16:00:13 <cdent> #startmeeting api-wg
16:00:13 <openstack> Meeting started Thu Feb 16 16:00:13 2017 UTC and is due to finish in 60 minutes.  The chair is cdent. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:14 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:16 <openstack> The meeting name has been set to 'api_wg'
16:00:21 <elmiko> hi
16:00:28 <edleafe> \o
16:00:35 <cdent> hello, who besides elmiko has shown up for today's exciting episode of api-wg?
16:00:50 <cdent> and mr leafe
16:00:55 <cdent> anyone?
16:01:14 <cdent> #link today's agenda https://wiki.openstack.org/wiki/Meetings/API-WG#Agenda
16:01:24 <cdent> #chair cdent elmiko etoews edleafe
16:01:24 <openstack> Current chairs: cdent edleafe elmiko etoews
16:01:40 <cdent> no action items from last meeting so we'll skip straight into new biz and open mic
16:01:41 <edleafe> the excitement is probably too much for people
16:01:45 <cdent> #topic open mic and new biz
16:01:47 <elmiko> lol
16:01:48 <edleafe> we should scale back
16:01:55 <cdent> it is pretty intense
16:02:16 <cdent> ptg planning
16:02:20 <cdent> #link ptg etherpad https://etherpad.openstack.org/p/ptg-architecture-workgroup
16:02:29 <cdent> we're sharing etherpad and space with arch-wg
16:02:37 <cdent> we have a room on monday now
16:02:43 <cdent> and there are many topics
16:02:53 <cdent> but it is also likely that thing we care about will overlap with other stuff on monday
16:02:55 <edleafe> I added a $BEVERAGE topic at the bottom
16:03:14 <elmiko> haha, ice
16:03:14 <cdent> ah, good one
16:03:18 <elmiko> nice even...
16:03:43 <cdent> I'm concerned about how we're going to manage rooms, topics, getting people's attention
16:03:45 <elmiko> that last question is like the eternal struggle
16:03:48 <cdent> but I guess we'll just wing it
16:03:53 <edleafe> Well , if IBM keeps on their current path, it might be only elmiko left
16:04:11 <elmiko> ALL THE POWA!!!!
16:04:14 <edleafe> cdent: the lack of room planning is making a lot of people nervous
16:04:27 <edleafe> not just our group
16:05:00 <cdent> edleafe: yeah, I guess lots of pinging in #openstack-ptg will be required, assuming the wifi works
16:05:15 <elmiko> what's the issue with the room planning?
16:05:40 <cdent> of course EmilienM and I have decided we're spending the entire week in a room with gin and tonic that we'll be distributing to people willing to hang out and talk with us
16:05:42 <stevelle> sorry, also here fwiw
16:05:57 <cdent> stevelle, hello, thanks for coming
16:05:58 <elmiko> lol, nice cdent
16:06:03 <elmiko> makes me want to attend
16:06:07 <cdent> elmiko: basically there are many topics and lots of room
16:06:10 <stevelle> double-booked meetings
16:06:14 <elmiko> ah, gotcha
16:06:17 <cdent> and unclear on where to go for anything
16:06:34 <edleafe> tonic?? bleh. Give me gin and soda
16:06:36 <edleafe> :)
16:06:53 <cdent> edleafe: that's perfectly acceptable
16:07:06 <edleafe> This is a summit-like week, but we don't have the sort of room slots assigned
16:07:07 <EmilienM> lol
16:07:13 <cdent> I think we might have some bourbon too
16:07:45 <EmilienM> does it make sense to ask api-wg to help out in organizing https://etherpad.openstack.org/p/ptg-pike-wsgi ?
16:08:12 <cdent> Anyway... the action here is for people to add topics on the etherpad (linked above) for next week. I will strive to be sure that there are reasonable summary notes for monday and tuesday (as I promised to do so in return for my travel funding)
16:08:52 <cdent> EmilienM: if you want, but in general the group tries to be deployment agnostic and think about the http not the server itself
16:09:11 <cdent> however, since I've already made a lot of noise in support of that one, I'll want to participate
16:09:33 <cdent> scottda you here today?
16:10:04 <EmilienM> cdent: it makes sense.
16:10:29 <cdent> I think we're going to have to wing it on the capabilities and service endpoints discussions, and that will just have to be fine. Unconferencing.
16:10:52 <cdent> Anybody have anything else they would like to say about the ptg?
16:11:10 <elmiko> i hope it's productive =)
16:11:26 <elmiko> but not too productive XD
16:11:43 <stevelle> nothing
16:11:46 <stevelle> to add
16:11:56 <cdent> heh
16:12:00 <cdent> okay moving on
16:12:06 <cdent> #topic tc's decision about glance
16:12:11 <cdent> #link irc logs: http://eavesdrop.openstack.org/meetings/tc/2017/tc.2017-02-14-20.01.log.html#l-93
16:12:16 <cdent> edleafe your floor
16:12:35 <edleafe> Oh, now you want me to sweep it?
16:12:51 <cdent> yes, but if you do, you'll end up meeting someone at the ball
16:13:04 <edleafe> Anyway, since Glance and QA were deadlocked even after out input, the matter was brought to the TC
16:13:34 <edleafe> The TC noted that both sides had done what was expected:
16:13:45 <edleafe> QA being conservative on change
16:14:03 <edleafe> Glance coming to us for API opinion, and then escalating to TC
16:14:34 <edleafe> Some concern about having to escalate was raised, but most people were "eh, what can you do? Disagreements happen"
16:15:20 <edleafe> The best part was that several TC members expressed the opinion that they would undercut the legitimacy of the API-WG if they ignored our recommendation
16:15:37 <edleafe> They really want us to be a voice for positive change in the OpenStack community
16:15:56 <edleafe> that's about all I have to add
16:15:59 * cdent kicks edleafe out of api-wg
16:16:04 <elmiko> that's encouraging, about the support from tc
16:16:09 <elmiko> lol, ouch
16:16:18 <cdent> if we need positive voice...
16:16:23 * cdent resigns from api-wg
16:16:25 <edleafe> aw, who wants to be in your stupid group anyway!
16:16:26 <elmiko> haha
16:16:30 * cdent looks at elmiko
16:16:50 <elmiko> well...
16:16:50 * cdent recovers
16:17:13 <cdent> Yeah, it does seem like a good outcome.
16:17:20 <cdent> There's concern about setting precedents, but I think that's just life.
16:17:55 <edleafe> I think that the precedent that was set was that the API-WG knows what it is doing :)
16:17:57 <cdent> mordred: here
16:18:01 * mordred waves
16:18:12 * edleafe waves back
16:18:21 <stevelle> I was against the change in my role inside Glance. I missed the chance to repeat that here. I was concerned about precedent too, but am hopeful based on the discussion re: re-validate api change guidelines
16:18:47 <stevelle> I feel like the new guidelines state clearly on the tin that this change voids the warranty
16:19:01 <cdent> stevelle: it's all a bit difficult when a project is not actively versioning
16:19:37 <cdent> the guidelines, as written old and new, kind of assume that as a foundation
16:19:44 <edleafe> stevelle: yeah, it's hard to take an absolute stand when the basic versioning support isn't there
16:19:51 <stevelle> well, that project tag will be unobtainable for glance w/o versioning
16:20:07 <stevelle> that says what needs to be said
16:21:29 * cdent nods
16:21:58 <edleafe> stevelle: the other thing is that we don't want Glance to feel like they can keep coming to us to "bail them out"
16:22:13 <stevelle> +100
16:22:16 <edleafe> stevelle: we'd really like to see a strong commitment to versioning by Glance
16:22:26 <stevelle> that may be a bridge too far, for now.
16:22:38 <edleafe> understood
16:22:43 <stevelle> that is a serious anchor to throw at a small team of contributors
16:23:15 <stevelle> and I project problems will continue for them in keeping contributors
16:23:36 <edleafe> unfortunately, that's true of just about every project
16:23:39 <stevelle> images doesn't attract a broad bench
16:23:58 <edleafe> .
16:24:02 <edleafe> doh!
16:24:22 <elmiko> a real bummer considering how crucial glance is
16:24:25 <mordred> ++
16:24:35 <edleafe> Well, if you called them "container images" people would get all excited!
16:24:39 <elmiko> LOL
16:24:51 <elmiko> "glance, now with more containers and versioning!"
16:25:03 <stevelle> this brings up a topic we can discuss more at ptg, but in terms of leveling the playing field the api stability project tag will be out of reach for a lot of teams given the high standard we are setting, but I don't know that can be helped.
16:25:05 * cdent weeps
16:25:24 <stevelle> it should be a high standard.
16:25:28 <elmiko> +1
16:25:29 <mordred> I think it's honestly better to have a tag that means something
16:25:31 <cdent> (at containers, not the bar)
16:25:38 <mordred> cdent: ++
16:25:58 <stevelle> Probably good to recognize it is such though
16:25:59 <edleafe> mordred: yes!
16:26:21 <cdent> stevelle: I think it is okay for the asserts tags to not be asserted
16:27:01 <edleafe> I added the interop topic to the etherpad because I think that this will have more of an effect on API stability guidelines than anything
16:27:02 <stevelle> I think some of the existing tags have quite high standards already, so this isn't a new precedent.
16:27:25 <elmiko> yeah, i'm all for tags that actually mean something
16:27:40 <edleafe> Tags were supposed to be a way of letting consumers what to expect
16:27:40 <mordred> yah - also I think everyone _wants_ to have and meet high standards - even if it'll be hard to get there for a bit
16:28:10 <cdent> raising standards ++
16:28:13 <edleafe> In this case, "can I be assured that my API calls won't change?"
16:28:39 <edleafe> "or do I have to be more defensive in my programming?"
16:28:41 <stevelle> thanks for confirming my thinking on that. I'm satisfied with the consensus so far.
16:29:17 <cdent> edleafe: or how about "or do I need to be casual about my api code and make it easy to change"
16:29:31 <cdent> (just for completeness)
16:29:39 <edleafe> cdent: that's more of a dev-centric viewpoint, no?
16:29:55 <edleafe> I'm thinking about people _using_ OpenStack
16:30:03 <edleafe> consuming the APIs, not writing them
16:30:17 <cdent> yes, that's what I mean s/api code/api client code/
16:30:34 <cdent> I'm not asserting that as a correct viewpoint, just that it _is_ a viewpoint
16:30:48 * mordred loves the people who consider the people consuming the APIs - hands everyone a pie
16:31:01 <edleafe> cdent: OIC. It's simply a matter of setting the proper expectation for consumers
16:31:30 <cdent> mordred: I'm in cornwall, can I have a pasty instead please?
16:31:31 <elmiko> mmm pie
16:31:39 <mordred> ++ - for instance - if we say "just check success is in the 200 range, we're loose with specific codes" - then people can know how to behave
16:31:45 <edleafe> It's the TC's way of saying "this project is ready for prime time" with regard to a particular thing
16:31:46 <elmiko> cdent: no fish pie in cornwall XD
16:31:48 <elmiko> ?
16:31:56 <cdent> stargazy pie
16:32:44 <cdent> any further points to make on this topic?
16:33:37 <cdent> k, any further open mic or new biz topics?
16:34:09 <stevelle> not me, I'm ready for that pie
16:34:13 <cdent> #topic guidelines
16:34:18 * edleafe has already finished his pie
16:34:21 <cdent> #link https://review.openstack.org/#/q/status:open+project:openstack/api-wg,n,z
16:34:55 <cdent> boolean names is loved, post freeze, so I think we can merge, yes?
16:34:57 <cdent> #link https://review.openstack.org/#/c/411529/
16:35:02 <elmiko> cdent: seems so
16:35:29 <edleafe> and it disagrees with mordred, so that's even better :)
16:35:35 <elmiko> ouch
16:35:48 <cdent> the other things are under construction, or idle (pagination)
16:35:55 <cdent> What to do about pagination?
16:36:23 <elmiko> that one has been stagnant for a while?
16:36:26 <edleafe> Is there any further disagreement?
16:36:30 <mordred> edleafe: boo. I lost
16:36:34 * edleafe hasn't read that one in a while
16:37:01 <edleafe> My only objection was the tone of the one paragraph, which I re-wrote
16:37:17 <elmiko> hmm, i guess we just need more reviews on it
16:37:44 <cdent> it has un-responded comments, I guess
16:38:22 <edleafe> So let's respond, and see what happens from there
16:39:19 <cdent> edleafe: are you in a position to consider yourself the author now or should we try to drag sigmavirus back into the flow
16:39:40 * sigmavirus hides
16:39:41 <elmiko> edleafe: +1
16:40:07 <edleafe> dunno, we may have scared sigmavirus away. He's awfully sensitive like that :)
16:40:18 <sigmavirus> :D
16:41:05 <elmiko> lol
16:41:35 <edleafe> cdent: I only wrote one paragraph of that, so co-author at best
16:41:47 <edleafe> cdent: I can be the steward, though
16:41:54 <cdent> shall we move on to bugs or do we need to counsel edleafe and sigmavirus to take some action
16:42:03 <cdent> edleafe: that's what I meant: somebody needs to ride it in
16:42:04 <edleafe> more pie would help
16:42:23 <cdent> I will get you pie next week. I think it is spelt gee eye enn
16:42:30 * mordred witholds pie from edleafe who witheld his support of mordred's view on boolean field naming
16:42:39 <cdent> HARSH
16:42:43 <edleafe> pie.enabled = False
16:42:48 <elmiko> haha
16:43:06 * edleafe thinks that's nicer than pie.is_enabled = False
16:43:08 <mordred> edleaf.has_pie = False
16:43:20 <mordred> edleafe.has_pie = False
16:43:22 <mordred> darned pep8
16:43:23 <cdent> mordred: clearly edleafe is the resource
16:43:34 <mordred> pie.has_edleafe = False
16:43:48 <mordred> which is very different than pie.is_edleafe = True
16:43:50 <edleafe> edleafe.has_hunger = True
16:44:07 <cdent> #topic bug review
16:44:09 <elmiko> yeah, this is not helping in a pre-lunch meeting
16:44:10 <cdent> there are no new bugs
16:44:19 <edleafe> elmiko: heh
16:44:31 <cdent> but there are many that still remain but we've all been too of late to do much with
16:44:46 <cdent> this touches on the question edleafe added to the ptg etherpad: new meat!
16:45:22 <cdent> comments questions on that topic?
16:45:33 <cdent> #topic weekly newsletter
16:45:39 <cdent> #link https://etherpad.openstack.org/p/api-wg-newsletter
16:47:49 <edleafe> wow, that quieted everyone!
16:48:05 * stevelle doesn't talk while he's eating pie
16:48:10 <elmiko> haha
16:51:04 <elmiko> is it worth mentioning the whole glance/tc thing in there?
16:51:17 <elmiko> you guys are having way too much fun with etherpad XD
16:51:40 <cdent> the newly published is what went from frozen to merged
16:51:46 <edleafe> elmiko: join us!
16:51:59 <elmiko> i'm watching =)
16:55:49 <elmiko> ship it!
16:55:50 <edleafe> Looks good to me
16:56:31 <edleafe> (as if anyone reads these things!)
16:56:53 * edleafe keeps reading about mascots and other critical issues
16:57:12 * cdent reconsiders edleafe as a force for positive change
16:57:53 * edleafe is surprised he fooled cdent for so long
16:58:29 <cdent> it's sort of enforced ignorance
16:58:39 <cdent> anyway, sent, and good timiing
16:58:49 <cdent> anyone want to fill hat last minute and a half?
16:59:05 <cdent> oh noes
16:59:08 <cdent> look what I did
16:59:15 <cdent> I left the "Subject:" in the subject
16:59:16 <cdent> :(
16:59:27 * cdent tenders resignation
16:59:33 <cdent> thanks everyone for coming
16:59:40 <edleafe> I did that too
16:59:47 <cdent> those of you will be there next week, see you there
16:59:56 <cdent> those of you who will not, you will be missed
17:00:03 <cdent> and thanks for coming today
17:00:05 <cdent> #endmeeting