13:05:06 <johnthetubaguy> #startmeeting nova-api
13:05:07 <openstack> Meeting started Wed Jul  6 13:05:06 2016 UTC and is due to finish in 60 minutes.  The chair is johnthetubaguy. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:05:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:05:10 <openstack> The meeting name has been set to 'nova_api'
13:05:36 <johnthetubaguy> #link https://wiki.openstack.org/wiki/Meetings/NovaAPI
13:05:51 <johnthetubaguy> so we are totally missing a few folks this week
13:06:01 <johnthetubaguy> alex_xu is probably at a bug smash event
13:06:07 <johnthetubaguy> sdague is on holiday, I think
13:06:20 <johnthetubaguy> #topic API Priorities
13:06:35 <johnthetubaguy> now the agenda has lots of notes, lets check whats new
13:06:49 <johnthetubaguy> #link https://review.openstack.org/#/q/topic:bp/policy-in-code
13:06:54 <johnthetubaguy> some good progress with policy
13:07:50 <johnthetubaguy> so alex pinged me about a bunch of new code
13:07:54 <johnthetubaguy> just trying to find it
13:08:50 <johnthetubaguy> #link https://review.openstack.org/#/q/topic:bp/deprecate-api-proxies
13:09:22 <johnthetubaguy> anyways, lots going on
13:09:27 <johnthetubaguy> #topic Open Discussion
13:09:32 <johnthetubaguy> so there are some things in the agenda
13:09:51 * johnthetubaguy alaski added this one: https://review.openstack.org/#/c/305369/3
13:10:23 <edleafe> johnthetubaguy: about the proxy deprecation
13:10:40 <edleafe> it's set up so that each patch is removing one thing
13:10:49 <edleafe> and then the final bumps the microversion
13:11:04 <johnthetubaguy> yeah, I think thats the intent
13:11:09 <edleafe> is that OK? mriedem doesn't think so, and I'm not sure either
13:11:27 <edleafe> IOW, when the first few merge, master will be in a weird state
13:11:51 <edleafe> and if another microversion bump merges, then those early ones will be wrong
13:12:08 <johnthetubaguy> yeah, it is funky
13:12:34 <johnthetubaguy> my personal preference is we review them separately, then if need to a final squash before merge
13:12:54 <edleafe> I know "small is better" with patches, but this is one case where a single patch would be better
13:13:38 <johnthetubaguy> better for merge, not for review really
13:13:39 <edleafe> So... something like, get the +2s for all, but -W them
13:13:52 <johnthetubaguy> so matt blocked the first patch
13:14:00 <johnthetubaguy> we can +W all the others, then squash them
13:14:08 <edleafe> Then when they are all +2, remove the -W and make them +W
13:14:26 <johnthetubaguy> just the one at the end is enough to block the follow ons
13:14:44 <edleafe> johnthetubaguy: true
13:15:15 <edleafe> I'm concerned about merging them un-squashed. The gate can make a smooth merge like that pretty uncertain
13:15:30 * edleafe thinks about the config option series that never merged together
13:15:42 <johnthetubaguy> so I am probably being unclear
13:16:02 <johnthetubaguy> my squash, I mean merge them all into a single patch before we merge, after having reviewed them as little chunks
13:16:08 <johnthetubaguy> s/my/by/
13:16:36 <edleafe> johnthetubaguy: I get what 'squash' means.
13:16:54 <edleafe> johnthetubaguy: I just wanted to make sure that that was the plan
13:17:32 <johnthetubaguy> that was my assumption of the plan, I added a comment to ask for clarification
13:17:33 <edleafe> IOW, we're faking a feature branch in git with a patch series in Gerrit
13:17:42 <johnthetubaguy> yep
13:17:47 <edleafe> ok cool
13:18:01 <johnthetubaguy> we could do a feature branch (its supported somehow), but it seems like more effort than its worth
13:18:41 * edleafe has never seen a feature branch in Gerrit
13:18:57 <edleafe> but that would be more intuitive for a git user
13:19:07 <johnthetubaguy> yeah, I think swift tried a few for something, I never managed to chase up how that works
13:20:07 <johnthetubaguy> so alaski raised this one: https://review.openstack.org/#/c/305369/
13:20:19 <johnthetubaguy> there is a discussion going on in there, following from a quick IRC sync up
13:20:23 <johnthetubaguy> feel free to join in
13:21:18 <edleafe> ok, I hadn't seen that one yet. Added to my queue
13:21:50 <johnthetubaguy> https://review.openstack.org/#/c/326326/ is another one thats been added
13:22:17 <johnthetubaguy> but I have a feeling thats now blocked by feature freeze
13:22:43 <edleafe> johnthetubaguy: yeah, it is
13:23:00 <edleafe> It's trickier than it appears
13:23:16 <johnthetubaguy> yeah, I think thats what was decided yesterday
13:23:29 <johnthetubaguy> anyways, any more for any more?
13:23:32 <edleafe> Probably a good topic for the midcycle
13:24:14 <johnthetubaguy> well, its something we probably need to actually agreed in the API wg
13:24:23 <johnthetubaguy> they have a draft, but its only partial right now
13:24:35 <johnthetubaguy> its would be nice to get that completed, then adopt the same patterns
13:25:18 <johnthetubaguy> alex is quite correct, the sort order should be maintained between versions, I think.
13:25:33 <edleafe> I think the sort order thing is specific to Nova; the pagination patterns are more general
13:25:38 <johnthetubaguy> anyways, we should probably all get back to code reviews with our spare time
13:25:44 <johnthetubaguy> edleafe: yep
13:25:52 * edleafe wonders what is the "spare time" thing?
13:25:54 <johnthetubaguy> edleafe: I was really meaning about the link debate
13:26:08 <edleafe> gotcha
13:26:11 <johnthetubaguy> edleafe: its whats about to happen when I end the meeting, kinda
13:26:18 <edleafe> :)
13:26:24 <johnthetubaguy> thanks all
13:26:26 <johnthetubaguy> #endmeeting