21:02:35 <mikal> #startmeeting nova 21:02:36 <openstack> Meeting started Thu Aug 14 21:02:35 2014 UTC and is due to finish in 60 minutes. The chair is mikal. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:37 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:39 <openstack> The meeting name has been set to 'nova' 21:02:41 <dansmith> o/ 21:02:43 <mriedem> hi 21:02:50 <adrian_otto> o/ 21:02:53 <mikal> Oh, we say hello in the logged bit do we? 21:02:57 <funzo> o/ 21:02:57 <mikal> :P 21:03:03 <dansmith> mikal: yes :) 21:03:06 <mriedem> it doesn't count otherwise 21:03:08 <mikal> Heh 21:03:17 <mikal> The agenda is at https://wiki.openstack.org/wiki/Meetings/Nova#Agenda_for_next_meeting as always 21:03:18 <mrda> o/ 21:03:28 <mikal> #topic Juno-3 status, Feature Proposal Freeze August 21st, September 4th Juno-3, FF and SF 21:03:34 <mikal> #note mikal sent an email out with review "priorities" an hour or so ago 21:03:38 <mikal> #link http://lists.openstack.org/pipermail/openstack-dev/2014-August/043098.html 21:03:50 <mikal> dansmith has already replied with some status updates there 21:04:03 * jogo walks in late 21:04:05 <mikal> But the punch line is that we need to spend some time reviewing important BPs for j-3 21:04:16 <mikal> The ironic one seems to be under control 21:04:24 <mikal> And the vmware one is suffering under the reign of minesweeper 21:04:39 <mikal> But helping out with reviews on those would be much appreciated 21:04:43 <dansmith> mrda and I are slaughtering the ironic one 21:04:47 <mriedem> i think the CI system formerly known as minesweeper is fixed 21:04:55 <mikal> dansmith: in a good way? 21:05:01 <dansmith> mikal: yes, like "killing it" 21:05:03 <mrda> thanks dansmith, your reviewing prowess is greatly appreciated! 21:05:04 <dansmith> hmm 21:05:09 <dansmith> yes, in a good way 21:05:17 <dansmith> guess I need some non-violent references 21:05:18 <mikal> dansmith: do you need another core to hand hold with you? 21:05:38 <tjones> im late i joined the open-tack meeting but no one was there 21:05:38 <dansmith> mikal: I can stir up some reviews once I think it's really ready if you want 21:05:49 <mikal> dansmith: that would work 21:05:49 * mriedem resists dance in gym class w/ sweaty hands joke 21:05:52 <dansmith> mikal: but if anyone else wants to jump on that's cool too 21:06:17 <mikal> Yeah, I think dansmith is doing a good job, but helping hijm out would be cool too 21:06:25 <mikal> So... I am going to regret this... 21:06:32 <mikal> What else is urgent that we should be reviewing for j-3? 21:06:43 <mikal> Are there things that should be high that we have the priority wrong on? 21:07:37 <mriedem> link to the bp list? 21:07:39 <mriedem> w/ priorities? 21:07:45 <mikal> #link https://launchpad.net/nova/+milestone/juno-3 21:07:50 <mriedem> thanks 21:07:53 <mikal> NP 21:08:01 <mikal> Maybe we can come back to that in open discussion 21:08:05 <mikal> To give people time to think it over 21:08:08 <mikal> Instead of stalling now 21:08:35 <mikal> The other thing that a few people have raised with me is the idea of bringing back some form of "review day" back 21:08:44 <mikal> Be it a day, or a week rotation 21:08:51 <adrian_otto> how does that work? 21:08:54 <mikal> I think its a good idea 21:09:09 <mikal> adrian_otto: there would be a roster of nova-cores, who would devote their time while rostered to just doing code reviews 21:09:22 <mikal> Its something we used to do when vishy was PTL, but it fell out of fashion 21:09:37 <adrian_otto> why? were the cores exhausted by it? 21:09:49 <mikal> I think we all just got busy 21:09:53 <mikal> And sort of forgot about it 21:10:01 <dansmith> I don't really find them all that useful to be honest 21:10:08 <mikal> I think it depends on your workflow 21:10:12 <adrian_otto> sounds like a good idea to me. 21:10:18 <mikal> Some cores do reviews every day 21:10:25 <mikal> So it wouldn't change much for them 21:10:33 <mikal> Other cores it helps to have the external motivation 21:10:38 <mikal> So... Let me rephrase. 21:10:48 <mikal> I think its a good idea, but I want someone else to run it 21:11:06 <mikal> So, does anyone think it s agood enough idea to volunteer to come up with a roster and encourage people to actually do reviews on their days? 21:11:38 <adrian_otto> be the official nag 21:11:48 <mikal> Yeah 21:11:59 <mikal> Perhaps report in this meeting who did a great job etc etc 21:12:01 <mriedem> so i don't get it, its' like 'office hours'? 21:12:10 <mriedem> so people can just bug the cores that are in during those office hours to review their patches? 21:12:16 <mikal> mriedem: not really 21:12:22 <mikal> Its more carving time out of your schedule to do reviews 21:12:29 <mriedem> oh ok 21:12:39 <mikal> If you were rostered today we'd expect you to clear your schedule for the day and just do code reviews all day 21:12:40 <adrian_otto> I thinkt aht should probably be an arrangement between the cores and the PTL 21:12:43 <mriedem> for us undisciplined folk 21:13:03 <mikal> Pretty much 21:13:10 <mriedem> not a bad idea for me, i've been pretty sporadic with reviews lately 21:13:25 <adrian_otto> rather than something published and trolled by all contributors to get things to land 21:13:26 <mriedem> so it might help forcing me personally 21:13:32 <mikal> Well, if you review the things from that email I sent out today, then that has a similar effect overall 21:13:45 <mriedem> yeha 21:14:00 <mikal> So, it sounds like no volunteer to run it though 21:14:20 <mikal> #action mikal to ask openstack-dev if anyone wants to give running a review day roster a go 21:14:26 <jogo> shocking no one wants to be the offical nag 21:14:32 <mikal> jogo: yeah, I know right? 21:14:41 <dansmith> I think you mean "shocking that jogo doesn't want to be the official nag" :P 21:14:50 <mikal> Ok, anything else on j-3 and release status? 21:15:31 <mikal> #topic Bugs 21:15:37 <jogo> when does the culling of the blueprints happen? 21:15:44 <jogo> proposal deadline 21:15:53 <mikal> jogo: bps without code should be being culled now 21:16:08 <mikal> jogo: bps with code will hit the FF eventually 21:16:23 <mikal> (Well, without code and with no chance of proposing all the code by FPF) 21:16:23 <jogo> sounds like an action item fo johnthetubaguy 21:16:34 <mikal> Yeah, he's at a cold meat themed music festival this week 21:16:40 <mikal> So I expect that to pick up next week 21:16:46 <mikal> Anyways, bugs 21:16:53 <mikal> It seems our bug fix rate has leveled out 21:17:00 <tjones> if sure has :-( 21:17:01 <mikal> I say based on zero data, just my impression 21:17:18 <mikal> tjones: do we need to start doing a weekly mail out of the most important bugs? 21:17:27 <mikal> tjones: like we are for j-3 bp reviews? 21:17:29 <jogo> http://54.201.139.117/nova-bugs.html 21:17:30 <tjones> yeah good idea 21:17:34 <tjones> i can start that 21:17:41 <jogo> 1059 open bugs 21:17:42 <mikal> tjones: that would be appreciated 21:17:49 <jogo> ++ 21:17:52 <mikal> That's a lot less than earlier in the cycle, but its still 1,000 too many 21:18:04 <mikal> So, if each of you closes 100 bugs over the weekend... 21:18:09 <mikal> ...then the code wouldn't merge 21:18:13 <tjones> there are 3 critical bugs - the 1st one has patches merged for ironic and oslo and could use revirews for nova https://blueprints.launchpad.net/nova/+spec/use-oslo-vmware 21:18:36 <tjones> its a test issue 21:18:41 <mikal> tjones: that's a blueprint not a bug? 21:18:52 <tjones> LOL 21:18:57 <mikal> Or is it a big bug or something? 21:18:58 <tjones> https://bugs.launchpad.net/nova/+bug/1328997 21:19:00 <uvirtbot> Launchpad bug 1328997 in nova "Unit test failure: openstack_citest" is being accessed by other users\nDETAIL: There are 1 other session(s) using the database." [Critical,In progress] 21:19:04 <mikal> Oh, ok 21:19:04 <tjones> sorry 21:19:14 <mikal> NP 21:19:14 <tjones> that was something else i wanted to discuss later 21:19:43 <dansmith> is there a nova patch to review for that? 21:19:47 <dansmith> or does it need to be created still? 21:19:57 <mikal> Yeah, I'm scrolling through looking for it 21:20:11 <tjones> um - wait i see 1 merged now 21:20:37 <dansmith> like a while ago 21:20:57 <mikal> So do we think its fixed and just in the wrong state? 21:20:59 <tjones> ok this is a left over that should be closed then. sorry was looking from the bottom 21:21:01 <tjones> yep 21:21:08 <mikal> Yay! One less bug! 21:21:09 <dansmith> excellent 21:21:11 <dansmith> I like those 21:21:22 <tjones> the other 2 arguably are not cirical as they only affect the vmwareapi 21:21:22 <jogo> tjones: is the source for your webpage open? 21:21:33 <jogo> tjones: because matbe we can move it to status.openstack.org somewhere 21:21:39 <dansmith> tjones: didn't we say that no bug can be critical if it's only affecting one driver? 21:21:45 <tjones> yes it is on gitup - i can end the link to you 21:21:55 <dansmith> gitup? 21:21:55 <tjones> dansmith: yes and i knew you were going to say that 21:21:57 <jogo> tjones: thanks that would be great 21:22:03 <tjones> github ;-P 21:22:05 <dansmith> is that like github for cowboys? 21:22:08 <mikal> dansmith: I assume it depends on how bad it is? 21:22:10 <tjones> yahooo! 21:22:16 <tjones> yes i will decrease prio 21:22:17 <mikal> dansmith: "deletes all instances" would probably still be critical 21:22:23 <dansmith> mikal: the definition of critical on the wiki says it has to affect all users I think 21:22:32 <dansmith> mikal: which by definition doesn't apply to a single HV issue 21:22:39 <mikal> Fair point 21:22:48 <mikal> But I suspect we'd argue about it at the time if it happened 21:22:58 <mikal> But that's a hypothetical, so let's not dig into it too hard 21:23:12 <mikal> tjones: any other bug stuff we should know apart from "please fix them"? 21:23:13 <dansmith> https://wiki.openstack.org/wiki/BugTriage#Task_2:_Prioritize_confirmed_bugs_.28bug_supervisors.29 21:23:18 <tjones> in any case - they could use a review 21:23:28 <tjones> yeah there are 179 bugs that need review 21:23:46 <mikal> Is there a dashboard of those? 21:23:54 <tjones> so when people are reviewing stuff - please review bugs too 21:24:12 <tjones> click the "ready for review" radio button on http://54.201.139.117/nova-bugs.html 21:24:25 <tjones> dims gave me a fix to sort the coumns which i'll put up today 21:24:26 <mikal> Cool. 21:24:39 <tjones> then you san see them ordered by age as well 21:24:40 <mikal> I will include that in the next review priorities email as well then 21:24:49 <tjones> great 21:25:12 <mikal> We should probably move on unless there's something else in bugs land 21:25:28 <tjones> that's pretty much it - there are about 30 that could be knocked off as they are abandoned. thats it from me 21:25:36 <mikal> #topic Gate status 21:25:40 <mikal> mriedem: did you add this item? 21:25:45 <mriedem> mikal: yeah 21:25:50 <mriedem> gate is more or less ok 21:25:50 <mikal> Want to talk to it then? 21:25:59 <mriedem> but, InstanceInfoCacheNotFound traces are pervasive 21:26:11 <mriedem> like hundreds of thousands per week, 90% successful runs 21:26:21 <mikal> That's... a lot 21:26:36 <jogo> sounds like a stacktrace we should fix 21:26:38 <mriedem> yeah i thought it was just stable/icehouse, so was glad to see this https://review.openstack.org/#/c/112520/1 21:26:47 <mriedem> but logstash says it's still on master 21:26:53 <mikal> Is there a bug filed for it? 21:27:09 <mriedem> the old bug was https://bugs.launchpad.net/nova/+bug/1304968 21:27:10 <uvirtbot> Launchpad bug 1304968 in nova "Nova cpu full of instance_info_cache stack traces due to attempting to send events about deleted instances" [High,Fix released] 21:27:13 <dansmith> so, 21:27:22 <dansmith> that error is just a symptom of an instance being deleted, 21:27:33 <dansmith> and the code looking it up not being careful about it 21:27:39 <dansmith> so that fix stands on its own, 21:27:48 <dansmith> but there are just other cases on master tha trigger the same behavior I think 21:28:08 <dansmith> what's the hit rate on master? less than icehouse, I hope/assume? 21:28:19 <mikal> dansmith: so its a case of working through them all and cleaning them up one at a time? 21:28:26 <mriedem> http://goo.gl/hHMGdO 21:28:31 <dansmith> mikal: yeah 21:28:48 <mikal> Kibana will mostly be master right? 21:28:51 <dansmith> mriedem: is that limited to master? 21:28:54 <mriedem> master is 212K in 7 days 21:28:56 <mikal> And it says 235,362 hits in the last seven days 21:28:56 <mriedem> dansmith: nope 21:29:21 <mriedem> dansmith: so we want this for stable/icehouse https://review.openstack.org/#/c/112520/1 21:29:23 <mikal> But stable runs are relatively rate comparatively, right? 21:29:40 <dansmith> mriedem: the first few I see are the lifecycle thing 21:29:41 <mikal> s/rate/rare/ 21:29:49 <dansmith> oh 21:29:54 <mriedem> hits on stable/icehouse in 7 days are 19K 21:30:03 <dansmith> but don't we see this for any grenade job as long as it's still present in icehouse? 21:30:04 <mriedem> so yeah, stable/icehouse is far less, and we probably just need to merge that backport 21:30:11 <dansmith> right 21:30:11 <mriedem> yeah 21:30:17 <mriedem> old side grenade logs are rife with these 21:30:20 <dansmith> as long as this isn't in icehouse, then we'll see it charged against master 21:30:22 <dansmith> yeah 21:30:24 <mriedem> which makes debugging grenade hard 21:30:36 <mikal> So we want two things? 21:30:38 <mriedem> so dansmith can you hit that backpor 21:30:39 <mikal> Merge that stble fix 21:30:45 <mikal> And then walk through master cleaning up the others? 21:30:50 <mriedem> yeah 21:30:59 <mriedem> if it's mostly grenade, the stable fix should bring the rate way down 21:31:00 <mikal> So... Who wants to give the master stuff a go? 21:31:08 <mikal> Oh, true 21:31:17 <mriedem> mikal: i can open a new bug for it on master to investigate 21:31:18 <dansmith> mriedem: done 21:31:26 <mikal> mriedem: that would be good 21:31:27 <mriedem> dansmith: cool, i'll bug mtreinish to hit the +A 21:31:29 <dansmith> lets see what this does to our rate when we get it merged 21:31:37 <dansmith> but obviously any that persist still need chasing 21:31:38 <mriedem> mtreinish: https://review.openstack.org/#/c/112520/1 21:31:40 <mikal> I can review the stable change after the meetnig if no one beats me to it 21:31:59 <mikal> Anything else on gate stuff? 21:32:01 <mriedem> nope 21:32:05 <mikal> Cool 21:32:14 <mikal> #topic Mid-cycle meetup summary 21:32:23 <mikal> This is mostly informational... I was asked to write a summary up. 21:32:28 <mikal> I've been doing that, but its long. 21:32:42 <mikal> So I've started posting the bits I've done, instead of blocking on perfection 21:32:42 <mtreinish> mriedem: done 21:32:50 <mikal> #link http://www.stillhq.com/openstack/juno/000005.html -- social stuff 21:32:53 <mikal> #link http://www.stillhq.com/openstack/juno/000006.html -- containers 21:32:56 <mikal> #link http://www.stillhq.com/openstack/juno/000007.html -- ironic 21:32:58 <mikal> ^-- what I've go done so far 21:33:10 <mikal> There will be more later today, but its going to take me a while to get through it all 21:33:21 <mikal> Topics I've written up but need proof reading before posting are db2 and cells, there will be more after that as well 21:34:04 <mikal> I think its likely that there are cases where my recollection of what we agreed differs from other people's, so feel free to ping me if you're concerned 21:34:31 <mikal> #topic Open Discussion 21:34:43 <mikal> mriedem: did you add the powerkvm thing? 21:34:48 <bknudson> so what happens with unapproved juno specs? 21:34:52 <mriedem> mikal: yeah, we sorted that out before the meeting 21:34:57 <mriedem> mikal: powerkvm ci was disabled on monday 21:34:59 <mikal> bknudson: please hold 21:35:03 <mriedem> krtaylor updated the pkvm ci wiki page 21:35:12 <mikal> mriedem: ok, so its mostly informational? 21:35:16 <mikal> mriedem: no action required? 21:35:18 <mriedem> mikal: yeah, we can move on 21:35:20 <mikal> Cool 21:35:29 <mikal> bknudson: so, there's been some talk in the release meeting about that 21:35:37 <bknudson> e.g., https://review.openstack.org/#/c/103617/ 21:35:43 <mikal> I think what is likely to happen is that K specs will open soonish 21:35:51 <mikal> But with the expectation of no review until we ship J 21:35:53 <bknudson> I didn't get around to working on it as I had hoped... too much other stuff 21:36:02 <mikal> So we'll be asking people to effectively rebase to the new directory 21:36:10 <mikal> People who don't rebase after a timeout we assume have lost interst 21:36:17 <mikal> And we can abandon those reviews 21:36:29 <jogo> mikal: oh I have one thing: https://review.openstack.org/114044 21:36:32 <jogo> adrian_otto: ^ 21:36:40 <bknudson> mikal: sounds easy enough 21:36:41 <mikal> #action mikal to propose a K spec plan on the mailing list 21:36:41 * adrian_otto perks up 21:36:51 <mikal> But I want to wait for John to get back so we're on the same page 21:37:03 <mikal> jogo: go! 21:37:06 <adrian_otto> jogo: I will cover that in the subteam update 21:37:20 <adrian_otto> but now it fine too 21:37:26 <adrian_otto> *is 21:37:28 <jogo> so adrian_otto posted https://review.openstack.org/114044 which is a kilo spec for a container service 21:37:32 <mikal> adrian_otto: we don't do subteams as an agenda item any more 21:37:36 <mikal> So do it now 21:37:42 <dansmith> yeah, I thought we were going to want to see this in stackforge first 21:37:52 <jogo> I thought at the mid-cycle we agreed to send this to stackforge first 21:37:54 <mikal> Oh, I didn't notice that 21:38:01 <mikal> We did agree to go to stackforge first 21:38:05 <mikal> Are you asserting that means no spec? 21:38:09 <adrian_otto> the rationale for starting it in nova-specs was to be sure the concept was well socialized among Nova devs 21:38:12 <mikal> Or that the plan in the spec needs to reflect that? 21:38:19 <adrian_otto> and that we will otu the compute program in a style consistent with Nova 21:38:49 <adrian_otto> I'd fine if we don't merge that spec 21:38:51 <jogo> adrian_otto: so I like that idea, do it in stackforge but have a design doc 21:38:58 <jogo> adrian_otto: cool, that works 21:39:00 <adrian_otto> but I'd like us to collaborate on it where we are comfortable 21:39:03 <mikal> So... I think we should have all the cmpute specs in one place 21:39:07 <mikal> To be less confusing to our victims 21:39:10 <adrian_otto> and since there is no compute program spec repo, that made the most sense 21:39:28 <jogo> adrian_otto: ++ sounds like we are all on the same page (I think) 21:39:40 <dansmith> how about this: 21:39:51 <dansmith> just put this doc in the new repo, formatted like our spec, 21:40:00 <dansmith> so that when we want to adopt it, we propose that doc as a spec 21:40:12 <dansmith> because the point of it being in stackforge, I thought, was to let it iterate quickly, 21:40:18 <dansmith> which might change some of this ascii art, I'd think :P 21:40:19 <jogo> dansmith: but we should also get nova feedback in the idea before they actaulyl do anything 21:40:25 <dansmith> sure, sure 21:40:36 <mikal> We could review a thing in stackforge though 21:40:38 <dansmith> just doesn't seem worth having an uncommitted spec for us 21:40:40 <jogo> so they interate in the right direction 21:40:44 <mikal> It would be good to not bring up anotehr specs repo right now 21:40:48 <mikal> Its a distraction 21:40:52 <jogo> mikal: works for me 21:40:55 <mikal> But if it was in the project repo that would work 21:41:16 <jogo> adrian_otto: thoughts? 21:41:22 <adrian_otto> I'm happy to land a derivative of this proposal in the project repo 21:41:28 <mikal> I think the mechanics of this is worth a mailing list thread to make sure people know what's happening 21:41:36 <adrian_otto> I think the important thing now is that we refine it together 21:41:39 <mikal> adrian_otto: do you want to do that or shall I? 21:41:47 <dansmith> yeah, and ping us when it's up so we can go review it over there 21:42:08 <adrian_otto> #link http://lists.openstack.org/pipermail/openstack-dev/2014-August/043113.html ML THREAD About Containers Service 21:42:21 <adrian_otto> mikal: done 21:42:27 <mikal> Heh, thanks 21:42:41 <jogo> adrian_otto: well that has the spec in our repo 21:42:52 <mikal> jogo: its ok, we can reply to the mail once its moved 21:43:02 <jogo> yup 21:43:09 <adrian_otto> so dansmith and mikal are apparently not aligned yet on where the spec should begin. 21:43:25 <dansmith> I thought we were 21:43:33 <mikal> I think we agreed in the end 21:43:40 <adrian_otto> if I submit a review for a stackforge repo it will take about 9 days for that to happen 21:43:42 <mikal> But I can reply to that email 21:43:47 <adrian_otto> and I want to make progress in the mean time 21:43:48 <mikal> And then Dan can argue with me if he disagrees 21:44:12 <mikal> adrian_otto: what's the slow bit? Infra getting to the review? 21:44:12 <jogo> adrian_otto: well no matter what the outcome you will need a stackforge repo 21:44:25 <adrian_otto> yes, I will make a project repo, and put the spec in there 21:44:40 <mikal> Cool 21:44:40 <adrian_otto> I am suggesting we iterate on https://review.openstack.org/114044 until then 21:44:46 <jogo> adrian_otto: thanks 21:44:53 <adrian_otto> with the understanding that I will abandon that 21:45:03 <mikal> So, we also deferred discussion of possible bp priority errors until now 21:45:04 <adrian_otto> and make a link to its sucessor 21:45:20 <mikal> adrian_otto: yep, but I think that one can stay up for discssion while you wait for your shiney new repo 21:45:34 <adrian_otto> mikal: tx. 21:45:50 <adrian_otto> I'll submit a review for the new repo 21:46:08 <mikal> adrian_otto: ping me with the details when you do that and I will see what I can do to speed it up 21:46:16 <adrian_otto> mikal: ok 21:46:23 <mikal> We should move on... 21:46:28 <dansmith> mikal: can you get me out of parking tickets too? 21:46:38 <mikal> So, does anyone feel we've gotten bp priority for j-3 horribly wrong? 21:46:42 <mikal> dansmith: yes, yes I can 21:46:46 <dansmith> sweet 21:47:10 <mriedem> there are 77 bp's and i haven't looked through the list, so no idea 21:47:13 <tjones> ok not horriblly - but this one https://blueprints.launchpad.net/nova/+spec/use-oslo-vmware is medium and it is very closely related to the spawn refactor which is high. they are both very critical for us 21:47:32 <mikal> tjones: does that one block the spawn refactor? 21:47:35 <mikal> tjones: or is just related? 21:47:44 <mikal> i.e. do we have a critical bp depending on a medium bp? 21:48:15 <tjones> they are very closely related - but not blocked. i almost would have said the oslo one is higher than the refactor in terms of cleaning up the dirver 21:48:32 <mikal> And the code is out for the oslo one? 21:48:34 <dansmith> tjones: I think what is holding things up is minesweeper 21:48:46 <tjones> is code is out. minesweeper is limping. 21:48:56 <tjones> the hardware gets here next week - hurrah! 21:49:08 <tjones> of course then it has to be deployed.... 21:49:10 <mikal> Stupid physical constraints 21:49:13 <tjones> lol 21:49:38 <mikal> Ok, so how about this... 21:49:49 <mikal> We can track that one as important, but let's leave it at medium for now 21:49:59 <mikal> Until we land some of the other open highs, I worry about diluting attention and getting nothing done 21:50:21 <tjones> ok sure 21:50:24 <tjones> thanks 21:50:33 <mikal> Thanks for being understadning 21:50:37 <tjones> np 21:50:46 <mikal> So... Anything else for open discussion? 21:50:51 <mikal> Or do you want an early mark? 21:51:40 <mikal> Going... 21:51:54 <mikal> ...going... 21:52:07 <mriedem> gone 21:52:08 <adrian_otto> see you next time 21:52:08 <mikal> ...gone 21:52:13 <mikal> Thanks for your time peeps 21:52:19 <mikal> #endmeeting