21:03:53 <ttx> #startmeeting project
21:03:54 <openstack> Meeting started Tue Jun  3 21:03:53 2014 UTC and is due to finish in 60 minutes.  The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:03:55 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:03:58 <openstack> The meeting name has been set to 'project'
21:03:59 <ttx> Agenda @
21:04:33 <ttx> #link http://wiki.openstack.org/Meetings/ProjectMeeting
21:04:39 <ttx> Due to travel we couldn't have 1:1s today, so if you have news you can shout them now
21:04:47 <ttx> notmyname: how is the storage policies merging going so far ?
21:05:34 <notmyname> ttx: really well. in fact I'm sitting in a room with a bunch of devs after having just done a 2-hour overview with the code on the screen. still probably a week or so left to get it merged, but great progress so far
21:05:50 <notmyname> (and really cool development today, swift will probably have two-dimentional time!)
21:06:13 <mikal> notmyname: ?!?
21:06:25 <ttx> ?
21:06:38 <zaneb> I definitely want to hear more about that :D
21:06:48 <notmyname> heh. ask me later. just a cool way to represent progress in the system without changing what the client sees as timestamps :-)
21:06:56 <ttx> notmyname: ok, teasing I see
21:06:58 <notmyname> not really important for this meeting :-)
21:07:01 <SergeyLukjanov> ttx, [sahara] I think we're going to start -specs pilot for selected bps at the end of week
21:07:01 <SlickNik> two dimensional time? I'm intrigued :)
21:07:02 <mikal> I'm still struggling with the linear flow of time I experience day to day
21:07:19 <dhellmann> mikal: in 2 dimensional time, you can go around meetings
21:07:20 <ttx> #info [sahara] to start -specs pilot for selected bps at the end of week
21:07:30 <mikal> dhellmann: omg, that would be awesome
21:07:40 * ttx uses 3D time. With glasses
21:07:43 <dhellmann> mikal: I know!
21:08:06 <ttx> notmyname: looks like a blogpost is in order
21:08:09 <mikal> ttx: nova is chasing juno-1, and de-ephasising specs approvals to focus reviewers on juno-1. We're also having a bug day today or tomorrow depending on your timezone.
21:08:25 <ttx> ok, that brings us to...
21:08:27 <ttx> #topic Juno-1 plans
21:08:33 <ttx> I looked up the juno-1 plans for all the projects tagging the milestone next week
21:08:38 <ttx> ceilometer, keystone, neutron, nova, sahara, trove plans all look good
21:08:42 <mikal> ttx: also, johnthetubaguy has done a really good job of shunting stuff to juno-2 that wont make it into juno-1
21:08:45 <ttx> heat, oslo, horizon may need a bit of polish/cleanup
21:08:52 <ttx> but overall plans look good
21:08:59 <ttx> I'll be in touch with you all over next week
21:09:02 <mikal> ttx: do you have advice on producing release notes?
21:09:06 <ttx> as we refine the juno-1 plans (tag due next week)
21:09:09 <mikal> ttx: should I be writing them progressively over the release?
21:09:27 <ttx> mikal: people usually do them at the very end, but i can only encourage frontloading
21:09:28 <mikal> ttx: i.e. should they be part of the juno-1 milestone?
21:09:29 <eglynn> relnotes are only for the final 2014.2 tag right?
21:09:39 <eglynn> ... i.e. not per-milestone, or?
21:09:49 <ttx> eglynn: yes, final only
21:09:49 <mikal> eglynn: sure, but how to you remember 6 months of work?
21:09:56 <mikal> s/to/do/
21:10:05 <eglynn> mikal: fair point :)
21:10:10 <zaneb> ttx: I did some tidying earlier today, not sure if you looked since then
21:10:21 <ttx> zaneb: depends on your definition of earlier
21:10:34 <zaneb> I forget what time
21:10:41 <ttx> there were a few undefined/nuassigned last time I looked
21:10:42 <eglynn> mikal: let LP remember for you ... no, actually, scratch that thought ;)
21:10:58 <ttx> I'll also reach out to decide if we are ready to enable the auto-kick-unprioritized-bps script
21:10:59 * dhellman_ may be the victim of a netsplit
21:11:00 <mikal> eglynn: I just don't want to have to read 2,500 commit messages in a big rush later
21:11:01 <zaneb> ttx: I think there are maybe 2 undefined ones left
21:11:04 <mikal> So any advice welcome
21:11:26 <ttx> (for the projects using -specs)
21:12:12 <ttx> #topic Other program news
21:12:16 <ttx> Infra, QA, Docs... anything you'd like to mention ?
21:12:31 <dhellman_> I have a couple of things for oslo
21:12:33 <mtreinish> well I forgot to mention we pushed the first tag for branchless tempest last week
21:12:54 <mtreinish> not that special though
21:12:59 <ttx> dhellman_: sure, you cane #info them yourself
21:13:15 <ttx> #info branchless tempest is now a thing
21:13:44 <dhellman_> #info Please remember as you set your priorities that the rpc code in the incubator will be removed at the end of this cycle, so reviews for work on oslo.messaging adoption should get some love. :-)
21:13:49 <dhellman_> I think only heat, neutron, and trove are left.
21:13:58 <dhellman_> also, I'm not seeing the participation I was hoping for from Oslo liaisons. Please remind your liaisons to attend the weekly oslo meeting https://wiki.openstack.org/wiki/Meetings/Oslo
21:14:02 <dhellman_> It's not like I'm calling roll, but I'm not hearing from many of them.
21:14:04 <mtreinish> ttx: heh, well it was already a thing because we didn't make an icehouse branch at the release
21:14:22 <mikal> dhellman_: also, log message hinting needs to not be forgotten, right?
21:14:29 <eglynn> mtreinish: guidance on the policy around skipping selected Juno tests against stable/icehouse would be welcome
21:14:45 <dhellman_> mikal: yes, I have a todo to write up some instructions as part of the oslo.i18n release
21:14:49 <ttx> mikal: about release notes - ideally we would autogenerate most of them from a specs blurb or some commit message parts -- but it's non-trivial to fetch those back
21:14:59 <SlickNik> dhellman_: Thanks for the heads up. Will follow up on it.
21:15:03 <dhellman_> mikal: assuming you mean translation?
21:15:16 <mikal> ttx: do we only include landed bps in release notes? Not important bug fixes?
21:15:17 <mtreinish> eglynn: guidance in what format?
21:15:18 <mestery> dhellman_: We're working on the oslo.messaging port in neutron right now.
21:15:20 <mikal> dhellman_: yeah
21:15:26 <ttx> #oslo not seeing the participation I was hoping for from Oslo liaisons. Please remind your liaisons to attend the weekly oslo meeting https://wiki.openstack.org/wiki/Meetings/Oslo
21:15:29 <dhellman_> mestery: excellent
21:15:31 <ttx> #info oslo not seeing the participation I was hoping for from Oslo liaisons. Please remind your liaisons to attend the weekly oslo meeting https://wiki.openstack.org/wiki/Meetings/Oslo
21:15:45 <eglynn> mtreinish: just in English is fine :)
21:16:01 <jogo> ohh oslo meetings almost forgot
21:16:02 <dhellman_> mikal: my goal is to have oslo.i18n ready for J1, and to include all of the guidelines in the docs
21:16:06 <ttx> mikal: we include "major features"
21:16:22 <eglynn> mtreinish: (... since I thought the discoverable API-driven approach was the initially agreed approach last week, but seems that config driven option has made a re-appearance?)
21:16:24 <mikal> ttx: ok
21:16:25 <ttx> mikal: it's that editorial part (the selection of what is major) that takes time
21:17:02 <ttx> OK, I think we can safely switch to...
21:17:03 <ttx> #topic Open discussion
21:17:10 <ttx> Anything else, anyone ?
21:17:14 <dolphm> yes...
21:17:22 <mtreinish> eglynn: it needs both. the discoverable api is for end users (and config generation and verification in tempest). But from the tempest side it still needs to be a config flag
21:17:30 <dolphm> regarding the auto-kick-unprioritized-bps script... that's a one time thing per project, run?
21:17:40 <ttx> mikal, mestery: wanted to ask you if you didn't end up with twice as many reviews to do with the whole -specs thing
21:18:02 <ttx> dolphm: it would run regularly. The current script runs every 2 hours
21:18:04 <mikal> ttx: could you rephrase the question please?
21:18:15 <mestery> ttx: It's true it has increased the review load, but it's helped as well because it's gotten discussions going earlier on some items.
21:18:27 <mestery> ttx: And having one place to go for people to participate and see other's comments is also very nice.
21:18:33 <eglynn> mtreinish: k, we can take the discussion off-line for later/tmrw (... prolly not relevant to the rest of this meeting)
21:18:36 <dolphm> ttx: so what does it mean to kick them? mark as obsolete or something?
21:18:39 <mikal> I don't see evidence that its reduced the load on nova reviewers
21:18:44 <mikal> We do have a lot of specs reviews open
21:18:48 <ttx> mikal: I think there is general agreement tat specs is a good thing... just want early feedback on the additional work it generates, if any
21:18:50 <mikal> So people doing those are pretty busy
21:18:57 <mikal> ttx: ahh, I see
21:18:58 <mtreinish> eglynn: fair enough :). I'll write up something in the tempest developer docs for this too
21:19:00 <mikal> I think its worth it
21:19:04 <dhellman_> dolphm: I think to just unset the target release
21:19:05 <mestery> I think the additional work it generates is worth it as well.
21:19:12 <mikal> There's definitely code we're -2'ing until we get a spec
21:19:17 <eglynn> mtreinish: cool, thanks!
21:19:18 <mikal> Which I think means we're landing better designed code
21:19:33 <dolphm> dhellman_: ah, thanks
21:19:35 <mikal> We're still pretty badly behind on reviews though (around 500 open last I looked)
21:19:36 <ttx> dolphm: that leans clearing the milestone target field
21:19:41 <ttx> means*
21:19:58 <dolphm> i'd like a script that kept *code* reviews referencing un-approved blueprints/specs to be maintained as Work In Progress automatically :)
21:20:01 <dhellman_> ttx: milestone or series?
21:20:09 <dhellman_> dolphm: +1
21:20:18 <SlickNik> dolphm: ++
21:20:24 <ttx> dhellman_: milestone. The series is autoadjusted to match the milestone (damn LP)
21:20:44 <ttx> dolphm: I think mikal has something
21:20:49 <dolphm> mikal: ooh?
21:20:52 <dhellman_> ttx: ok, I just removed a bunch of blueprints with juno as the series but without specs (I only had a few, so I did them by hand)
21:21:07 <mikal> dolphm: we're doing that manually
21:21:14 <dolphm> mikal: same here
21:21:20 <mikal> Well, I have a script, but then I do a lot of hand checking
21:21:27 <dolphm> mikal: link?
21:21:29 <mikal> johnthetubaguy might have something better
21:21:32 <ttx> dhellman_: you shouldn't care about the series -- the script maes sure it stays coherent with the milestone (since Lp doesn't)
21:21:37 <ttx> makes*�
21:21:46 <dhellman_> ttx: ok, I was using the series for planning
21:21:53 <dhellman_> before the summit
21:22:00 <mikal> dolphm: to my script? Let's talk about that in PM, its tied up in a bigger, not very general script for my personal code reviews
21:22:07 <dolphm> mikal: ack
21:22:09 <ttx> dhellman_: the problem is it can get out of sync with the milestone
21:22:20 <dhellman_> ttx: both are blank now, so they're in sync :-)
21:22:24 <ttx> dhellman_: so we use "priority" and "milestone target" only
21:22:27 <dhellman_> ok
21:22:40 <ttx> series is autoadjusted (with the current script)
21:23:17 <ttx> the new script will additionally makes sure only prioritized stuff can be listed in a milestone, effectively restricting that "blessing" task to project drivers
21:23:48 <dhellman_> I now have a bunch of prioritized bps without milestones, so I guess I need to set those?
21:23:50 <ttx> script leaves a gentle comment in the whiteboard when it kicks things out of milestones
21:24:19 <ttx> dhellman_: yes, ideally. You can use the future/next things
21:24:27 <dhellman_> ttx: ok
21:24:31 <ttx> if that's not juno material
21:25:00 <ttx> anything else before we close ?
21:26:14 <ttx> I guess that's a no
21:26:26 <ttx> Talk to you all later!
21:26:30 <ttx> #endmeeting