Thursday, 2011-05-05

*** GasbaKid has quit IRC00:03
*** chmouel has quit IRC00:07
*** chmouel has joined #openstack-meeting00:08
*** adjohn has joined #openstack-meeting00:15
*** jaypipes has quit IRC00:17
*** deshantm has quit IRC00:17
*** jaypipes has joined #openstack-meeting00:17
*** deshantm has joined #openstack-meeting00:19
*** dendrobates is now known as dendro-afk00:27
*** dragondm has quit IRC00:43
*** ewindisch has quit IRC04:11
*** dendro-afk is now known as dendrobates05:14
*** dendrobates is now known as dendro-afk05:43
*** zul has quit IRC07:11
*** zul has joined #openstack-meeting07:12
*** adjohn has quit IRC09:29
*** cbeck1 has joined #openstack-meeting10:05
*** cbeck has quit IRC10:05
*** ovidwu has quit IRC10:43
*** ovidwu has joined #openstack-meeting10:43
*** adjohn has joined #openstack-meeting11:02
*** mancdaz has quit IRC11:31
*** dendro-afk is now known as dendrobates12:38
*** mray has joined #openstack-meeting12:57
*** mray has quit IRC12:59
*** mray has joined #openstack-meeting13:00
*** adjohn has quit IRC13:08
*** edconzel has joined #openstack-meeting13:49
*** edconzel has quit IRC13:53
*** edconzel has joined #openstack-meeting13:59
*** jkoelker has joined #openstack-meeting14:10
*** mancdaz has joined #openstack-meeting14:36
*** mancdaz1203 has joined #openstack-meeting14:36
*** mancdaz1203 has quit IRC14:36
*** troytoman-away is now known as troytoman14:56
*** larzy has joined #openstack-meeting15:07
*** troytoman is now known as troytoman-away15:19
*** troytoman-away is now known as troytoman15:19
*** blamar has joined #openstack-meeting15:21
*** med_out is now known as medberry15:57
*** blamar has quit IRC16:12
*** dendrobates is now known as dendro-afk16:19
*** johnpur has joined #openstack-meeting16:22
*** dragondm has joined #openstack-meeting16:34
*** RobertLaptop has joined #openstack-meeting16:51
*** RobertLaptop has left #openstack-meeting16:54
*** dendro-afk is now known as dendrobates16:55
*** dragondm has quit IRC17:11
*** troytoman is now known as troytoman-away17:18
*** cbeck1 is now known as cbeck17:27
*** dendrobates is now known as dendro-afk17:33
*** medberry is now known as med_out17:37
*** JordanRinke has joined #openstack-meeting18:00
*** dendro-afk is now known as dendrobates18:08
notmynameI won't be in the PPB meeting today. I'll be int he channel, so I'll have a record of the meeting, but I won't be active18:13
*** dragondm has joined #openstack-meeting18:18
jaypipesnotmyname: no worries.18:18
*** troytoman-away is now known as troytoman18:26
*** troytoman is now known as troytoman-away18:34
*** jbryce has joined #openstack-meeting18:34
*** dendrobates is now known as dendro-afk18:39
*** jkoelker has quit IRC18:55
*** jkoelker has joined #openstack-meeting18:55
*** eday has joined #openstack-meeting18:57
*** ewanmellor has joined #openstack-meeting18:59
ewanmelloro/ to you too18:59
openstackMeeting started Thu May  5 19:00:13 2011 UTC.  The chair is jbryce. Information about MeetBot at
openstackUseful Commands: #action #agreed #help #info #idea #link #topic.19:00
jbrycehello ttx and ewanmellor19:00
jbrycewho else is here?19:00
* ttx hopes it will stay short, so that he can go back to his vacation schedule :)19:00
jbryce - agenda items19:01
jbrycettx: we can try!19:01
ttxdo we have quorum ?19:01
*** anotherjesse has joined #openstack-meeting19:01
ttxthose extra three should help.19:02
jbrycewe've got 7 now, so we can get started19:02
jbryce#topic OpenStack project model19:02
*** openstack changes topic to "OpenStack project model"19:02
jbryceeday did you see the 2011 scope and charter doc that johnpur had put together previously?19:02
edayjbryce: I don't think so19:03
edayoh wait, perhaps I did see that19:03
anotherjessejbryce: I think it is missing identity/authorization19:04
jbrycei think the intent is similar to your definition of integrated/loose project-level controls19:04
*** namaqua has joined #openstack-meeting19:05
ttxI think we are a bit between Linux and Ubuntu in your descriptions.19:06
ttxi.e. a bit more controlled than Ubuntu.19:06
johnpurSomehow we need to maintain a degree of control, while allowing max transparency19:07
edayjbryce: yeah, it does seem to cover some of it19:07
ttxI can tell from experience that complete lack of upstream control is a problem, and full control is probably not desirable.19:07
johnpurwhat is not clear is who has decision making authority, to stop the endless community debate on everything19:07
ttxSo "tightly integrated with some level of control" :)19:07
jbrycettx: that's how i see it19:08
jbrycejohnpur: it depends on where the debate happens19:08
edayjohnpur: I don't think the control/integration properties will affect transparency in any way... but it is something we need to emphasize :)19:08
ttxPTLs have, with appeal to PPB. Unless for cross-project decisions, where PPB has ?19:08
jbrycethat should be the model now19:08
johnpurttx: yes for technical project issues19:08
ttxjohnpur: define non-technical project issues ?19:09
ttxsummit location ?19:09
johnpurdiscussion channels, splitting dev/user discussions, what forum software19:09
vishyIMO, that control will have to increase a bit over time.  It is clear that projects aren't going to agree on certain things that are really important to be standardized.  At some point we have to realize that for OpenStack to succeed as a complete platform, projects are going to have to stop trying to adopt the "best" solution, and start focusing on adopting the same solution.19:09
ttxvishy: I love you. Will you marry me ?19:10
johnpursummit location is in the purview of the community manager19:10
ttxttx: err. I'd have to divorce from my wife first.19:10
jbrycevishy: i agree19:10
edayvishy: this is exactly what I was trying to get at. should we be enforcing openstack-comon things, or just suggesting them? If so, I think the release changes we decided on at the summit may have been a step in the wrong direction (for integration)19:11
johnpurttx: many user/devops/operations issues will not be "technical"19:11
jbrycejohnpur: the non-technical items kind of bleed into the second agenda topic19:11
johnpurjust saying :)19:11
vishyeday: I think if we spend some time winning developers over to the idea that we need commonality will be a lot more successful than trying to enforce it19:12
edayif we really want a well integrated stack, this means a lot of common work needs to be done with existing projects, and the bar is raised quite a bit for new projects coming in (as far as integration requirements)19:13
vishys/ will/, it will19:13
vishyeday: +119:13
edayvishy: agree, that's more of an implementation detail :)19:13
ttxeday: At some point we may need to enforce it. Though I hope we can convince people to do the right thing instead.19:14
edayie, gflags, logging, common conf, etc all need a fair amount of work with existing projects if this is the direction we want to head, and afaict, no resources have been allocated for those19:14
jbrycewith the PTLs now in place and members of the policy board, i think it makes it easier to decide on things like what eric is talking about. if openstack-common should be suggested or required use for it's functionality. but i also think that requirements like that need to be phased in.19:14
ttxeday: to that effect, we need to state that commonality is a goal.19:14
anotherjessethings like API/auth/user models should be shared19:14
vishythe only thing we agreed on standardizing is a library for swift-init19:15
edayttx: yup, so far it's not been clear of suggestion vs requirement long-term19:15
jbrycein other words, maybe the PTLs come up with a recommendation about the end state, steps to get there, and their buy-in to lead their projects there over x releases19:15
ttxeday: I think it's required, though options parsing doesn't have the same prio as identity service, obviously19:15
vishyso does that mean notmyname, jaypipes, myself and eday should discuss it?19:16
edayvishy: I'm not a PTL :)19:16
johnpurwhen do we bring in the other projects to the discussion?19:16
*** dendro-afk is now known as dendrobates19:17
edayjohnpur: I don't think we should consider new projects until we decide how the current ones should play together and can offer the common requirements list19:17
jaypipesjohnpur: I've already brought keystone into the discussion to change their direction to utilize some of the common code around wsgi/daemon stuff.19:17
edayif thats the direction we want to go, anyways19:17
vishyeday: well i was thinking your input would be good as a representation of future projects trying to integrate19:17
jbrycevishy: i think we can all discuss it, but i think PTLs saying "yes, i believe this is the best thing for my project and i'm going to help us get there" is better than just a mandated policy19:17
johnpuri just want to give early guidance, if we can, so they don't get too far off track19:17
edayvishy: sure, would be happy to contribute where I can19:17
vishyi'm 100% for integration19:18
jbryceand considering that you all know what is going to be really painful or really useful, i think it makes more sense to have it generated from your side than from a theoretical policy side. make sense?19:18
johnpurjbryce: +119:18
edayjohnpur: agreed, but I'd hate to tell new projects to use something the PTLs will decide against later. we shoudl sort this out asap19:19
vishywe have had requests from the new projects for a kind of standard template for creating a new service.  Something like that would be incredibly useful...19:19
johnpurearly guidance, but too early?19:19
vishy(since we have 3+ starting right now)19:19
edayvishy: agreed19:19 it sounds like most of us who have been talking agree that we see openstack as a collection of tightly integrated projects with some level of control to ensure common behaviour is consistent, uses shared technology, etc.19:20
edayvishy: do you want to coordinate a ptl meeting to make a requirements list and plan existing projects to switch to them?19:20
jbryceeday: that was going to be my next suggestion. thanks!19:20
vishywhere are the resources coming from? Just hoping the community will step up here?19:21
jbrycei think the agreement on what will be commonly required is more important up front19:21
jbryceso we can give that early guidance to all the new projects that want to get started19:22
vishyok i'll send out an email and we'll try and get on a call/irc19:22
edayvishy: I'm guessing if we mail out the plan, someone will step up19:22
jbrycewe can figure out the process to get legacy projects switched over, but it would be great if we could keep new "legacy" projects from getting started19:22
jbryce#action vishy to schedule meeting with PTLs and other interested parties to discuss common requirements across projects19:23
jbrycevishy: can you just keep the whole PPB list in the loop on that as well so anyone can participate as they desire?19:23
jbryceany other discussion on this item?19:24
edaywe should also have a short, concise summary about what OpenStack is somewhere too (controlled+integrated)19:24
edaymaybe making a comparison to a model most people will understand (like Linux/subsystems)19:24
vishyshould i make the meeting public so everyone can come share their opinion? or keep it small?19:24
ttxI think we'll be exercising less control than the Linux model.19:25
ttxso "somewhat-controlled + tightly-integrated"19:25
edayttx: ok19:25
jbryceeday: do you want to draft something along those lines for us to look at?19:26
ewanmellorI'd rather see a description of our behaviour, than a reference to other people's models.  "The Linux model" is probably something that people understand.  "The Apache model" is so meaningless to me, it's not a good analogy.19:26
ewanmellorDitto "the Ubuntu model".19:26
ttxI can work with eday on that.19:27
ttxnext week.19:27
jbrycevishy: i could see it either way in terms of the group to discuss. might be nice to start smaller.19:27
jbryce#action ttx, eday to draft a short description of the OpenStack project model19:28
jbryce#topic shared OpenStack resources19:28
*** openstack changes topic to "shared OpenStack resources"19:28
vishyjbryce: agreed, we'll start small PTLs + anyone who is interested from the PPB, then expand to greater community when we have some consensus19:29
jbrycethis is an interesting one for us to discuss because i'm not sure the PPB actually has the responsibility for some of these resources19:29
edayjbryce: such as?19:30
jbryceuse forums for instance19:30
ttxI guess the problem is, the only Openstack-representative body currently is the PPB, so it makes sense that it's called to decide on things19:30
vishyjbryce: agreed, but i think we can request from the current owners that they attempt to be as open about the process as possible19:30
edaywell, if we host it, we do :) unofficial not at all19:30
jbrycei agree with both of you19:30
ttxeven though the trademark is Rackspace property.19:31
ttxI meant that last sentence as a question, not a statement.19:31
vishyso the two main points of contention are and launchpad groups right?19:31
jbrycebut i think where we actually have a mandate to set policy is around projects and their ability to be recognized and associated with openstack and make use of openstack resources19:32
ewanmellorI think we have a mandate to make decisions on too.  I know that Rackspace own the trademark and the domain, but they (Lew) have made it clear that they are holding these for the community, not as Rackspace assets.19:32
ewanmellorAnd we're the only representative board for the community.19:33
vishyewanmellor: I don't really want to make decisions on openstack.org19:33
jbrycewhen we say are we talking website or domain name19:33
ewanmellorvishy: Someone has to.19:33
edayso,,, etc, are PPB controlled, so to speak19:33
johnpurand we are back to my point19:33
ttxewanmellor: what about the advisory board, with PPB having veto power ?19:33
jbrycefor domain name, i think we should definitely lay out what the requirements are for a project to make use of subdomains, etc19:33
ewanmelloreday: Yes, that's what I mean.  I think we should own those.19:33
ttx(whenever the advisory board will be set up)19:34
edayand with the tightly integrated model, we should probably be requiring these assets be used for approved projects19:34
johnpurwe need a mechansm to make community driven decisions that rae not "technical" about the project19:34
ttxthen vishy doesn't have to care about it, until it causes problems.19:34
edaythe Q is what about non-approved projects. should go away for now?19:34
vishyewanmellor: if we do have some kind of control here (which I'm not totally sure about), I'd prefer to delegate management of these things to someone else19:34
jbrycevishy: +119:34
jbrycethierry is the release manager19:35
johnpurdelegate to whom?19:35
johnpurnot ttx19:35
jbrycehe makes decisions on feature freeze exceptions, etc19:35
vishyI think in general the current team is doing quite well19:35
jbrycehold on a minute and let me finish my point. = )19:35
vishyaside from a few screwups around too many people having access19:35
ewanmellorvishy: I didn't mean that we would have to be sysadmins.  We can certainly delegate.  But it's up to us to decide when a thing should be blessed with the OpenStack name, and when it shouldn't.19:35
jbrycebut ttx has a clearly defined process and it happens in public19:35
ewanmellorAnd by "decide" I mean "set clear, transparent policies, and then arbitrate if necessary"19:36
jbrycei think the issue with some of the non-technical assets is they are a little black box right now19:36
ttxjbryce: right, I can always be vetoed by the PPB. That's delegation.19:36
ttxjbryce: so we could also delegate some parts of user-oriented stuff19:37
ewanmellorI think we've seen recent instances where things have gone up, and people have objected, and had to come down again.  We should avoid that.  The PPB shouldn't have to veto after the fact.19:37
johnpuri believe we need to have community project management as a defined role, just as we have project release management19:37
jbrycejohnpur: agreed. and we have stephen in a community manager role right now who does some of these things19:38
edayewanmellor: agreed, and I think this will be fine once we have the process laid out. right now it's just been 'find ant on IRC and ask hime to change DNS', no rules around it19:38
johnpurjbryce: operative word is "some".19:38
*** dubsquared1 has joined #openstack-meeting19:38
ewanmelloreday: Yes.  Part of this is making sure that Ant and Stephen know that they shouldn't just do something because Joe Random asked them to.19:39
ewanmellorThey're being too helpful, which isn't normally a criticism ;-)19:39
johnpurewanmellor: agree19:39
edayI think we already have delegates, we just need to let them know whats ok or not :)19:39
ttxdeciding on a forums system in less than 24 hours, for example, is not.19:40
vishyand make sure that the process to contact them is public19:40
jbrycevishy: agreed. so i asked todd who has been managing the website to document it in the wiki.19:40
jbrycei think we basically need that for the other non-technical things too19:41
jbryceand then we need to make sure that they understand some basic ground rules for handing things out19:41
vishyjbryce: +1 that all seems reasonable to me19:42
jbrycebut the reality is that this all doesn't actually fall under what the ppb is supposed to be responsible for. the advisory board has the brand guidelines in their mandate. in the absence of the advisory board existing, though, i think it's fine for us to make some recommendations and encourage the people who are doing this stuff right now to follow some principles19:42
edaywe just need a comprehensive list and policies around each19:43
jbrycei think what we SHOULD define some specific policies around is how projects can make use of the openstack name, brand, logo, domains, etc19:43
jbryceas it stands we have 2 designations of projects--related and core--and we want to add a 3rd, incubated19:43
edayjbryce: there are some technical issues too, ie, should burrow get jobs on Jenkins? should it get for docs even though it's not a project yet? I ask because I'm sure reddwarf/atlas/etc will want the same soon19:43
johnpuri thought the use of the logo, etc. was already defined?19:44
*** troytoman-away is now known as troytoman19:44
jbrycejohnpur: to some extent, but not specifically around how projects can use it19:44
jbrycerelated is unofficial, no access to openstack resources19:44
ttxjbryce: if we had an openstack foundation, we could have a management board and a technical board. That would hepl in shoving non-technical stuff over the wall :)19:45
jbrycettx: = )19:45
jbryceincubated, i would say is an official recognition, you can began to use community infrastructure like jenkins, you can make some use of the brand, but cannot present yourself as an "openstack project"19:45
jbrycealso, i think incubated projects should live under a single subdomain like or and inside of a separate launchpad group19:46
johnpurttx: i am suggesting an intermediate step, separating management of the OpenStack project from the technical realities of building the component the re19:46
jbrycecore projects are in the main openstack launchpad group, can have their own subdomains, get full use of the openstack name and brand, their own sections on, etc.19:46
edayjbryce: my issue with that is when the graduate to a real project, links/groups/etc need to change everywhere, and it may be pretty mature19:46
edayjbryce: I'd rather have them start in their final place, but with a giant banner or somethings saying it's incubated19:47
jbrycegood point19:47
edayit's easy to drop dns entries if things don't work out :)19:47
jbrycethen i would say we require some kind of "incubation" badge on whatever is hosted on those subdomains19:48
ttxI think that's workable, with the intermediate "incubation" state19:50
ttxthat allows us to control what gets those resources19:50
vishyjbryce: +1 on all points from me19:51
jbryceok...i will collect those designations in a doc and send it around for everyone to review in a little more coherent format19:51
jbryce#action jbryce to publish project designations and usage rights19:51
ttxyou just can't be called "OpenStack Blah" until you get core status19:51
ttxyou only live by your code name until then.19:52
jbryce#action jbryce to track down community asset managers and get published processes for use19:52
edayI guess I need to update burrow :)19:52
johnpurthere are implications on packaging and distribution as well between core and incubated projects19:52
ttxright -- for example it enters the realm of my release management only when it's core.19:53
johnpurttx: correct19:53
ttxbut my point is that "Burrow" becomes "OpenStack Simple Queue (Burrow)" when it's in core.19:53
vishyI think that the goal is to strive to help the incubated projects join the automated packaging and testing systems we have in place, and help make it easy to install them as an add-on to core components19:53
ttxnext topic ?19:55
jbrycewe can definitely help them and they should make use of automated systems, but ttx should not be bird-dogging them to get their release out19:55
johnpurvishy: agree, but the responsibility is on the project team to produce their packages and make them available19:55
jbryce#topic incubation application process19:55
*** openstack changes topic to "incubation application process"19:55
jbrycedid anyone get a chance to review prior to the meeting?19:56
ttxjbryce: maybe add "language" to
jbrycegood one19:56
johnpurjbryce: yes. question on a ppb sponsor?19:56
johnpurwhat is the responsibility of the sponsor to the incubated project?19:57
jbryceyour guess is as good as mine19:57
jbryce= )19:58
jbrycepoint of contact, helping hand, connect them to the right people to use jenkins19:58
ttxan advocate on the PPB ,maybe19:58
johnpurdoes a project need to get a ppb sponsor *prior* to being submitted, ie getting someone to advocate for the new project?19:58
ttxthough I'd prefer them to present their case directly19:58
vishyjbryce: I did, i didn't see anything that i thought needed to be changed19:59
jbrycettx: i like direct access too19:59
edayppb sponsor=SPoF, why not just keep it general?19:59
ttxyes, I don't really like the sponsor idea19:59
ttxI mean, the PTL for the incubated project might become a PPB regular in a near future, it's good to have him present in the meeting to defend his case ?19:59
jbryceit's not about the ppb meetings as much as it is making sure they have a point of contact in the community who can help them find their way around20:00
johnpuri don't disagree with the intent, I have been trying to do this for the new RAX projects, anotherjesse has done it for Keystone, etc.20:00
jbryceand give them advice and direction on things like...wording announcement emails20:00
ttxIf they ask for incubation, they should have a pretty good idea who to network with ?20:00
ttxI mean, if not as such our desk door was closed all day :)20:01
jbrycei think if it's just general access, they're less likely to reach out20:01
edayit seems  wiki page should suffice for most things a PPB sponsor would provide (and if not, just provide an email to the ppb list so anyone can respond)20:02
ttxI just don't want a bad sponsor to drown a project -- they should be able to contact any of us20:02
ttxand might engage more freely in the absence of a designated person20:03
jbrycejohnpur: what do you think?20:03
johnpuri'm on the fence20:03
jbryceme too20:03
ttxi.e. the PPb in it's entirety is the sponsor.20:03
jbrycei can see that side, but having been involved with some of the groups who are getting started, i think that having someone they are more in direct contact with would be more effective for them20:04
edayif we have one, then we need mroe rules around what happens when a PPB drops off before the project goes core. etc20:04
ttxI don't really mind either way... just saying that eday has a point with his SpoF comment20:04
vishyshould we change it to "the project may request a sponsor from the ppb"?20:04
jbrycei think the asf actually gives you multiple mentors20:04
johnpuri think the role is needed, the new projects have a lot of potential pitfalls they can trip over (as we saw with the RAX proposed projects). They do need guidance from someone who is plugged into the curren state of the project.20:04
jbrycemaybe that's a better approach20:05
edayjohnpur: definitely agree, it just seems the entire PPB should provide that role, not just one person.20:05
ttxeday: maybe asking everyone to follow the state of every incubated project is a bit too much20:06
edayin any event, seems like a pretty minor point20:06
jbrycettx: i concur20:06
ttxright, we can try and change if that doesn't work so well20:06
johnpurthe key point is that the project needs to reach out at the right time, before the bloodletting commences from the outraged villagers :)20:06
vishydoes the project weigh as much as a duck?20:07
jbrycei'm going to change it to may assign one or more mentors for the project20:07
jbrycethen we can try it out and see how it goes20:08
johnpurvishy: exactly!20:08
ttxjbryce: +120:08
jbryceeric obviously hasn't needed a mentor or sponsor for burrow20:08
vishyjbryce: +120:08
jbrycesome projects may need more or less guidance20:08
johnpurjbryce: mentor is a much better term than sponsor20:08
jbrycethis thing is running long...sorry guys20:08
ttxnext topic, then :)20:08
edaydo I get one when I apply for incubation? :)20:09
vishyeday's mentor is... eday20:09
jbryce#topic forums.openstack.org20:09
*** openstack changes topic to ""20:09
jbrycementor recursion20:09
* jbryce opens can of worms20:09
johnpurjbryce: not exactly true on burrow. we had much conversation prior to starting the project. eday was "sponsored" he just didn't need guidance on the open source community part :)20:10
jbrycei honestly am not even sure where we are with the issue20:10
vishyfirst question: do we want to make the decision about forums? or do we delegate it?20:10
edaywe have a Q&A site now on LP. Is it really insufficient?20:10
ttxeday: I think it's a bit insufficient, yes.20:10
vishyeday: IMO LP Q&A fails in basically every respect20:11
johnpurthe reality is that a forum site (or 2 or 3) is going to be set up regardless of the official stance the ppb takes.20:11
ttxeday: I think if we want to do Q&A, we should do it properly :)20:11
vishyeday: no community building, hard to search, no reputation, etc..20:11
jbrycejohnpur: +120:11
jbryceforums are also different20:11
edayttx: ok, so the consensus seems to be around stackexchange style forums, so lets set one up on forms. and call it good?20:11
johnpurthe question we need to talk about is how we want the "official" communication channels to be set up20:11
jbrycewe may not read them or use them but other people will20:11
vishyjohnpur: I think that is true, so we should probably have an official forum20:11
jbryceso the question really is the official nature20:11
edayand by this I mean delegate to Jordan :)20:11
johnpureday: or chad keck20:12
ttxeday: I think Jordan wants a discussion-oriented forums system. The devs hate that, but kinda like the Q&A-oriented forums systems20:12
johnpurttx: do the devs want to abandon Answers in favor of Q&A?20:13
jbrycei have to say that i've heard requests for discussion oriented forums from the non-dev community members as well20:13
vishyi would prefer to abandon answers for stackexchange style q&a20:14
ttxjohnpur: I don't like to talk for everyone -- I'd do, if we were eto use OSQA or StackExchange20:14
jbryceit seems like there was general consensus on the list to that approach as well20:14
vishyand I'm ok with forums as well, especially if we put a link to the answers site as a sticky :)20:14
ewanmellorCould we set up a poll?  It seems like the only way that we'll get an agreement that everyone will live with.20:15
johnpurmy 2 cents: suggest we transition Answers to OSQA for dev oriecnted discussion, set up "" for sysadmin/devops/users. Delegate moderation of the forums to jordan and chad.20:15
ttxIf we do discussion forums, the devs will ignore them. Maybe that's not too bad. If we do Q&A forums, then the devs would more likely participate to them...20:15
ewanmellorIt doesn't seem like the devs should care much -- this is a user-to-user thing, isn't it?20:15
jbrycethe problem is that a poll sent out to the current openstack list will be skewed like the mailing list discussion20:15
jbryceewanmellor: yes20:15
johnpurUse OSQA because it is open source and fits with our project values.20:16
jbrycei'm fine with johnpur's proposal20:16
ewanmellorjbryce: No, the mailing list discussion is full of mad screaming.  A poll would be out of my way and much quieter!20:16
jbryceewanmellor: good point!20:16
vishyFWIW, we got denied by stackexchange, so osqa!20:17
johnpuri think this is an area we can exert a little community leadership20:17
ewanmellorjbryce: And I think that if we did a poll, and left it up for a week, say, then everyone has plenty of time to make a decision.20:17
vishyjohnpur +120:17
ttxI'm ok with johnpur's suggestion, though I would not rush the OSQA transition. We are not in a hurry, I think20:17
vishypeople can always link from forums to QA20:17
johnpurttx: agree20:17 becomes discussion forums20:17
ttxThat's dilution, but I guess we can live with that20:17
ewanmellorWouldn't the right open source thing be to write our own forum software from scratch, because we're smart and can do it better than everyone else?20:18
jbrycejordan and chad and others can moderate20:18
ttxewanmellor: we are indeed !20:18
jbryceand then we will move questions to OSQA at some point?20:18
jbryceis that where we've landed?20:18
johnpurjbryce: +120:18
vishyewanmellor: +1 OpenStack Forums (code name NIH)20:18
*** rnirmal has joined #openstack-meeting20:18
ttxhaving a Q&A system should limit people asking technical questions on the discussions site.20:18
ewanmellorDid you just make a decision on this with a consensus of three people?20:19
jbrycei think it was actually only 220:20
jbryceno decision was made20:20
vishyewanmellor: what would the options on the poll be?  I'm not sure if it will solve anything?20:20
johnpurewanmellor: good point... ppb members please vote... now.20:20
ewanmellorI think, given that there are emails coming in about this even now, that us making a decision would be heavy-handed.20:21
ttxewanmellor: you mean we could let more time pass for discussion before making that decision ?20:21
johnpuri think the emails are rehashing and posturing. i havent seen a new argument today.20:22
jbrycethere are going to be forums...the question is really do they get to use forums.openstack.org20:22
ttxI'd fully support that20:22
ewanmellorPoll options would be: OSQA, the three least horrible web-BBS technologies, and stay with Launchpad.20:22
vishyI think going with both will make 95% of people happy20:22
jbryceewanmellor: i think one of the problems is that somehow this got pushed into our lap and we were supposed to have a stance on it today20:23
johnpur95% is above our target goal of 92%.20:23
alekibangoewanmellor: which web-bbs technologies are not horrible? lol20:23
ttxjbryce: we can decide to decide next week, though.20:23
ewanmellorttx: +120:23
vishysomeone is going to make a forum anyway...20:24
johnpurttx: my view is that waiting is not going to change the outcome.20:24
alekibangojohnpur: +120:24
* ttx is returning to his well-deserved vacation time in ~6 minutes.20:24
jbrycei don't think the list discussion is going to result in agreement20:24
ttxso I'll stop arguing :)20:24
ewanmellorjbryce: I agree with that!20:24
* johnpur wants to be ttx and on vacation20:24
jbrycelet's take an actual vote and see if we're just completely deadlocked here, then we can push it back.20:25
ewanmellor-0: Don't care enough to veto.20:25
jbrycewho thinks we should recommend that some piece of discussion forum software gets to live forums.openstack.org20:25
ewanmellor+1 for whatever it is appearing on  That's fine.20:26
alekibango+ 0.1 (... i am not in PPB:)20:26
jbrycei think we're all running low on gas20:26
jbrycethis has been another marathon meeting20:26
alekibangowould be 120:26
ttxcould we push the last item to next week ?20:26
edaylets put it on it's own VM though, when it gets hacked, I don't want the other domains to go down :)20:26
jbrycettx: yes20:27
jbrycesorry for bleeding into more of your vacation time20:27
ttxeday: +100020:27
JordanRinkeCan delegate out to Stephen and me to work out with the commmunity what bbs goes there?20:27
ttxjbryce: my fault, I could have ignored it completely :)20:27
jbrycewell add an extra 1.5 hours on the end20:27
johnpurJordan: work with Chad Keck as well.20:27
edayJordanRinke: sounds good to me20:27
jbryceJordanRinke: i have no desire to dictate forum software20:28
ttxJordanRinke: no problem with that.20:28
JordanRinkeExcellent, thx all20:28
jbrycelet's end this thing20:28
jbrycethank you guys!20:28
edayyay for delegation!20:28
*** openstack changes topic to "Openstack Meetings: | Minutes:"20:29
openstackMeeting ended Thu May  5 20:28:58 2011 UTC.  Information about MeetBot at . (v 0.1.4)20:29
openstackMinutes (text):
*** eday has left #openstack-meeting20:29
*** ewanmellor has quit IRC20:29
alekibangoi love those20:31
alekibangowe might end up writing our own lol20:32
*** anotherjesse has quit IRC20:32
*** dendrobates is now known as dendro-afk20:33
*** troytoman is now known as troytoman-away20:34
vishymonty, soren: here?20:46
*** larzy has left #openstack-meeting20:52
*** anotherjesse has joined #openstack-meeting21:10
*** dubsquared1 has left #openstack-meeting21:42
*** namaqua has quit IRC21:45
*** rnirmal has quit IRC21:56
*** edconzel has quit IRC22:01
*** mray has quit IRC22:14
*** dendro-afk is now known as dendrobates22:17
*** dendrobates is now known as dendro-afk22:37
*** jkoelker has quit IRC22:45
*** blamar has joined #openstack-meeting23:03
*** adjohn has joined #openstack-meeting23:35
*** dendro-afk is now known as dendrobates23:39
*** dragondm has quit IRC23:42
*** adjohn has quit IRC23:59

Generated by 2.14.0 by Marius Gedminas - find it at!