13:00:41 <alex_xu> #startmeeting nova api
13:00:42 <openstack> Meeting started Wed May 11 13:00:41 2016 UTC and is due to finish in 60 minutes.  The chair is alex_xu. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:00:43 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:00:45 <openstack> The meeting name has been set to 'nova_api'
13:00:54 <alex_xu> who is here today?
13:01:29 * andrearosa lurks
13:01:49 <jichen> o/
13:01:51 <alex_xu> emm...strange, nobody at here today
13:02:06 <alex_xu> as i know, ken'ichi on the travel
13:02:24 <sdague> o/
13:02:36 <johnthetubaguy> o/
13:02:40 <sdague> sorry, was generating new stats for the sprint to put on email
13:02:42 <Kevin_Zheng> o/
13:02:48 <alex_xu> sdague: np
13:03:00 <alex_xu> ok, we have people now, let's start the meeting :)
13:03:27 <alex_xu> #topic API priority
13:03:44 <alex_xu> let's talk api ref first
13:04:25 <alex_xu> sdague: i guess we are good, and maybe we should keep today's meeting short, and back to doc sprint?
13:04:52 <sdague> alex_xu: yeh, the doc sprint is pretty much the priority this week. I haven't even looked at any of the specs
13:05:12 <sdague> http://burndown.dague.org/ it's coming along, but it will take a bit
13:05:17 * edleafe shows up late...
13:05:48 <alex_xu> ah, that diagram is cool
13:05:59 <sdague> there are a bunch of changes up for review, we should probably just fix/update reviews that are close when we find issues
13:06:09 <johnthetubaguy> a have a quick random question from a review, but I can leave that till open
13:06:11 <sdague> I was going to go back through the list and do that after the meeting
13:06:12 <alex_xu> ++
13:06:21 <sdague> johnthetubaguy: no, lets do it now
13:06:43 <alex_xu> johnthetubaguy: yea, please go ahead
13:06:45 <johnthetubaguy> https://review.openstack.org/#/c/314924/2/api-ref/source/os-hosts.inc@73
13:06:50 <johnthetubaguy> sorry creating link
13:06:55 <johnthetubaguy> we have two host params
13:07:00 <johnthetubaguy> its kinda nested
13:07:08 <johnthetubaguy> https://git.openstack.org/cgit/openstack/nova/tree/doc/api_samples/os-hosts/host-get-resp.json#n7
13:07:29 <johnthetubaguy> as its an admin API, I am fine ignoring the problem
13:07:38 <sdague> oh, host / host
13:07:40 <johnthetubaguy> just not see that happen before, so its a heads up
13:07:44 <johnthetubaguy> yeah
13:07:44 <johnthetubaguy> host and host
13:07:52 <sdague> so... I think you can actually specify it twice
13:07:59 <alex_xu> the host at line 68 looks like useless
13:08:01 <sdague> it's a list
13:08:07 <jichen> maybe hosts?
13:08:13 <johnthetubaguy> oh right, good point
13:08:29 <johnthetubaguy> jichen: we can't change the representation, otherwise I would be tempted
13:08:31 <sdague> I almost anticipated this issue because of nested resources
13:08:40 <johnthetubaguy> sdague: :)
13:08:45 <jichen> johnthetubaguy: ok :)
13:08:46 <sdague> what we probably should do is actually the following
13:08:49 <sdague> resource.x
13:08:52 <alex_xu> that is just part of structure, no exactly meaning, we needn't list that in the parameters list
13:08:54 <sdague> for all of these
13:08:57 <johnthetubaguy> yeah, I was just thinking about that
13:09:04 <johnthetubaguy> they are all the same then
13:09:05 <sdague> that's what I was doing for complex data in servers
13:09:12 <sdague> like security_groups.name
13:09:30 <sdague> because, otherwise, name being in the list twice is super confusing
13:09:44 <johnthetubaguy> yeah, that works, when its needed
13:10:57 <johnthetubaguy> so, I am happy, lets fix that up
13:11:20 <sdague> johnthetubaguy great
13:11:20 <alex_xu> yes, that is also works.
13:11:33 <sdague> johnthetubaguy: thanks for catching that
13:12:13 <johnthetubaguy> sdague: np
13:12:24 <alex_xu> anymore question on the api-ref?
13:12:31 <alex_xu> s/anymore/any more/
13:13:03 <johnthetubaguy> FWIW, I thought this was terrible to review, until you have the code open for the API and the samples in another window, and its so much easier
13:13:26 <alex_xu> yea
13:13:40 <sdague> yeh, servers is still a pain, because of all the extensiosn
13:13:46 <sdague> I can't wait until that's gone
13:13:52 <johnthetubaguy> sdague: true, +100
13:14:28 <johnthetubaguy> just a note to the docs team, I have no idea how the docs are so good as they are right now!
13:14:36 <sdague> :)
13:14:59 <alex_xu> :)
13:15:07 <alex_xu> ok, i think we can move on
13:15:45 <alex_xu> for legacy v2 api removing, looks good. just the patches need rebase
13:16:13 <alex_xu> after the v2 code removed, we also can begin to clean the policy for legacy v
13:16:36 * mriedem joins late
13:16:42 <johnthetubaguy> ah, so looking at this, we totally have a task to clean up our policy rules later on
13:16:50 <sdague> alex_xu: yeh, let me know when it looks like we're going to be close on the legacy v2 side
13:17:01 <johnthetubaguy> the admin_or_owner stuff is very... misleading in how it works today
13:17:04 <alex_xu> sdague: ok, got it
13:17:10 <sdague> as I've got that -2 on the final remove to let everyone pile on for +2
13:17:25 <sdague> johnthetubaguy: well, that's going to come much later
13:17:36 <johnthetubaguy> yeah, once the defaults are in code
13:18:40 <sdague> ok, next topic?
13:18:40 <alex_xu> yeh, as mentioned policy...for the policy discovery, i didn't notice that this week
13:18:59 <sdague> right, I think that all the specs work is sort of on hold this week to get through as much of the api-ref work as possible
13:19:07 <sdague> keep folks focussed
13:19:18 <alex_xu> yea
13:19:25 <sdague> next week we should revisit the various specs
13:19:32 <johnthetubaguy> sdague: +1
13:19:42 <sdague> and probably plan to spend this meeting mostly on outstanding specs, including the deprecations ones
13:20:06 <alex_xu> +1
13:20:09 <johnthetubaguy> do we want to defer any +W on them, till next week?
13:20:48 <johnthetubaguy> actually ignore that, regular view process should be fine, I was being over-optimistic
13:20:56 <mriedem> they have -1s anyway
13:21:03 <johnthetubaguy> yeah, that would do it
13:21:17 <alex_xu> ok, i think we can just to open now
13:21:22 <alex_xu> #topic Open
13:21:42 <alex_xu> there is one item about "Add support for user defined stop instance"
13:21:50 <Kevin_Zheng> I can wait for next week
13:22:03 <Kevin_Zheng> as there are more priority work
13:22:50 <alex_xu> people want to talk about it today? or back to doc sprint this week first?
13:22:57 <sdague> lets focus on doc sprint
13:23:01 <johnthetubaguy> the second one
13:23:04 <alex_xu> sdague: ok
13:23:08 <alex_xu> Kevin_Zheng: thanks
13:23:24 <Kevin_Zheng> NP, I will try to help on doc tomorrow
13:23:25 <mriedem> keep in mind non-priority spec freeze is in 3 weeks
13:23:47 <alex_xu> mriedem: got it, thanks
13:23:50 <alex_xu> Kevin_Zheng: thanks again
13:23:59 <alex_xu> so we can close the meeting now?
13:24:11 <alex_xu> 3...
13:24:14 <alex_xu> 2..
13:24:18 <alex_xu> 1.
13:24:22 <alex_xu> thanks all!
13:24:25 <jichen> thanks~
13:24:26 <alex_xu> #endmeeting