19:02:59 <mtaylor> #startmeeting
19:03:00 <openstack> Meeting started Tue Jan 10 19:02:59 2012 UTC.  The chair is mtaylor. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:03:01 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic.
19:03:46 <mtaylor> ayoung: do we need to add you to keystone-drivers team?
19:04:09 <jeblair> proposed item: vishy suggested:
19:04:10 <jeblair> 22:15 < vishy> jeblair: also is it possible to just put a note on the merge prop instead of blocking propsing  for people that haven't signed the cla
19:04:24 <mtaylor> hrm
19:04:53 <mtaylor> I think honestly we might want to get a legal opinion from someone ...
19:05:17 <jeblair> here's why i don't want to do it that way:
19:05:22 <mtaylor> AIUI, the CLA is a requirement for contributing, and it could be argued that uploading that patch to gerrit is a contribution, whether it gets merged or not
19:05:39 <jeblair> 1) it's very clear upfront what is required _before_ you submit something
19:06:04 <jeblair> (what you said is a similar thought, i think, but even stronger)
19:06:27 <mtaylor> should we get a legal opinion from someone who isn't NAL?
19:06:33 <jeblair> 2) it's an extra hurdle people have to go through before they start using our resources
19:06:59 <jeblair> ie, i feel a little more comfortable about running pre-commit jobs on things if everyone who uploads has gone through the cla process
19:07:13 <mtaylor> that's a good point too
19:07:25 <mtaylor> and we don't need a lawyer for that one
19:07:46 <mtaylor> also - it's only really a bitch right now while we're cleaning up the cla membership for folks who are already contributors
19:07:48 <jeblair> i mean, it's no guarantee about their intentions, but it's obvious cost prohibitive to run a spambot operation using our jenkins slaves if you have to commit purjury every time you want to do it.
19:07:56 <jeblair> i agree with that as well.
19:08:00 <mtaylor> I doubt it's going to be ongoingly annyoing
19:08:28 <jeblair> 3) other large projects with CLAs get them out of the way up-front; eg, android
19:09:13 <mtaylor> jeblair: also, from that timestamp - you were working late last night :)
19:09:16 <jeblair> 3b) the built in gerrit method of handling CLAs, which I think is fairly friendly, and we should move to if we want to keep the cla in the long term (perhaps after the lawyers have a break after setting up the foundation) also does CLA upfront
19:09:30 <jeblair> (end of list)
19:09:39 * mtaylor agrees with jeblair's list
19:09:56 <jeblair> yeah, devstack gating job wasn't behaving well:
19:09:57 <jeblair> https://review.openstack.org/#change,2929
19:10:02 <jeblair> that should help
19:10:24 <jeblair> (or more specifically, RS public cloud wasn't behaving well, that should work around the problem)
19:11:10 <mtaylor> looks good
19:11:51 <mtaylor> so - speaking of requests ... we got a request from notmyname which I filed a bug for
19:12:10 <mtaylor> which is to not have votes generate automatic comment text
19:13:00 <mtaylor> I think most specifically -1 and -2 were the ones that initially generated annoyance, and the original request was to change their text ... but then as the conversation went on, the request changed to just having them not generate text at all
19:13:34 <jeblair> for jenkins, or for people?
19:14:03 <mtaylor> for people
19:14:06 <mtaylor> 10:54 <notmyname> " I don't really understand why I need/want an automated comment for my comment field I'm filling out."
19:15:03 <jeblair> does he just not want email, or also doesn't want it to show up in the web history?
19:15:36 <mtaylor> I think he doesn't want the text "I would prefer you didn't submit this" to be pre-pended to the comments that the reviewer actually writes
19:16:00 <mtaylor> so that the review would actually just be the words that the reviewer typed
19:16:49 <jeblair> ok.  so those serve an important function in that they record the voting history for a review
19:17:09 <jeblair> without them, you would not be able to see that someone had, say, voted -1, then voted +2 later
19:17:22 <mtaylor> isn't that recorded by the actual +1/-1 vote? ... AH
19:17:25 <mtaylor> ok, gotcha
19:17:34 <jeblair> however, that text is easily configurable.
19:18:26 <jeblair> http://paste.openstack.org/show/4214/
19:18:40 <jeblair> ^ current values.
19:18:40 <uvirtbot> jeblair: Error: "current" is not a valid command.
19:19:40 <jeblair> mtaylor: bug number?
19:19:49 <mtaylor> jeblair: ok - I'll try to get some feedback on what "better" values might be
19:20:16 <mtaylor> 914431
19:20:31 <mtaylor> #link https://bugs.launchpad.net/openstack-ci/+bug/914431
19:20:32 <uvirtbot> Launchpad bug 914431 in openstack-ci "remove automatic comment text from reviews" [Low,Triaged]
19:22:17 <jeblair> i updated the bug with my comments here.
19:22:27 <mtaylor> awesome
19:22:39 <mtaylor> #topic New Team Member
19:22:49 <mtaylor> for anyone who happens to be lurking - I'd like to welcome LinuxJedi to the team
19:22:50 <heckj> ?? cool
19:22:59 <LinuxJedi> hi :)
19:23:07 <LinuxJedi> mtaylor: thanks :)
19:23:16 <jeblair> welcome!
19:23:18 <heckj> welcome!
19:23:35 <mtaylor> LinuxJedi and I worked together back at MySQL AB, and then as core devs on Drizzle. he's awesome, you're all going to love him
19:24:22 <LinuxJedi> lol, not sure about that, but I'll do my best to bring some awesomeness to our team :)
19:24:29 <mtaylor> :)
19:25:01 <mtaylor> and on that note ...
19:25:10 <mtaylor> #topic expiring stale reviews
19:25:24 <heckj> +1-00!
19:25:25 <LinuxJedi> ah, my new pet script
19:25:53 <mtaylor> LinuxJedi has written some code to expire stale reviews ... we're currently going with stale == no activity for 2 weeks OR bad review plus no activity for 1 week
19:26:02 <mtaylor> anybody have any dissent on those timeframes?
19:26:20 <jeblair> i think the script is looking great so far.  i'm not sure if i have dissent, as such...
19:26:22 <jeblair> but i am worried.
19:26:47 <LinuxJedi> jeblair: sorry about expiring almost everything on the dev site ;)
19:26:50 <heckj> that timeframe seems pretty reasonable to me - there's a lot of "old crap" out there now.
19:26:50 <jeblair> mtaylor: i'm not sure our patches to quantum will land in that timeframe, for instance.
19:27:00 <jeblair> LinuxJedi: no! great! that saves me so much clicking!
19:27:14 <heckj> I thought one of you hit the mailing list and received some consensus on that timing too...
19:27:18 <mtaylor> jeblair: no?
19:27:27 <mtaylor> heckj: did we?
19:27:33 <jeblair> but since you can always 'un-abandon' a review, so it's not like too much is lost.
19:27:48 <LinuxJedi> mtaylor: jaypipes updated the review to say he was happy now I think
19:27:51 <heckj> Am i recalling just an IRC conversation then?
19:28:02 <mtaylor> heckj: might have been IRC convo
19:28:02 <jaypipes> LinuxJedi: I'm always happy with you :)
19:28:12 <LinuxJedi> lol! :)
19:28:48 <mtaylor> jeblair: I agree re: un-abandon - I think it's safe enough to roll out and then deal with carnage as it happens, yeah/
19:28:54 <jeblair> but right now, gerrit does have the useful feature of showing me patches that i've submitted that people are slacking off on.  in the future, people will have to watch out for reviews being expired and take action.
19:29:36 <LinuxJedi> mtaylor: I need to do one more minor update to it today anyway (info logging goes to stdout at the moment so you will get lots of cron mail) so if timing needs to be changed then tonight is a good time to do it
19:29:39 <mtaylor> jeblair: we could add a new status :)
19:29:58 <mtaylor> jeblair: "stale" :)
19:30:16 <heckj> can we hide that state by default?
19:30:35 <mtaylor> LinuxJedi: awesome. I think 2 weeks/1 week is fine for a start
19:30:48 <jeblair> mtaylor is being slightly facetious -- adding a state is hard, but we're looking into adding "work in progress" nonetheless.
19:30:51 <mtaylor> heckj: oh, I was mostly kidding about that - jeblair is already working on adding a "work in progress" state
19:30:57 <mtaylor> jinx
19:31:03 <jeblair> i think stale is overkill, abandon is probably the right thing here.
19:31:04 <heckj> Oh - okay :-)
19:31:04 <LinuxJedi> mtaylor: cool, very easy to tune later.  I can always update the script at a later date to use a config file to set it
19:31:18 <mtaylor> LinuxJedi: good point
19:31:45 <mtaylor> LinuxJedi: hey, as soon as you're happy with that script you get to write a puppet config patch! yay for first-week learning curve! :)
19:32:17 <LinuxJedi> mtaylor: yep, I was just about to look at that before I had to leave earlier, will try and get a peak tonight
19:33:11 <mtaylor> LinuxJedi: you're going to love it
19:33:33 <LinuxJedi> mtaylor: and learning a new programming language to write the thing was more learning that I expected to do already this week, let alone everything else ;)
19:34:04 * mtaylor has a new goal - make LinuxJedi learn a new langauge every week ...
19:34:12 <jeblair> sounds groovy.
19:34:16 <LinuxJedi> haha :)
19:34:18 <mtaylor> hahah
19:34:40 * LinuxJedi re-writes mtaylor's scripts in brainf**k
19:34:54 <mtaylor> awesome
19:35:49 <mtaylor> jaypipes: jeblair  just filed a bug for it, but we're working on your glance config files for S3 testing and whatnot
19:36:39 <jaypipes> mtaylor: rock on brotherman.
19:37:28 <mtaylor> I think that's all we've got for this week
19:37:49 <mtaylor> next week jeblair and I will be in Australia, where this meeting starts at 6am
19:37:55 <mtaylor> so I'm not promising that it's going to happen :)
19:39:03 <mtaylor> #endmeeting