17:00:00 <jroll> #startmeeting ironic
17:00:01 <openstack> Meeting started Mon Oct 17 17:00:00 2016 UTC and is due to finish in 60 minutes.  The chair is jroll. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:03 <devananda> o/
17:00:04 <openstack> The meeting name has been set to 'ironic'
17:00:06 <mat128> o/
17:00:07 <vdrok> o/
17:00:07 <JayF> o/
17:00:08 <rpioso> o/
17:00:10 <mjturek> o/
17:00:16 <yuriyz|2> o/
17:00:17 <bfournie> o/
17:00:18 <krtaylor> o/
17:00:22 <lucasagomes> o/
17:00:36 <mariojv> o/
17:00:36 <jroll> as always, agenda is here:
17:00:38 <hshiina> o/
17:00:39 <jroll> #link https://wiki.openstack.org/wiki/Meetings/Ironic#Agenda_for_next_meeting
17:00:46 <jroll> #topic announcements / reminders
17:00:57 <jroll> so the summit is next week
17:01:21 <rloo> o/
17:01:30 <jroll> accordingly, let's not have this meeting next week
17:01:38 <jlvillal> o/
17:01:41 <jroll> don't see any reason not to have it the following week, though
17:01:42 <TheJulia> o/
17:02:06 <jroll> also, I mostly won't be here for the rest of this week
17:02:06 <lucasagomes> jroll, it's halloween no ?
17:02:13 <milan> o/
17:02:21 <dtantsur> o/
17:02:29 <rloo> lucasagomes: we can have a ghostly meeting?
17:02:34 <jroll> I'll be here a bit today and tomorrow and then gone
17:02:39 <jroll> hit me on hangouts if anything urgent :)
17:02:40 <lucasagomes> rloo, ++ let's do it heh
17:02:44 <sambetts> o/
17:02:44 <jroll> lucasagomes: tis, but that's not a 'day off' holiday anywhere that I know of
17:02:50 <jroll> I'll wear a costume
17:03:02 <mat128> videoconference? :)
17:03:06 <lucasagomes> I think it is in ireland, but for another reason
17:03:07 <lucasagomes> bank holiday
17:03:09 <lucasagomes> anyway...
17:03:12 <jroll> ah
17:03:28 <jroll> anyone have other announcements or reminders?
17:03:47 <rloo> jroll: reminder for folks attending summit read specs/prepare beforehand
17:03:54 <jroll> ++
17:03:57 <jroll> all of my ++
17:04:17 <rloo> jroll: maybe we should send an email out about that.
17:04:47 * rloo volunteers after i see what the state of the specs, etc are...
17:04:53 <jroll> rloo: thanks :)
17:05:10 <milan> rloo ++ mail w/ specs to read would be nice :)
17:05:11 <jroll> I'd be surprised if people that don't know they should read background info will actually read the ML, but we shall see
17:05:40 <JayF> rloo: ++ and also good for folks not attending to get input in the spec before others talk about it in the summit
17:06:00 <milan> JayF, good point
17:06:25 <jroll> anything else?
17:06:40 <rloo> jroll: oh, any meetup at summit?
17:06:46 <rloo> jroll: i mean, evening gathering
17:06:56 <jroll> rloo: lucasagomes was going to look into a thing but as of yet nothing planned
17:07:00 <lucasagomes> rloo, I want to book a restaurant
17:07:05 <lucasagomes> we need to decide the day
17:07:10 <rloo> lucasagomes: thx for volunteering!
17:07:11 <vdrok> and are we going to have an arrival list as the last time?
17:07:15 <lucasagomes> let's do it at the end of the meeting ? Or should I send an email to the ML ?
17:07:26 <lucasagomes> rloo, I will try, my spanish is not that great either
17:07:26 <jlvillal> vdrok: https://etherpad.openstack.org/p/ironic-ocata-summit-people  ???
17:07:49 <jroll> TIL
17:07:52 <vdrok> oh, thanks jlvillal :)
17:07:56 <krtaylor> meetup++
17:07:57 <mat128> jroll: might be worth a #link
17:08:10 <jroll> #link https://etherpad.openstack.org/p/ironic-ocata-summit-people
17:08:22 * krtaylor missed planning that for ocata
17:08:33 <rloo> could we use that etherpad to indicate availability for dinner?
17:08:51 <dtantsur> rloo, I'd prefer doodle, it's easier to analyze
17:09:00 <rloo> dtantsur: ok, that works too
17:09:04 <lucasagomes> yeah I think doodle is easy to visualize
17:09:07 * lucasagomes creates a doodle
17:09:15 <jroll> alright, shall we move on?
17:09:21 <rloo> although wasn't it doodle that the/some asians couldn't access?
17:09:31 <rloo> + movin' on
17:09:45 <jroll> can always add in the email "if you can't get to doodle, reply here"
17:09:54 <jroll> #topic subteam status reports
17:10:03 <jroll> as always, those are here
17:10:05 <jroll> #link https://etherpad.openstack.org/p/IronicWhiteBoard
17:10:12 <jroll> line 80 this time
17:10:22 <jroll> woo, serial console is finally done \o/
17:10:37 <jroll> thanks to folks that were involved there
17:10:41 <rloo> jroll: ++. the reviews by the nova folks were ... interesting...
17:10:43 <jroll> root device hints also done, except docs \o/
17:10:44 <jlvillal> +1!
17:10:51 <lucasagomes> o/
17:12:01 <jroll> and our first real notifications were approved today, woo
17:12:06 <mariojv> \o/
17:12:14 <jroll> yay we did stuff between release and summit
17:12:14 <rloo> lots of woo's today :)
17:12:18 <milan> :)
17:12:57 <rloo> dtantsur: wrt the work around for neutron pool issue -- whoever is going to remove the workaround, I suggest do it early this week, or wait til after summit.
17:13:20 <jroll> so if people need things to work on this week beyond prepping for summit, looks like BFV specs, driver comp specs, portgroups patches
17:13:28 <dtantsur> we can wait, I'll just leave this bug as critical as a reminder :)
17:13:59 <rloo> dtantsur: it isn't critical any more though but i guess it is fine to leave it as that. you're the only one that looks :D
17:14:23 <dtantsur> sigh...
17:14:24 <dtantsur> :)
17:14:54 <jroll> heh
17:15:01 <jroll> anything else on this topic?
17:15:29 <jroll> k
17:15:43 <jroll> #topic Where should Ironic keep it's canonical admin guide until the docs team supports writing the admin guide in-tree?
17:15:48 <jroll> #link http://lists.openstack.org/pipermail/openstack-dev/2016-October/105493.html
17:15:58 <jroll> someone was bad and didn't put their name here, but I suspect this is a JayF topic?
17:16:09 <JayF> Whoops!
17:16:12 <JayF> Yeah,
17:16:28 <JayF> the basic outline is this: The docs team does not want to do admin-guides in-tree this cycle (even if I was willing to do the work)
17:16:39 <JayF> so we have to figure out what we want to do for admin-guide stuff moving forward:
17:16:44 <JayF> 1) Keep it in the docs repo, and contribute to it
17:17:10 <JayF> 2) Replace the minimal information in the existing admin-guide (which appears to have been copied from our dev docs) with a link back to our dev docs and put the admin guide in our dev docs
17:17:17 <dtantsur> for me, it sounds like something to try to discuss with docs folks on the summit...
17:17:31 <dtantsur> otherwise, I'd prefer #2
17:17:32 <mat128> JayF: is there a 3?
17:17:37 <JayF> I don't anticipate we'll do a lot of admin guide work this cycle, but it's getting hard to draw the line between "admin" stuff and "install" stuff
17:17:51 <jroll> 3) continue ignoring the concept of an admin-guide until we can do it in-tree :)
17:17:54 <JayF> mat128: I don't really think so, I mean, there's an option to do nothing implied
17:18:08 <rloo> dtantsur: what sort of discussion at summit, are you thinking of? JayF already had a discussion with them in the ML
17:18:25 <JayF> jroll: I don't think that's really a reasonable option; information on how to administrate ironic has to go /somewhere/ and I'm trying to avoid us overloading our install-guide in the short term.
17:18:27 <mat128> JayF: move current admin-guide to our install-guide as the line is so thin anyway?
17:18:40 <dtantsur> rloo, well, get to know their motivation better, maybe future plans. maybe show them a PoC.
17:18:45 <rloo> so the thing with ignoring (status quo) is that there IS ironic stuff in the admin-guide, but it is a subset of what we have in our dev docs
17:19:11 <dtantsur> rloo, iirc it's pretty miserable, but it's linked from the openstack main docs page.. so that's what people find first
17:19:13 <jroll> dtantsur: they plan to do it in pike or later, fwiw
17:19:34 <rloo> dtantsur: JayF won't be at the summit. and based on the replies in the ML, i doubt that anything will happen before pike.
17:19:36 <JayF> My opinion is that we should do #2; put an admin-guide in-tree, with a specific mind to organizing it so it's easy to migrate to an in-tree admin-guide later
17:19:36 <dtantsur> jroll, that's strange that they don't accept help for that..
17:19:44 <JayF> dtantsur: that's exactly what I thought too
17:19:44 <jroll> dtantsur: ¯\_(ツ)_/¯
17:19:46 <mariojv> 2) sounds ok. that way we can get something useful out quicker, and maybe move it to the docs repo later once we have something high quality
17:20:01 <rloo> I vote for 2)
17:20:02 <JayF> dtantsur: I basicaly offered to do admin-guide intree for them, and they turned me down
17:20:15 <dtantsur> hmmmm.. ok, #2 then.
17:20:23 <jroll> JayF: yeah, indeed, I'd love for the later migration to be mv ironic/doc/source/admin-guide ironic/admin-guide
17:20:30 <dtantsur> if they're so busy, they probably won't be pleased with a spam of our patches as well
17:20:49 <jroll> so yeah, I'm good with #2
17:20:50 <JayF> I mean, they said they're willing to take our patches for the out-of-tree admin guide
17:21:01 <JayF> I'm just not keen on that as we tend to like having our stuff in-tree
17:21:17 <JayF> especially since the relationship between dev docs / install guide / admin guide is ... not well established yet for ironic
17:21:18 <dtantsur> well, yeah, we can provide much more valuable review IMO
17:21:26 <rloo> dtantsur: so folks should read the entire thread. Basically, Lana said: "However, I want to collect data based on that experience before we begin, rather than barrelling on down that road. Additionally, I don't feel as though the Install Guide is absolutely complete until we've got the index page done. As I said earlier, there is also the consideration of the short release cycle for Ocata."
17:21:27 <JayF> It sounds like consensus is go with #2?
17:21:39 <mat128> jroll, JayF: I like the idea of having the admin-guide in-tree but published inside of the developer docs and/or install-guide
17:22:12 <jroll> JayF: sounds like it to me
17:22:28 <rloo> if we eventually want to have the admin guide intree, i don't think it makes sense to move it outoftree and then back in again
17:22:36 <dtantsur> rloo++
17:22:42 <jroll> +1
17:22:55 <JayF> So I'll take that action item, and will work to have patches to eliminate the partial information from the out-of-tree guide and link it to the in-tree dev docs by the time you all are back from summit
17:23:08 <rloo> thx JayF!
17:23:08 <mat128> JayF: keep me in the loop, not going to the summit either :(
17:23:37 <jroll> JayF: awesome, thanks
17:24:08 <jroll> any other comments here?
17:24:43 <jroll> #topic MoltenIron as a sub project
17:24:49 <jroll> #link https://review.openstack.org/#/c/387418/
17:25:00 <jroll> mjturek: krtaylor: hamzy: this is you
17:25:15 <mjturek> So, we committed a blueprint today for MoltenIron with the intention of getting it under Ironic's governance.
17:25:24 <krtaylor> sure, molten iron is a tool we wrote to manage pools of baremetal test targets, we've talked about it before and at last summit
17:25:55 <mjturek> it's currently living in the third-party-ci-tools repo
17:26:22 <krtaylor> we were wanting to know what ironic wanted us to do with it, we'd love to get input from others, I know several vendors are interested
17:26:28 <devananda> do any of the CI teams outside of IBM use this?
17:26:54 <krtaylor> yes, Dell I believe is one
17:27:00 <jroll> just to be clear - the question at hand is "should molteniron become part of the ironic umbrella (under ironic governance)?"
17:27:16 <krtaylor> there are a few others that were interested if it had featureXYZ
17:27:25 <rloo> have you discussed with CI, about the 'shortcomings' of nodepool? I mean, what about changing nodepool to do what you want?
17:27:35 <devananda> krtaylor: that's encouraging.
17:27:55 <mat128> krtaylor, mjturek: we have (had) this requirement but at a much larger scale - nodes, racks, servers, pdus, etc.
17:28:07 <devananda> rloo: last time I spoke with the CI team, they said that nodepool v3 would include functionality to do this sort of thing, but that was ~6mo ago
17:28:34 <rloo> devananda: oh, maybe they have it done now :D (j/k)
17:28:34 <JayF> Can I ask the flip-side question: Why wouldn't we want molteniron under the Ironic umbrella?
17:28:40 <krtaylor> yes, but there were reasons it wouldnt work, for us anyway
17:28:48 <JayF> It's used by ironic 3rd party UI, if it gets superceded later it gets superceded
17:28:53 * krtaylor looks at mjturek
17:29:01 <devananda> JayF: my only reason to object would be "only one vendor is using it". but that's not the case, apparently :)
17:29:03 <JayF> I'm not sure I understand any scenario we'd say "no" to the question at hand
17:29:16 <rloo> i think it makes sense to talk to the CI team, if they are going to do something similar soon, why not just help them get that going, instead of duplicating.
17:29:20 * dtantsur is still confused why we say yes
17:29:33 <dtantsur> JayF, it depends on what we imply under "taking under our governance", I guess
17:29:36 <jroll> right, so the tradeoff is "core team is responsible for it", but it means molteniron is able to publish to docs.o.o
17:29:36 <mat128> krtaylor, mjturek: is this only for CI or it's a full fledged cmdb?
17:29:56 <devananda> rloo: ++
17:29:56 <dtantsur> JayF, e.g. what prevents us taking ironic-staging-drivers under our governance?
17:30:00 <jroll> (it could have a separate/additional core team, of course, but the ironic team would be responsible for it)
17:30:09 <mjturek> matt128 currently just CI
17:30:17 <sambetts> mat128: it is literally a tool for reserving a node to run tests against
17:30:33 <sambetts> mat128: then giving it back for the next test run
17:30:39 <JayF> dtantsur: I really wouldn't be opposed to that either, honestly
17:30:57 <krtaylor> yes, just a tool for managing reservations for baremetal targets
17:31:02 <rloo> I feel like this falls under CI, not ironic
17:31:09 <devananda> taking it under our governance means the PTL is taking responsibility for it, we need to give time in meetings to the project, and granting election voting rights to contributors to that project
17:31:10 <dtantsur> JayF, me neither. we don't include it, because we don't want to assume the same level of support as for our production drivers.
17:31:18 <krtaylor> rloo, it is very ironic specific
17:31:24 <jlvillal> maybe related. Today there was a thing with nodepool using ironic discussion: http://eavesdrop.openstack.org/meetings/third_party/2016/third_party.2016-10-17-15.00.log.html
17:31:38 <jroll> btw, this is the molteniron repo:
17:31:40 <jroll> #link https://github.com/openstack/third-party-ci-tools/tree/master/nodepool/molteniron
17:32:10 <jroll> roughly 1500-2000 LOC, seems pretty simple
17:32:16 <krtaylor> the third-party-ci-tools repo doesnt require tests,so we are looking to have automated tests,etc so a new project or a subproject is the question
17:32:23 <rloo> krtaylor: you're going to be at the summit -- would you be willing to talk to CI folks about molteniron, see where/if nodepool works or if there is common ground?
17:32:40 <dtantsur> I guess I'm just -0 to all inclusions, until we figure out what "being under our umbrella" fully requires and implies
17:33:00 <jroll> I seem to remember quite a few talks between these folks an infra folks, but maybe I'm thinking of something else
17:33:07 <rloo> dtantsur: to future inclusions? since we already have a few things under the ironic umbrella :)
17:33:29 <krtaylor> yes, there have been several discussions, nodepool and also zuul were proposed
17:33:43 <JayF> dtantsur: what devananda said above matches what I was thinking it meant; and tbh I don't like splitting a hair to say if you work on something like molteniron you get less rights than someone who works on inspector, bifrost, ipa, ironic, etc
17:33:45 <dtantsur> rloo, well, we can always reconsider them too.. I don't remember anything feeling foreign, but maybe it's only me
17:33:53 <JayF> dtantsur: I guess I'm just more +0 by default :D
17:34:19 <rloo> dtantsur: i agree with you wrt knowing more specifically, what it means for things under the ironic umbrella
17:34:26 <dtantsur> yeah, I'm only asking because we have some problems with ironic-staging-drivers, which could be worked around by it being under an official project
17:34:49 <jroll> dtantsur: what sort of problems?
17:34:51 <dtantsur> and this is not something out of blue, that's our drivers, developed by ironic contributors and reviewed by ironic core reviewers (a subset of them)
17:35:05 <dtantsur> jroll, no docs, no release support, no i18n (to name a few)
17:35:13 <jroll> nod
17:35:13 <lucasagomes> a better guide on what should be under the Ironic umbrella and what not would be very welcome
17:35:19 <dtantsur> I've gone through all this with ironic-discoverd  /me is nostalgic
17:35:45 <lucasagomes> but I also think that we probably should check nodepool first, if they are willing to extend it to support baremetal nodes we probably should help 'em out
17:35:46 * mat128 doesnt understand why projects not under governance can't have docs, but here is not the place for this
17:35:50 <rloo> seems like openstack should have a general guide, i don't know that ironic is special from other umbrellas
17:35:53 <dtantsur> so, I'm not against "why not" approach, I just want it stated officially :)
17:36:06 <krtaylor> re: nodepool, we could bring that up again, but generally there has not been much traction for third-party (non-infra) CI test changes
17:36:13 <lucasagomes> mat128, it can, but not official under openstack (we use http://readthedocs.org)
17:36:13 <JayF> mat128: being in the "big tent" gives you access to shared resources, such as docs
17:36:24 <jroll> nodepool is also designed to only speak to nova, AIUI (intentionally such)
17:36:46 <TheJulia> My understanding is that will be changing over the next six months or so
17:36:48 <devananda> jroll: IIUC, that's changing in v3...
17:36:50 <mjturek> right, the changes don't necessarily fit there. It's something we could propose, but we have a working solution sitting here
17:37:03 <jroll> ok, I had not heard that yet
17:37:13 <sambetts> so the problem with nodepool is that it is designed to solve a different problem, e.g. its designed to provide nodes to Zuul as test slaves, but we want to provide BM resources for the next layer down, i.e the instances created during the tests
17:37:18 <jroll> but yes, in the meantime molteniron is useful for third party CI people
17:37:41 <krtaylor> yes, thats what we've found
17:37:54 <devananda> so we have two discussions going in parallel: a) should this CI tool be embraced, or deffered until nodepool gets better? b) should it be part of the ironic project team umbrella, and thereby get access to shared resources, give contributors voting rights, etc
17:38:05 <mat128> sambetts: so molteniron is about delegating hardware resources to a test run so you don't have resource clash
17:38:11 <sambetts> mat128: yes
17:38:22 <mat128> sounds like our day to day problems from 7-8 years ago :)
17:38:26 <mat128> with a physical lab
17:38:26 <krtaylor> exactly, and as late as possible
17:38:27 <mat128> got it then
17:38:38 <devananda> I think that, regardless of (a), if molteniron is used today by two of our driver maintainers (ibm and dell)
17:38:53 <JayF> devananda: and maybe c) What exactly does it mean to be a part of Bare Metal governance, and if we need clear guidlines about what is in/out
17:38:58 <devananda> and the designated driver maintainers are the ones working on molteniron already
17:39:21 <devananda> it's already being developd by the ironic project team and I don't see a reason not to include the project
17:39:31 <mariojv> i think i'd be +1 to this if the mission of MoltenIron were something more general, instead of solely testing/ci. this would definitely be easier to decide if we had clearer definitions of what is/is not under our umbrella
17:39:44 <devananda> JayF: what it means is quite clear, from a TC / Foundation perspective
17:39:50 <dtantsur> devananda, do you propose that as a general guidance?
17:40:12 <devananda> JayF: voting rights in the PTL and TC election, access to cross-project resources, access to summit / PTG time (as determined by the PTL)
17:40:14 <mat128> mariojv: "Delegate parts of a physical lab to an Ironic installation" sounds generic enough
17:40:22 <rloo> it may be clear from a TC/Foundation perspective but it isn't clear to us :-(
17:40:29 <JayF> devananda: I agree with your logic about why it should be included. My question would then be: why wouldn't something like ironic-staging-drivers be included under the same logic?
17:40:30 <devananda> rloo: ^ does that help?
17:40:39 <jroll> devananda: the TC/foundation doesn't mandate what is part of a project team's governance, though, that's completely up to the project team
17:40:52 <devananda> JayF: because, AIUI, the core team prevoiusly said "we do not want to accept responsibility for maintaining these drivers"
17:41:00 <devananda> jroll: indeed
17:41:01 <rloo> krtaylor: so you want to be under the ironic umbrella why again?
17:41:06 <jroll> some of the core team said that, yes
17:41:49 <lucasagomes> yeah, all cores in the staging-drivers are also cores in ironic (/me don't want to diverge this conversation to it tho)
17:41:51 <rloo> krtaylor: i mean, molteniron is being used as-is now. what will change (what do you want to have changed) if it is under our umbrella
17:41:57 <sambetts> For those interested, https://www.openstack.org/summit/barcelona-2016/summit-schedule/events/16921/cross-project-workshops-where-to-draw-the-line-for-proprietary-code-with-drivers
17:42:01 <sambetts> dtantsur:  ^
17:42:07 <krtaylor> rloo, hehheh, we are splitting out of third-party-ci-tools repo, so either a nongoverned project or the ironic umbrella made sense to us
17:42:21 <dtantsur> devananda, well, then "accept responsibility" is also an argument, right?
17:42:27 <rloo> krtaylor: and i may have missed it. why are you splitting out? getting kicked out?
17:42:29 <dtantsur> do we want to accept responsibility for this new thing?
17:42:46 <krtaylor> rloo, no, not kicked out, we want to add tests, docs, ect
17:43:03 <mjturek> right, it's hard for us to manage enhancements  under it as well.
17:43:04 <krtaylor> rloo, third-party-ci-tools repo is tests noop
17:43:23 <rloo> krtaylor: ok. i think this is worth talking to infra folks about first.
17:43:26 <dtantsur> fwiw you can add tests (even dsvm) for non-official projects
17:43:32 <jroll> to be clear, you can use openstack... what dtantsur said
17:43:36 <devananda> krtaylor: splitting the project code into a separate repository seems orthogonal to whether that is under ironic project team governance
17:43:38 <dtantsur> ironic-staging-drivers has a dsvm job, for example
17:43:56 <mat128> devananda: ++
17:43:58 <devananda> krtaylor: you can easily create your own project, add tests, publish docs to readthedocs, etc
17:44:01 <jroll> so maybe we should do this:
17:44:14 <jroll> molteniron friends can work on making their own repo with tests running and such
17:44:17 <devananda> and then very easily be added to ironic governance at any later time
17:44:27 <jroll> while we go to the summit and bikeshed about the governance topic
17:44:30 <krtaylor> devananda, agreed
17:44:39 <devananda> jroll: ++
17:44:40 <mjturek> jroll: sounds reasonable to me!
17:44:43 <TheJulia> jroll: ++
17:44:45 <JayF> I mean, I disagree
17:44:47 <sambetts> sounds good
17:44:48 <jroll> and send summary to the ML for those not at the summit
17:44:49 <dtantsur> jroll++
17:44:50 <rloo> jroll: and talk to infra folks about molteniron
17:44:55 <JayF> because I'd like people not at the summit to have input into the governance topic
17:44:57 <jroll> and continue the conversation there
17:45:07 * jroll will talk in full thoughts
17:45:09 <lucasagomes> jroll, agree... maybe we should talk there and then ML the outcome
17:45:25 <dtantsur> maybe we can start with the ML, then continue in person, then again the ML?
17:45:28 <mat128> JayF: we should have a virtual summit :)
17:45:31 <dtantsur> to not exclude JayF and mat128 from the topic?
17:45:37 <jroll> rewritten: molteniron friends can work on making their own repo with tests running and such, while we go to the summit and bikeshed about the governance topic, and send summary to the ML for those not at the summit, and continue the conversation there
17:45:51 <krtaylor> that seems reasonable, mjturek you agree?
17:45:57 <mjturek> +1
17:46:01 <lucasagomes> jroll, +1
17:46:05 <jroll> dtantsur: we could, yeah
17:46:18 * rloo still wants to know what infra is thinking wrt v3, molteniron
17:46:19 <mat128> jroll: I'm fine with that, but unsure if this is what JayF meant. Will let him answer though
17:46:31 <dtantsur> starting with the ML could help us get ideas from other teams
17:46:42 <dtantsur> e.g. neutron folks definitely have a different view ;)
17:46:44 <JayF> I just wanna make sure folks not at the summit don't end up having zero input beyond just reading a consensus developed without them
17:46:50 <jroll> dtantsur: heh
17:47:01 <JayF> I know how well we take notes for these sort of ... "bikeshed about X" discussions at the summit
17:47:02 <jroll> JayF: we won't stamp anything
17:47:07 <JayF> and traditionally most of the decision is made there
17:47:09 <mjturek> heh
17:47:12 * rloo glad that JayF isn't going to the summit so he can stand up for those that don't attend
17:47:14 <jroll> but we need to document it post-summit
17:47:32 <krtaylor> agreed
17:47:33 <JayF> I am -1 to a solution that ends in "talk about it at the summit".
17:47:34 * dtantsur prefers to see summits as "make up our own minds" rather than "set something in stone forever"
17:47:49 <jroll> JayF: which solution ended with that?
17:47:54 <mat128> dtantsur: nothing is ever set in stone :)
17:48:05 <dtantsur> mat128, yeah, but I think you get the idea :)
17:48:21 <JayF> Literally what you just said; the discussion happens at the summit with a summary to the ML. Myself and others not in Barcelona are going to be missing a whole truckload of context.
17:48:31 <JayF> I mean, if that's what the majority says, it is what it is
17:48:37 <jroll> JayF: "and continue the conversation there" (there == ML)
17:48:41 <jroll> ...
17:48:52 <JayF> but especially for governance things we need to ask how open it is to have the bulk of the conversation in aplace that requires $$$$ to attend
17:48:53 <dtantsur> my version: molteniron friends can work on making their own repo with tests running and such, while we go to the summit and bikeshed about the governance topic, and propose a summary for further discussion to ironic-specs
17:49:18 <mat128> ^ i like that wording
17:49:20 <rloo> JayF: what's the use of going to the summit then, if not to have discussions etc?
17:49:23 <mat128> less interepretation
17:49:29 <jroll> rloo++
17:49:42 <jroll> JayF: it sounds like you're saying we shouldn't discuss anything at the summit
17:49:52 <jroll> or anything important, at least
17:50:07 <mariojv> i'm +1 to dtantsur's suggestion, as someone not going to the summit. we can still have input there at least
17:50:08 <JayF> I'm saying, especially for governance stuff, we should be careful to ensure the conversations happen in a place where everyone has an equal voice.
17:50:23 <JayF> I clearly wouldn't feel so strongly this way if I was going to the summit
17:50:27 <mariojv> if it's something really important that's being discussed like governance, maybe require a stricter threshold for passing it, like 4 +2s
17:50:39 <JayF> but I'm glad to have this experience to consider next time I *am* there hiow folks who are excluded might feel
17:50:39 <devananda> JayF: it's the nature / purpose of the summit that important discussions happen
17:51:10 <rloo> look, nothing gets totally passed at the summit. or shouldn't. i hope whatever was discussed/decided upon, is reflected in some patch that people can comment on post-summit
17:51:19 <devananda> rloo: ++
17:51:22 <lucasagomes> rloo, yeah
17:51:23 <jroll> +1
17:51:29 <dtantsur> that's why I propose an ironic-specs patch (or docs or whatever) later on
17:51:32 <rloo> and if that isn't the case, please point it out
17:51:34 <devananda> I don't think anyone has said we're going to decide once-and-for-all on anything at the summit
17:51:36 <jroll> we can't just restrict discussions because folks may miss context, IMO
17:52:07 <jroll> now, I'm totally +2 on starting the discussion now, in the ML
17:52:19 <jroll> but I don't want to be the one to start that thread and then go on vacation for the week
17:52:29 <devananda> but we can't NOT use that time to discuss important things -- it is the time with peak mass of people involved in the project
17:53:01 <devananda> even this IRC meeting doesn't have every contributor, and for some years, we didn't even have every core here, because of timezone differences
17:53:21 <jroll> (and still don't)
17:53:50 <devananda> this is why TC uses gerrit
17:54:06 <JayF> This IRC meeting is at least logged in its entirety. That's not true of summit sessions. At this point, it doesn't matter though, the path forward is decided. I just don't like it, and that's OK.
17:54:08 <devananda> and requires every motion be posted for a certain number of days, and discussed in at least one meeting, before being passed
17:54:17 <JayF> devananda: ++ tjat
17:54:31 <JayF> *That's what I care more about; discussions at the summit are more opaque
17:54:37 <jroll> JayF: like I said, we can start now in the ML, if somebody is willing to do that
17:54:52 <jroll> other than that I see no alternatives
17:54:52 <mjturek> jroll: I can start it up
17:54:58 <mjturek> if that's preferred
17:54:59 <JayF> thanks mjturek
17:55:07 <jroll> mjturek: thanks
17:55:09 <jroll> however
17:55:22 <TheJulia> 5 minute warning
17:55:24 <jroll> I would like the ML conversation to be around "what should and should not be in ironic's governance"
17:55:30 <jroll> not "should molteniron be there"
17:55:36 <mjturek> ahhh sorry
17:55:54 <jroll> mjturek: you're still welcome to start that discussion, of course
17:56:04 <jroll> just pointing out that's the real question here
17:56:15 <jroll> going to move to open discussion in case there's anything else, thanks TheJulia
17:56:19 <mjturek> jroll: sure, I'll post to the ML later
17:56:20 <jroll> #topic open discussion
17:56:23 <jroll> mjturek: thanks
17:56:36 <lucasagomes> so the doodle about the meetup in barcelona is here: http://doodle.com/poll/56qkr5rsxu447acr
17:56:41 <lucasagomes> should I send it to the ML as well ?
17:56:51 <jroll> lucasagomes: can't hurt
17:56:59 <lucasagomes> jroll, ack, will do after the meeting
17:57:09 <krtaylor> lucasagomes, nice
17:57:17 <lucasagomes> so ppl willing to attend the meeting, please vote for the best date at
17:57:18 <lucasagomes> #link http://doodle.com/poll/56qkr5rsxu447acr
17:57:33 <dtantsur> lucasagomes, I'd wish it has "tentative" option
17:57:50 * lucasagomes is not doodle expert
17:57:57 <lucasagomes> dtantsur, I can edit it and try to add
17:58:11 <jlvillal> lucasagomes: Is that assuming in the evening?
17:58:11 <dtantsur> lucasagomes, yeah, it's somewhere in "advanced" options
17:58:20 <lucasagomes> jlvillal, yes
17:58:26 * lucasagomes should add it to the description
17:58:29 <lucasagomes> dtantsur, I will take a look
17:59:01 <lucasagomes> that said, it's not guaranteed that we get a venue. I got some indications of possible ones and I will try to make it happen
17:59:12 <lucasagomes> if not we just gather together at the summit and go to a pub or something :-)
17:59:20 <dtantsur> ++
17:59:20 <jroll> ++
17:59:57 <jroll> alright, that's time, thanks everyone
17:59:59 <jroll> #endmeeting