16:00:14 <cdent> #startmeeting api_sig
16:00:14 <openstack> Meeting started Thu Jun 28 16:00:14 2018 UTC and is due to finish in 60 minutes.  The chair is cdent. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:15 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:18 <openstack> The meeting name has been set to 'api_sig'
16:00:18 <cdent> #chart dtantsur edleafe elmiko
16:00:22 <elmiko> o/
16:00:23 <cdent> oops
16:00:29 <cdent> #chair dtantsur edleafe elmiko
16:00:30 <openstack> Current chairs: cdent dtantsur edleafe elmiko
16:00:40 <cdent> #link agenda https://wiki.openstack.org/wiki/Meetings/API-SIG#Agenda
16:00:43 <dtantsur> o/
16:00:46 <edleafe> \o
16:00:57 <cdent> #topic old biz
16:01:01 <cdent> #link last minutes http://eavesdrop.openstack.org/meetings/api_sig/2018/api_sig.2018-06-21-16.00.html
16:01:09 <edleafe> I was just typing #startmeeting when you did. Damn you fast typists!
16:01:16 <cdent> action items from edleafe (huzzah!)
16:01:23 <elmiko> huzzah!
16:01:30 <cdent> I think he did both
16:01:44 <cdent> the graphql people are aware of storyboard
16:01:57 <cdent> and the links on the agenda are correct
16:02:03 <elmiko> cool
16:02:14 <cdent> I learned a thing about storyboard: if you want gerrit to update it, you have to list both a story and a task
16:02:18 <cdent> just a story won't do it
16:02:25 <dtantsur> yep
16:02:27 <edleafe> #link email to the GraphQL folk: http://lists.openstack.org/pipermail/openstack-dev/2018-June/131756.html
16:02:31 <dtantsur> because you can have many tasks
16:02:44 <dtantsur> (you actually should, if you use it properly)
16:03:38 <cdent> yeah, the paradigm has shifted and does not match launchpad bugs well
16:04:02 <dtantsur> yep
16:04:13 <cdent> any other old biz?
16:04:31 <edleafe> nopd
16:04:32 <cdent> #topic new biz
16:04:34 <edleafe> nope, even
16:04:55 <cdent> any topics (not related to pending guidelines) people would like to bring up?
16:05:18 <elmiko> as edleafe pointed out in -sdks, we have a monday slot for ptg
16:05:23 <elmiko> just a heads up
16:05:40 <cdent> our usual then. I guess that's good
16:06:30 <elmiko> ++
16:07:02 <cdent> anything else?
16:07:25 <elmiko> nothing from me
16:07:39 <dtantsur> nothing
16:07:41 * edleafe yawns
16:07:46 <cdent> #topic guidelines
16:07:47 <cdent> #link https://review.openstack.org/#/q/status:open+project:openstack/api-wg,n,z
16:07:47 <cdent> #link https://review.openstack.org/#/q/status:open+project:openstack/api-sig,n,z
16:08:02 <elmiko> don't get me started edleafe !
16:08:03 <cdent> dtantsur: raised a good point on my new review
16:08:05 <elmiko> XD
16:08:14 <dtantsur> :)
16:08:17 <cdent> #link error links https://review.openstack.org/#/c/578369/
16:08:51 <cdent> when I initially wrote the thing I was trying to go down the road of being prescriptive about where the links were supposed to go, but I guess I overplayed my hand with the commit message, and maybe we should?
16:09:16 <elmiko> i like dtantsur's suggestion
16:09:37 <cdent> it is something that has come up before
16:09:48 <cdent> but has devolved into super upper ontology mess
16:09:58 <elmiko> i can see that
16:10:11 <elmiko> i just like that at the least it gives the individual projects the power to control their own docs
16:10:18 <elmiko> but, at the cost of centralization
16:10:40 <dtantsur> well, we've learned already that a centralized docs team does not scale well enough for our growth
16:10:51 * elmiko nods
16:11:13 <edleafe> So the other expense is duplication
16:11:25 <elmiko> i do kinda favor the idea of giving projects some guidance on /where/ they might place their error docs
16:11:28 <edleafe> IOW, each project can define their own errors, and they can overlap
16:11:31 <elmiko> that could help with centralization some day
16:11:38 <elmiko> edleafe ++
16:11:47 <dtantsur> edleafe: we prefix errors with service name, no?
16:12:02 <cdent> type
16:12:05 <edleafe> why would we do that if we aren't centralized?
16:12:11 <cdent> not name
16:12:19 <dtantsur> what I'm worried is compute.server-not-found verses baremetal.not-found.node vs image.image-was-not-found
16:12:33 <dtantsur> cdent: what was the TC position about nitpicks? ;)
16:12:56 <cdent> dtantsur: just trying to make sure we don't raise the ghost of christmas mordred
16:13:05 <dtantsur> well, now you did :D
16:13:08 <cdent> :)
16:13:13 * mordred throws a chicken at cdent
16:13:16 <dtantsur> but yes, I did mean type
16:13:25 * cdent cooks and eats chicken
16:13:27 <dtantsur> mordred: baked, raw or alive?
16:13:30 <mordred> dtantsur: yes
16:13:46 <dtantsur> lol
16:13:54 <elmiko> lol
16:14:01 <cdent> okay, if people could add any additional thoughts on the review, I'll cook a new version soonish
16:14:10 <edleafe> will do
16:14:12 <elmiko> ack
16:14:19 <dtantsur> as to "why prefix" - for easier googling?
16:14:39 <cdent> namespacing, basically
16:14:50 <cdent> which does also mean easier googling
16:15:17 <edleafe> I can just see the Nova cores being thrilled to have to add 'compute.' to everything :)
16:15:39 * elmiko chuckles
16:15:44 <dtantsur> hehe
16:15:44 <cdent> they have to add support for errors at all first
16:15:58 <elmiko> zing!
16:16:04 <edleafe> heh
16:16:05 <dtantsur> does anybody implement our error spec at all?
16:16:11 <cdent> placement has started to
16:16:16 <dtantsur> neat
16:16:23 <cdent> and then mordred shamed me into making it even better
16:16:33 <dtantsur> with a chicken?
16:16:40 <cdent> something foul
16:16:46 <cdent> anyway
16:16:55 <cdent> this ogther errors guideline needs some eyes too
16:17:05 <cdent> #link error code clarity: https://review.openstack.org/#/c/577118/
16:17:07 <elmiko> ouch, that was bad
16:18:13 * dtantsur wonders which part of our conversation elmiko considers bad; presumably cdent's spec
16:18:24 <dtantsur> cdent: I have no concerns with this one
16:18:33 * edleafe is reading it
16:18:41 <elmiko> "something foul" after the chicken comment XD
16:19:06 <elmiko> that review lgtm
16:19:12 <cdent> I couldn't help msyelf
16:19:17 <elmiko> i know =)
16:19:53 * dtantsur suspects some euphemisms he does not know
16:20:07 <cdent> dtantsur: fowl and foul
16:20:26 <cdent> is _really_ bad and lame
16:20:27 <dtantsur> lol ok
16:20:30 <elmiko> haha
16:20:49 <cdent> so no immediate readiness on guidelines then, yes?
16:21:09 <elmiko> maybe that second one you linked, seemed low risk
16:21:14 <dtantsur> I's say the latter is ready
16:21:15 <elmiko> i added my +1
16:21:23 * cdent looks at edleafe
16:21:43 <edleafe> Just one thought, you gave an example, but not the example codes to show the difference
16:22:09 <cdent> true
16:22:18 <cdent> that's certainly easy to fix if we want
16:22:25 <edleafe> So it feels like an incomplete example
16:22:48 <cdent> k, put a -1 on it so I can fix that and I will (otherwise I'll forget)
16:23:05 <edleafe> I'm adding the comment now
16:23:08 <cdent> rad
16:23:14 <elmiko> agreed with edleafe, that's a nice upgrade
16:23:19 <cdent> #topic bug review
16:23:20 <cdent> #link https://storyboard.openstack.org/#!/project/1039
16:24:00 <cdent> based on those two changes is the bug at https://storyboard.openstack.org/#!/story/1593327 nearly addressed, or do we want more?
16:24:23 <elmiko> hmm
16:24:39 <elmiko> i think so
16:24:49 <cdent> I've added some stuff, but not "vision" but I'm not sure "vision" is really required
16:25:05 <elmiko> i mean, we could add more. but at the least this gives some insight
16:25:22 * cdent nods
16:25:36 <elmiko> i'm ok with closing that bug
16:25:41 <edleafe> me too
16:26:08 <cdent> I've made a "board" to experiment with that, but not learned the details of how such things work, yet
16:26:12 <cdent> #link storyboard board https://storyboard.openstack.org/#!/board/91
16:26:56 <cdent> i'm unclear on who can see that or change it or how to manipulate it
16:26:56 <edleafe> sweet
16:27:05 <cdent> but we'll figure it out I suppose
16:27:11 <elmiko> i can see it, not sure about manipulate, i'm not that bold
16:27:17 <dtantsur> ditto
16:27:26 <dtantsur> I think the purpose of boards to do things like release tracking
16:27:29 <edleafe> I don't see any way to change it
16:27:36 <dtantsur> I'm not sure if it's so valuable for us here
16:28:24 <cdent> yeah, I'm not sure either
16:28:33 <cdent> I can add individuals as "users"
16:28:36 <cdent> and "owners"
16:28:42 <cdent> we'll work it out, no rush
16:28:44 <cdent> could be useless
16:29:42 <edleafe> Seems more like a kanban-ish sort of thing
16:29:48 * cdent nods
16:29:52 <elmiko> yea
16:29:52 <dtantsur> like trello
16:30:14 <cdent> anything else on bugs?
16:31:13 <elmiko> nothing from me
16:31:14 <cdent> #topic weekly newsletter
16:31:14 <cdent> #link https://etherpad.openstack.org/p/api-sig-newsletter
16:31:31 <cdent> It's probably either me or dtantsur
16:32:00 <elmiko> i could do it if everyone is busy
16:32:08 <dtantsur> as long as somebody cleans up my English afterwards :)
16:32:25 <cdent> if one of you is keen that would be great, as I'm in a yaml-related hole
16:32:32 <elmiko> yeesh
16:32:36 <elmiko> i'll take it
16:32:36 <dtantsur> yaml-related hole, sounds scaaary
16:32:45 <edleafe> I'll be the English editor :)
16:32:55 <elmiko> ++
16:33:38 * dtantsur hears crickets
16:33:44 <elmiko> i'll ping in sdks when i've got something
16:33:54 <edleafe> ack
16:33:55 <elmiko> maybe cdent got trapped by the yaml?
16:34:02 <elmiko> XD
16:34:11 <dtantsur> that's bad, we need to throw some toml at him
16:34:23 <elmiko> haha
16:34:26 <elmiko> or some json?
16:34:40 <dtantsur> json is yaml's lil brother, no?
16:34:49 <dtantsur> I'd not trust him
16:34:51 <cdent> PyYAML 4.x changes things in a way that breaks gabbi tests
16:34:53 <elmiko> hahaha
16:35:03 <elmiko> oh, bummer cdent =(
16:35:18 <dtantsur> cdent: parser or library API?
16:35:21 <cdent> sorry, that was ambiguous
16:35:39 <cdent> there are tests in gabbi itself, which are testiing yaml "safe" features which break with 4.x
16:36:01 <cdent> and fixing it a change in behavior
16:36:15 <cdent> it's weird and my brain is struggling: https://github.com/cdent/gabbi/pull/252
16:36:16 <dtantsur> fun
16:36:48 <cdent> it's also freaking hot which is not helping.
16:36:59 <elmiko> =(
16:37:15 <dtantsur> I feel sorry for you
16:37:21 <edleafe> cdent: what is "hot" in your neck of the woods?
16:37:46 <edleafe> Currently 31C here
16:37:54 <dtantsur> ouch
16:37:58 * elmiko faints
16:38:17 <cdent> not there yet edleafe
16:38:33 <dtantsur> you'll cool down on the Berlin's summit :D
16:38:33 <edleafe> Oh, it's still early. Going up to 37C later today
16:38:38 <dtantsur> oh no
16:38:47 <cdent> not actually sure what the temp is, but with no ac, no fans, and currently no wind either it is ... weird
16:38:49 <edleafe> Welcome to South Texas!
16:38:59 <elmiko> yikes
16:39:02 <dtantsur> speaking of the Berlin summit
16:39:04 <edleafe> My A/C is going almost continuously
16:39:07 <elmiko> 37 is just not right
16:39:14 <dtantsur> anyone wants to co-present something about our beloved microversions?
16:39:28 * cdent shrugs
16:39:32 <edleafe> That will draw a hostile crowd
16:39:32 <dtantsur> or maybe some other topics, but this is probably the most interesting slash confusing slash contentions
16:39:57 <elmiko> i doubt i will make it to berlin
16:40:01 <edleafe> I'd do it, but I have no idea if IBM will send me
16:40:06 <cdent> I try to avoid doing any presentations at summit, but something on microversions might be useful, so I'd encourage it to happen
16:40:09 * dtantsur summons mordred again
16:40:13 <edleafe> I'm still not even approved for the PTG
16:40:28 <dtantsur> I'm not going to the PTG, but the summit happens 40 minutes by s-bahn from my house, sooo
16:40:56 <cdent> especially if it was flavored towards "fun and profit with using microversin against openstack" rather than "here's yet another explanation of how to add microversions to your project"
16:41:18 <dtantsur> yeah, more from client/consumer side
16:41:22 <edleafe> Slide 1: What is a Microversion?
16:41:29 <dtantsur> Slide 2: OH GOD WHY
16:41:37 <edleafe> heh
16:41:42 <elmiko> lol
16:41:49 <cdent> on that note shall we call it a day?
16:41:55 <elmiko> ++
16:41:55 <dtantsur> ++
16:42:06 <cdent> great. thanks for coming.
16:42:11 <cdent> #endmeeting