13:13:39 <ttx> #startmeeting ptl_sync
13:13:40 <openstack> Meeting started Tue May 27 13:13:39 2014 UTC and is due to finish in 60 minutes.  The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:13:41 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:13:43 <openstack> The meeting name has been set to 'ptl_sync'
13:13:47 <ttx> #topic Sahara
13:13:56 <ttx> SergeyLukjanov: o/
13:14:07 <SergeyLukjanov> ttx, hey
13:14:18 <ttx> Looking at https://launchpad.net/sahara/+milestone/juno-1
13:14:48 <ttx> You have 2 undefined there -- you should set a priority for them, or remove the milestone target
13:15:13 <ttx> otherwise are all those 17 relevant for J1 ? Or do you still need to review them all ?
13:15:23 <SergeyLukjanov> ttx, it's not fully completed, today is a bug triage day, so, it should looks better tomorrow
13:15:28 <ttx> #link https://launchpad.net/sahara/+milestone/juno-1
13:16:02 <SergeyLukjanov> ttx, at least several blueprints already going to j2
13:16:21 <ttx> OK
13:16:24 <SergeyLukjanov> ttx, oh, and we're decided to have a pilot of -specs
13:16:36 <ttx> ah? Starting when ?
13:16:58 <ttx> J1 or J2 ?
13:16:59 <SergeyLukjanov> ttx, starting from several huge blueprints and then decide to continue using it or not
13:17:15 <ttx> #info Sahara will use a spec repository
13:17:31 <SergeyLukjanov> ttx, -specs repo creation is on review, so, we'll try it right after it'll be ready
13:17:38 <SergeyLukjanov> so, I think it'll be near j1
13:18:04 <ttx> SergeyLukjanov: Does that mean we should no longer use blueprints as the entry point for feature submission ?
13:18:14 <SergeyLukjanov> it'll be pilot to decide how it works for us
13:18:25 <ttx> should we autoclean blueprints that are targeted but not prioritized, like other spec-using projects ?
13:18:55 <SergeyLukjanov> ttx, not for j1 I think, we'd like to see how it's working, because many of us aren't sure that it'll work good
13:19:16 <ttx> hmm, ok
13:19:39 <ttx> Also make sure you include blueprints to cover for what got done since icehouse
13:19:43 <SergeyLukjanov> ttx, we have no big problems with tracking incoming features, so, profit is about ensuring good description
13:19:56 <ttx> i.e. early features that landed when juno dev was opened
13:20:09 <SergeyLukjanov> ttx, yup, I'll check it
13:20:46 <ttx> OK thx
13:21:04 <ttx> That's all I had. Anythink you'd like to discuss at meeting today ?
13:21:31 <SergeyLukjanov> ttx, sorry that we aren't sure about moving to specs...
13:21:56 <SergeyLukjanov> ttx, I think no more news/questions from me
13:22:10 <ttx> SergeyLukjanov: : a problem, we'll act as if you didn't yet :)
13:22:15 <ttx> ok, great
13:22:25 * ttx bbls, back in a few
13:22:52 <SergeyLukjanov> ttx, thank you
13:34:41 <ttx> back
14:00:13 <ttx> dolphm: around?
14:00:22 <dolphm> ttx: o/
14:00:24 <ttx> #topic Keystone
14:00:48 <ttx> #link https://launchpad.net/keystone/+milestone/juno-1
14:01:08 * dolphm failed to shuffle blueprints around last week
14:01:10 <ttx> So... 2 undefined there, one unassigned
14:01:30 <ttx> You can prioritize or bump the two undefined
14:01:42 <ttx> And find an assignee for the unassigned
14:02:17 <ttx> Does that reflect everything major that happened since juno dev opened end of March ?
14:04:22 <dolphm> no - but i can't think of anything else that i expect to land in j1 either
14:05:51 <ttx> what about strong-password-enforcement ? Keep it in (give it a prio ?)
14:06:23 <ttx> guess that's a "no"
14:06:26 <dolphm> ttx: just untargeted that one :)
14:06:36 <ttx> OK, looks good
14:07:02 <ttx> Just keep an eye open on new last-minute submissions :)
14:07:18 <ttx> Anything you'd like to discuss at meeting ?
14:07:27 <ttx> large tokens ?
14:08:03 <ttx> or is that one solved by forever supporting old-style tokens
14:08:05 <dolphm> you mean compressed tokens?
14:08:17 <dolphm> oh yeah, we'll be supporting uncompressed tokens forever
14:08:46 <ttx> yeah, I mean the thread we had about with keystone increasing payload on data-oriented APIs
14:09:38 <dolphm> ah, i don't know there's much more to discuss that wasn't mentioned in that thread (?)
14:09:38 <ttx> seems like the most blatant fears were addressed in that thead
14:09:43 <ttx> thRead*
14:09:49 <ttx> yep
14:09:52 <ttx> OK then
14:10:00 <ttx> talk to you later!
14:10:02 <dolphm> we'll continue working on narrowing the gap though
14:10:17 <dolphm> worth noting in the agenda that we have our specs repo up?
14:10:36 <ttx> #info Keystone spec repo is now up
14:10:43 <ttx> will now appear on the summary
14:10:44 <dolphm> /salute
14:10:57 <ttx> jgriffith: ready when you are
14:33:01 <ttx> david-lyle: o/
14:33:09 <david-lyle> ttx: o/
14:33:10 <ttx> #topic Horizon
14:33:37 <ttx> #link https://launchpad.net/horizon/+milestone/juno-1
14:33:43 <ttx> That's a pretty busy list
14:33:59 <ttx> A few unassigned in there
14:34:07 <ttx> A lot of undefined priorities too
14:34:30 <david-lyle> planning on working through those this week
14:34:32 <ttx> david-lyle: did you go through the list yet and refine it to make reasonable targets for Juno-1 ?
14:34:41 <ttx> Juno-1 in June 12 :)
14:34:59 <david-lyle> a fair number of those are carry-over from Icehouse
14:35:16 <ttx> Right, so you might want to push them back to other milestones, now that they are created
14:35:19 <david-lyle> I'll move several to Juno-2
14:35:39 <ttx> The other trick is to make sure you reflect the significant things that were introduced since juno dev opened
14:36:16 <ttx> So you can retroactively create blueprints to cover them
14:37:05 <ttx> We'll review the resulting list next week
14:37:19 <ttx> david-lyle: remember you are invited to the TC meeting today
14:37:34 <david-lyle> yes
14:37:38 <ttx> david-lyle: anything you'd like to add to the release/cross-project meeting agenda ?
14:37:47 <david-lyle> not this week
14:39:07 <ttx> #info List will be refined by next week
14:39:13 <ttx> david-lyle: thx!
14:39:18 <david-lyle> ttx: thank you
14:39:21 <ttx> jgriffith: still not around ?
14:40:28 <ttx> mestery: ready when you are
14:40:33 <mestery> ttx: o/
14:40:50 <ttx> #topic Neutron
14:41:12 <ttx> #link https://launchpad.net/neutron/+milestone/juno-1
14:41:36 <mestery> I went through and cleaned up the Neutron Juno-1 BPs this morning.
14:41:53 <mestery> What is in there is fairly reasonable, though some of it may slip over the next week or so.
14:42:06 <ttx> mestery: how is the -specs effort going ?
14:42:23 <ttx> are most of those targeted BPs backed with an approved spec yet ?
14:42:28 <mestery> Very well! The hardest part has been encouraging reviewer and submitter turnaround, but otherwise it's working out great.
14:43:05 <mestery> Most of them, though some are not. I have indicated to submitters those would be removed from Juno-1 this week if the spec isn't approved.
14:43:14 <ttx> mestery: sounds good
14:43:52 <ttx> my autokick script still needs to be written, but in your case it would not autokick anything out of the milestone, since everything is properly prioritized
14:44:29 <ttx> mestery: does the j1 list also reflect anything major that would have landed in Juno since the master branch was unfrozen early April ?
14:44:30 <mestery> ttx: OK, and thanks! I'm going to work hard to keep this up to date and factual to the best of my knowledge. :)
14:44:40 <ttx> mestery: thx, that helps :)
14:44:49 <mestery> ttx: IT should, but I will take an action to verify that.
14:45:15 <ttx> mestery: ok, great. Anythig you'd like to add to the agenda of the cross-projetc meeting for today ?
14:45:30 <mestery> ttx: Nothing for today, no.
14:45:33 * ttx is fine with a short meeting for today anyway :)
14:45:47 <ttx> mestery: ok, great. talk to you later then
14:45:48 * mestery agrees. :)
14:45:53 <mestery> Thanks ttx!
15:00:06 <dhellmann> ttx: let's give it a week; the team only agreed to use the specs repo a day or two before the summit
15:23:30 <ttx> jgriffith: o/
15:26:12 <ttx> I guess I lost him
15:26:19 <ttx> notmyname: ready when you are
15:34:54 <jgriffith> ttx: sorry
15:35:05 <ttx> #topic Cinder
15:35:08 <ttx> jgriffith: o/
15:35:13 <jgriffith> ttx: 0/
15:35:17 <jgriffith> sorry about all that
15:35:30 <ttx> #link https://launchpad.net/cinder/+milestone/juno-1
15:35:41 <ttx> The milestone is still a bit of a mess a tthis point
15:35:50 <jgriffith> ttx: yes, I'll clean it up this morning
15:35:52 <ttx> jgriffith: how is your use of specs coming up ?
15:35:59 <jgriffith> ttx: it's finally landed
15:36:06 <jgriffith> ttx: folks just started using it
15:36:13 <jgriffith> ttx: everything new I'm pointing them to it
15:36:38 <jgriffith> https://review.openstack.org/#/q/status:open+cinder-specs,n,z
15:36:47 <jgriffith> harlowj is my only victim thus far
15:36:55 <jgriffith> but I have some of my own to submit today as well
15:37:03 <ttx> OK, so feel free to curate a list of j-1 blueprints to match
15:37:11 <ttx> ideally next week it should be clean
15:37:18 <ttx> so that i could enable the autokick script
15:37:27 <ttx> that would reject random targeting automatically
15:37:35 <jgriffith> ttx: so you want me to have a spec for each item that's in arleady?
15:37:42 <ttx> jgriffith: not really
15:37:52 <ttx> jgriffith: anything you set a priority for will be golden
15:38:06 <ttx> then YOU decide if you want a spec or not for each of those prioritized BPs :)
15:38:12 <jgriffith> ttx: ahh
15:38:14 <jgriffith> ok
15:38:23 <ttx> jgriffith: oh right, you might have missed last week
15:38:34 <jgriffith> I may have been interpreting the use of specs incorrectly
15:38:39 <jgriffith> ttx: yes, I missed last week
15:38:55 <ttx> jgriffith: the idea is to use the milestone page as a communication tool from PTL to release management and beyond
15:39:07 <ttx> ratherthan as an entry point for new features
15:39:15 <ttx> since that's what you'll use the specs repo for
15:39:18 <jgriffith> ttx: right, that's what I had though
15:39:29 <jgriffith> ttx: so step 1: submit a specs patch
15:39:42 <jgriffith> ttx: step 2: on approval target bp and prioritize
15:39:44 <jgriffith> in LP
15:39:47 <jgriffith> correct?
15:39:50 <ttx> yes
15:40:00 <jgriffith> cool, we're on the same page
15:40:09 <ttx> You can add extra BPs without a spec for corner cases
15:40:15 <jgriffith> the only confusing thing that I'd like to suggest a change for is the specs template
15:40:21 <jgriffith> it has a BP entry/link
15:40:28 <jgriffith> ttx: ok, fair enough
15:40:34 <ttx> I'm fine as long as you give them a priority, which is the proected field we'll use to distinguish between blessed things and random proposals
15:40:46 <jgriffith> ttx: yep, I will make sure that happens
15:40:57 <ttx> then I'll have a script automatically remove milestone target for unprioritized stuff
15:41:06 <ttx> so that whatever you put there is not polluted
15:41:07 <jgriffith> ttx: NICE!
15:41:12 * jgriffith will like that
15:41:33 <ttx> we work around LP limitations, but it should result in an.. easier time for all of us
15:41:45 <jgriffith> sounds like it will work well
15:41:50 <ttx> basically you express what you think will get done on the milestone page
15:41:59 <ttx> and the rest of the world consumes that
15:42:07 <jgriffith> excellent
15:42:14 <ttx> and we don't use BPs anymore as the entry point, but use specs instead
15:42:27 <ttx> that's the process we'll follow for every -spec using project out there
15:42:29 <jgriffith> Yep, I'm with you... all sounds great t me
15:42:41 <jgriffith> s/t me/to me/
15:42:49 <ttx> so spec is for suggestion/approval, BPs is to track progress
15:43:03 <ttx> and we'll just ignore untargeted stuff.
15:43:15 <jgriffith> always try to anyway :)
15:43:39 <ttx> jgriffith: ok, the other trick with J1 is to catch the work that was merged since Juno opened
15:43:48 <ttx> so feel free to retroactively add BPs to show for them
15:43:58 <ttx> (if any)
15:44:09 <jgriffith> ttx: will do
15:44:14 <ttx> jgriffith: anything you'd like to discuss at cross-project meeting today ?
15:44:24 <jgriffith> ttx: no thanks
15:44:52 <ttx> cool, ttyl
15:44:58 <ttx> notmyname: still not around?
15:45:00 <jgriffith> ttx: thanks
15:45:01 <jgriffith> cya
15:45:10 <ttx> zaneb: ready when you are
15:45:15 <zaneb> o/
15:45:23 <ttx> #topic Heat
15:45:51 <ttx> #link https://launchpad.net/heat/+milestone/juno-1
15:46:00 <zaneb> btw we decided at the meeting on Wednesday that we will move to a specs repo
15:46:07 <ttx> OK, was about to ask
15:46:20 <ttx> will that be for J1, or mostly for J2 ?
15:46:29 <zaneb> #link http://eavesdrop.openstack.org/meetings/heat/2014/heat.2014-05-21-20.00.html
15:46:44 <ttx> #info Heat will use a spec repo as well
15:46:54 <zaneb> I think it will take a while to get migrated
15:47:08 <zaneb> I'm still not 100% clear on how the process is going to work
15:47:16 <ttx> zaneb: OK, maybe use taditional milestone targeting for J1, then ?
15:47:21 <ttx> traditoinal*
15:47:23 <ttx> arh
15:47:24 <zaneb> but we want to follow the other projects while keeping it lightweight
15:47:33 <zaneb> that makes sense to me
15:48:11 <zaneb> AIUI the repo doesn't really replace launchpad so much as replace the wiki as a place to discuss designs?
15:48:18 <ttx> zaneb: did you follow the discussion about using the milestone page a bit more aggressively (no longer as the feature entry point) ?
15:48:41 <zaneb> no, I missed that I think
15:48:51 <ttx> ok, let me explain quick
15:49:09 <ttx> The main difference with before is that the first step is no longer "file a blueprint"
15:49:18 <ttx> the entry point for new features becomes te -spec repo
15:49:34 <ttx> we still use blueprints to track implementation and target milestone
15:49:41 <ttx> but not that much to track spec approval
15:50:09 <ttx> so tha tlets us use blueprints as a normalized tracking tool to keep track of what is likely to be in the milestone and therefore in the final release
15:50:51 <ttx> since we don't use BPs as an entry point anymore, we can use it as a clean slate for PTLs and other drivers to list stuff that they think will make it
15:51:32 <ttx> the issue being, anyone can set the target milestone so that clean slate becomes polluted very fast
15:51:39 <ttx> so we use a trick
15:51:57 <ttx> anything that doesn't have a priority set should be removed from hte target milestone automatically
15:52:09 <ttx> (priority is set by the drivers group in LP)
15:52:23 <ttx> that lets us use the milestone pages as real communication tools
15:52:35 <ttx> rather than as a workflow for people to suggest features
15:52:47 <ttx> (we use -specs for tha tnow anyway)
15:52:54 <ttx> zaneb: does that make sense ?
15:53:16 <zaneb> yep, makes sense
15:53:17 <ttx> 1- people file specs, they are discussed, some of them get approved
15:53:31 <ttx> 2- PTLs and drivers bbuild a list of stuff that will land soon in milestone pages
15:53:49 <ttx> note that I plan to write a tool to help create a blueprint corresponding to a spec
15:54:06 <ttx> to facilitate that transitoin
15:54:16 <zaneb> what's the migration plan in the other direction?
15:54:23 <ttx> something like "spec2bp my-spec juno-2"
15:54:27 <zaneb> i.e. bps that are already filed in lp
15:54:55 <ttx> zaneb: there is no plan, so you might need to engage with authors there
15:55:01 <zaneb> ok
15:55:07 <ttx> you can still reuse those BPs and just prioritize/target them
15:55:19 <ttx> but in some cases a spec wouldn't hurt
15:55:37 <ttx> note that I don't expect people to forget "file a BP" antime soon, so there will be some issues
15:55:50 <zaneb> ok, that makes sense
15:56:17 <ttx> my goal is to make sure we don't do everything twice, once in -specs and once in BPs
15:56:42 <ttx> reduce friction to make the tools as integrated as they can be at this point
15:56:48 <ttx> anyway
15:57:03 <ttx> for J1 you should probably do it old style
15:57:24 <ttx> zaneb: you should curate that J1 list to match what was done in J1 so far, what is likely to be done for J1, and push back the rest
15:57:36 <ttx> Whatever you keep in the list, you should have assignees and priorities
15:57:56 <ttx> the trick being, without the autokick script enabled, people will keep adding stuff there ;)
15:58:27 <ttx> zaneb: questions?
15:58:29 <zaneb> yeah, that will be unavoidable
15:58:44 <ttx> zaneb: I expect the new process to be much less pain
15:58:50 <ttx> zaneb: at leas tfor us both
15:59:01 <zaneb> I hope so :)
15:59:11 <zaneb> I am keen to make the PTL's job smaller ;)
15:59:12 <ttx> zaneb: in previous cycles we would take a lot of time fixing that list
15:59:35 <ttx> at least with the autokick enabled it doesn't degrade all by itself
15:59:51 <ttx> it shall stay how you write it :)
16:00:01 <ttx> zaneb: anything you'd like to add to today's meeting agenda ?
16:00:04 <zaneb> tbh, except to the extent that other openstack projects are depending on a blueprint, I don't really care what milestone they land in
16:00:34 <ttx> zaneb: i's still an important information to try to communicate outside of the project
16:00:53 <ttx> zaneb: some downstream stakeholders do not just follow releases
16:01:05 <ttx> and appreciate a slightly more educated guess
16:01:17 <ttx> zaneb: but then it's far from being exact
16:01:28 <ttx> markwash: around ?
16:01:37 <zaneb> yes, I think the circumstances vary from program to program
16:01:39 <ttx> zaneb: all set ?
16:01:44 <markwash> ttx: hello
16:01:45 <zaneb> yep :)
16:01:53 <ttx> zaneb: ok, ttyl
16:01:56 <ttx> #topic Glance
16:02:01 <ttx> markwash: hola!
16:02:32 <ttx> markwash: how is your spec effort going so far ?
16:02:36 <ttx> #link https://launchpad.net/glance/+milestone/juno-1
16:03:10 <ttx> milestone page looks a bit dirty still. Good thing my autokick script is not ready yet :)
16:03:34 <markwash> needing to land https://review.openstack.org/#/c/94464/
16:03:57 <ttx> markwash: so you will probably handle J1 targeting the old way ?
16:04:29 <markwash> ttx: hmm, unfortunately it seems like we might need to?
16:04:35 <markwash> but we'll probably have some specs that show up as well
16:04:51 <ttx> OK, so try to clean up the list to come up with assigned, prioritized stuff that may actually make it
16:05:13 <ttx> also if there was anything significant landed since Juno was oened, maybe retroactively create a BP to cover for it
16:05:17 <ttx> opened*
16:05:34 <markwash> kk
16:06:14 <ttx> Then when you're ready to switch, make the -spec proposal the feature entry point
16:06:25 <ttx> and deprectae use of LP for that purpose
16:06:42 <ttx> hopefully I shall have my conversion and autokick script done soon
16:06:53 <markwash> okay
16:07:07 <ttx> markwash: think you'll have the list refined by next week ?
16:07:08 <markwash> is there a way to protect something from autokicking? just adding a spec entry for it?
16:07:14 <ttx> (J1 is June 12 fwiw)
16:07:25 <ttx> markwash: no, just set a priority for it
16:07:31 <markwash> gotcha
16:07:36 <ttx> the script won't look at the spec at all
16:07:53 <ttx> so you're free to add stuff withut a spec, as long as it's prioritized, it's golden
16:08:03 <markwash> okay cool, so we could turn on the autokick script as soon as we have specs turned on and the current entries prioritized
16:08:15 <ttx> we just work around the fact that milestone targeting is a free-for-all field in LP
16:08:28 <ttx> markwash: exactly
16:08:41 <markwash> as for prioritized by next week, seems possible
16:08:56 <markwash> so yes
16:08:56 <ttx> #info Glance shall have a J1 plan by next week
16:09:01 <markwash> :-)
16:09:06 <ttx> ok great
16:09:17 <ttx> anything you'd like to add to meeting agenda for today ?
16:09:41 <markwash> nope
16:10:09 * ttx expects that the new process will result in less BP maintenance pain
16:10:19 <ttx> or at least that's my goal
16:10:25 <ttx> markwash: ok thx! ttyl
16:10:30 <ttx> SlickNik: ready when you are
16:10:45 <ttx> notmyname: if you are around, same
16:15:16 <SlickNik> ttx: here now
16:16:27 <ttx> #topic Trove
16:16:29 <ttx> SlickNik: o/
16:16:33 <SlickNik> o/
16:16:47 <ttx> #link https://launchpad.net/trove/+milestone/juno-1
16:17:07 <ttx> Looks pretty clean. Only one is unassigned
16:17:16 <ttx> Does that reflect what you think might make J1 ?
16:17:35 <SlickNik> Yes, I've been working on regular triage to keep it up-to-date. :)
16:17:41 <ttx> does that include the work that was merged since Juno opened ?
16:17:47 <SlickNik> Yes, it does.
16:17:52 <ttx> ok, good
16:18:21 <SlickNik> Will follow up on status for that 1 bp that was unassigned.
16:18:22 <ttx> #info J1 plans are up and current
16:18:48 <ttx> OK, I guess we can close this one fast then
16:18:48 <SlickNik> There's a couple of BPs that I don't think will make J1.
16:18:59 <ttx> Anything you'd like to discuss at meeting later ?
16:19:01 <SlickNik> I'll talk to the owners to confirm, and move them out.
16:19:12 <ttx> sure, just defer them to J2
16:19:24 <SlickNik> Yup, will do that.
16:19:41 <ttx> and try to keep the implementation status in sync with current state
16:19:50 <ttx> (assignees can help by updating that themselves)
16:20:07 <SlickNik> Yes, will do. (And will remind assignees to do it themselves)
16:20:21 <ttx> OK then, anything else ?
16:20:26 <SlickNik> I didn't really have anything special for the meeting later.
16:20:33 <ttx> good good
16:20:38 <SlickNik> Nope, pretty simple this week. Thanks!
16:20:42 <ttx> talk to you later then!
16:20:50 <SlickNik> Thanks. Later!
16:21:07 <ttx> notmyname: last call to be included in this weeks notes
16:22:19 <ttx> #endmeeting