21:04:23 #startmeeting 21:04:24 Meeting started Mon Dec 12 21:04:23 2011 UTC. The chair is vishy. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:04:25 Useful Commands: #action #agreed #help #info #idea #link #topic. 21:04:37 #topic First Nova Subteam Meeting 21:04:41 welcome everyone 21:05:03 first item on the agenda is figure out how we don't have to do this 21:05:25 #topic Best method of communication 21:05:41 #info Goals for communication: 21:06:08 #info Efficiently managing and targeting blueprints 21:06:26 #info Addressing shared blueprints and work across teams 21:06:34 #info Ensuring we aren't duplicating work 21:06:42 #info Ensuring we aren't blocking each other 21:06:49 #info Staying abreast of decisions made by teams 21:06:58 #info Staying aligned with the release schedule 21:07:14 soren sugested that this could all be done via email 21:07:33 <_0x44> Should there be a sub-team mailinglist? 21:07:47 this seems like a possibility but we all have to be better at emailing updates 21:08:10 _0x44: based on the relative value of the other mailing lists I'm not sure it is necessary 21:08:28 we could just hit openstack mailing list with a [subteam] heading if necessary 21:08:39 other options are: 21:08:39 vishy: I like that approach 21:08:49 regular meetings (maybe biweekly) 21:08:52 <_0x44> I personally think a lot of the db discussions would have been better done on a mailing list instead of an IRC meeting as the only people who attend have been myself, bcwaldon, and monsyne dragon. 21:08:58 shared place to put updates 21:09:14 I bet we would get better participation through email anyways 21:09:30 _0x44: agreed. I was trying to avoid the bikeshedding that sometimes occurs on the ml 21:09:41 by shrinking the commenters to people who are actually going to work on it 21:09:47 but I'm not sure if that really worked 21:09:58 i think it has been pretty effective for nova-api 21:10:05 (which had the most bikeshedding imo) 21:10:17 I know there are people that work on the nova apis that don't come to nova-api team meetings, but I'm not sure if that's because they dont care about doing it right or what 21:10:48 #idea the team leads are probably enough to prevent bikeshedding 21:11:14 vishy: yeah, as long as we empower the team leads enough to shut down bikesheders 21:11:15 #idea so we can return to ML discussions if the team leads are willing to keep the discussions focused and directed. 21:11:38 <_0x44> vishy: +1 to that. 21:11:46 does anyone have alternative proposals or other suggestions? 21:12:15 vishy: how about we encourage email correspondence through the list and say meetings can be scheduled when its absolutely necessary 21:12:28 bcwaldon: I like it 21:12:30 but to include those in other timezones, we absolutely need to shoot for email 21:12:46 when I say *other*, I mean non-US timezones 21:13:07 I would love to see if participation picks up from non-US timezones if we eradicate all these meetings 21:13:08 so cross team communication can include multiple headers or will be with the [subteam] heading if it applies to everyone 21:13:29 so If i'm emailing the team leads, I'm going to use the [subteam] heading, so eyes out for those 21:13:41 <_0x44> I like that, especially since the person doing the most work on db clean up right now isn't able to attend the -db meeting. 21:14:08 vishy: does it make sense to have team-specific mailing lists 21:14:09 #action vishy to email the list explaining use of the list by subteams. 21:14:13 I know we've already gone back and forth on that one 21:14:23 bcwaldon: no, we have those and i think it is not necessary 21:14:30 can you explain why 21:14:49 bcwaldon: my last email suggested using [api] header for example for nova-api related discussion 21:14:59 bcwaldon: it isolates information 21:15:09 kind of like the meetings do 21:15:17 bcwaldon: people can ignore messages they don't care about 21:15:18 so what's the point of having logical teams on launchpad? 21:15:28 I've got ~50 people on nova-api that have never said a word 21:15:37 bcwaldon: perhaps there isn't a lot of value 21:15:43 bcwaldon: aside from having a team lead 21:15:58 that's what I'm thinking 21:16:09 bcwaldon: I don't know that we need to remove the teams though 21:16:23 we can just shut down the ml after everyone has switched over to using the main ml 21:16:28 I think what's most important here is that people that care about different topics can have a way to communicate with one another 21:16:34 bcwaldon: actually 21:16:43 bcwaldon: the subteams on launchpad are great for tracking blueprints 21:16:49 vishy: good point 21:16:51 bcwaldon: so i think they still have value 21:16:55 absolutely 21:17:30 #info subteams will not use separate mailing lists, but still have value for blueprint tracking 21:17:38 ok next topic 21:17:53 #topic Blueprint management and targetting 21:18:16 #info we need to get better at assigning and targetting blueprints 21:18:38 there is a lot of stuff still assigned to teams with no actual person assigned 21:18:43 how can we improve that 21:18:45 suggestions? 21:18:55 so the problem I see a lot on the project is that we're great at reacing to what features are proposed to be added 21:19:04 but when we have a feature that a specific company isn't driving, it just sits 21:19:15 since we don't have resources to allocate to those kinds of features 21:19:34 bcwaldon: I agree. Unfortunately a lot of the stuff is important to others 21:19:42 bcwaldon: for example separate-volume-api 21:19:48 <_0x44> I have to run 21:19:52 bye _0x44 21:19:53 ! 21:19:54 <_0x44> Sorry 21:20:00 _0x44: ok thx, will send out update 21:20:22 vishy: so we're back to more of a technical-debt problem 21:20:31 how can a product organization prioritize technical debt 21:20:37 when there are so many tempting features 21:20:51 bcwaldon: I don't think there are that many tempting features left 21:21:04 bcwaldon: we don't have that much feature work targeted 21:21:13 maybe at this point 21:21:19 this feeling has been around a while 21:21:21 I'm digging down deep 21:21:22 bcwaldon: so i feel like we should all be moving to technical debt, cleanup, bugfixing, perf testing 21:21:29 vishy: absolutely! 21:23:11 bcwaldon: looks like it is me and you now 21:23:16 just the two of us 21:23:18 we are nova 21:23:24 haha 21:23:57 so yes, we do need to get better at assigning blueprints 21:24:11 how about I go over the nova-api blueprints and send out an email to the team 21:24:18 that would be cool 21:24:31 I will take a look at the state of the other ones 21:24:57 ok 21:25:00 what other problems can I solve for you 21:25:36 i guess that is it 21:25:53 actually 21:25:55 let me ask a question 21:25:57 #action bcwaldon to clean up nova-api blueprints and email about targeting 21:26:17 for the openstack-api-ssl bp, there has been a review in gerrit forever 21:26:26 #action vishy to look over other blueprints and send out specific email to [subteam] highlighting issues 21:26:28 how would you suggest we move forward with that 21:26:35 bcwaldon: is that an hp one? 21:26:40 sometimes you have to kick them 21:26:50 i kicked them about the vm state one and they updated it 21:26:51 vishy: we're in a public forum! 21:26:56 ok, I'll ping them 21:27:11 do you want to be my assignee for separate-nova-volumeapi 21:27:14 vishy: can I kick anybody for you? 21:27:40 bcwaldon: I guess it should be me, but I want to talk with you about where exactly we should move the parts 21:27:54 bcwaldon: we discussed it briefly but i will probably bug you about it more 21:27:58 yeah, I need to talk to you about adminapi 21:28:03 we can hop on skype 21:28:06 lets finish this meeting 21:28:10 yes 21:28:12 #topic open discussion 21:28:17 any questions comments? 21:28:20 thoughts? 21:28:22 complaints? 21:28:22 mtaylor: this is directed at you 21:28:26 is anyone even here? 21:28:28 :) 21:28:43 just the two of us 21:28:43 o/ 21:28:46 we can make it if we try 21:29:08 * mtaylor is just lurking 21:29:18 #endmeeting