16:00:15 #startmeeting api-wg 16:00:16 Meeting started Thu May 25 16:00:15 2017 UTC and is due to finish in 60 minutes. The chair is elmiko. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:18 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:19 ha 16:00:19 or elmiko 16:00:20 The meeting name has been set to 'api_wg' 16:00:21 win 16:00:23 o/ 16:00:25 #chair cdent edleafe 16:00:26 Current chairs: cdent edleafe elmiko 16:00:37 ok, how does this thing work again 16:00:41 * elmiko cracks fingers 16:00:42 sorry, was busy in -nova 16:00:58 s/was/am 16:01:16 #link https://wiki.openstack.org/wiki/Meetings/API-WG#Agenda 16:01:26 #topic previous meeting action items 16:01:38 i think the main one was reading mordred's prs 16:02:21 o/ 16:03:06 has everyone had a chance to read the prs starting with https://review.openstack.org/#/c/459405 16:03:10 #link https://review.openstack.org/#/c/459405 16:03:22 * mordred hopes everyone enjoys long novels 16:03:31 new versions were just posted, so I'm not up to date with today 16:03:41 cdent: mostly tiny changes from last versions 16:03:50 on the whole they look good to me 16:03:53 1% of very big is big 16:03:55 yeah, I'm behind in my reading due to PyCon travel 16:03:59 cdent: I fixed a typo you pointed out and changed the URL suggetsed for service-types publication 16:04:08 there were some minor questions, but it seems like a win for me 16:04:19 * dtantsur is half way through the novel 16:04:27 dtantsur made a good suggestion on the first one based on the recent ironic thread, so I'm going to try to incorporate that today 16:04:43 sweet, i think that was one of the main questions 16:04:55 cool 16:04:58 but I'll wait a bit to see if there are any other feedbacks on the stack so we can take care of them all in one pass 16:05:20 mordred: +1 16:06:03 also - related but not related - I pushed up a patch to system-config for the creation of a new mailing list for SDK authors and API consumers: 16:06:06 https://review.openstack.org/#/c/468046/ 16:06:22 interesting... 16:06:36 this stems from the fact that David Flanders and I have an email chain going with a bunch of the SDK authors ... and maybe getting that intoa mailng list is better than private emails 16:06:44 that's a good idea 16:06:52 nice 16:07:05 i'm glad to see that "everyone has to be on os-dev or not at all" is finally dying 16:07:06 especially to cover less-known SDK authors 16:07:09 they're all, so far, excited about us reaching out and interested in working more closely 16:07:11 huge +1 to get that activity on a list somewhere 16:07:13 dtantsur: ++ exactly 16:07:27 did anyone know there are 2 active java sdks? 16:07:39 i didn't 16:07:39 * dtantsur did not 16:07:42 or that the php sdk is actuall pretty solid? 16:07:48 me either 16:08:33 * dtantsur is interested what kind of applications people create with the PHP SDK. hopefully, some additions to Nextcloud or something like that. 16:08:41 mordred: I left a comment: the name of the list is no good 16:09:21 cdent: awesome 16:09:24 with that name the audience is huge, and that appears contrary to what you're going for 16:09:32 dtantsur: I was on the SDK team at Rackspace years ago. The PHP SDK was pretty popular 16:09:41 interesting 16:09:51 cdent: openstack-api-abusers? 16:09:51 dtantsur: I believe owncloud is using it as well 16:10:21 cdent: well - so definitely some help on name/description is needed here ... 16:10:23 \o/ 16:10:30 i'm both scared and impressed about the php sdk 16:10:32 * cdent will think 16:10:52 I think the target audience is "anyone interacting with the REST API directly rather than via an SDK - mostly relevant for SDK authors" 16:11:10 openstack-not-only-in-python@lists.o.o (though it will attract swift folks too) 16:11:10 or "SDK authors - but people with a deep interest in consuming the REST API directly may be interested" 16:11:25 dtantsur: LOL yes! 16:11:32 some people consume API directly, they should be included too 16:11:51 maybe just openstack-sdk-authors? and if you're a power-REST consumer you can join too 16:12:03 i think that's reasonable 16:12:07 openstack-sdk-authors++ 16:12:16 if you consume the api raw, you are your own sdk-author 16:12:19 if you write an app that consumes the API directly, you're writing an SDK with a user count of 1 16:12:23 jinx 16:12:26 ish 16:12:32 haha 16:12:36 :) 16:12:39 heh 16:12:52 openstack-sdk-authors seems reasonable to send up the flag pole 16:13:47 ok, moving along 16:13:49 #topic open mic and new biz 16:14:06 i think we just talked about mordred's changes, anything more on that? 16:14:20 more reviewing to happen, but seems we are getting close 16:14:20 so 16:14:22 yay 16:14:43 #action everyone re-review mordred's changes begining with https://review.openstack.org/#/c/459405 16:14:51 #undo 16:14:51 Removing item from minutes: #action everyone re-review mordred's changes begining with https://review.openstack.org/#/c/459405 16:14:56 #action everyone re-review mordred's changes beginning with https://review.openstack.org/#/c/459405 16:15:00 thank you all again for reviewing such long and involved things. I really appreciate it and recognize just how much content it is 16:15:17 no no, thanks to you. that's a boat load of work =) 16:15:43 is there any other new business folks would like to bring up? 16:16:04 are people aware of the request-id changes that sdague has in progress? 16:16:11 * dtantsur is not 16:16:16 neither am i 16:16:32 vaguely 16:16:33 i assume this is a result of the logging discussions from summit? 16:16:34 ttps://gist.github.com/sdague/eb91caee9ce27c4d0ae913e5743b4416 16:16:37 it's going to make it possible and easy to pass a request id from one service to another so that logs have both a global and local request id 16:16:48 neat 16:16:49 thanks sdague 16:17:06 ++ 16:17:11 also http://lists.openstack.org/pipermail/openstack-dev/2017-May/117367.html 16:18:19 sdague: this might be off topic - or not relevant ... 16:18:40 mordred: go for it 16:18:43 sdague: but in putting that together did you happen to look at the opentracing/openzipkin work out there? 16:18:52 no 16:19:17 it seems very close at least problem-space-wise - but it's also possible that impl details make it irrelevant 16:19:17 there is already osprofiler if you are going that deep 16:19:36 this is litterally make grep and ELK work 16:19:47 totally 16:20:02 I was mostly thinking on the terminology/concepts front - not code 16:20:07 sure 16:20:23 but like I said - maybe not relevant - just wanted t bring it up in case it was 16:20:26 well, the ship has sailed on all of those parts as it's half landed 16:20:30 kk 16:20:54 the oslo infrastructure is done, it's just consumption, which is a bunch of small patches 16:20:56 what's been done is really nicely simple and hugely valuable 16:21:22 sweet 16:22:01 elmiko, edleafe: i have to go, let me know if I end up needing to do something 16:22:09 it's all mostly a project management task actually, which is why I've got the big list of steps 16:23:53 anymore on this ? 16:24:24 ok, then 16:24:26 #topic guidelines 16:24:34 #link https://review.openstack.org/#/q/status:open+project:openstack/api-wg,n,z 16:24:44 i think the only new stuff are mordred's changes 16:24:51 is there anything else we need to address? 16:25:00 not from me 16:25:36 #topic bug review 16:25:44 #link https://bugs.launchpad.net/openstack-api-wg 16:26:21 doesn't seem like anything new there either 16:26:37 we should probably do a triage at some point 16:27:34 if there is nothing else, i guess we'll move on to the wrap 16:27:40 #topic weekly newsletter 16:27:52 #link https://etherpad.openstack.org/p/api-wg-newsletter 16:28:01 edleafe: are you still available for the newsletter? 16:28:46 I'm running out for a couple of hours. If it can wait, I'll do it when I return 16:29:21 i can do it, i'll ping on -sdks when i've got it ready 16:29:46 great. I won't be able to review it for a while (see above) :) 16:29:54 np 16:30:34 last call... 16:30:53 thanks everyone for attending! 16:30:59 #endmeeting