20:03:06 <ttx> #startmeeting tc
20:03:07 <openstack> Meeting started Tue Oct 28 20:03:06 2014 UTC and is due to finish in 60 minutes.  The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:03:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:03:11 <openstack> The meeting name has been set to 'tc'
20:03:20 <ttx> Our agenda for today:
20:03:27 <ttx> #link https://wiki.openstack.org/wiki/Governance/TechnicalCommittee
20:03:34 <ttx> #topic Board/TC joint meeting agenda
20:03:46 <ttx> Our first item is the agenda for the Board/TC joint meeting at 2:30pm on Sunday
20:03:49 <vishy> o/
20:03:55 <ttx> The agenda proposed by the board is:
20:04:01 <ttx> - Introductions
20:04:06 <ttx> - Followup from last joint meeting (travel support program & code review backlog)
20:04:15 <ttx> - The need for project structure reform (Ring 0 and the big tent approach)
20:04:22 <ttx> - Tempest and it's use to drive interoperability and DefCore
20:04:28 <ttx> - mid cycle meetups - is there a way to reduce travel strain?
20:04:37 <ttx> I called for topics at http://lists.openstack.org/pipermail/openstack-tc/2014-October/000847.html
20:04:47 <ttx> dhellmann wondered if we should put the CLA question on the meeting agenda
20:04:57 <ttx> jaypipes suggested we discuss how the process by which we check for name/copyright issues for new projects can be streamlined
20:04:59 <mikal> I think its more important than some of the other things listed there
20:05:08 <mikal> The CLA that is
20:05:20 <ttx> After a few checks I'm not sure the legal namechecks one is relevant to the board, since this is currently handled by Foundation staff, so it may be better to talk to them about it first
20:05:28 <dhellmann> I'd like an update on the CLA, but that doesn't have to happen in person.
20:05:37 <ttx> we can discuss it in a TC meeting directly with them
20:05:40 <sdague> yeh, the CLA seems like a thing that we need the board to take an action on, and it's been 7 or 8 months since this conversation started
20:05:42 <jaypipes> dhellmann: ++
20:05:57 <jeblair> since we passed a resolution about the cla, we should probably put it on the agenda to continue to indicate we are interested in it, even if we do so in a way structured to avoid having it take the whole meeting :)
20:06:04 <sdague> yeh
20:06:08 <dhellmann> jeblair: ++
20:06:11 <ttx> jeblair: we can place it as the last item
20:06:20 <ttx> so that it's mentioned but doesn't take the whole time
20:06:23 <sdague> mostly, I think bringing it up will demonstrate seriousness that the TC thinks it's important
20:06:25 <jeblair> ttx: or timebox it
20:06:35 <ttx> yes
20:06:42 <mikal> I am a bit confused by some of the other board items there to be honest
20:06:45 <ttx> I proposed a discussion of how much OpenStack (in general) should embrace Containers/Docker, as well as an update on the proposed bylaws change
20:06:57 <ttx> Thoughts on what we should push forward ?
20:06:58 <jeblair> sdague: yeah, and i think that's a point worth making since the board explicitly wondered whether that was the case (thus prompting the resolution)
20:07:03 <ttx> (The meeting is only 3 hours so we might need to prioritize)
20:07:16 <mikal> Like, do we really think mid cycle meetups is one of the biggest problems we face at the moment?
20:07:17 <russellb> only 3
20:07:18 <sdague> ttx: the containers thing seems odd to me for this forum.
20:07:27 <russellb> sdague: that's how i felt, too
20:07:34 <sdague> it's good beer chat, but does it really need to be a TC / Board thing?
20:07:42 <ttx> sdague: maybe not
20:07:47 <dhellmann> mikal: who is complaining, the person doing the travel or the person paying?
20:07:55 <ttx> I mean, I won't push it unless someone else at the TC wants it
20:08:01 <jaypipes> I think it does, because it goes to the general direction that we wish OpenStack to drive towards.
20:08:03 <mikal> dhellmann: I dunno, its just on the boards list up above
20:08:16 <jaypipes> and I'm not just talking about technical direction.
20:08:22 <dhellmann> mikal: yeah, this isn't the first time I've seen it come up
20:08:27 <ttx> jaypipes: yes, that was my idea
20:08:39 <ttx> jaypipes: checking for general alignment on where we should go
20:08:41 <russellb> i don't mind the business side input on direction
20:08:46 <jaypipes> containers bring with them a paradigm shift that matters more than technology. it shapes the way app developers visualize and work with the cloud.
20:08:47 <russellb> yeah, that seems sensible
20:09:24 <jaypipes> everybody run.. markmcclain is here.
20:09:36 <russellb> run ... up to markmcclain and say hi!
20:09:41 <ttx> jaypipes: are you fine with tabling the legal namechecks discussion to a future TC meeting with Foundation staff ? The board is not really involved with those
20:10:19 <jaypipes> ttx: isn't Mark Radcliffe involved in those discussions? I remember Mark being at the last Board meeitng...
20:10:36 <ttx> jaypipes: no, the lawyer doing those checks is not Radcliffe. I fear we won't get clarification, but more confusion
20:10:37 <jaypipes> board/TC meeting that is.
20:10:50 <markmcclain1> jaypipes: haha
20:10:51 <jaypipes> ttx: ok, no problem. i don't want more confusion.
20:11:26 <ttx> so, I'll ask for CLA timeboxed
20:11:42 <ttx> Maybe a temperature check on the Docker world
20:12:00 <ttx> anything else ?
20:12:06 <russellb> may need to timebox docker too
20:12:15 <dhellmann> ++
20:12:15 <ttx> ack
20:12:35 <mikal> ttx: I think we should drop the mid-cycle thing, or at least move it later in the list
20:12:39 <mikal> It just doens't seem that important to me
20:12:54 <ttx> mikal: I'll suggest that. It's not on my part of the agenda :)
20:13:04 <mikal> ttx: ok
20:13:20 <ttx> Note that we can probably adjust the agenda in person when the meeting starts if we havbe a brand-new idea
20:13:20 <sdague> ttx: was that board added?
20:13:20 <mikal> Do we need daa for the code review backlog thing?
20:13:26 <mikal> data even
20:13:27 <ttx> sdague: yes
20:13:33 <sdague> ttx: ok
20:13:37 <ttx> mikal: it wouldn't hurt
20:13:39 <mikal> Its just a silly item
20:13:41 <jeblair> i do see this summit as an experiment in how we can reduce the need for midcycles; i'm not sure we're ready to discuss it again until after this summit
20:13:43 <russellb> sdague: rob h. brought it up at a recent board meeting
20:13:46 <russellb> i'm not sure who raised it to him
20:13:55 <ttx> at least from Nova / Neutron who had the largest backlog
20:13:58 <russellb> jeblair: ++
20:14:16 <mikal> Its also just penny pinching
20:14:22 <mikal> Travel is a tiny fraction of salary spend
20:14:23 <sdague> jeblair: yeh, agreed, so maybe just make sure the time box is small on it
20:14:28 <mikal> Sure, we should be frugal
20:14:29 <russellb> mikal: it's more than the $
20:14:33 <mikal> But if its efficient, we should do it
20:14:44 <russellb> but let's not get into it here
20:14:56 <ttx> mikal: being all relocated in the same office would be efficient, but we don't do it either
20:15:01 <russellb> right
20:15:09 <jeblair> i second us not getting into it here :)
20:15:10 <jaypipes> ttx: one thing that might be interesting to get the BoD feedback on is the recent Neutron core team's proposal to do peer reviews for core committers.
20:15:26 * markmcclain resorts to using crappy irc client on his phone
20:15:35 <jaypipes> ttx: just to get their feedback, nothing more. I'm just curious what they think.
20:15:38 <mikal> jaypipes: it would be interesting to get feedback on how that is working out before going too far with it
20:15:54 <dhellmann> mikal: +1
20:15:55 <ttx> yeah, I agree with mikal, let's let the experiment play out
20:16:03 <mikal> jaypipes: as in, does Kyle feel is working? Was the feedback useful?
20:16:10 <ttx> We can give feedback on various experiments like specs now
20:16:14 <jaypipes> mikal: they haven't started it yet :)
20:16:16 <russellb> definitely not something i see we should push more broadly until we see how it works for them
20:16:23 <ttx> as part of the "review backlog" discussion I guess
20:16:25 <jeblair> with a time constraint on the agenda, i don't think that's the kind of feedback we need from the board
20:16:33 <jaypipes> mikal: I'm more interested in just the board members thoughts on it (same with the user committee's thoughts)
20:16:36 <russellb> ttx: sure
20:16:36 <sdague> yeh, I'm also a little concerned with just bringing up technical policy stuff to the board to get their view on if we don't have enough time to set context
20:16:41 <jaypipes> mikal: but, fine, it's a beer topic :)
20:16:43 * devananda pops in while waiting for boarding to start
20:17:11 <ttx> ok, let's move on
20:17:33 <ttx> If you sudenly have new ideas, push them to the thread. I'll compose an answer to Alan tomorrow morning
20:17:48 <ttx> #topic Cross-project workshop final agenda
20:17:59 <ttx> russellb, markmcclain: care to present the short list ?
20:18:03 <russellb> sure, i'll cover this
20:18:08 <russellb> so we have a couple of things to cover on this
20:18:14 <russellb> first, review the list of sessions and the proposed schedule for them
20:18:30 <russellb> and if that looks good, we need to collect session leads and have the leads write up a session description that can be used on sched.org
20:18:35 <russellb> proposed schedule:
20:18:37 <russellb> #link http://kilodesignsummit.sched.org/overview/type/cross-project+workshops#.VE_2XHVGjUY
20:18:48 <dhellmann> russellb: I went through and did the math for the votes so far
20:18:51 <russellb> and then for collecting leads and descriptions:
20:18:53 <russellb> #link https://etherpad.openstack.org/p/kilo-crossproject-summit-topics
20:19:01 <russellb> dhellmann: yeah we did all that this morning
20:19:22 <russellb> and by we, i mean myself, markmcclain, and ttx
20:19:24 <dhellmann> oh, I didn't see it in the etherpad
20:19:25 <russellb> based on action from last week
20:19:31 <russellb> nah, it was done in another spreadsheet
20:19:38 <dhellmann> k
20:19:39 <russellb> sorry
20:19:51 <mikal> So, that cells one confuses me
20:19:52 <dhellmann> meh, I was bored and curious
20:19:58 <russellb> so first, does anyone see any obvious issues with the list of sessions or conflicts in schedule?
20:20:03 <russellb> mikal: yeah, it made the cut *shrug*
20:20:17 <mikal> russellb: so, that makes three cells sessions a tthe summit
20:20:19 <russellb> i think the cells one was to talk about how it impacts other projects
20:20:21 <mikal> Which seems like too many to me
20:20:26 <russellb> not just the nova part
20:20:28 <mikal> Especially for a feature nova might drop
20:20:38 <ttx> mikal: the idea was to give some outlet for epople to discuss various ways to scale out in 40min. hopefully they will meet each other and stop inventing parallel ways ?
20:20:41 <sdague> mikal: well not if it means people actually work on it :)
20:20:42 <russellb> another part of that was merging in the huawei thing
20:21:01 <mikal> Ok, so rename it then
20:21:01 <russellb> there's a double session in the nova track too, i know
20:21:15 <ttx> all the similar topics were merged, so I don't expect it to be a cells thing. More of a Cells vs. Cascade vs. Z
20:21:18 <mikal> I'd be ok with it if it was called "scaling openstack" or something
20:21:30 <mikal> But its named "...with cells"
20:21:34 <mikal> Which is the bit that confuses me
20:21:37 <ttx> yes, name should change
20:21:43 <russellb> k
20:22:05 <ttx> I'd say "Scaling out: Cells, Cascading, Alliance, etc"
20:22:15 <mikal> Or name none
20:22:26 <mikal> "Proposals for scaling out OpenStack"
20:22:27 <russellb> Approaches for Scaling Out ?
20:22:36 <ttx> russell++
20:22:37 <mikal> Yeah, that would do
20:22:41 <markmcclain> Rather name none
20:23:10 <russellb> k, sched updated
20:23:21 <ttx> well, sched caches, so ymmv
20:23:23 <mikal> Ta
20:23:28 <russellb> i had john g. listed as leading that, since he proposed the cells one in the etherpad
20:23:36 <russellb> he's also listed on at least one other, though
20:23:41 <ttx> who can lead the API WG one ? jaypipes ?
20:23:59 <mikal> russellb: John would do a good job of it, depends if we think two is too many for one person
20:24:14 <ttx> though you already have the technbical debt one I suspect
20:24:15 <russellb> mikal: ok, I can ask him what he thinks, will CC you
20:24:23 <jaypipes> yes, though I'd love it if the part 1 and part 2 were separated... I really wanted to attend the functional tesst in projects session.
20:24:24 <mikal> russellb: works for me
20:25:00 <jaypipes> heck i'd love to go to the common approach to ha one too :)
20:25:18 <russellb> jaypipes: yeah, scheduling this stuff is fun.
20:25:26 <jaypipes> russellb: yeah, i know :)
20:25:44 <russellb> if anyone has proposed leads for any of the stuff on the etherpad, fill it in
20:25:55 <russellb> some may have it later in the etherpad and i just haven't gone back to find it yet
20:26:06 <jaypipes> russellb: suggestion... perhaps we could swap the api wg part 2 with the growth challenges part 1?
20:26:23 <jeblair> i don't think the intent there is to actually have two parts, but rather combine them into one long session
20:26:36 <jaypipes> russellb: and have a morning and afternoon session on those topics?
20:26:44 <russellb> yeah, was thinking it was one long session
20:26:50 <jaypipes> jeblair: yes, I know... but ... just offering an alternative.
20:26:50 <ttx> yes, it's one long session
20:26:53 <jeblair> i think breaking them up would disrupt the idea of being able to have longer-form sessions on some topics
20:26:58 <sdague> jeblair: ++
20:27:00 <ttx> 90min instead of 2x40min
20:27:20 <ttx> I think it needs 90min in one block
20:27:32 <russellb> we could maybe move "moving testing ot projects" elsewhere
20:28:02 <russellb> it's a bad conflict no matter where it moved to
20:28:19 <devananda> I'm wondering who else feels that the necesity of attending the cross project sessions pretty nearly trumps all the sessions and panels on Tuesday
20:28:36 <russellb> devananda: that's my general feeling, yes
20:28:37 <ttx> russellb: maybe swap it with the reaquirements session ?
20:28:41 <sdague> devananda: yep
20:28:44 <jeblair> devananda: i feel that way for the whole summit, but yes :)
20:28:45 <devananda> thankfully i don't have any speaking commits on tuesday
20:28:55 <devananda> but a bunch of panels that I really want to attend are on tuesday ...
20:29:10 <russellb> jaypipes: if we do the swap ttx suggests, it conflicts with the "scaling out" thing
20:29:23 <russellb> and docs
20:29:27 <russellb> jaypipes: that better?
20:29:29 <jaypipes> devananda: unfortunately, I have to give a presentation at the conf side on tues...
20:29:35 <jaypipes> and monday, and wednesday...
20:29:39 <russellb> jaypipes: sucker
20:29:41 <russellb> :-p
20:29:57 <ttx> jaypipes: one would think you would know to avoid the CFP by now
20:30:01 <jaypipes> russellb: lookign...
20:30:03 <devananda> 16:40 gerrit third-party CI && growth challenges p1.
20:30:08 <russellb> honestly, it's a shame more people from dev community don't present because of the pressure we have to attend design summit stuff
20:30:14 <devananda> are there people who need to be in both of those?
20:30:28 <sdague> russellb: well that was one of the reasons for the offset
20:30:38 <russellb> sdague: true, speak on monday
20:30:57 <fungi> gerrit third-party ci does in fact seem related to growth challenges
20:31:07 <jaypipes> russellb: yeah, unfortunately I'd love to be at the scaling out session :) it's ok, i'll just deal with it. I do every time anyway.. we all do. :)
20:31:15 <jaypipes> try and catch a bit of each.
20:31:17 <sdague> jaypipes: ++
20:31:51 <russellb> depends on what the 3rd party meetup is about
20:31:58 <jaypipes> ttx: I'm happy to lead the first part of the API WG, and maybe etoews will take up the latter part so I can catch some of the func testing session
20:31:58 <russellb> but yeah, definite overlap
20:32:23 <devananda> none of the other design things on tuesday stand out as single-person conflicts within the design track
20:32:24 <ttx> fungi: unfortunately there is little ways to avoid that one.
20:32:40 <devananda> *to me
20:32:59 <ttx> fungi: We want that one at a moment where only one other workshop runs. That's either at the start or the end of the day
20:33:01 <fungi> fair enough, just agreeing with devananda's point
20:33:05 <russellb> who would lead "moving tests to projects" ?
20:33:10 <ttx> I run that session and I also need to be at the Design Summit 101
20:33:20 <devananda> fungi: ttx: could we move gerrit 3rd party one slot earlier?
20:33:46 <devananda> russellb: mtreinish or sdague?
20:33:55 * annegentle lurks after appointment and catches up
20:33:57 <devananda> that seems related to tempest-lib IIUC
20:34:10 <ttx> devananda: then we have 3 parallel workshops at the same time as the beginning of the growth discussion
20:34:10 <ttx> devananda: what would you swap it with ?
20:34:16 <jaypipes> russellb, ttx: crap... no, that won't work. I'm giving my talk on tuesday at 11:15 :( so I would not be able to attend the API WG session.
20:34:24 <russellb> jaypipes: :(
20:34:34 <russellb> yeah, definitely hoped you could be at that one
20:34:36 <jaypipes> russellb: sorry: https://openstacksummitnovember2014paris.sched.org/event/0a602a57f19a73c9995dfb88f22a1538#.VE_90R8aekA
20:34:37 <ttx> jaypipes: you can still ask to move to another time on the conf side
20:34:38 <fungi> devananda: i think the only critical timing concern on third-party testing cross-project discussion was that it come before project-specific third-party testing sessions, but that will be the case regardless
20:34:47 <russellb> ttx: ++ :)
20:34:52 <annegentle> jaypipes: I can help with API WG and ask Everett Toews if he's available.
20:34:53 <mtreinish> devananda: the functional testing one, I think dkranz put it in the list
20:34:55 <jaypipes> ttx: k, lemme see what I can do.
20:34:55 <ttx> like on Monday or Wednesday
20:34:59 <devananda> time for me to board a plane ...
20:35:09 <russellb> devananda: o/
20:35:17 <jaypipes> annegentle: that would be swell. i can make the second half of the API WG session (or most of it at least)
20:35:32 <sdague> yeh, do we have existing session leaders for most of these? (defaulting to proposer I'd assume)
20:35:33 <ttx> fungi: I think anteaya wants the 3rd-party CI discussion as a feedback session
20:35:50 <russellb> sdague: trying to sort that out
20:35:53 <russellb> etherpad is kind of a mess
20:36:00 <sdague> ok
20:36:04 <annegentle> jaypipes: okay, sounds good, what's the time again? for API WG?
20:36:05 <russellb> most of them don't have a proposer listed
20:36:13 <sdague> russellb: oh, bummer
20:36:15 <ttx> Anyway, I propose we push that schedule publicly and we see what conflicts people start to report
20:36:18 <russellb> basically i added a new section at the top of the pad to try to clarify it
20:36:22 <russellb> still missing a bunch
20:36:31 <ttx> because I expect some PTLs to complain too
20:36:39 <jaypipes> annegentle: 11:15am on Tues
20:36:53 <sdague> I'm pleading ignorance as this is second day back, so still getting up to speed on the schedule :)
20:37:14 <russellb> sdague: wb!
20:37:27 <jaypipes> indeed, wb sdague
20:37:30 <russellb> so, should we sort out session leads async?  i can post to -dev list asking for volunteers ...
20:37:35 <russellb> or we can try to burn through it now
20:37:37 <annegentle> shoot I'm in the Design Summit 101 also with ttx
20:37:45 <jaypipes> russellb: I think that's the best bet.
20:37:46 <annegentle> guess it's on Everett
20:37:48 <ttx> annegentle: yay!
20:37:49 <sdague> yeh, ML might be good to flush out the original proposers
20:37:54 <russellb> sdague: ++
20:38:14 <russellb> ttx: can you raise this at the next meeting, too?
20:38:16 <jaypipes> russellb: so did the scheduler/gantt proposed session get nixed? not enough interest? :(
20:38:20 <sdague> and if there are still gaps tomorrow this time, play a game of who's it
20:38:23 <ttx> russellb: sure
20:38:24 <russellb> jaypipes: score of 0
20:38:50 <russellb> jaypipes: lack of interest, and a few -1s ... mainly that for now, it's a nova concern
20:39:08 <ttx> I think Nova scheduled two sessions on that topic
20:39:09 <russellb> get the split done ... too many people only want to talk about steps 12-18, and not the work that needs to get done first
20:39:13 <russellb> IMO
20:39:20 <sdague> russellb: +1000
20:40:02 <annegentle> only steps 1-7 may be discussed :)
20:40:08 <ttx> OK, let's move on -- russellb: post the selection to the ML and ask for leads ?
20:40:11 <jaypipes> russellb: k, understood (and I agree with you on that)
20:40:19 <russellb> cool
20:40:29 <russellb> ttx: wfm
20:40:31 <jeblair> after that, any session without a lead gets mordred as the lead
20:40:37 <jaypipes> russellb: course, that doesn't mean I won't be inundated with NFV people asking me about it... but whatevs :)
20:40:46 <russellb> jaypipes: ha, i feel your pain.
20:40:50 <ttx> #topic Governance for the openstack-specs repo
20:41:04 <ttx> A few weeks ago we allowed the creation of an openstack-specs repo for truly cross-project specs, and it was created as a TC-owned repository
20:41:16 <ttx> The question left is... how should we manage this repo ? who should get +1/+2... ?
20:41:28 <jeblair> well, everyone should get +1 i think :)
20:41:34 <russellb> jeblair: +1
20:41:38 <annegentle> great starting point jeblair
20:41:42 <annegentle> :)
20:41:45 <russellb> I think starting with TC as the group with +2 seems sane
20:41:47 <russellb> and see how it goes
20:41:49 <jeblair> maybe even -1? :)
20:42:13 <sdague> yeh tc as -core group seems sane
20:42:17 <russellb> not that we have *ALL* the expertise, but can work toward consensus, take all the inputs, etc, just like other cross project issues
20:42:23 <sdague> and normal review criteria for everyone else
20:42:53 <jeblair> should we put specs on the agenda before final approval?
20:42:59 <jogo> there is also the question of what it means for a spec to be approved there. Do all projects have to implement it now? etc.
20:43:00 <dhellmann> yeah, we may also want a different approval policy for that repo
20:43:26 <ttx> jogo: I guess it depends on the content of the spec
20:43:27 <mikal> I like the idea of an in-meeting check point before approving
20:43:29 <jeblair> like "this spec is ready for approval, put it on the tc meeting agenda for a final call for tc review, and approve in more or less the normal tc way?"
20:43:33 <russellb> a doc on expectations around this sounds like a ncie idea
20:43:37 <russellb> perhaps even a spec on specs!
20:43:41 <sdague> is that something we can have in the specs cross project session ?
20:43:54 <russellb> sdague: ++
20:44:08 <mtreinish> we try to do the meeting check in for qa specs, it's only moderately successful there...
20:44:17 <jaypipes> why can't we handle it similarly to how we handle openstack/governance?
20:44:39 <dhellmann> jaypipes: yeah, that's more or less how we do oslo specs
20:44:41 <annegentle> one final vote?
20:44:46 <sdague> jaypipes: and have ttx approve it all?
20:44:51 <jaypipes> and have ttx (or someone else?) be the only +2/+W'er, and have a policy of quorum or consensus that is enforced by ttx.
20:44:52 <dhellmann> sdague: right
20:44:52 <ttx> jaypipes: depends on how "similar" it is
20:44:57 <jaypipes> sdague: yeah.
20:44:59 <ttx> someone else++
20:45:07 <jaypipes> doesn't have to be ttx of course.
20:45:16 * jaypipes nominates annegentle
20:45:18 <sdague> I kind of like the idea that cross project specs have the ability for any community member to +1/-1 feedback
20:45:33 <jaypipes> sdague: of course, that wouldn't change...
20:45:36 <annegentle> heh
20:45:42 <sdague> well it's not the way governance works
20:45:50 <jaypipes> sdague: ? sure it is.
20:45:52 <russellb> nah
20:45:55 <sdague> nope
20:45:55 <dansmith> nope
20:45:57 <ttx> jaypipes: only TC members can +1
20:45:58 <russellb> only TC can +1/-1
20:46:00 <annegentle> it has to get as least as many votes as the TC election :)
20:46:03 <ttx> and I just tally the votes
20:46:04 <dansmith> I was disappointed I couln't +1 things in governance
20:46:05 <jaypipes> oh...
20:46:09 <jaypipes> well I be darned.
20:46:09 <dhellmann> sdague: it's more like oslo-specs, where anyone can vote, cores can vote 2, and by agreement only I approve things
20:46:16 <jeblair> anyone can leave a message of course
20:46:18 <russellb> we could revisit that idea i guess
20:46:27 <russellb> but anyway, i like everyone having +1/-1 on the specs
20:46:29 <annegentle> I think for cross project you're gonna need to gather more consensus
20:46:35 <jaypipes> well, sorry about that. I was misinformed.
20:46:35 <ttx> jeblair: could we have TC members do +2 and I would tally +2 with Workflow+1 instead ?
20:46:43 <russellb> in theory governance has the same cross project impact
20:46:47 <annegentle> so everyone +1/-1 is a good thing
20:46:47 <ttx> that would allow "normal people" to +1
20:46:50 <sdague> ttx: yeh, that might be fine
20:46:57 <jeblair> ttx: -2 is a problem
20:47:01 <sdague> jeblair: is it?
20:47:02 <ttx> sdague: I think we suggested that back then
20:47:07 <russellb> anyone remember why we did that for governance?
20:47:07 <mikal> I think we could all just agree that only ttx +A's
20:47:16 <jeblair> may i have the floor? :)
20:47:30 * ttx passes  the mike to jeblair
20:47:34 <russellb> jeblair takes the meeting lock
20:47:48 * markmcclain1 wonders why that is not a feature of meetbot
20:47:55 <jeblair> part of why governance is structured the way it is is due to limitations in a previous version of gerrit, and partially to make sure that the voting (since we _actually_ vote) is clear
20:48:07 <jeblair> since we have upgraded gerrit, we do have more options than before
20:48:28 <vishy> hot spec on spec action: http://cdn.vectorstock.com/i/composite/66,23/spectacles-vector-6623.jpg
20:48:36 <jeblair> so if we would like to enable +/-1 for everyone while maintaining the ability to discern tc votes, i believe we can come up with a proposal for that
20:48:36 <annegentle> vishy: SNORT
20:48:55 <jeblair> it will still be something different than the process used for other repos, but i think we can work it out
20:49:05 <jeblair> is that something the tc would like us to do?
20:49:08 <russellb> new column for TC votes or something?
20:49:13 <sdague> jeblair: so I think it's also OK to use normal core rev rules. TC members -1 things for 'needs more work' and -2 for 'hell no'
20:49:17 <ttx> being able to record random +1/-1 sounds like a good thing, yes
20:49:18 <jeblair> russellb: that may be an option
20:49:22 <jeblair> sdague: -2s block
20:49:25 <sdague> jeblair: yes
20:49:31 <jeblair> sdague: no tc member gets a veto
20:49:34 <russellb> i think so yes, i'd love to see a proposal for it
20:49:37 <sdague> for specs?
20:49:49 <jeblair> sdague: i believe governance is the repo under discussion?
20:49:50 <dhellmann> yeah, I think in this case blocking is a feature
20:49:50 <ttx> sdague: we are discussing governance now
20:49:54 <sdague> oh... sorry
20:50:03 <sdague> I missed a context switch
20:50:07 <jaypipes> jeblair: I'm certainly interested in a proposal to do that.
20:50:21 * dhellmann missed the context switch to
20:50:33 <fungi> -1..+1 for peanut gallery, -1..+2 for tc members, 0..+1 workflow for tc chair?
20:50:34 <russellb> yeah, tangent
20:50:39 <ttx> we can workl something out for governance, but that's pretty off-topic
20:50:50 <ttx> fungi: something like that yes
20:50:56 <jeblair> fungi: that still doesn't work
20:50:58 <sdague> yeh, governance repo discussion to tc list?
20:51:07 <jeblair> sdague: please no
20:51:08 <fungi> meh, right, tc -1 is different from peanut gallery -1
20:51:09 <russellb> because tc -1 is different than general -1
20:51:15 <annegentle> it's pretty relevant though to cross project and api wg repos
20:51:31 <jeblair> what i would like to do is come back to the tc with a proposal that meets the requirements
20:51:38 <jeblair> rather than designing something in this meeting
20:51:40 <ttx> so.. back to openstack-specs
20:51:52 <ttx> jeblair: can you propose something for that as well ?
20:51:59 <russellb> what are the requirements for specs though
20:52:02 <markmcclain1> jeblair: +1
20:52:06 <dhellmann> I'm happy with the way the governance repo works now, and I think we can manage the openstack-specs repo by consensus that only the appointed person approves
20:52:09 <russellb> liek any other repo?
20:52:17 <russellb> and tc as -core ?
20:52:18 <jeblair> yeah, i think i know the requirements for governance, but i don't know them for specs
20:52:31 <russellb> i think ^^^ works for me for specs
20:52:37 <ttx> works for me too
20:52:38 <dhellmann> russellb: +1
20:52:41 <sdague> russellb: yeh, +1
20:52:57 <jeblair> that gives tc members veto
20:53:01 <ttx> maybe with some rule that we want more than 2 approvers, but taht can be figured out as we go
20:53:05 <dhellmann> jeblair: yes, that's what we want for specs
20:53:14 <ttx> jeblair: sounds reasonable for specs
20:53:38 <jeblair> sounds like a starting point i guess :)
20:53:43 <ttx> so we don't really need a secretary to tally votes like for govrenance
20:53:44 <russellb> yeah
20:53:53 <sdague> yep
20:53:53 <russellb> (re: starting point)
20:54:02 <jeblair> i do still think that final approval should happen after it hits the meeting agenda
20:54:04 <ttx> just do the usual core dance
20:54:06 <russellb> jeblair: ++
20:54:08 <dhellmann> jeblair: ++
20:54:11 <markmcclain1> jeblair: ++
20:54:15 <russellb> but only 2 +2s before +W?
20:54:16 <russellb> or more?
20:54:34 <annegentle> seems like you need +2 from affected PTLs?
20:54:36 <russellb> or just let people use good sense?
20:54:37 <dhellmann> russellb: I look for broad consensus on oslo-specs before I approve them
20:54:43 <jeblair> annegentle: affected ptls may not have +2
20:54:52 <annegentle> but that's gaterhing consensus
20:54:52 <dhellmann> annegentle: yeah, I'd look for a +1
20:54:56 <ttx> maybe +1s forom affected ptls
20:54:59 <jeblair> yeah
20:55:14 <annegentle> okay that works too as long as they're aware, voting, we know there is consensus gathered
20:55:28 <ttx> jeblair: you up for proposing a technical solution for both ?
20:55:46 <sdague> or you just talk to people and be reasonable
20:55:53 <russellb> sdague: that.
20:55:56 <sdague> which is typically what happens in the specs today
20:55:58 <dhellmann> sdague: we could do with more of that
20:56:03 <jaypipes> sdague: be reasonable? what's wrong with you! :P
20:56:04 <jeblair> ttx: it sounds like we have what we need for specs; or do you want me to write the process doc?
20:56:14 <jeblair> ttx: and yes, i will make a proposal for governance
20:56:16 <russellb> yeah, sounds like specs is just like everything else
20:56:28 <ttx> #action jeblair to propose some novel Gerrit rules solution for governance repo
20:56:39 <ttx> jeblair: maybe write it up ?
20:56:43 <jeblair> ttx: can do
20:56:45 <russellb> but we could probably still use a general guidelines doc on cross project specs to clarify for people what kind of consensus we'll be looking for
20:56:46 <ttx> (the openstack-specs rules)
20:57:01 <jeblair> yep.  should i propose that to openstack-specs or governance?
20:57:11 <ttx> #action jeblair to write up proposed Gerrit rules for openstack-specs repo
20:57:15 <ttx> jeblair: openstack-specs I'd say
20:57:17 <russellb> cool.
20:57:21 <annegentle> jeblair: ttx: okay to also consider the api wg repo as well in your proposal? Seems more efficient
20:57:23 <ttx> self-documentation
20:57:44 <dhellmann> annegentle: I thought we said the wg would have +2 in that repo
20:57:51 <russellb> api wg repo deserves its own consideration
20:57:56 <russellb> maybe we can just cover that next week?
20:57:58 <ttx> annegentle: api wg is a bit of another beast, I'd like them to self-organize first
20:58:07 <russellb> ttx: ++
20:58:13 <ttx> no need to impose process externally
20:58:23 <ttx> let's see what they come up with
20:58:39 <ttx> ok, let's run through the other items in agenda
20:58:46 <ttx> #topic Other governance changes
20:59:10 <ttx> I have two changes from mordred that are blocked, couldn't really get hold of him, but I'll soon corner him physically
20:59:15 <ttx> * Remove support for vendor extensions from our code (https://review.openstack.org/122968)
20:59:18 <ttx> * Add a docs environment to the testing interface (https://review.openstack.org/119875)
20:59:23 <ttx> #topic Open discussion
20:59:29 <ttx> Anything urgent to discuss, anyone ?
20:59:38 <russellb> looking forward to seeing you all in person soon!
20:59:42 <ttx> Question about upcoming BoD/TC meeting, Design Summit etv ?
20:59:43 <dhellmann> same!
20:59:44 <ttx> etc
20:59:46 <sdague> yep, definitely
21:00:00 <markmcclain1> yeah.. will be good to get in same room
21:00:10 <russellb> esp for the big governance changes discussion ...
21:00:13 <ttx> around the same dinner table too
21:00:15 <russellb> really need some real-time chat time on that
21:00:32 <markmcclain1> food for thought… but don't want to discuss now: should we be considering an in person TC meetup 2015 Q1?
21:00:33 <ttx> ok then, safe travels
21:00:35 <jeblair> see you in paris!
21:00:36 <annegentle> where are the pods ttx?
21:00:36 <dhellmann> russellb: maybe we should plan to have lunch together once or twice so we can cover that?
21:00:45 <dhellmann> (everyone, not just the 2 of us :-)
21:00:47 <ttx> annegentle: in the Design Summit space (Le Meridien)
21:00:50 <mikal> markmcclain1: I like the idea
21:00:51 <annegentle> okay thanks
21:00:52 * jaypipes will come to TC dinner dressed in a giant tent.
21:01:00 <ttx> #endmeeting