21:02:01 <ttx> #startmeeting project
21:02:01 <openstack> Meeting started Tue Oct 28 21:02:01 2014 UTC and is due to finish in 60 minutes.  The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:02:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:02:04 <openstack> The meeting name has been set to 'project'
21:02:04 <dhellmann> marun: if it was only 2 projects, it wouldn't have made the list anyway
21:02:10 <ttx> Our agenda for today:
21:02:15 <ttx> #link http://wiki.openstack.org/Meetings/ProjectMeeting
21:02:25 <ttx> #topic Final Design Summit scheduling
21:02:28 <russellb> dhellmann: likely, yes ... though those 2 projects usually have packed agendas and a hard time getting together
21:02:33 <ttx> We seem to have everything on the agenda at this point, except...
21:02:36 <russellb> so would have been worth considering ...
21:02:40 * ttx refreshes
21:02:44 <dhellmann> russellb: we should call it "all projects" instead of "cross projects"
21:02:49 <SergeyLukjanov> ttx, I've pushed sahara sessions
21:02:50 <russellb> dhellmann: heh
21:03:02 <ttx> except TripleO
21:03:05 <russellb> dhellmann: cross project, with preference for all project
21:03:12 <russellb> doesn't have a good ring to it
21:03:19 * russellb gets out of ttx's meeting ...
21:03:39 <ttx> Note that I changed a few titles so that it's clearer which project each session belongs to
21:03:52 <ttx> since they won't appear in cheerful colors in the mobile app
21:04:00 <dhellmann> ttx: I have to drop off, but I'll look at the logs tonight in case there's something you need from me
21:04:10 <ttx> dhellmann: sure!
21:04:11 <jeblair> ttx: thank you for that
21:04:17 <ttx> One session is unassigned in QA
21:04:21 <SergeyLukjanov> ttx, yeah, it's a good idea, last time there were some very unclear titles (one of them were sahara session :) )
21:04:30 <ttx> we just published the tentative cross-project workshops
21:04:55 <jeblair> ttx: mtreinish and i have a plan for that
21:04:59 * asalkeld goes to look at cross projects
21:05:00 <ttx> mtreinish: do you plan to make use of that unassigned session ?
21:05:03 <ttx> jeblair: oh
21:05:05 <david-lyle> o/
21:05:10 <notmyname> ttx: thanks for fixing the titles
21:05:23 <jeblair> ttx: iiuc, we have put a joint infra/qa topic on formalizing gating strategies in there
21:05:26 <ttx> notmyname: feel free to refix them if they look bad
21:05:30 <russellb> cross project stuff: http://kilodesignsummit.sched.org/overview/type/cross-project+workshops#.VFAFFXVGjUa
21:05:36 <notmyname> ttx: looks fine from my perspective
21:05:54 <eglynn> hmmm, the cross-project session on notifications made the cut
21:05:56 <eglynn> http://kilodesignsummit.sched.org/event/0b6e7e23da3d4fdcdc5cec0777b92c6a#.VFAE-opziVo
21:05:57 <jeblair> ttx: it was contingent on functional testing being approved as a cross project (which it was), so i believe it's spoken for and mtreinish should be able to push the updated schedule
21:06:01 <ttx> everyone: note that if you want to edit scheduled sessions in my crappy UI, you'll have to unschedule slots before you can edit them
21:06:06 <ttx> then schedule them again and push to sched
21:06:18 <russellb> eglynn: and still need a session lead for that btw
21:06:28 <mtreinish> ttx: yeah I'm going to update the schedule nowish
21:06:28 * eglynn wonders if we need the ceilometer-specific session on notifications-as-a-contract also
21:06:40 <russellb> eglynn: would you be interested in leading it?
21:06:40 <eglynn> russellb: jd__ is leading the ceilo session
21:06:51 <russellb> ah OK
21:06:55 <thingee> ttx: I have verified firefox works for me, but not the latest chrome. re: my email earlier about pushing the sched
21:06:58 <asalkeld> russellb, who leads these?
21:07:06 <russellb> asalkeld: still trying to figure that out
21:07:12 <ttx> We need leads for all the cross-project sessions, to make sure they are going somewhere
21:07:13 <russellb> most sessions didn't have obvious proposers listed
21:07:19 <asalkeld> I fleshed 2 out
21:07:20 <russellb> asalkeld: i have you down for the upgrades one, as i think you proposed it?
21:07:25 <ttx> russellb: etherpad link for that ?
21:07:27 <asalkeld> ha and upgrades
21:07:31 <russellb> https://etherpad.openstack.org/p/kilo-crossproject-summit-topics
21:07:32 <eglynn> russellb: I'll chat with jd__ about it tmrw, we could decide to "repurpose" the ceilo session
21:07:33 <asalkeld> but big topics
21:07:47 <asalkeld> happy for others to get involved
21:07:50 <russellb> asalkeld: i know dansmith is willing to help on the upgrades one, to give info on what nova has been doing, and some related things coming
21:07:58 <asalkeld> great
21:08:29 <ttx> So, please check the proposed schedule for conflicts, maybe we can move things around to facilitate your life
21:08:31 <asalkeld> it's a shame beekhof is not coming to summit
21:08:41 <russellb> asalkeld: indeed, i've never met him in person
21:08:56 <jeblair> ttx: if mtreinish doesn't update the qa slot, here's text that could go into it: http://paste.openstack.org/show/126134/
21:09:10 <ttx> Although the cross-project workshops current schedule is alkready the result of a lot of tweaks and we have trouble changing it at this point without breaking someone
21:09:31 <mtreinish> jeblair: I just pushed the update
21:09:34 <ttx> jeblair: mtreinish said he would update nowish
21:09:34 <eglynn> breaking someone == causing conflicts?
21:09:40 <ttx> eglynn: yes
21:09:49 <eglynn> cool, got it
21:09:54 <asalkeld> russellb / ttx, will there be some kind of ml thread to figure out who is leading the cross project sessions
21:09:58 <jeblair> ttx, mtreinish: cool, thanks.  sorry i missed that, it wasn't in yellow.  :)
21:10:22 <ttx> asalkeld: we have an etherpad, and russellb will push a thread about it
21:10:31 <ttx> #link https://etherpad.openstack.org/p/kilo-crossproject-summit-topics
21:10:34 <asalkeld> cool, thx
21:10:35 <russellb> asalkeld: yes, starting a thread now-ish
21:10:55 <russellb> for the future, should really have some more formatting around idea proposals
21:10:58 <russellb> including a lead for every proposal
21:11:02 <russellb> this was a mess
21:11:08 <asalkeld> yip
21:11:19 <russellb> downside of using an etherpad instead of the system from before
21:11:20 <asalkeld> so I happy to help with both upgrades and ha
21:11:24 <ttx> russellb: yes, the etherpad-driven thing works well until you reach critical mass of proposers
21:11:32 <russellb> asalkeld: anyone else that should be listed for HA?
21:11:38 <eglynn> also, how about including the PTL group in the session voting?
21:11:46 <eglynn> (as opposed to just the TC)
21:12:04 <asalkeld> russellb, you could ask for interested parties
21:12:10 <russellb> OK
21:12:31 <asalkeld> lots of RHT and miratis folk are interested in HA
21:12:38 * russellb nods
21:12:44 <ttx> eglynn: yes, I agree
21:12:53 <ttx> no need to limit voting
21:13:00 <russellb> eglynn: +1
21:13:12 <russellb> asalkeld: fabio will be there
21:13:18 <asalkeld> nice
21:13:36 <eglynn> yeah just gonna suggest fabio also
21:13:36 <jeblair> i don't think there's a need to limit feedback.  i do think there's a need to limit voting.  we are asking the tc to exercise some judgement and discretion here.
21:14:37 <eglynn> jeblair: well, I think the PTLs have plenty of judgement and discretion
21:14:41 <ttx> jeblair: the selection is not necessarily the highest votes, depends on what gets selected and what can be merged, too
21:15:05 <ttx> so we exercise discretion when we make the final selection, not necessarily when we vote :)
21:15:11 <jeblair> ttx: you said "no need to limit voting" i was responding to 'voting' :)
21:15:30 <jeblair> ttx: that works :)
21:15:40 <jeblair> and is in fact what we did in infra
21:16:05 <jeblair> i also don't believe ptls or anyone was excluded this time
21:16:16 <jeblair> i believe i saw quite a bit of feedback on the etherpad
21:16:41 <ttx> the preamble at the top discouraged non-TC-members from voting though
21:16:51 <ttx> which is I think what eglynn is referring to
21:17:22 <ttx> heck, we should use Gerrit for this
21:17:31 <ttx> :)
21:17:39 <eglynn> yeah I assumed it was a TC-only deal on the voting, apols if I misread that
21:17:44 <ttx> anyway, that's a tangent again
21:18:01 <jeblair> i have this natural inclination not to call something "voting" if the tally of results doesn't count.  ttx, i think we're getting hung up on terminology :)
21:18:11 <ttx> any other question/ urgent action to take wrt: Design Summit ?
21:18:24 <ttx> jeblair: that's a thing we do indeed
21:18:50 <mtreinish> ttx: is it a good time to bring up scheduling conflicts?
21:18:58 <ttx> mtreinish: as good as any
21:19:10 <eglynn> ttx: when does the schedule need to be absolute finalized?
21:19:12 <SergeyLukjanov> ttx, do we have a venue map? (what's the distance between design summit sessions and summit itself)
21:19:25 <ttx> it's literally across the street
21:19:35 <ttx> I don't have a handy map
21:19:46 <mtreinish> well I haven't seen mikal, but there is overlap with nova's functional testing and the qa track
21:19:47 <jeblair> i could only find one for the conference
21:19:48 <SergeyLukjanov> ttx, ok, across the street is enough to know, thx
21:20:01 <jeblair> mtreinish: mikal said he had to drop
21:20:04 <ttx> eglynn: in theory, today, but if we push changes they should be picked up until the end of the week
21:20:20 <mtreinish> jeblair: ah, ok
21:20:28 <ttx> the mobile app does async updates with sched
21:20:44 <ttx> so the sooner we are final the less confusion we generate
21:20:52 <eglynn> cool
21:21:07 * ttx looks for a venue map
21:21:22 <jeblair> https://www.openstack.org/summit/openstack-paris-summit-2014/venue-maps/ seems to be conference only
21:21:30 <SergeyLukjanov> ttx, jeblair, I'll join infra/qa/release meetup only afternoon, so, I hope not to miss something important
21:21:51 <jeblair> SergeyLukjanov: good to know, thanks
21:22:15 <ttx> SergeyLukjanov: ok
21:22:54 <SergeyLukjanov> (due to the sahara meetup before afternoon)
21:24:04 <curtis_p> pwd
21:24:28 <asalkeld> /tmp
21:24:32 <asalkeld> :)
21:24:46 <ttx> ok, any other question ?
21:25:04 <jeblair> ttx: i guess that's still a no on the design summit map?
21:25:05 <asalkeld> all good, looking forward to getting to paris
21:25:12 <ttx> jeblair: asking
21:26:02 <jeblair> ttx: also, is there an indoor passage between the two?
21:26:15 <ttx> jeblair: no indoor passage for sure
21:26:21 <ttx> it's two separate buildings.
21:26:38 <jeblair> so bring our umbrellas to lunch :)
21:27:01 <ttx> you can survive crossing the street in open air, I think :)
21:27:05 <ttx> ok, next topic
21:27:12 <ttx> #topic Explicitly state which projects can add requirements
21:27:16 <ttx> #link https://review.openstack.org/#/c/130245
21:27:30 <ttx> Not exactly sure what we need to discuss there
21:28:15 <jeblair> dhellmann, sdague: ?
21:28:19 <asalkeld> so "including stackforge" is new?
21:28:24 <jeblair> it's a major proposed policy change, yes
21:28:25 <ttx> I think the idea was to check it with all PTLs
21:28:34 <jeblair> it even has a cross-project summit session
21:28:44 <ttx> before it gets accepted and it's a pain to rollback
21:28:57 <thingee> :q
21:29:18 <sdague> there is also some infrastructure being added to devstack to make it probably not needed, so that should postpone
21:29:25 <SergeyLukjanov> IMO due to the crossproject session it should be postponed to summit / after summit
21:29:26 <jeblair> i think it would have less impact on ptls than the requirements reviewers and distros, whare are actually the primary audience for the requirements repo
21:29:31 <ttx> maybe we should all do our homework and read that, and keep the discussion about that for the cross-project session ?
21:29:44 <asalkeld> yeah
21:29:51 <asalkeld> big impact to distros
21:29:51 <SergeyLukjanov> +1
21:30:15 <asalkeld> also do we have a policy on what gets accepted
21:30:25 <asalkeld> or is it open
21:30:37 <jeblair> asalkeld: https://wiki.openstack.org/wiki/Requirements
21:31:11 <asalkeld> jeblair, so that part is staying the same then
21:31:12 <SergeyLukjanov> one more issue - for now if we have some lib in global req. than we're trying to avoiding alt. lib too instead of using the first one
21:31:36 <SergeyLukjanov> but in case if we'll open global reqs. it'll start looking like pypi.python.org eventually
21:32:33 <SergeyLukjanov> I'm feeling myself disconnected from IRC
21:32:39 <eglynn> don't we already accept requirements from stackforge projects? ... is this just making that policy explicit?
21:32:56 <asalkeld> eglynn, not to my knowledge
21:33:05 <jeblair> eglynn: we do not accept requirements from stackforge, and for good reason
21:33:38 <asalkeld> jeblair, are you against this then?
21:33:53 <asalkeld> just seeing the "and for good reason"
21:34:02 <jeblair> asalkeld: yes, i agree with sdague, both in the review, and in the alternate proposal to resolve the issue
21:34:11 <asalkeld> ok
21:34:32 * jeblair reviews that change :)
21:34:33 <eglynn> jeblair: I was thinking specifically of posix_ipc for stackforge/tooz
21:34:45 <eglynn> jeblair: ... but I see now that was originally added for oslo lockutils
21:35:25 <asalkeld> i like the "soft updating"
21:35:30 <jeblair> i reviewed the other ones (that implement soft update)
21:35:45 <jogo> I wonder how global-requirements would work in the big tent proposal
21:36:33 <asalkeld> jogo, "big tent" still doesn't mean totally unfiltered surely
21:36:53 <eglynn> jeblair: similarly for dependency version lower bounds that are motivated by stackforge usage?
21:36:55 <jeblair> jogo: yeah, definitely worth talking about.  i believe the main thing we get from it is the ability for distros to sanely, well, distribute "openstack"
21:37:13 <ttx> Part of the issue is that we use global-requirements for multiple things
21:37:20 <asalkeld> hopefully it is sensible but more open acceptance
21:37:24 <ttx> one of them is general dependency convergence
21:37:28 <eglynn> (... where the same dependency is also used by another project under openstack/ )
21:37:35 <ttx> others are more integration/testing related
21:37:37 <jeblair> so i think the answer to that is probably something like "it's small, for the stuff we think is really the core of openstack" or "it's big and anyone who wants to be part of the big tent uses it"
21:38:18 <jeblair> ttx: i think integration/testing is still part of the primary use -- dependency convergence + compatability
21:38:41 <ttx> FYI Maps of Le Meridien shall be uploaded tonight, so that "venue maps" thing should hopefuly look better tomorrow
21:38:51 <jeblair> it's not to select a version of something we test with, it's so that we're testing with the set of dependencies we have selected for the project as a whole
21:40:13 <jeblair> ttx: thanks!
21:40:20 <asalkeld> unless everyone uses docker this is going to get messy
21:40:36 <jeblair> asalkeld: what's going to get messy?
21:40:36 <asalkeld> (big tent and global requirements)
21:40:43 <dims__> asalkeld: fyi, heat already uses docker-py, just not in requirements
21:41:07 <asalkeld> dims__, i mean to distribute openstack services
21:41:20 <asalkeld> docker-py stuff is in contrib/
21:41:26 <dims__> asalkeld: ack
21:41:38 <ttx> OK, so I think the next step on that one is the cross-project workshop on requirements at the summit
21:41:57 <jeblair> asalkeld: maybe?  or maybe part of the big tent is taking a hard line on requirements to prevent it from getting messy... enough non-openstack projects want requirements enforcement that it seems like a possibility to me.
21:42:03 <ttx> I'll leave a note on the review
21:42:13 <asalkeld> jeblair, maybe
21:42:14 <eglynn> jeblair: just to clarify on the version lower bounds ... not kosher to require redis>=2.10.0 if motivated by tooz usage of redis, but kosher if say motivated by zaqar?
21:43:19 <jeblair> eglynn: correct.  (and that's why we're talking about tents)
21:43:46 <ttx> #topic Open discussion
21:43:51 <ttx> Anything else, anyone ?
21:44:06 <asalkeld> nope
21:44:08 <fungi> er, python 3.4!
21:44:23 <ttx> We should almost all be together in a few days
21:44:29 <fungi> #link https://etherpad.openstack.org/p/py34-transition
21:44:44 <fungi> we're waiting for a couple of bugs to clear through an ubuntu sru to trusty
21:45:05 <fungi> but also i'm worried glanceclient won't get their tests working in time for us to cut them over from 3.3 to 3.4
21:45:18 <asalkeld> seeing lots of client related patches
21:45:26 <fungi> #link http://lists.openstack.org/pipermail/openstack-dev/2014-October/048659.html
21:45:41 <fungi> #link https://launchpad.net/bugs/1382582
21:45:43 <uvirtbot> Launchpad bug 1382582 in python-glanceclient "untestable in Python 3.4" [Medium,Confirmed]
21:46:03 <fungi> asalkeld: yes, the glanceclient patches are great. i've already tested them and they're working
21:46:05 <ttx> nikhil_k: ^
21:46:09 <fungi> er, heatclient
21:46:21 <asalkeld> cool
21:46:25 <ttx> Looks like Glance is on the critical path
21:46:29 <fungi> so i think heatclient is on track
21:47:05 <fungi> but yeah, glanceclient may end up delaying our switch off the old py3k-precise nodes, or we might have to temporarily drop py3k testing for that project
21:47:10 <ttx> fungi: not sure nikhil_k is around. Looks like you should pay a visit to the Glance crew next week
21:47:17 <jeblair> or we enforce python3.4 testing for it :)
21:47:23 <fungi> i'll lurk in their team pod menacingly
21:47:43 * fungi makes sure to pack his most menacing hawaiian shirt
21:48:23 <asalkeld> :-)
21:48:25 <ttx> ok, one more minute for questions before I close the shop
21:48:34 * nikhil_k jumping between channels
21:48:48 <ttx> fungi: grab nikhil_k while he is here!
21:49:08 <fungi> nikhil_k: just asking for the glanceclient devs to bump priority on https://launchpad.net/bugs/1382582 if possible
21:49:10 <uvirtbot> Launchpad bug 1382582 in python-glanceclient "untestable in Python 3.4" [Medium,Confirmed]
21:49:31 <nikhil_k> fungi: noted
21:49:36 <fungi> and planning for contingencies if needed
21:49:54 <nikhil_k> fungi: would that be sufficient on master or need a version?
21:50:02 <fungi> nikhil_k: master will be fine
21:50:20 <fungi> just want to make sure we can get it passing 3.4 tests before we drop 3.3 tests in infra
21:50:50 <nikhil_k> fungi: I'm sure I'd your email on my to-do list
21:51:06 <fungi> nikhil_k: thanks!
21:51:08 <nikhil_k> will bring this up in the meeting
21:51:21 <ttx> alrighty then. See you all in a few days!
21:51:24 <ttx> #endmeeting