16:00:22 <krotscheck> #startmeeting StoryBoard
16:00:22 <openstack> Meeting started Mon Nov 24 16:00:22 2014 UTC and is due to finish in 60 minutes.  The chair is krotscheck. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:23 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:27 <openstack> The meeting name has been set to 'storyboard'
16:00:58 <krotscheck> Agenda: https://wiki.openstack.org/wiki/StoryBoard#Agenda
16:01:26 <krotscheck> Although to be honest, that agenda is starting to get a bit out of hand. I’m going to rework the wiki a tad.
16:01:32 <ttx> o/
16:01:39 <krotscheck> #action krotscheck Reorganize agenda wiki.
16:01:56 <krotscheck> #topic Actions from last week: Migration
16:02:18 <krotscheck> Ok, so migration bugs were fixed, infra was migrated, email was sent, and things are more or less wrapped on that topic.
16:02:27 <ttx> woohoo
16:02:52 <krotscheck> Indeed :)
16:03:02 <krotscheck> We’ve also got… two confirmed downstream users of storyboard now.
16:03:13 <krotscheck> So no more breaking downstream.
16:03:20 <NikitaKonovalov> o/
16:03:30 <NikitaKonovalov> finally back from summit and vacation
16:03:30 <jeblair> yay!
16:04:09 <krotscheck> Woo vacation!
16:04:12 <ttx> NikitaKonovalov: welcome back!
16:04:28 <krotscheck> #topic Urgent Items
16:04:31 <krotscheck> Anything urgent?
16:04:48 <NikitaKonovalov> nothing from me
16:05:04 <krotscheck> I have a not-really-urgent-but-omg-annoying thing.
16:05:29 <krotscheck> The webclient is super chatty - as in it makes a lot of unecessary API calls.
16:05:34 <krotscheck> I did some work last week to fix that.
16:05:53 <krotscheck> Or, at least, mitigate it.
16:05:54 <krotscheck> ANd those reviews are up
16:05:55 <yolanda> the caching changes? i'm still not very confident about that, but it can help
16:06:32 <krotscheck> Ok, let’s talk about that in discussion, if you have concerns let’s raise ‘em.
16:07:05 <krotscheck> Any other urgent items?
16:07:12 <krotscheck> Anything from our users? (jeblair that’s you)
16:07:50 <jeblair> krotscheck: i haven't been back enough from my own vacation to be on top of that feedback.  i will endeavor to be so asap.  :)
16:08:03 <krotscheck> jeblair: You got it :)
16:08:21 <krotscheck> We got a bunch of comments in the channel on wed/thur, and those have spawned a few stories.
16:08:29 <krotscheck> But nothing super crazy
16:08:38 <jeblair> awesome, just what we wande
16:08:40 <krotscheck> More I want features, When will this happen, etc kind of questions.
16:08:40 <jeblair> wanted
16:08:47 <krotscheck> Anyway
16:08:58 <krotscheck> #topic Discussion: Core Reviewers
16:09:23 <fungi> it's been working great so far
16:09:47 <krotscheck> So, this topic was raised by yolanda last week. She’s been doing a lot of code reviews, and has been asking a lot of really good questions, and she was curious as to what it takes to be a storyboard core.
16:09:57 <krotscheck> And I didn’t have an answer!
16:10:12 <yolanda> yes, i've been really interested in the project, i love working on it and i really want to be core of it
16:10:34 <krotscheck> So what I’d like to do is figure out what we as a team want a storyboard core to be.
16:10:43 <jeblair> krotscheck: one good first order approximation is this: when you see a +1 or -1 from a person, do you start to treat it as a +/-2 in your mind
16:11:13 <ttx> then you bend the spoon
16:11:17 <krotscheck> jeblair: That’s a good one.
16:11:20 <ttx> then there is no spoon
16:11:28 <krotscheck> jeblair: What about seeing a +2 and treating it like a +1? ;)
16:12:00 <jeblair> are you confident that person understands the code well enough to identify problematic code before it merges, understands the road map well enough to take it in the direction the project should go, and engaged enough that the investment will be worthwhile
16:12:25 <krotscheck> Mind if I copy that verbatim into the wiki?
16:12:54 <jeblair> krotscheck: not at all
16:13:04 <krotscheck> Coool
16:13:07 <jeblair> krotscheck: and about the +2 as a +1, that's a concern too.  most people with +2 should know when they are reviewing something out of their area of expertise
16:13:15 <jeblair> krotscheck: and should generally downgrade their own votes accordingly
16:13:35 <krotscheck> jeblair: Well, we’re in a weird position on that one, since there’s a lot of +2’s that we inherit from infra.
16:13:36 <jeblair> krotscheck: so actually, understanding that that person knows their own limits is a good thing to look for in a potential core reviewer as well
16:13:57 <jeblair> krotscheck: if you feel like we are overstepping bounds, i'm happy to talk about it
16:14:16 <ttx> also we don't have that many cores, so if I started to say +1 when I'm not sure, we would stall :)
16:14:19 <krotscheck> jeblair: If it becomes a thing I’ll let you know :)
16:14:34 <krotscheck> Yeah, that too.
16:14:47 <fungi> yeah, i've so far tried to only +2 storyboard changes i feel like i understand the ramifications of
16:14:49 <ttx> that said once we have a solid team, I'm happy to un-core myself and become a very opinionated stakeholder
16:15:17 <ttx> since I can't seem to get enough time to become angular-fluent
16:15:20 <krotscheck> ttx: Well, I think at that point we’ll be able to ACL prioritization on storyboard, meaning you can go make changes there :)
16:16:08 <krotscheck> Ok, so what I’ve got so far:
16:16:33 <krotscheck> - Demonstration that the individual understands the codebase well enough to identify problems.
16:16:55 <krotscheck> - Demonstration that the individual understands and contributes to the roadmap.
16:17:07 <krotscheck> - Demonstration of engagement
16:17:14 <krotscheck> - Demonstration that the person knows their limits.
16:17:27 <krotscheck> These all feel quite fluffy to me.
16:17:36 <krotscheck> a little subjective, but I do like them.
16:17:51 <jedimike> it is subjective though
16:18:17 <SergeyLukjanov> sounds fair
16:18:23 <yolanda> yes, these are subjective criteria but if it's a shared opinion it can become more valid
16:18:32 <NikitaKonovalov> sounds fair for me as well
16:18:33 <jeblair> krotscheck: i have semi-objective ways of evaluating those that i'm happy to share (but this may not be the best time)
16:18:55 <jeblair> but yeah, the subjectiveness is a feature :)
16:19:02 <SergeyLukjanov> btw there is a nova core team page - https://wiki.openstack.org/wiki/Nova/CoreTeam with some formal workds
16:19:04 <krotscheck> Bah. Why can’t humans be binary.
16:19:19 <jedimike> yolanda, that's why cores are voted in :)
16:19:31 <SergeyLukjanov> krotscheck, ternary can
16:19:34 <krotscheck> SergeyLukjanov: Thanks for that :)
16:19:54 <krotscheck> I don’t see any disagreement on those four, and it seems like a good starting point.
16:20:37 <krotscheck> So de/elevation of cores then can be a consensus thing, since everything in openstack is consensus.
16:21:16 <krotscheck> Consensus among cores?
16:21:25 <krotscheck> Storyboard cores + infra cores?
16:21:29 <krotscheck> Storyboard cores + infra PTL?
16:21:35 <krotscheck> All the cores in openstack?
16:22:10 <SergeyLukjanov> IMO infra cores + sb cores works ok
16:22:31 <NikitaKonovalov> krotscheck: as StoryBoard is part of an Infra programe Infra cores should vote
16:22:57 <SergeyLukjanov> for me, for example, I'm not able to contribute something to code directly, but I'm able to verify and review at list server (and then python client) side
16:23:32 <krotscheck> I disagree on that. I haven’t seen any infra cores actively contribute to the codebase, though they’ve been useful to get reviews when people go on vacation.
16:24:02 <krotscheck> To me, infra core feels more like an advisory panel of experts than people who meet the four storyboard core criteria.
16:25:01 <krotscheck> jeblair: Opinions?
16:25:06 <jeblair> krotscheck: what are you disagreeing with?
16:25:09 <yolanda> well, i think the most people having an opinion, the better should be. Allowing everyone to vote cannot be bad
16:25:25 <yolanda> everyone that knows the project and have a relevant opinion
16:25:52 <krotscheck> jeblair: I disagree with SergeyLukjanov and NikitaKonovalov’s desire to make elevation of cores a SB Core + All Infra Core consensus.
16:26:41 <krotscheck> I personally prefer SB Core + PTL.
16:27:37 <krotscheck> Also, because I have vested interest in having Mirantis dedicate more of NikitaKonovalov’s time to the project :)
16:27:51 <NikitaKonovalov> krotscheck: if an Infra core member does not have an opinion, they should not be forced to vote
16:27:51 <jeblair> krotscheck: i think infra-core has quite a bit to contribute here.  just as people contribute differently to different kinds of reviews and changes, there's an opportunity to involve infra-core, but not in a way that oversteps bounds.
16:28:22 <krotscheck> So it sounds to me like it’s more of a straight vote than  consensus?
16:28:24 <jeblair> krotscheck: i'd prefer the consensus process for core to be inclusive, but i suspect that people in infra-core who are not equipped to evaluate a member will not obstruct the process
16:28:38 <jeblair> krotscheck: here's the general process:
16:29:37 <jeblair> krotscheck: nomination to list; invitation to +1 or -1, stays open for a week.  people chime in with opinions, after a week if no significant opposition is observed, it happens
16:29:53 <jeblair> krotscheck: so it's not really a formal vote
16:30:09 <jeblair> krotscheck: it's more of, who's the group from which you are willing to consider objections
16:30:14 <SergeyLukjanov> yup, and -1 is veto
16:30:56 <jeblair> that pretty much never happens by the time we actually get to the proposal stage
16:31:34 <jeblair> everyone wants more cores, and we try not to propose people who aren't ready (handling a rejection like that could be awkward)
16:31:56 <krotscheck> Ok, so how does that work for de-elevation?
16:32:01 <krotscheck> It all sounds good to me :)
16:32:23 <yolanda> cool, sounds fair to me
16:32:44 <jedimike> +1 from me
16:32:52 <jeblair> usually the ptl makes the decision and announces a de-elevation
16:33:05 <jeblair> generally that only happens because someone has stopped reviewing code
16:33:16 <jeblair> and it's usually apparent to everyone that it's time
16:33:40 <jeblair> but nonetheless, we don't ask people to publicly "+1" removing someone from core
16:33:47 <jeblair> because again, that could be awkward
16:34:26 <jeblair> i think there may be some morke text about this on the wiki somewhere
16:35:08 <krotscheck> Righto.
16:35:31 <krotscheck> Ok, so let’s stay with Infra + SB cores for now and see how that works out for us.
16:35:45 <jeblair> oh yeah, it's in that nova core team page
16:35:53 <jeblair> 2nd par of 'adding or removing members'
16:36:17 <krotscheck> yolanda: As far as I’m concerned, you have my +1, but you’ll have to go talk to others :)
16:36:22 <krotscheck> next topic?
16:37:00 <krotscheck> #agreed Core elevation to be “consensus” of Infra + Storyboard Cores.
16:37:14 <krotscheck> #topic Story Types.
16:37:16 <yolanda> how is the proposal to the mailing list sent? who needs to send that? the interested in being core?
16:37:27 <krotscheck> oops
16:37:47 <krotscheck> #topic Discussion: Core Reviewres
16:37:58 <yolanda> sorry, typed too slow
16:38:15 <krotscheck> I’d say, pick whatever communication medium you prefer (IRC, Email, etc)
16:38:31 <krotscheck> And at the very least get existing SB Cores on board.
16:38:32 <jeblair> yolanda: actually...
16:38:41 <krotscheck> And then talk to jeblair
16:39:42 <jeblair> yolanda: by convention the proposal usually comes from the ptl; i think in this case we could say it's okay for a proposal to come from a storyboard core though, yeah?
16:40:14 <jeblair> but i don't think we should ask people who want to be core to publicly propose themselves.  i think it's better for the core team to make the invitation
16:40:24 <krotscheck> I agree
16:40:39 <jeblair> but people interested in being core should definitely talk to those on the core team to help get them ready
16:40:50 <krotscheck> That too.
16:40:51 <jeblair> yolanda: so you're definitely doing the right thing here :)
16:41:13 <krotscheck> Personally, what I love about yolanda’s reviews is that she’s EXTREMELY cautions about things.
16:41:27 <ttx> indeed!
16:41:42 <krotscheck> She’s more or less the opposite of me.
16:41:44 <krotscheck> :D
16:42:15 <yolanda> heh, that can be good and bad :)
16:42:15 * krotscheck has also been noticing wendar doing a lot of reviews for us.
16:42:26 <jeblair> yay!
16:42:45 <krotscheck> yolanda: So did that answer your question?
16:42:51 <yolanda> yes, sure
16:43:25 <krotscheck> Awesome.
16:43:32 <krotscheck> #topic Discussion: Story Types
16:43:38 * krotscheck has not reviewed that spec.
16:43:42 * krotscheck goes to sit in a corner.
16:44:02 <ttx> I won't write another spec until that one is reviewed :)
16:44:14 * ttx can't only handle one spec at a time :)
16:44:23 <ttx> Also nothing like a good excuse for not doing it
16:44:34 * krotscheck thinks that ttx’s reaction is the right one.
16:45:20 <krotscheck> Can we get everyone to reviewing https://review.openstack.org/#/c/129267/ this week?
16:45:21 <jedimike> what's the link to the spec? I can't see it linked from the wiki
16:45:33 <krotscheck> jedimike: ^^
16:45:37 <jedimike> aha :)
16:45:37 * krotscheck is updating the wiki
16:46:46 <jeblair> i think this looks like a good one for the broader infra audience to see...
16:46:58 <jeblair> i can highlight it in the infra meeting if you like
16:47:05 <krotscheck> jeblair: That’d be awesome.
16:47:21 <krotscheck> #action krotscheck review https://review.openstack.org/#/c/129267/
16:47:30 <krotscheck> (Others can add their own actions)
16:47:39 <krotscheck> Ok, moving on:
16:47:47 <krotscheck> #topic Discussion: Documentation
16:47:51 <krotscheck> persia: You here?
16:48:47 <krotscheck> Hrm. Ok, so it’s been 3 weeks with noshows from him. I’ll ping him directly.
16:49:03 <krotscheck> #topic Product Evangelism
16:49:14 <krotscheck> Ditto on rainya
16:49:34 <krotscheck> #action krotscheck contact rainya and persia about the items they agreed to at the summit.
16:49:44 <krotscheck> #topic MVP 1.2
16:50:16 <NikitaKonovalov> I've started the python client task
16:50:20 <krotscheck> Yay!
16:50:34 <krotscheck> NikitaKonovalov: You saw the user token issue api that I built, yes?
16:50:38 <NikitaKonovalov> As soon as the repo and all infra around is ready
16:50:45 <NikitaKonovalov> krotscheck: I've seen that
16:51:01 <NikitaKonovalov> and it will be very usefull
16:51:05 <krotscheck> Is that an acceptable way of having the python client authenticate, or do you want something more sophisticated?
16:51:26 <SergeyLukjanov> I've created client on pypi and CR for adding new repo is near to be finished
16:51:27 * krotscheck is thinking of allowing OAuth password flow in addition to launchpad.
16:52:02 <krotscheck> #topic MVP 1.2: Client Library
16:52:08 <NikitaKonovalov> I had an idea of a client making a request to a public endpoint, which will then result in a question to a user "Allow this client to authenticate or not"
16:52:10 <SergeyLukjanov> probably use the same approach of approving tokens in Web UI like launchpad?
16:52:28 <SergeyLukjanov> it's the same as NikitaKonovalov said
16:52:37 <NikitaKonovalov> I think I'll describe the full flow in a spec
16:52:45 <krotscheck> Good idea.
16:52:59 <krotscheck> That feels more like traditional OAuth flow, good thinking.
16:53:03 <NikitaKonovalov> but for now I'll make it work with a provided token
16:53:09 <SergeyLukjanov> krotscheck, +1
16:53:16 <NikitaKonovalov> both as a python module and CLI
16:53:24 <krotscheck> Cool.
16:53:36 <krotscheck> Does openstack have a standard for how to structure the CLI arguments?
16:53:36 <jeblair> ++spec.  fwiw, that approach is great for real users and really hard to use with non-interactive accounts
16:54:02 <SergeyLukjanov> jeblair, provided token for non-interactive?
16:54:08 <jeblair> er, are we sure we want to conflate the cli and python api client?
16:54:09 <NikitaKonovalov> btw, oslo apiclient appeared to be pretty handy
16:54:15 <jeblair> i think that has worked out very poorly for openstack
16:54:32 <NikitaKonovalov> jeblair: openstack clients usually do
16:54:48 <SergeyLukjanov> jeblair, I think for us it'll not be an issue mix them in one repo
16:55:06 <jeblair> which is why currently the python-*clients are heading in the direction of being internal api only, the "sdk" project is heading toward being a python api client, and the 'openstackclient' project is aiming for a command line client
16:55:52 <NikitaKonovalov> I think CLI is low priority for SB
16:56:05 <NikitaKonovalov> so there is time to figure out
16:56:10 <jeblair> NikitaKonovalov: agreed; i think the api client is much more important
16:56:16 <jeblair> (at the moment)
16:56:22 <ttx> most cli users would be apiclient users anyway
16:57:08 <jeblair> SergeyLukjanov: [going back a bit] yeah, a way to have a provided token for non-interactive is better (something we can put in hiera for puppet, etc)
16:57:37 <SergeyLukjanov> jeblair, yup
16:57:45 <krotscheck> I think we’re all on the same page.
16:57:57 <krotscheck> And we’re getting low on time.
16:58:59 <krotscheck> Most of my work has been on fixing annoying UI things.
16:59:05 <krotscheck> Like inline editing of tasks.
16:59:18 <krotscheck> So not really roadmap things, so I don’t have anything to contribute here.
16:59:37 <krotscheck> Well, that’s a lie. The reason I fixed that is because I was trying to get the 1.2 things into storyboard and the UI made my eyes blee.d
16:59:44 <krotscheck> Anyway: Anything else?
16:59:49 <krotscheck> jedimike? yolanda?
16:59:54 <krotscheck> You’ve got less than a minute :)
17:00:01 <krotscheck> (sorry)
17:00:15 <jedimike> did we come to a solution with results paging?
17:00:38 <jedimike> (sorry for the highly controversial thread ;)
17:00:41 <krotscheck> jedimike: Ugh. No. I don’t think we have the time to discuss that in the meeting though, bounce over to channel?
17:00:45 <jedimike> cool
17:00:52 <krotscheck> Alright, thanks everyone!
17:00:52 <krotscheck> #endmeeting