16:00:15 <etoews> #startmeeting api wg
16:00:16 <openstack> Meeting started Thu Nov  5 16:00:15 2015 UTC and is due to finish in 60 minutes.  The chair is etoews. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:17 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:19 <openstack> The meeting name has been set to 'api_wg'
16:01:16 <edleafe> o/
16:01:20 <etoews> anyone around? i wouldn't be surprised if there isn't...
16:01:23 <etoews> ohai
16:01:33 <edleafe> hey - missed you in Tokyo
16:01:50 <etoews> ya. couldn't make it out there.
16:01:54 <etoews> it was great
16:02:22 <etoews> i'm briefly distracted here...
16:02:25 <edleafe> I did have fun hanging out at the RAX booth
16:07:34 * cdent is suffering at the hands of a world wide conspiracy
16:09:19 <etoews> cdent: edleafe: considering it's the week after the summit there are many folks around, is there anything in particular you want to discuss?
16:09:48 <cdent> elmiko had some pretty exciting ideas from our session
16:09:49 <etoews> s/are/aren't/
16:09:57 <cdent> which was kind of small because nobody could find it
16:10:21 <etoews> cool. too bad about the smallish turnout.
16:10:25 <edleafe> I feel like I lost touch with all the summit craziness, and wanted to see where the wg was
16:11:36 <cdent> It was in building number #5. Attendees: elmiko, lucasgomes, cdent, kenichi, and I think that was miguelgrinberg
16:12:41 <cdent> The main point of discussion were ways to create things that demonstrate best practice as a way of getting more active contributions and discussion
16:17:59 <etoews> that sounds like a good way to go.
16:21:42 <cdent> elmiko promised to make with some mail and assorted other writings when he returned (he was taking an extra week in japan)
16:22:37 <etoews> good for him. let's catch up with him when he gets back.
16:23:18 * etoews takes a look at miguelgrinberg's action guideline https://review.openstack.org/#/c/234994/2/guidelines/actions.rst
16:32:07 <sdague> so, something of note, I know at least on the Nova front we're kind of in hold on a couple of things until the error spec and the filter spec land
16:32:47 <sdague> I know there are lots of parallel threads in the api-wg, but I think it would be good to focus on a few long standing bits that should get finalized
16:33:00 <sdague> the error spec in particular
16:37:39 <cdent> sdague is right
16:37:54 <etoews> sdague: thanks for the feedback! i wasn't aware of the hold up for nova.
16:37:58 <cdent> one of the other things we discussed was a desire to use the mailing list more to get people involved
16:38:06 <cdent> enhance visibility
16:39:29 <sdague> so... I think at some point I lost the irc channel in bouncer upgrades. Is there a reason we don't pile this group into the sdk channel?
16:39:41 <sdague> because it seems like it would gather more of all the right people in one place
16:39:53 <sdague> and maybe help on critical mass bits
16:40:47 <etoews> i'd be amenable to that as -sdks is my main hangout anyway.
16:41:22 <etoews> we certainly don't have quorum here to make such a decision though
16:41:28 <edleafe> sdague: I think that would work
16:42:47 <etoews> sdague: regarding the error guideline, i'm willing to push past the insistence that that guideline also be responsible for the definition of the Error Codes. the only other thing holding it up is the linking format.
16:43:33 <etoews> currently the guideline just has a href in the body of the json response
16:43:59 <sdague> etoews: yeh, so how about gain quorum about irc change via ML
16:44:05 <etoews> _elmiko rightfully pointed out that that really isn't inline with any of the other linking formats used in openstack.
16:44:45 <etoews> he was going to come up with a guideline for a linking format and then i'd follow that format.
16:44:52 <etoews> then the summit happened.
16:44:57 <sdague> on the error guideline, I'll take another look this week, but we do need to come up with a thing there, because Nova team would like to implement that as a microversion probably very early in N (this cycle is more about docs)
16:45:34 <sdague> but to land it then, it's going to be a bunch of run up work, as we're going to want to bring that in for all the resource errors in one go
16:45:39 <etoews> looks like he did some analysis here https://wiki.openstack.org/wiki/API_Working_Group/Current_Design/Links
16:46:33 <etoews> i'm willing to move forward on the error guideline in short order.
16:47:36 <sdague> hmmm... that's not quite right, the nova version doc links to docs.openstack.org
16:48:25 <sdague> ok, I'll comment again on this.
16:49:34 <etoews> the predominant style seems to be of the form "links": [{"href": "", "rel": "self"}]
16:50:01 <etoews> sdague: comments are appreciated.
16:52:18 <sdague> right, sure, except when referencing docs
16:52:30 <sdague> that's a resource link, not a doc link
16:52:45 <etoews> true enough
16:53:23 <etoews> although a doc can certainly be considered a resource. just one that's outside the domain of the api...
16:55:05 <sdague> that seems to be spliting hairs in very delicate ways
16:55:31 <sdague> ok, looking at the error spec, I'm still not happy with it because of the uuid ids
16:55:44 <sdague> because that's just busy work that everyone's going to have to look up
16:56:23 <etoews> the id field? it's just the Openstack-Request-Id.
16:56:57 <sdague> hmmm
16:57:20 <etoews> well, anyway...best to move this discussion into the review comments or ping me in -api or -sdks for sync discussion.
16:57:24 <sdague> yep
16:57:56 <etoews> i'm going to end the meeting a couple min early as i have another meeting to attend. wheeeeee!
16:58:01 <sdague> etoews: lets follow up now in -sdk just so I can correctly collect my thoughts
16:58:15 <etoews> sure
16:58:16 <etoews> #endmeeting