16:00:47 #startmeeting api-sig 16:00:48 Meeting started Thu Aug 24 16:00:47 2017 UTC and is due to finish in 60 minutes. The chair is elmiko. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:49 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:52 The meeting name has been set to 'api_sig' 16:01:38 #chair edleafe cdent 16:01:38 Warning: Nick not in channel: cdent 16:01:40 Current chairs: cdent edleafe elmiko 16:01:45 o/ 16:01:57 hey 16:02:00 o/ 16:02:14 sorry, was busily trying to convince other TC that code complexity is a real problem 16:02:19 #link https://wiki.openstack.org/wiki/Meetings/API-SIG#Agenda 16:02:21 no worries 16:02:42 wow, that sounds like an intense discussion with lots of code samples 16:02:58 we got an edleafe ? 16:03:15 haven't heard from him, but he's around somewhere 16:03:57 maybe give him a minute or 2? 16:05:18 #topic previous meeting action items 16:05:25 #link http://eavesdrop.openstack.org/meetings/api_sig/2017/ 16:05:50 so, i did my email thingie 16:06:22 only got 2 responses, one was an out-of-office notice, the other was a suggestion about hitting more PTLs 16:06:45 did you not hit all the ptls? 16:06:58 but, i did get into a chat with the PTL for sahara and he was considering if they should do a review 16:07:06 i *think* i did 16:07:13 i used the list from the governance rpo 16:07:17 *repo 16:07:39 I'm still interested, but I did not have any time to wrap my head around it 16:08:07 well, seeing as we don’t have a ton of topic prepare for the ptg, we may be able to take walkins 16:08:14 yeah 16:08:19 this was the list i used 16:08:20 https://github.com/openstack/governance/blob/master/reference/projects.yaml 16:08:31 #link ptg etherpad https://etherpad.openstack.org/p/api-ptg-queens 16:08:34 there was some concern that maybe the queens PTLs had not been added yet 16:08:47 ack, i can send the email again just to be certain 16:10:08 anything else about the ptl list or action items? 16:10:17 a valid point. but I also think rc2 is an issue. and also, for many this api stuff is not really all that relevant in the grand scheme 16:10:26 great point 16:10:54 so I think if we get one or two respondents _this year_, that’s a success 16:10:57 (unfortunately) 16:11:01 agreed 16:11:10 and i can send the email again when we are like 1 week out from the ptg 16:11:15 it’s useful to have the service written up, something we can point to 16:11:18 that’s a good idea 16:11:35 #action elmiko send ptl email again 1 week before ptg 16:11:40 just so i don't forget 16:11:41 (I’ve remembered where edleafe is: he had an appointment and will not be here today, I’m supposed to assign all actions to him) 16:11:49 ✔ 16:11:49 LOL 16:11:52 :D 16:12:04 #action edleafe take care of stuff 16:12:11 +2W 16:12:15 hahaha 16:12:17 and so it shall be 16:12:24 moving along 16:12:24 #topic open mic and new biz 16:12:45 anything here to review? 16:12:47 ed has the ball on SIG namign stuff 16:12:52 I don’t know the state 16:12:57 ack 16:13:06 for posterity 16:13:08 #link https://etherpad.openstack.org/p/api-ptg-queens 16:13:33 elmiko are you signed up for ptgbot: http://ptg.openstack.org/ptg.html 16:13:40 I think we got edleafe signed up 16:14:02 i looked at it, don't think i'm authorized to do anything though 16:15:28 do i need to make a request in the ptg channel? 16:15:30 you have to ask in #openstack-prg 16:15:37 sorry #openstack-ptg 16:15:51 ack 16:15:53 thanks 16:16:14 elmiko: I can get you access in a bit once you hop in the channel. 16:16:34 elmiko: you’ll like the bot because it says “ack” a lot 16:16:36 ah, sweet 16:16:41 =) 16:16:50 gotta ack 16:16:59 diablo_rojo_phon: i just dropped a message in there 16:17:14 ok, anything else we should cover here? 16:17:27 actually, i just thought of something 16:17:31 api-sig diner 16:17:34 dinner 16:17:41 I’m thinking monday is best 16:17:49 did we come up with any sort of agenda for talking with the sdk folks? 16:17:57 agreed, monday works for me 16:18:21 as far as I can tell flanders put more stuff on the sdk etherpad (linked from that series of email) but I got no real sense of next actions 16:18:22 currently the ironic folks also seems to prefer monday for our dinner.. 16:18:32 but this may change (and you may ignore me anyway) 16:19:14 aw, it wouldn’t be the same without you 16:19:25 maybe sunday night? 16:19:27 but I think elmiko may be leaving on tuesday? and I’m busy sunday 16:19:36 yeah, i'm out tuesday afternoon 16:20:18 ok, well if you don't make the dinner dtantsur, maybe you can make the api-sig barcrawl 16:20:19 monday it is 16:20:36 okay, maybe it's not going to end up on Monday. the ironic party, I mean.. 16:20:39 various emails suggest that where the hotel is is a bit of a wasteland 16:20:54 but we’ll cross that bridge closer to the time 16:21:02 #agreed api-sig ptg dinner on monday night 16:21:10 sounds good 16:21:20 it looked a wasteland on the google street view indeed 16:21:24 speaking of: I will miss the next two meetings. I’ll be flying the next one, and at some “edge” conference the next one. 16:21:40 noted 16:21:46 #topic guidelines 16:22:00 #link https://review.openstack.org/#/q/status:open+project:openstack/api-wg,n,z 16:22:07 #link https://review.openstack.org/#/q/status:open+project:openstack/api-sig,n,z 16:22:27 the extensions peice is coming along nicely 16:22:32 the only one in any reviewable state is mine and I think that needs more local review before being frozen 16:22:34 jinx 16:22:35 should we wait for more +1s before going to freeze? 16:22:36 haha 16:22:41 i 16:22:48 i'm good with waiting for more reviews 16:23:26 and i guess at some point we'll need to circle back around on mordred's PRs? 16:24:22 yes, I was thinking about that this morning and wondered whether one of us needs to pick them up, but then decide there was too much other stuff on my radar to think about that 16:24:27 (so stopped) 16:24:53 well, at the least we could talk about it at the ptg (assuming no movement by then) 16:24:59 tru 16:25:05 i fear i'm in the same boat as you, work-wise 16:25:47 speaking of ptg topics, I wrote this as sort of a get my thoughts aligned for when people ask about them: https://anticdent.org/microversion-conversion-therapy.html 16:26:17 very groovy 16:26:34 should we post a sign about "microversion therapy $0.05" ? 16:27:00 sure, why not 16:27:41 ok, so that can also be a possible topic if we get enough people interested 16:27:56 yeah it’s already on the etherpad 16:28:06 awesome 16:28:22 #topic bug review 16:28:32 #link https://bugs.launchpad.net/openstack-api-wg 16:28:38 #link https://bugs.launchpad.net/openstack-api-sig 16:28:41 I closed a bug (the placement needs a service type one) 16:28:45 and I added some comments on https://anticdent.org/microversion-conversion-therapy.html 16:28:49 whooops not that 16:28:56 hehe 16:28:57 #link singular resource https://bugs.launchpad.net/openstack-api-wg/+bug/1593310 16:28:58 Launchpad bug 1593310 in openstack-api-sig "guidance required on the basic form for representing a singular resource" [Undecided,New] 16:29:06 yeah, you've been burning down our tech debt 16:29:09 cdent++ 16:29:26 that one may need some discussion, there are effectively three options: top level, nested by type, something like jsonapi 16:29:42 lifting and shifting people to jsonapi is likely impossible (also it is weird) 16:29:50 we have a mixture of the other two options throughout openstack 16:29:52 yeah, i could see that 16:30:23 when you say "nested by type", that is a resource contained in a json object with it's key being the type? 16:30:50 {"mytype": ....} 16:30:50 yeah top-level: /foo/id -> {‘name’: ‘barney’} 16:30:57 ack 16:31:08 i really like your note about getting some outside input 16:31:10 nested: /foo/id -> {‘foo’: {‘name’: ‘barney’}} 16:31:33 i have a feeling i know what i'd push as a guideline, but i like hearing other opinions 16:31:37 jsonapi: /foo/id -> {data: {‘type’: ‘foos’, ‘name’: ‘barney’}}\ 16:32:01 (and there are other keys besides data 16:32:06 right 16:32:38 i've looked at jsonapi, like you mention it would be nearly impossible to get that across the projects 16:32:40 I don’t like nested because I think that’s an indicator that you must have bad URIs, but then again, it is all over the place 16:33:00 interesting 16:33:11 i thought nested would be better as it's more explicit 16:34:11 I think it probably really comes down to what I’m used to/used before 16:34:20 gotcha 16:34:23 and no _real_ reason 16:34:44 yeah, i've heard explanations about why nested might be preferred, but never about top level 16:34:59 and the nested explanations probably also suffered from perception bias 16:36:10 probably strong typers! 16:36:21 exactly 16:36:23 * elmiko chuckles 16:36:53 anyway, I’ll link this to that bug, so we have some additional info 16:36:59 great, thanks 16:37:01 and try to incorporate what I hear from others 16:37:18 if we get lonely at ptg, we can start an argument about it XD 16:37:38 arguments++ 16:37:46 hehe 16:38:08 ok, then, moving along 16:38:13 I suspect it is my turn for the newsletter? 16:38:30 #topic weekly newsletter 16:38:40 #link https://etherpad.openstack.org/p/api-sig-newsletter 16:38:48 or we can dice off for it or something? 16:39:21 heh, unless you for some reason really want to do it I’ll get it this time since I’ll be gone for the next two 16:39:44 sounds good 16:39:45 thanks 16:40:18 well then, safe travels cdent and we'll see you in denver =) 16:40:20 i’ll ping you when I need a proof 16:40:25 yeah, looking forward to it 16:40:26 thanks dtantsur 16:40:29 great cdent 16:40:35 #endmeeting