16:00:15 <etoews> #startmeeting api wg
16:00:21 <openstack> Meeting started Thu Aug 25 16:00:15 2016 UTC and is due to finish in 60 minutes.  The chair is etoews. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:23 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:25 <openstack> The meeting name has been set to 'api_wg'
16:00:32 <etoews> #chair cdent elmiko etoews
16:00:33 <openstack> Current chairs: cdent elmiko etoews
16:01:07 <etoews> anyone around today?
16:01:13 <etoews> i know elmiko is out
16:01:19 <cdent> o/
16:02:01 <etoews> #link https://wiki.openstack.org/wiki/Meetings/API-WG#Agenda
16:02:19 <etoews> #topic previous meeting action items
16:02:27 <etoews> #link http://eavesdrop.openstack.org/meetings/api_wg/2016/
16:03:12 <cdent> Hmmm I had an action, not listed there, to get launchpad and gerrit hooked up
16:03:24 <etoews> and i saw you had something for that right?
16:03:50 <cdent> that's been done, but we won't know if it is working until we create a _new_ review (the hooks only work when they are there from the start, as I recall)
16:05:01 <etoews> gotcha. that'll happen...someday.
16:05:30 <etoews> #topic ongoing tracking of capabilities and discovery
16:06:17 <etoews> did you want to chat about that cdent ?
16:06:18 <cdent> nothing new on that
16:06:52 <cdent> it's just something we need to be aware of, but as we are nearing the end of the cycle it is unlikely that either nova or cinder will have time to talk to one another
16:07:14 <cdent> What I'm hoping is that we can make sure it is consistent. That's all.
16:07:32 <etoews> i was a bit surprised i didn't see you reply on the ml on that thread...unless i missed it.
16:08:18 <cdent> I've been utterly swamped on resource provider/placement api stuff
16:08:38 <cdent> there's been tons of things go by that I'd like to be a part of
16:08:49 <etoews> shall we reply to say, please bring your common understanding to the api wg?
16:08:51 <cdent> or at least have an opinion on
16:09:03 <cdent> that's a good idea
16:09:26 <etoews> i can reply unless you prefer to
16:09:31 <cdent> go for it
16:09:42 <etoews> i'll do it right now
16:10:26 <cdent> awesome
16:22:37 <etoews> sent
16:22:57 <etoews> cdent: you know that means the weekly newsletter is yours right? ;)
16:23:06 <cdent> that seems fair
16:23:09 <etoews> (if it's worthwhile to send one)
16:23:59 <etoews> #topic open mic
16:24:09 * etoews points mic at the crickets.
16:24:18 <cdent> they make a glorious chirping
16:24:42 <etoews> #topic guidelines
16:24:51 <etoews> #link https://review.openstack.org/#/q/status:open+project:openstack/api-wg,n,z
16:24:58 <cdent> I tuned up https://review.openstack.org/#/c/322194/ I reckon it can be refrozen
16:25:46 <cdent> gerrit very slow for me at the moment
16:25:51 <etoews> yep
16:26:09 <cdent> other two probably freezable as well
16:27:13 <etoews> > Any resource should only have one URI. Where possible do not provide multiple ways to reference the same thing. Use HTTP redirects to resolve indirect requests to the correct canonical URI and content negotiation to request different representations.
16:27:30 <etoews> cdent: i take it you don't care for the version number in the url
16:28:08 <etoews> it seems like openstack is moving away from having version number or tenant in the url
16:28:27 <cdent> tenant in the url is especially bad
16:28:46 <cdent> version number is probably bad, but I can see situations where it is hard not to
16:29:03 <cdent> but really: decide on your resources, stick with them. Need versions? version your media types
16:29:19 <etoews> or microversion your api ;)
16:29:28 <cdent> no, microversions are horrible
16:29:32 <etoews> haha
16:29:32 <cdent> but they are what we've got
16:30:08 <cdent> microversions enable and encourage change
16:30:23 <cdent> and make simple clients harder
16:30:30 <etoews> yep
16:30:38 <cdent> and make server code complex
16:30:47 <cdent> but we are saddled with them so at least we can make them work
16:31:13 <etoews> gerrit is having a bad day
16:31:20 <cdent> it was fine about an hour ago
16:32:30 <etoews> froze the uri guideline
16:33:22 <cdent> 👍
16:33:40 <etoews> cdent: go ahead and workflow+1 the backslash one
16:34:00 <cdent> roger
16:34:12 <etoews> i'll freeze version string and auto-pester
16:34:25 <cdent> roger
16:38:07 <etoews> cdent: your comment on the links guideline
16:38:07 <etoews> > We probably want another one for _why_.
16:38:22 <etoews> do you mean another PR for the _why_?
16:38:42 <cdent> yeah, a different patch, at some later date, for completeness
16:39:17 <etoews> sure
16:40:28 <etoews> i'm thinking i should really link to the links guideline from other places in the guidelines too.
16:40:44 <cdent> yeah, that's probably true
16:40:50 <cdent> hypermedia everywhere!
16:41:27 <etoews> it's the HT in HTML and HTTP!
16:41:39 <cdent>16:44:29 <cdent> that  covers it for reivews (I'll link them up to the newsletter when I send it) it seems
16:44:36 <etoews> i'm just workin on links for links.
16:44:47 <cdent> apiimpact is on hold (I do look every week but don't actually do anything)
16:44:57 <cdent> #topic bug review
16:45:21 <cdent> nothing changed since last week
16:45:56 <cdent> I have an intention to take one each week, but see above about swampedness
16:46:10 <cdent> I'm at least looking each week to triage anything new
16:46:36 <etoews> yep. links was my take. i can't commit to one a week. just one whenever....
16:48:48 <cdent> shall I freeze links?
16:49:18 <etoews> let me get one more commit in there and then yes
16:49:29 <etoews> cdent ^
16:49:35 <cdent> roger
16:55:37 <cdent> #link: https://etherpad.openstack.org/p/api-wg-newsletter
16:55:41 <cdent> newsletter in progress ^
16:57:28 <etoews> go ahead and freeze/auto-pester on https://review.openstack.org/#/c/354266
16:57:39 <etoews> i've got to run to the next thing.
16:57:45 <cdent> roger
16:57:48 <etoews> mind if i end the meeting?
16:57:48 * cdent waves
16:57:51 <cdent> go for it
16:57:55 * etoews waves back
16:58:02 <etoews> #endmeeting