17:10:01 <Oxf13> #startmeeting Fedora Release Engineering
17:10:01 <zodbot> Meeting started Fri May 28 17:10:01 2010 UTC.  The chair is Oxf13. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:10:01 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:10:06 <Oxf13> #meetingname fedora-releng
17:10:06 <zodbot> The meeting name has been set to 'fedora-releng'
17:10:10 <Oxf13> #topic roll call
17:10:32 <Oxf13> ping notting jwb lmacken wwoods spot poelcat rdieter dgilmore
17:11:07 <rdieter> here
17:11:26 <dgilmore> present
17:11:51 * spot is here
17:11:55 * notting is here
17:13:28 <Oxf13> Ok.
17:13:42 <Oxf13> #info present are rdieter, dgilmore, spot, notting, and Oxf13
17:13:49 <Oxf13> #topic Fedora 13 Release
17:13:54 <Oxf13> so yeah, that happened.
17:14:07 <Oxf13> I thought it went pretty smooth, not a lot of last minute fires.
17:14:21 <dgilmore> seemed to go very smoothly
17:14:39 <Oxf13> of course, slipping a week for a non-bug just gave us extra time...
17:14:42 <notting> there was the spins thing, but that was a different process issue.
17:14:55 <Oxf13> yeah
17:15:11 <Oxf13> #info A spin was missed due to process breakdown, but has been generated and posted.
17:15:56 <Oxf13> I don't really have any plans to make adjustments to how or schedule and freezes and stuff work this release.  It seemed pretty good in F13, and I want to give F14 a chance to show a pattern or not
17:16:12 <Oxf13> #info no real schedule or process changes planned for F14
17:16:41 <Oxf13> The schedule is up, just need to make some milestones and file some tickets for them.
17:16:58 <dgilmore> Oxf13: id like to finsih writting a proposal for blockers by flags and implement for F-14
17:17:07 <Oxf13> ah yes
17:17:28 <Oxf13> #info One process change would be using bugzilla flags for blocker management, needs more planning and a proposal posted.
17:17:42 <Oxf13> dgilmore: we'll cover that topic next I think.
17:17:55 <Oxf13> Anybody with anything else to say about the Fedora 13 release?
17:17:56 <dgilmore> ok
17:18:30 <Oxf13> one interesting metric, my chiropractor told me that this time last year after the F11 release I was a real mess stress wise.  This year I'm in far better condition.  So whatever we did, keep doing it (:
17:19:21 <Oxf13> alright, moving on.
17:19:30 <Oxf13> #topic Blockers via Flags
17:19:46 <Oxf13> #info I asked dgilmore to follow up on my mailing list proposal to use flags to manage blockers
17:19:52 <Oxf13> dgilmore: take it away...
17:19:55 <dgilmore> ok
17:20:11 <dgilmore> so we have a few options available to us to use flags
17:20:24 <dgilmore> the one that makes most sense to me  is
17:20:29 <dgilmore> use 3 flags
17:20:36 <dgilmore> one for dev one qa and one releng
17:20:52 <dgilmore> without all 3 being + a bug is not considered a blocker
17:21:12 <dgilmore> the dev would be set by the bug owner who is saying yes its a blocker and yes ill fix it
17:21:23 <dgilmore> one from qa saying yes its a blocker and ill test it
17:21:38 <dgilmore> and one from releng saying yes its a blocker and we will take the fix
17:21:51 <notting> i'm not sure a releng ack is needed in such a situation
17:22:14 <notting> if the process is clear enough, it should be obvious
17:22:17 <dgilmore> notting: its mostly to make sure we are on board and get the fix included
17:22:34 <dgilmore> we could drop to 2 dev and qa
17:22:58 <dgilmore> but it would mean that releng needs to track the qa flags  to include things
17:23:08 <dgilmore> and if we are ok with that then thats fine
17:23:35 * wwoods lurking
17:23:47 <Oxf13> I suggested 3 acks, since that's essentially what we've been doing in the blocker bug meetings already
17:23:51 <dgilmore> we will need to sync some groups from fas to bugzilla
17:24:08 <Oxf13> I'm not opposed to dropping releng ack
17:24:13 <notting> dgilmore: surely releng would just include things that ended up with a final 'blocker+' flag
17:24:20 <notting> after whatever acks are granted.
17:25:04 <Oxf13> notting: I think the issue is whether or not releng gets an opinion on if something is a blocker or not.  Right now we do, and if we go with flags and don't have a releng flag, then we would no longer have an opinion.
17:25:21 <Oxf13> however I think that's fine, if we have an override or some ability to re-visit an issue
17:26:04 <dgilmore> notting: releng kinda acts like pm now and  would take the role of the pm flag in other land
17:26:16 <notting> right. i'm just saying that it's sort of an odd place for that to coalesce. i think we would have an opinion as 'experienced pepole in fedora', not necessarily 'as a member of release engineering'
17:26:23 <notting> dgilmore: so we're a bot?
17:26:39 <Oxf13> yeah I don't want to duplicate the silly pm bot thing.
17:26:54 <dgilmore> notting: yeah there will be no bots
17:27:24 <notting> ... except maybe one that tracks bugs attached to F14Blocker or blocker?, and sets the appropriate ? flags for devel/qe?
17:27:43 <dgilmore> the only bot i could see doing is setting all the blocker flags to ? if one is set and the others are not
17:28:28 <dgilmore> notting: we wont use a F14Blocker keyword
17:28:55 <notting> ok. given the bug exists now, we'll probably have to at least keep an eye on it to redirect things
17:29:12 <dgilmore> though i guess we could monitor for it for backwards compatability
17:29:15 <Oxf13> yeah
17:29:22 <Oxf13> there will have to be some sort of transition help
17:29:29 <dgilmore> but the reporter would set the 3 flags to ?
17:29:49 <dgilmore> then it would be considered a blocker when it gets 3 +
17:30:14 <dgilmore> or 2 ? + transistions if we want to go that route
17:30:54 <Oxf13> hrm, I don't think we'd want the reporter to set 3 flags.  The reporter would set the one blocker flag to ?, and that would make the ack flags show up pre-set to ?
17:30:57 <dgilmore> right now i just need to write it all down as a concrete proposal
17:31:01 <notting> dgilmore: just clarifying since they aren't here - are you working with at least one of jlaska/wwoods/adamw on the proposal?
17:31:01 <Oxf13> or rather the 2
17:31:46 <dgilmore> notting: not yet, ive just spoken with bugzilla admins on what our options are
17:31:57 * adamw perks up
17:32:03 <notting> ok
17:32:09 <dgilmore> I was goingto run it by them when we get something concrete
17:32:17 <adamw> sounds fine. *perks down again
17:32:22 <dgilmore> make sure they were ok with it
17:32:27 <notting> cool. just with my fesco hat on, i think it needs at least signoff from QE before it goes to fesco.
17:32:46 <Oxf13> I can't recall if percocet is an upper or a downer...
17:33:48 <dgilmore> downer
17:34:04 <Oxf13> alright, thanks for the update dgilmore
17:34:12 <Oxf13> anything else on this topic before we move in?
17:34:16 <dgilmore> nope
17:34:57 <Oxf13> #topic dist-git
17:35:11 <Oxf13> now that the 13 fire is out, I can switch gears and start working on dist-git again.
17:35:26 <dgilmore> do we want to try land it for F-14
17:35:30 <Oxf13> I honestly have little idea of where I left this off, and need to spend a day or so catching back up and getting my brain back in gear.
17:35:30 <dgilmore> or F-15?
17:35:42 <Oxf13> dgilmore: if I can get things in place before the branch of F-14, I"d like to do it for F-14
17:35:50 <Oxf13> but if I miss the branch date, we'll have to again wait.
17:36:23 <notting> i'd say if we can't land this in a month or so, we postpone.
17:37:01 <notting> Oxf13: is at least the koji side done? (building from git tag)
17:37:40 * poelcat present
17:37:44 <Oxf13> notting: sortof.  THere was code already in koji for building from git, it just needs to be modified to work with how dist-git is laid out
17:37:53 <Oxf13> #info poelcat is present
17:38:08 <Oxf13> notting: I spent most of last release tuning how the git repo would look, and then working on fedpkg
17:38:17 <Oxf13> fedpkg is done enough that we can do the koji build from work
17:38:19 <dgilmore> notting: main thing koji needs is to make some commands configurable
17:39:22 <dgilmore> koji right now expects a makefile to downlaod the source from lookaside cache
17:40:13 <dgilmore> we need to make that configuable  so we can have tell koji to run the fedpkg command to do it
17:40:27 <dgilmore> but not break backwards compatability
17:41:04 <dgilmore> and tell it to not try and checkout the common module
17:42:10 <Oxf13> alright
17:42:18 <Oxf13> #info weekly reports of dist-git progress will be made
17:42:31 <Oxf13> #info concentrating next on getting koji to build from a dist-git repo
17:42:48 <Oxf13> #info Targeting completion before F-14 branch, but may have to wait until post-F-14
17:42:55 <Oxf13> #topic open floor
17:43:01 <Oxf13> I got nothing else to talk about...
17:45:26 <poelcat> dgilmore: i'm glad to help with the bugzilla/flags stuff
17:45:51 <poelcat> the other PMs i work with configure the bot, etc. (though i realize we may not want that here)
17:47:05 <Oxf13> poelcat: we may need some form of bot, so your help and experience will be useful
17:47:39 <notting> poelcat: "Since this issue was entered in bugzilla, the blocker flag has been set to ? to ensure that it is properly evaluated for this release." <shoots self>
17:48:06 <poelcat> notting: i'll make sure that is possible ;-)
17:48:28 <poelcat> i'm also glad to help draw process flow, etc. I like that kind of stuff :)
17:51:38 <Oxf13> ok, thanks for coming all, I'm going to wrap it up.
17:51:40 <Oxf13> #endmeeting