:00:35 <david-lyle> #startmeeting horizon
12:00:35 <openstack> Meeting started Wed Mar  2 12:00:35 2016 UTC and is due to finish in 60 minutes.  The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot.
12:00:36 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
12:00:39 <openstack> The meeting name has been set to 'horizon'
12:00:42 <robcresswell> o/
12:00:55 <r1chardj0n3s> o/
12:00:56 <doug-fish> \o
12:01:26 <mrunge> o/
12:02:47 <david-lyle> well, we are a day away from M-3
12:03:08 <david-lyle> how are we looking?
12:03:33 <mrunge> tired?
12:03:36 <mrunge> :D
12:03:44 <david-lyle> indeed
12:03:48 <doug-fish> lol, yeah that's true
12:03:50 <masco> :)
12:03:50 <robcresswell> I'd still quite like to flip the switch on Launch Instance. The patch to fix the theming has had a few iterations and reviews now.
12:03:59 <robcresswell> Swift has its first 2 patches in I think
12:04:04 <r1chardj0n3s> yep
12:04:11 <r1chardj0n3s> the meat is still yet to come tho
12:04:18 <tsufiev> o/
12:04:29 <r1chardj0n3s> launch instance is looking great
12:04:45 <doug-fish> lbaas-dashboard is looking pretty good - so I guess angular plugin support in horizon must be in good shape too
12:05:06 <betherly> o/ sorry im late!
12:05:16 <robcresswell> There is a floating IP panel and some cinder code I wanted to review too
12:05:35 <david-lyle> and I need to review the domain scoped token patch again
12:05:55 <betherly> ditto to what doug-fish said for the ironic-ui. angular plugin support is good!
12:06:04 <david-lyle> seems like we're close on a lot, but likely won't land any by M-3
12:06:20 <masco> QoS patch also almost ready in angular some reviews are most welcome
12:06:28 <david-lyle> so then we have to start assessing what is FFE worthy
12:06:48 <itxaka> what does _FFE_ means, sorry?
12:06:57 <robcresswell> Feature Freeze Exception
12:07:05 <itxaka> gotcha, thanks o/
12:07:12 <robcresswell> Permission to keep working on your blueprint :)
12:07:17 <r1chardj0n3s> M-3 is a feature freeze point for Mitaka
12:07:49 <david-lyle> well you can always keep working on your bp, just may be blocked from merging until N
12:08:05 <robcresswell> yeah
12:08:29 <david-lyle> if things are in the gate by tomorrow, I think that's an easy call to make i
12:08:31 <david-lyle> t
12:09:03 <david-lyle> I also need to try and release d-o-a again
12:09:10 <robcresswell> I think LI should be. It's way more consistent now, and I'd like to start putting some angular content in front of people.
12:09:10 <david-lyle> bug fixes and 1.9 support
12:09:50 <robcresswell> yep
12:10:15 <david-lyle> ok, once LI changes are merged, I think we can move forward on making it default
12:10:55 <robcresswell> Sounds good
12:11:28 <david-lyle> any other pressing M-3 items
12:11:30 <david-lyle> ?
12:12:14 <tsufiev> unfortunately, I haven't managed to refresh patches for AVAILABLE_REGIONS setting renaming :(
12:12:24 <tsufiev> so I guess they are left for N
12:14:23 <itxaka> ok, so if Im understanding this correctly, whatever does not get in by tomorrow, its left for next release rigth (sorry, release noob!)
12:14:38 <david-lyle> itxaka: for new features yes
12:14:48 <david-lyle> there will be a few exceptions
12:14:50 <doug-fish> unless it's given an exception
12:14:55 <mrunge> Just looking at https://launchpad.net/horizon/+milestone/mitaka-3
12:15:00 <robcresswell> Any blueprinted features. Bugfixes are normally fine, unless they involve new strings.
12:15:25 <mrunge> there are 3/8 blueprints implemented
12:15:53 <robcresswell> I think that LBaaS is redundant, no? I thought it was just a plugin now
12:16:06 <itxaka> so if there is a bp with target release mitaka instead of mitaka-m3, what does that entail actually?
12:16:10 <doug-fish> "just" a plugin?
12:16:18 <doug-fish> I thought we were the stars of the show now.
12:16:26 <r1chardj0n3s> the theme blueprint needs updating - theme switching landed last week
12:16:26 <robcresswell> haha sorry doug. So much more than a plugin :)
12:16:45 <doug-fish> That's the same thing - we must have opened a horizon blueprint
12:17:02 <robcresswell> The cinder consistency groups blueprint is the one I mentioned earlier
12:17:19 <robcresswell> That needs some review attention, it was mentioned at the midcycle I believe.
12:18:32 <tsufiev> I think Dynamic themes can be marked as Implemented
12:18:45 <david-lyle> yes and we finally landed cinder v2 or v1 support patch, which makes me happy
12:18:46 <tsufiev> there is only an integration test for that is still pending
12:19:45 <itxaka> thinking about this 2 bp and how do they fit in this: https://blueprints.launchpad.net/horizon/+spec/configurable-boot-sources and https://blueprints.launchpad.net/horizon/+spec/allow-launching-ports
12:20:13 <itxaka> both are targeted for mitaka and are new features but if m-3 is FF...then I dont really get it :)
12:20:22 * r1chardj0n3s just hoping no-one has noticed there's no swift blueprint listed
12:21:18 <robcresswell> itxaka: So at this point, if you think the code is stable and ready, you can ask for an FFE; in that case, it can still merge in the next week or so.
12:22:04 <robcresswell> Generally speaking, FFEs are only given to very desirable patches or those that have just missed landing (i.e. failed jenkins on the final day or something silly)
12:22:05 <david-lyle> r1chardj0n3s: oh it's been noticed
12:22:17 <r1chardj0n3s> https://blueprints.launchpad.net/horizon/+spec/angularize-swift
12:22:19 <doug-fish> r1chardj0n3s: so what's the story there?
12:22:35 <r1chardj0n3s> the patches don't reference it :/
12:22:37 <r1chardj0n3s> my bad
12:23:10 <robcresswell> Well, its good to know people aren't merging code without any reference to LP.
12:23:13 <robcresswell> Oh wait... :p
12:24:03 <david-lyle> well at least r1chardj0n3s is new to the project and didn't know better
12:24:04 <r1chardj0n3s> yeah. again, my bad. not sure how the bp reference got lost along the way
12:24:12 <tsufiev> lol )
12:24:16 <r1chardj0n3s> le sigh
12:24:26 <r1chardj0n3s> I'll show myself out, shall I? :-P
12:24:44 <robcresswell> Bet you're glad you stayed up for this meeting now, ha
12:24:59 <r1chardj0n3s> I have some generalised regrets, yes
12:25:05 <david-lyle> the bp is now on the M-3 milestones
12:25:12 <doug-fish> r1chardj0n3s: you know, at least one of your patches does reference the blueprint
12:25:12 <david-lyle> or just the one
12:25:22 <doug-fish> r1chardj0n3s: you're better behaved than you realize
12:25:44 <doug-fish> https://review.openstack.org/#/c/259231/ right?
12:26:23 <r1chardj0n3s> yep, that's the next one in the list, but the subsequent patch loses the ref :/
12:27:12 <david-lyle> robcresswell: your theme-ability of LI seems conspicuously absent from the list as well
12:27:22 * robcresswell ducks
12:27:30 <r1chardj0n3s> HA!
12:28:16 <robcresswell> I'll put up a bp and target it then
12:28:24 <robcresswell> or a bug?
12:28:32 <david-lyle> you don't even have a bp?
12:28:36 <robcresswell> Oh yeah I've tagged it to a bug
12:28:52 <david-lyle> ok, leave it as a bug
12:29:22 <robcresswell> No I didn't blueprint it. Broken theming seemed like a bug rather than a feature. Up to you how you want to handle it.
12:29:38 <r1chardj0n3s> smells like a bug to me
12:29:39 <david-lyle> your bug does not target m-3 either
12:30:12 <robcresswell> https://bugs.launchpad.net/horizon/+bug/1538491 ?
12:30:14 <openstack> Launchpad bug 1538491 in OpenStack Dashboard (Horizon) "Angular Launch Instance should use adhere to Horizon theming standards" [Wishlist,In progress] - Assigned to Rob Cresswell (robcresswell)
12:30:14 * david-lyle thinks it was easier when it was all in launchpad and not random sources
12:32:25 <tsufiev> robcresswell, have you booked our CSS Guru in advance ;)?
12:33:01 <robcresswell> She's had a look over the patch once
12:33:19 <david-lyle> patch isn't linked to bug?
12:33:44 <robcresswell> The Closes-Bug is on there, but no it hasn't linked for some reason.
12:34:04 <robcresswell> I've left a comment on the bug report.
12:35:11 <david-lyle> great, maybe posted during a gerrit snafu period
12:38:44 <tsufiev> did I get it right that the general consensus is to enable Angular LI in M by default?
12:39:14 <david-lyle> tsufiev: I think that was the call
12:39:18 <robcresswell> It seems so. I'd like to toggle it and see how it goes over the RC period.
12:39:26 <tsufiev> ok, good to know :)
12:39:30 <doug-fish> integration tests?
12:39:34 <doug-fish> won't those break?
12:39:40 <robcresswell> Its trivial to swap back if we find a blocking issue or instability.
12:40:03 <tsufiev> doug-fish, they'll certainly will if we don't disable test for an old LI
12:40:19 <tsufiev> that's not very good, because we still don't have tests for Angular LI
12:40:22 <david-lyle> operators will have the option of using old for now, although when we make it default, we should probably mark the old as deprecating
12:40:41 <david-lyle> or at least frozen
12:41:27 <tsufiev> doug-fish, I think we need to rethink our priorities in new tests writing in case Angular LI is going to become a default
12:41:40 <tsufiev> also... does it make sense to run tests for both LI at gate?
12:42:23 <doug-fish> if we can test the new one (and not the old) I'd be satisfied ... though running both might be better
12:42:35 <doug-fish> tsufiev: do you have a vision for how to do that?
12:42:49 <tsufiev> doug-fish,  enable both buttons in devstack
12:42:59 <david-lyle> it's just a settings override
12:43:17 <doug-fish> oh for tests, right? there are separate settings
12:43:31 <doug-fish> I was thinking both buttons would be enabled for users have a default install
12:43:36 <doug-fish> devstack users
12:43:42 <doug-fish> if anyone actually does that.
12:44:34 <robcresswell> There's just a settings toggle right now to swap between them
12:44:46 <robcresswell> So you could just do that for the integration tests
12:46:27 <david-lyle> I think we're lined up for M-3
12:46:34 <david-lyle> #topic summit
12:47:02 <david-lyle> we've been asked for a summit session allocation request
12:47:29 <david-lyle> how many fishbowl, working sessions, meetups we want
12:48:08 <mrunge> same as in Tokyo?
12:48:16 <david-lyle> I think we were stretching to fill two fishbowls last time
12:48:16 <mrunge> maybe a working session less?
12:48:41 <david-lyle> also we didn't use the second half day of meetup in Tokyo
12:49:06 <david-lyle> part of the context is there are 11 more projects needing space too
12:49:46 <david-lyle> I'm tempted to go 1,7,1 on the allocation
12:49:49 <doug-fish> I think it feels offhand like we have fewer topics to cover than we did in Tokyo
12:49:59 <mrunge> rught
12:50:04 <david-lyle> we were 2,8,2 in Tokyo
12:50:04 <mrunge> uhm, right
12:50:07 <robcresswell> In Tokyo it was 2,8,2, so I think 1,7,1 is reasonable.
12:50:19 <doug-fish> yeah that sounds good to me
12:51:02 <david-lyle> one learning from the summits is you can always find space to talk about something that needs to be talked about
12:51:24 <doug-fish> that is to say, there are plenty of bars in Austin
12:51:27 <david-lyle> so if we find ourselves short on room times, we can make something up
12:52:04 <david-lyle> I would also like some flexibility to attend other projects' sessions
12:52:14 <david-lyle> we've lost that in the recent past
12:52:21 <mrunge> yes!
12:52:51 <mrunge> do we have to ability to opt on time-slots?
12:52:58 <tsufiev> speaking of integrations and new LI... sorry, was distracted by a parallel meeting
12:53:20 <david-lyle> mrunge: very little
12:53:30 <tsufiev> so, the setting is in settings/local_settings.py, but as doug-fish pointed, integration tests cannot override it as unit-tests do
12:53:52 <david-lyle> key speaker conflict and project conflict/overlap are about the only reasons
12:54:02 <david-lyle> too hard otherwise
12:54:06 <doug-fish> tsufiev: actually I'd amend that - they don't today, but they could have their own settings file
12:54:19 <doug-fish> just update tox.ini
12:54:57 <david-lyle> anyone have a problem with 1,7,1 ?
12:55:06 <betherly> not me
12:55:14 <robcresswell> Sounds good to me david-lyle
12:55:18 <doug-fish> +2
12:55:27 <mrunge> ack
12:55:30 <tsufiev> +1
12:55:49 <david-lyle> ok, will pass on, thanks
12:56:04 <tsufiev> doug-fish, that essentially means that horizon settings.py should import integration tests settings somehow
12:56:06 <david-lyle> #topic Open Discussion
12:56:35 <tsufiev> (when it's in DEBUG/SOME_NEW_TESTING mode)
12:56:57 <tsufiev> that seems totally doable to me
12:57:17 <doug-fish> for the unit tests we have a parallel settings.py, right?
12:57:21 <david-lyle> I would like to thank all who could attend the midcycle sprint. Overall I felt it was very productive. The items discussed can be found: https://etherpad.openstack.org/p/horizon-mitaka-midcycle and the accompanying notes: https://etherpad.openstack.org/p/mitaka-horizon-priorities-notes
12:57:22 <tsufiev> yes
12:57:29 <doug-fish> that replaces (and hopefully imports) our normal default one
12:57:55 <david-lyle> we had 13 people from 4 countries attending
12:58:10 <tsufiev> doug-fish, the tricky part here is that no new django server is being launched by integration tests themselves
12:58:22 <doug-fish> oh. that.
12:58:24 <doug-fish> hmm.
12:58:31 <tsufiev> they just run against the Horizon under Apache running in a Devstack
12:58:43 <doug-fish> right ... that's a problem.
12:59:03 <david-lyle> tsufiev: file change, apache restart?
12:59:03 <doug-fish> now I'm less sure how to have separate settings for integration test
12:59:21 <tsufiev> so... import test_settings approach seems more feasible to me...
12:59:44 <tsufiev> david-lyle, yeah, this could be done as well just before running tests. Would include some sed black magic :/
13:00:04 <tsufiev> but is still possible
13:00:04 <david-lyle> tsufiev: yeah it won't be pretty
13:00:24 <david-lyle> or we just disable old LI test
13:00:30 <tsufiev> hm...
13:00:36 <david-lyle> based on  setting
13:00:40 <tsufiev> this way we won't know if we broke it
13:00:43 <tsufiev> one day
13:00:54 <robcresswell> time - we should continue in #openstack-horizon
13:01:01 <robcresswell> :)
13:01:04 <david-lyle> thanks everyone
13:01:08 <david-lyle> #endmeeting