20:01:33 <ttx> #startmeeting tc
20:01:33 * flaper87 is sure fungi didn't notice
20:01:34 <openstack> Meeting started Tue Jul 11 20:01:33 2017 UTC and is due to finish in 60 minutes.  The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:01:35 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
20:01:37 <openstack> The meeting name has been set to 'tc'
20:01:39 <ttx> Hi everyone!
20:01:42 <flaper87> yooooooooooo
20:01:49 <smcginnis> Howdy
20:01:55 <EmilienM> hello!
20:01:57 * rockyg grabs a box pof popcoren for the show
20:02:14 <ttx> Our agenda for today is at:
20:02:17 <ttx> #link https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee
20:02:20 <ttx> Remember that you can all use #info #idea and #link to help build a more readable summary
20:02:34 <ttx> Two topics to discuss today
20:02:36 <ttx> #topic Final selection of Queens goals
20:02:45 <ttx> Let's start wit hthe less open-ended one
20:02:51 <ttx> we have a proposed set of goals:
20:02:55 <ttx> * Split Tempest plugins (already approved)
20:02:59 <ttx> * Policy and docs in code: https://review.openstack.org/469954
20:03:04 <ttx> In the ML thread champions volunteered to drive those:
20:03:09 <ttx> - Chandan Kumar for split-tempest-plugins
20:03:13 <ttx> - lbragstad for policy-in-code
20:03:23 <ttx> Are there objections to those being our two Queens goals ?
20:03:42 * mordred welcomes our new Queens goals overlords
20:03:47 <ttx> If not I'll approve 469954 and the two follow-ups, then abandon the others until they can be re-proposed
20:03:51 <dims> no objections
20:03:53 <ttx> I'll submit a follow-up change to add the champion names to those, if approved
20:04:02 <smcginnis> They seem enough to try to tackle.
20:04:14 * flaper87 gives our new Queens goals overlords a standing ovation
20:04:21 <ttx> I was wondering if mtreinish or sdague would shadow Chandan and make sure he is successful
20:04:28 <EmilienM> no objection
20:04:34 <dhellmann> yes, I was just thinking along the same lines, ttx
20:04:58 <ttx> Since those are our first champions, I don't want to drop him in the dark. That said very happy to see new leaders emerging
20:05:07 <dhellmann> ++
20:05:15 <dhellmann> let's help them actually emerge :-)
20:05:24 <cdent> I’m fine with the goals. I stil think that tempest’s asserition that some projects are more equal than others is a problem, especially in the context of the goal, but I don’t want to stop progress because of that
20:05:28 * dims visualizes a butterfly
20:05:45 <dhellmann> cdent ++
20:05:48 * edleafe wanders in late
20:05:53 * mordred is happy to provide grumpy old man services to Chandan or lbragstad as needed
20:06:17 <lbragstad> mordred: :)
20:06:18 <ttx> alright then. I'll clean everything up tomorrow
20:06:25 <cdent> hurrah
20:06:31 <ttx> thingee: if you have comments/objections please let me know
20:06:33 <dims> lbragstad, Chandan, i can make some time to help if you need
20:06:52 <fungi> i still think that tempest is getting unfairly painted as giving preference to projects which just happen to still have their tests in the tempest tree (interop/refstack tests aside)
20:06:54 <lbragstad> dims: thank you
20:07:01 <ttx> Chandan is chandankumar on irc
20:07:10 <ttx> (present in #openstack-qa at least)
20:07:37 <EmilienM> (note he's in India and offline at this time)
20:07:45 * flaper87 knows chandankumar, he's an awesome human
20:07:52 <flaper87> also, what EmilienM said
20:07:54 <mtreinish> ttx: I can probably shadow him if needed
20:08:00 <flaper87> mtreinish: ++
20:08:04 <mtreinish> I was gonna watch that goal pretty closely anyway
20:08:08 <ttx> #info ttx to approve policy-in-code early tomorrow and clean up other proposals, and approve subsequent typo patches, and propose a change adding champion names
20:08:15 <EmilienM> mtreinish: thanks
20:08:16 * mtreinish might have a vested interest in it
20:08:27 <flaper87> mtreinish: u think? :P
20:08:33 <ttx> #topic TC 2019 vision next step
20:08:41 <fungi> also, the in-tree tempest tests are orthogonal to the plugins issue. most of the tempest plugins are coming from repos with stable branches, which is the crux of the problem. tempest is branchless and so therefore are its in-tree tests
20:08:42 <ttx> OK, now the big piece
20:08:52 <ttx> fungi: ++
20:09:05 <ttx> We have three subsequent stages of drafting up for review:
20:09:09 <ttx> #link https://review.openstack.org/453262
20:09:14 <ttx> #link https://review.openstack.org/#/c/473620/
20:09:17 <ttx> #link https://review.openstack.org/#/c/482152/
20:09:27 <ttx> The question is, what's the next step ?
20:09:37 <ttx> With a short horizon for this vision (2019) I feel like we should approve it soon rather than wait for end of year and be too close
20:09:52 <cdent> Is johnthetubaguy here?
20:10:01 <ttx> Unfortunately has a concert tonight
20:10:02 <flaper87> can we just merge #1 in #2 and just merge things?
20:10:20 <dhellmann> I think we should land all of these patches and continue to refine the language if we need to. eglynn's comments about applicability aside, this is what we wrote at a point in time, and we expect to write new visions as we move ahead.
20:10:22 <flaper87> I'm happy with the current version, fwiw
20:10:23 <cdent> Okay, in that case I can explain why things are structured like that, and also related some john’s concerns.
20:10:28 <ttx> flaper87: I think they are subsequent patches, so we can approve them all
20:10:30 <EmilienM> I think we've done enough iterations to have a first text in place, and we can keep iterating on it later
20:10:40 <EmilienM> dhellmann: yes
20:10:52 <flaper87> ttx: yeah, I was more worried about the -1 in the first patch but what dhellmann said makes sense to me
20:10:59 <cdent> dtroyer, johnthetubaguy and I decided we wanted to present a second draft as a second change to keep it clear from the initial feedback, and have an evident diff
20:11:03 <dhellmann> I also don't think we want to refine this to perfection. It's meant to be a guide.
20:11:08 <cdent> the third change is just typos on the second
20:11:12 <flaper87> cdent: ++
20:11:13 <ttx> Yeah, it's not policy
20:11:16 <smcginnis> I think it would be good to merge them, then iterate as we go as needed.
20:11:17 <ttx> notr governance
20:11:28 <ttx> it's vision. It doesn't have to be perfect
20:11:29 <cdent> I agree that merging would be good/nice
20:11:32 <ttx> It's not a contract
20:11:49 <smcginnis> If we had perfect vision, we wouldn't all need to be here.
20:11:50 <flaper87> I guess we should all also vote on the first draft and let the merging countdown begin
20:12:04 <cdent> John raised some concerns that since the version was written before the latest reality check that it may be perceived as even more out of line with reality than it was at first (by some respondents)
20:12:19 <ttx> If you agree to merge them, please pile up +1 votes on the review. I can announce a "will merge in 3 days" on the next TC status update
20:12:20 <dhellmann> cdent : there's a date on it, right?
20:12:27 <persia> If the consensus is that vision is not to be perfect, it is worth asserting that the TC will review the vision every cycle, or some similar language as a resolution.
20:12:30 <ttx> on the reviews* I mean
20:12:35 <cdent> but said “doing the process again woul be wasteful” (paraphrase)
20:12:37 <dhellmann> cdent : maybe we can add a date in the title to address that
20:12:43 <rockyg> persia, ++
20:12:51 <cdent> dhellmann: that feels like a good idea
20:13:12 <ttx> Yes, it should have a date anyway
20:13:15 <dims> ++ cdent dhellmann
20:13:21 <ttx> And be filed as a resolution
20:13:22 <dhellmann> it has a "for" date but not an "on" date
20:13:22 <cdent> i like the idea of promising a regular review
20:13:23 <flaper87> done
20:13:37 <flaper87> cdent: johnthetubaguy thanks for working on this
20:13:47 <ttx> I would say revision after each election
20:13:47 <dhellmann> I'll submit that one
20:14:06 <ttx> dhellmann: I think it should be a resolution rather than a reference doc
20:14:14 <ttx> that way it's dated
20:14:20 <ttx> and not a living document
20:14:21 <fungi> well, the normal expectation for this vision pattern seems to be that it's reviewed at its expiration date and superseded by a new vision, but i guess there's nothing stopping us doing an intermediate vision course correction if we have the time
20:14:26 <ttx> and we can post new versions of it
20:14:36 <dhellmann> ttx: good point. Shall I move the file around?
20:14:46 <ttx> dhellmann: I think that would be a good idea
20:14:46 <persia> fungi: And nothing wrong with a future TC rubber-stamping it if they wish to do so.
20:14:51 <dhellmann> ok, let me do that quickly
20:14:54 <ttx> (and add date to it)
20:14:55 <fungi> i agree that as it's a vision of the people who were in office at the time it was drafted, revising after each election makes some sense too
20:15:03 <dims> yep fungi
20:15:12 <ttx> If a future TC feels like teh old vision is off, they can produce a new one
20:15:14 <rockyg> ++ fungi
20:15:20 <flaper87> ttx: indeed
20:15:33 <ttx> I don't think they HAVE TO revise it, but they sure should have that option
20:15:39 <ttx> by publishign a new resolution
20:15:45 <flaper87> I like the idea of revisiting the vision on every election. It should help bringing new TC members up-to-speed too
20:15:49 <fungi> as with any of our governance documents really
20:15:52 <cdent> flaper87++
20:16:03 <smcginnis> flaper87: Good way to make sure the vision is kept in sight.
20:16:11 <flaper87> smcginnis: right
20:16:43 <ttx> OK, so if I summarize...
20:16:45 <fungi> i do wonder whether spending a couple days together every cycle dedicated solely to vision drafting is the best use of the tc's time, but am open to convincing
20:17:21 <fungi> a more lightweight process might be called for
20:17:29 <ttx> - Add the date, and a blurb about next TC having the option to revisit it and publish their own corrected vision if needed
20:17:32 <smcginnis> fungi: That may be overkill, but spending at least a meeting time block to check in would be good.
20:17:33 <persia> fungi: I would argue that it is usually *not* a good use of time, excepting when it is. and that a future TC is better placed to decide when that time happens.
20:17:43 <ttx> - Make it a resolution rather than a reference doc
20:18:06 <ttx> (that way it will have a date since all resolutions are dated
20:18:08 <ttx> )
20:18:10 <mordred> yah- I think the future tc doing a rubber stamp - unless the new tc is now strongly opposed at which point time needs to be spent, yeah?
20:18:17 <dhellmann> ttx: https://review.openstack.org/482686 moves the vision doc
20:18:17 <ttx> did I miss something
20:18:21 <mordred> ttx: ++
20:18:36 <fungi> persia: fair point. no decision we make about this is binding for a future tc makeup (or even on the current tc roster)
20:18:38 <ttx> mordred: they should definitely look at the last vision and check if they feel it's completely off
20:18:38 <smcginnis> ttx: That looks like a good summary from my understanding.
20:18:44 <cdent> (thanks dhellmann )
20:18:44 <mordred> ttx: ++
20:18:52 <ttx> off enough to jutify them going through the process of replacing it
20:18:52 <dhellmann> oh, I missed the bit about adding a blurb
20:19:02 <mordred> ttx: but they shodl not feel compelled to change it just because
20:19:05 <dhellmann> do we really need to say that? isn't everything up for revision by future tcs?
20:19:06 <ttx> right
20:19:21 <ttx> dhellmann: fair point
20:19:23 <mordred> saying "yes, this is still the vision" is a fine thing - and actually the preferred thing unless there is a problem or clarification, yeah?
20:19:27 <persia> dhellmann: They may not remember that.  Precedence is a powerful thing.
20:19:29 <fungi> dhellmann: which was why i said "as with any of our governance documents really"
20:19:30 <cdent> dhellmann: I think the reason for doing it, if we want to, is to make it clear that we _encourage_ revision
20:19:41 <cdent> and at least some of the people here this evening (me, flaper87 , ?) want that
20:19:46 <dhellmann> I think you're all over thinking this, but ok.
20:19:48 <flaper87> cdent: ++
20:19:59 <ttx> dhellmann: maybe just say "this is the vision of the TC from that date. Future TCs may publish new visions for the future.
20:19:59 <dhellmann> give me words to use
20:20:00 <rockyg> cdent, ==
20:20:04 <dhellmann> ok
20:20:15 <fungi> i'm cool with that phrasing
20:20:18 <mordred> cdent: I want to encourage revisoin, as long as we dont' encourage revision so stronlgy that a future tc feels compelled to revise to show they're doing their job
20:20:28 <mordred> ttx: ++
20:20:29 <persia> mordred: ++
20:20:32 <dims> agree mordred
20:20:34 <rockyg> mordred, ++
20:20:49 <ttx> I don't thgink we should change the vision every 6 months. Every two years we definitely should. Truth probably lies between those two extremes
20:20:55 <mordred> ttx: ++
20:21:02 <flaper87> ttx: yup
20:21:07 <cdent> mordred: fair point, but I’d argue that if we’re not changing the vision at least once a year, we’re not paying enough attention to the world around us
20:21:08 <cdent> it’s not a mission
20:21:09 <edleafe> if in the future something about it doesn't seem to fit, someone should bring that up and propose an update
20:21:10 <mordred> minor course corrections as we learn thing are likely quite frequent
20:21:10 <dhellmann> updated
20:21:14 <cdent> it’s a vision
20:21:27 <mordred> cdent: well - we may be quibbling over the word "change" :)
20:21:29 <cdent> but yeah, it doesn’t need to be heavyweight, or required, or onerous, or ….
20:21:31 <ttx> cdent: from the zingerman's experience you have to keep the vision for "some time"
20:21:45 <ttx> enough to be able to judge your ability to reach those goals
20:22:10 <ttx> basically We shouldn't replace it because it starts to be late or anything. We should replace it when it's no longer a good vision
20:22:23 <smcginnis> ++
20:22:26 <ttx> It's ok to get to 2019 and fail to deliver on some
20:22:28 <flaper87> ok, let's move this forward and also learn from the experience. I think it's fair to encourage revision but not force it to avoid (as mordred pointed out) encouraging the wrong behavior
20:22:37 <flaper87> we're not going to get everything right this time around
20:22:41 <mordred> golly no
20:22:42 <ttx> It's not OK to get to 2019 and chase goals that we no longer think are good
20:22:51 <mordred> flaper87: :)
20:22:51 <dims> this is the first, so let's roll the dice :)
20:22:56 <cdent> having not had the pleasure of doing the zingerman thing, my attitude towards this thing is mostly as reader from the communit, so take my input with that grain of salt
20:23:15 <flaper87> mordred: :D
20:23:19 <mordred> cdent: I think that input is hugely helpful, since _most_ of the folks in openstack haven't been to ann arbor
20:23:30 <EmilienM> flaper87: well, said, let's move forward now
20:23:37 <cdent> sideways!
20:23:43 <edleafe> flaper87: even if you get everything right today, it may not be right next year
20:23:44 * mordred starts floating
20:24:02 <ttx> anyway, I think creating a new vision is enough work to deter casual changing of the vision
20:24:09 <flaper87> edleafe: prolly right but that's too deep of a thought for my poor, night mode, brain :P
20:24:25 <dhellmann> I have to drop off -- if you need more changes to that patch, someone please take it over
20:24:26 <edleafe> flaper87: :)
20:24:38 <fungi> thanks dhellmann!
20:24:40 <ttx> OK, I think we have a way forward. Merge early, stamp it as our vision at this date.
20:24:42 <flaper87> dhellmann: ++
20:24:49 <flaper87> ttx: deal
20:24:51 <ttx> Because frankly, we are already using that vision in our decisions right now
20:24:57 <EmilienM> dhellmann: thanks
20:25:09 <ttx> so there is little point in holding it up
20:25:31 <ttx> Anything else on that topic before we move on to open discussion ?
20:25:55 <dims> So do the minimal changes needed and get it out right? who has the pen ttx ?
20:25:56 <cdent> ttx: yes, as far as I can tell there has been a very positive reaction to the constellation idea, and lots of positivity around other stuff too
20:26:24 <ttx> dims: a patch from dhellmann, votes from TC members on all the patch chain
20:26:36 <ttx> then I'll announce the 3-day merge warning
20:26:40 <dims> sounds like a plan
20:27:04 <ttx> #topic Open discussion
20:27:11 <ttx> We've been looking for election officials to handle the PTL election
20:27:19 <ttx> We have two volunteers currently: persia and diablo_rojo
20:27:21 <ttx> yay!
20:27:25 <ttx> any objection to them running the show ?
20:27:34 <dims> thanks persia and diablo_rojo
20:27:36 <cdent> +many
20:27:41 <ttx> (Most of the other usual suspects may run for PTLship)
20:27:42 <smcginnis> None from me.
20:28:19 <ttx> How do you think our effort to rely less on meetings is going ?
20:28:28 <ttx> Feels like office hours have low attendance, but I feel like having designated time slots still helps
20:28:38 <cdent> I think it is good for us, but potentially bad for the community
20:28:46 <ttx> cdent: expand?
20:28:57 <smcginnis> Office hours seem to be picking up activity lately.
20:28:59 <EmilienM> no objection at all, thanks a lot
20:29:06 <cdent> I think we are disconnected somehow, I’m not sure I can put my finger on it.
20:29:20 <cdent> But if office hours pick up, it might be okay.
20:29:31 <ttx> Less outside participation on office hours than on meetings ?
20:29:33 <cdent> I certainly like the fact that for me, at least, having office hours means I’m talking to the rest of you more often
20:29:36 <EmilienM> ttx: low attendance on which side? I saw most of TC members around (if not all)
20:29:43 <fungi> persia: diablo_rojo: thanks for volunteering. get up with me if you have any questions on the tooling (and i'll be syncing up with you anyway on some development work i've got underway for it this week)
20:29:43 <ttx> Like rockyg attendingt every meeting but none of the office hours ?
20:29:48 <dims> EmilienM : i have been lax attending them so far
20:29:51 <rockyg> Hey!!!!
20:29:54 <cdent> but I’m not feeling observed/participated by/with the community.
20:29:58 <cmurphy> fwiw i usually follow the office hours discussions
20:30:05 * edleafe is in the same boat as rockyg
20:30:26 <ttx> edleafe, rockyg: do you know when the office hours are ?
20:30:28 <fungi> *reminder* office hours at 01:00 (in roughly 3.5 hours)
20:30:30 <ttx> (this is a test)
20:30:40 <fungi> heh, oops
20:30:47 <fungi> i blew that test, sorry :/
20:30:48 <persia> cdent: Some of us are observing you.
20:30:48 <rockyg> I don't have great IRC access, so meetings are currently on my calendar, but I haven't gotten the office hours on yet.
20:30:53 <persia> fungi: Thanks.
20:31:06 <cdent> persia: creepy?
20:31:07 <cdent> :)
20:31:11 <edleafe> ttx: nope. I should have written them down somewhere. :(
20:31:23 <ttx> I wish office hours were more "ask TC anything" rather than cocktail hour between us, but that may happen in the future :)
20:31:32 <jeblair> ttx: i've been regularly missing office hours because i'm bad at whatever skills are required in showing up at them
20:31:34 <dims> ttx : could do we do a ping list for the office hours too?
20:31:36 <EmilienM> #link https://governance.openstack.org/tc/#office-hours
20:31:42 <edleafe> and 0100 UTC is not when I'm near a compute
20:31:42 <fungi> i will say, our 01:00-02:00 wednesday office hour seems to be very under-attended even by the tc. most weeks it's been just me, though dtroyer showed up for one
20:31:47 <rockyg> I am reading the summaries from both cdent and ttx, and following the office hours announcements, but....
20:31:49 <ttx> dims: we could
20:31:49 <jeblair> ttx: but i found the retroactive "we had an important discussion" announcement you did *very* useful
20:31:53 <edleafe> or even a computer
20:32:15 <jeblair> ttx: so please keep doing those as warranted :)
20:32:32 <diablo_rojo> fungi, sounds good :)
20:32:37 <ttx> I feel like between teh new status tracking system / weekly updates + office hours + occasional meetings we make good progress
20:32:39 <rockyg> jeblair, ++
20:33:07 <dims> cdent : ttx : many many thanks for the summaries
20:33:13 <ttx> I won't hide that it's a bit of work for the chair to keep track of everything (rather than rely on that hour in the week to sync) but worth it imho
20:33:17 <fungi> the 09:00-10:00 tuesday office hour seems to be the most active, which i'm usually asleep during
20:33:39 <ttx> It's becauise europeans are so cool
20:33:54 <dims> :)
20:34:01 <jeblair> (the especially weird thing about that was that i was sitting at a terminal eating a bowl of cereal during that discussion unaware it was going on, then saw the email, then read the backscroll, then everyone was gone and i had an empty bowl)
20:34:05 <rockyg> I think that's a good sign of the other parts of the world wanting to participate
20:34:05 <ttx> Once flaper87 gets rid of the meeting channel check in irc-meetings, we'll be able to add them to the meeting calendar
20:34:19 <dims> jeblair : LOL
20:34:20 <ttx> might increase visibility
20:34:20 <fungi> if i didn't feel compelled to be up for the 01:00 the "same day" (from my waking hours perspective) i might try to be awake for the 09:00 one
20:34:34 <ttx> I also dropped pointers to the office hours all over the Project Team Guide
20:34:42 <ttx> which I hope everyone reads of course
20:34:44 <EmilienM> jeblair: valid use case :)
20:34:53 <fungi> but staking awake from 09:00-02:00 utc for me once a week is a little bit of a stretch
20:34:59 <fungi> s/staking/staying/
20:35:15 <ttx> The wednesday morning one is for reaching out to China/APAC
20:35:53 <ttx> It's under-attended because we have a lot of work to do on raising leaders there
20:35:54 <dims> so, have folks read the diversity report yet? http://superuser.openstack.org/wp-content/uploads/2017/07/OpenStack-Gender-Diversity-Report_Apr2017.pdf
20:36:03 <dims> http://superuser.openstack.org/articles/bitergia-intel-report/
20:36:13 <ttx> A lot of us did, yes
20:36:26 <ttx> Some good insights, although the contributor data looks questionable
20:36:29 <persia> Would it be worth moving one of the 9:00 and 2:00 so that they are not on adjacent days?
20:36:31 <fungi> dims: yeah, it got discussed at length in #openstack-tc
20:36:42 <ttx> persia: that was suggested yes
20:36:55 <ttx> We accept patches
20:36:56 <fungi> persia: yeah, that was an early suggestion of mine when going over the final scheduling
20:37:21 <fungi> but the counterargument was that i should just avoid trying to be awake for all three office hours and not worry about it
20:37:27 <ttx> http://git.openstack.org/cgit/openstack/governance/tree/doc/source/index.rst#n38
20:38:02 <smcginnis> Yeah, for me spreading out days won't make a difference for a bad time slot.
20:38:04 <ttx> I'm fine with considering another layout, as long as we manage to reach the three TZs sometimes in the week where people in those TZ actually work :)
20:38:26 <ttx> which generally means targeting Tue-Thu
20:39:08 <fungi> there are at least several of us on the current tc lineup for whom the tuesday slot is 5am local and the wednesday slot is 9pm tuesday local
20:39:31 <fungi> so attending both isn't a stretch if they didn't fall within 24 hours of one another
20:39:34 <rockyg> So, you could move the APAC to a Monday timeslot.
20:40:02 <rockyg> Their Tuesday is mostly the NA Monday
20:40:39 <ttx> Like Tuesday 1:00utc ?
20:40:47 <EmilienM> fungi: 2am local now :D
20:40:56 <rockyg> Yup.
20:41:19 <ttx> we can certainly move the 9am one to Wednesday to spread them out
20:41:33 <rockyg> Well, maybe I take that back.  China is +8
20:41:42 <fungi> EmilienM: yeah, but dims, dhellmann and i are still in the eastern usa timezone
20:42:14 <cdent> maybe we should instead of trying to find a good slot for westeners to be in chinese time we should try harder to create some chinese tc members?
20:42:25 <fungi> not that i'm going to pretend to know whether those timeslots are as convenient for them as they are for me
20:42:34 <EmilienM> fungi: right, I think 5am is also too early
20:42:35 <ttx> Another topic to quickly discuss: we have two items on our "Top 5 help wanted list". While I think we don't have to allocate every single spot, there is room for a few more if you see critical areas
20:43:03 <ttx> This list will be heavily promoted in an effort to get contributors
20:43:20 <mordred> cdent: I think the worry is that if we can't start by finding a good slot shared by western and chinese fokls it will be hard to develop chinese tc members
20:43:25 <ttx> (in OpenSTack Days, in Summit keynotes, at Board meetings...)
20:43:27 <EmilienM> ttx: I thought about it and I think we already have 2 things, it would be great to not having too much areas and focus on the 2 for now
20:43:29 <cdent> (another potential topic before we close, if want to go there, is glance/glare)
20:43:43 <cdent> mordred: yes, I know, I was sort of being snarky.
20:43:44 <fungi> cdent: the goal of having a known time for north americans to be around while apac tz'ers are awake is to encourage adding more apac tc members
20:43:46 <mordred> cdent: :)
20:43:47 <ttx> EmilienM: sure, I just don't want people to hold up or anything
20:44:00 <ttx> If you see an area needs urgent help, don't be shy
20:44:01 <mordred> cdent: glance/glare I think has some specific technical questions pending before it's actionable
20:44:10 <mordred> cdent: there is an assertion that needs to be backed up
20:44:30 <cdent> that’s just it, I think we’ve kicked that assertion dead
20:44:31 <mordred> which I _think_ is going to spur an excellent discussion
20:44:33 <fungi> (well, and south american community members too; we don't have any of them on the tc either)
20:44:33 <EmilienM> ttx: we could have proposals at any time I guess, but maybe we could keep the list of active areas "short"
20:44:33 <dims> mordred : REST API equivalence and data migration path?
20:44:34 <ttx> glance/glare... probably the topic for our next "occasional" meeting
20:44:52 <mordred> dims: api equiv - I think he claims data migration path is the easy part
20:44:58 <mordred> and I _think_ it should be easy to do api equiv
20:44:59 <ttx> At this stage the thread is still evolving and useful to raise the right issues
20:45:02 <mordred> from looking at it
20:45:10 <ttx> so not stuck yet
20:45:15 <mordred> it just needs to be declared important and then done
20:45:18 <smcginnis> mordred: But not there yet, from my understanding.
20:45:21 <ttx> I think the critical point is the defcore stuff
20:45:22 <mordred> smcginnis: yah
20:45:37 <dims> ttx : are we saying that they can be under governance only IF they are able to do this?
20:45:53 <ttx> We don't have a mechanism to replace something that is in the defcore set
20:45:58 <mordred> I'm not - but it would certainly help if we had one and only one project that did images
20:46:02 <ttx> that goes beyond us
20:46:08 <mordred> since we're having issues staffing the one we have
20:46:08 <fungi> i would suggest that we shouldn't be encouraging competing implementations of something we also insist needs to be cross-deployment interoperable
20:46:17 <ttx> dims: doing what ?
20:46:27 <mordred> this is one of the few cases where I think the internet snark about our interests being too diverse would actually be appropriate
20:46:32 <rockyg> So, defcore would still require glance.  Even if glare were there.
20:46:55 <dims> ttx : seems like the thread is saying, we can only add glare to governance only if you can be a replacement for glance
20:46:56 <rockyg> Doesn't mean it has to be used much, but it has to be accessible to use.
20:47:14 <mordred> rockyg: I think if we wanted to propose replacing glance with glare we'd need to propose a plan to defcore and get everyone's buy in
20:47:19 <ttx> rockyg: If you still need to run Glance, the pain/value proposition of adding Glare is limited
20:47:25 <smcginnis> rockyg: I think that glare can be dropped in instead of glance and things still work.
20:47:41 <cdent> smcginnis: s/can/could with some changes/
20:47:48 <mordred> I think IF glare can be dropped in in place of glance and things still work, THEN we can propose replacing one with the other
20:47:50 <edleafe> smcginnis: if that's truly the case, then fine. But I really doubt they are there
20:47:56 <mordred> if they cannot, that is a dead discussion
20:48:03 <edleafe> mordred: jinxish
20:48:06 <mordred> edleafe: :)
20:48:12 <smcginnis> cdent: Preface my statement with "It would need to be able to be dropped in..." :)
20:48:15 <rockyg> Yeah.  Until the glance *code* is removed from being required, it would still need to be there.  Board decision that is suggested to them by defcore
20:48:15 <mordred> BUT - there is no reason that technical state cannot be achieved
20:48:23 <smcginnis> I agree, they are not there yet.
20:48:25 <fungi> if glare were decided on as a successor to glance and we just called it glance, then there's no defcore/interop issue
20:48:26 <rockyg> There is a deprecation process in defcore, though.
20:48:40 <mordred> rockyg: I agree - and I think that's a thing we can work through with the board/defcore
20:48:48 <ttx> mordred: so are you saying that perfect compat is necessary for us to add them ? Or that if they say they will pursue it we should add them ?
20:48:50 <mordred> but I tihnk we're not even to the point of starting that conversation yet
20:48:55 <mordred> because right now it's unpossible
20:49:03 <rockyg> And, remember, defcore is like 2+ releases behind where you guys are.
20:49:13 <mordred> ttx: I'm saying that THEY are claiming that they provide the value of being able to replace glance
20:49:27 <mordred> I'm currently judging them on that standard since it's part of what they claim
20:49:31 <EmilienM> mordred: I liked your request of the devstack job to test glare instead of glance and see results
20:49:38 <mordred> I do not think it needs to be true for them to be added to governance
20:49:41 <mordred> but right now it's a claim
20:49:44 <mordred> so I'm judging it
20:49:45 <fungi> merge a change to glance that replaces the underlying implementation with one ported from glare (or turns it into a thin api shim) and defcore requirements are basically sidestepped. but i agree it's still tricky and assumes basically completely identical behavior
20:49:54 <EmilienM> mordred: the answer would be binary and would confirm if yes/no we can replace it
20:50:09 <mordred> EmilienM: exactly - or we, can confirm whether it's worth people's time to start the discussion with the board
20:50:13 <mtreinish> mordred: fwiw, I'm like 100% sure it will fail tempest horribly if you point it at glare and say its glance
20:50:23 <mordred> right now it's not worth bothering the defcore wg
20:50:30 <jeblair> fungi: it would not be the first time we have done that
20:50:30 <dims> mtreinish : hey
20:50:31 <mordred> mtreinish: I know for a fact that it will
20:50:34 <mordred> I went and read the aPI
20:50:37 <mordred> it is not compatible
20:50:44 <fungi> jeblair: this is true!
20:50:45 <mordred> but - an endpoint could be added that would be
20:51:06 <ttx> So it's a complex discussion, and I fear it will take time for us to reach a decision here. One issue is that Glare would like to join the PTG, which is in theory limited to official projects. Would you be happy with giving them a free ride ? If only because we could have those discussions at the PTG
20:51:10 <mordred> and I'm happy to work with them on defining what that would need to look like
20:51:25 <fungi> anyway, i still believe that if they declare replacing glance as a primary goal of their project and we don't have a solid plan for making that possible then we shouldn't be making glare an official project team
20:51:43 <ttx> free ride = give Glare a PTG room
20:51:46 <mordred> ttx: I would be happy to discuss this topic with them at the PTG - and it would not bother me to have them there this time
20:51:49 <dims> ttx : i am ok to give them the free ride to encourage them
20:51:50 <smcginnis> :q
20:51:51 <rockyg> mordred is right.  Not worth bothering InteropWG (defcorc) right now.  That's at least a year, more likely 1.5 or 2 out.
20:51:52 <smcginnis> Oops
20:52:05 <dims> smcginnis : trying to get out of here :)
20:52:10 <edleafe> rockyg: mordred: +1
20:52:10 <smcginnis> dims: Hehe
20:52:11 <fungi> well, glare said they'd love to "present a session at the ptg" which suggests to me that they're still pretty disconnected from the community at large and don't have any context as to what the ptg is
20:52:19 <ttx> FWIW We've been doing that at the past PTG as well -- give rooms to projects in the pipeline
20:52:33 <ttx> (if we still have space)
20:52:38 <dims> ++
20:52:40 <mordred> yah- so let's get them there so they can see it - set up a time for them to talk witha few of us that have some good context on what this might look like
20:52:49 <smcginnis> I think it would be worth having some time for them to pitch their case there.
20:52:55 <ttx> ok
20:52:59 <mordred> and we should be ableto come out of that with a "this is a possible plan" or "wow, this is never going to work"
20:53:05 <smcginnis> Or maybe even a glance time slot if they're willing to work it out.
20:53:06 <ttx> diablo_rojo: ^ we should add Glare if we can
20:53:15 <ttx> (to the PTG)
20:53:28 <dims> mordred : i'd like to know how many folks are actively engaged in Glare and want to pursue this goal as well
20:53:28 <rockyg> I think InteropWG will also be there.
20:53:28 <fungi> it sounds more like they want to present a proposal at the forum
20:53:32 <mordred> dims: I agree
20:53:33 <smcginnis> Based on the current Glance participation, I don't see them filling a room for a few days straight.
20:53:35 <ttx> rockyg: yes!
20:53:45 <mordred> dims: I think there are many things I'd like to dig in to on that topic :)
20:53:58 <dims> for sure :)
20:54:00 <ttx> smcginnis: but since they are on the top 5 list, millions of devs will join!
20:54:03 <rockyg> ttx, gotta double check, but I'll specifically ask tomorrow.
20:54:22 <smcginnis> ttx: I admire you're optimism. Or snark. Either one. :)
20:54:31 <ttx> 11pm optisnark
20:54:35 <smcginnis> *yoru
20:54:37 <diablo_rojo> ttx, there is space since Karbor dropped out.
20:54:46 <ttx> diablo_rojo: deal.
20:54:55 <diablo_rojo> I will update the spreadsheet :)
20:54:57 <dims> diablo_rojo : ++
20:55:08 <ttx> Also some teams may drop if they realize they won't have enough people joining
20:55:41 <ttx> OK, anything else, anyone ?
20:56:10 <dims> that's it. good night ttx
20:56:22 <fungi> thanks for a productive meeting, ttx!
20:56:34 <ttx> Thanks everyone for attending
20:56:38 <smcginnis> Thanks
20:56:41 <ttx> #endmeeting