10:00:08 <tonyb> #startmeeting requirements
10:00:09 <openstack> Meeting started Wed Mar  1 10:00:08 2017 UTC and is due to finish in 60 minutes.  The chair is tonyb. Information about MeetBot at http://wiki.debian.org/MeetBot.
10:00:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
10:00:12 <openstack> The meeting name has been set to 'requirements'
10:00:51 <prometheanfire> o/
10:01:13 <toabctl> hi
10:01:13 <tonyb> #topic rollcall
10:01:19 <prometheanfire> o/
10:01:22 <toabctl> hi
10:01:33 <tonyb> sigmavirus,  number80, dirk, coolsvap
10:01:42 <tonyb> prometheanfire, toabctl hey
10:03:17 <tonyb> #topic where are we at with requirements reviews ....
10:03:54 <prometheanfire> fairly caught up :D
10:03:58 <tonyb> So I asked that we wait a little while before we merged stuff that bumped minimes in g-r
10:04:20 <tonyb> I'd like to keep that up for another week for the cycle-trailing projects
10:04:30 <tonyb> but u-c and !='s shoudl be fine
10:04:56 <tonyb> then once the cycle-trailing have branched we're all systems go!
10:05:04 <prometheanfire> sure
10:05:31 <tonyb> #topic Any controversies in the Queue?
10:06:02 <tonyb> for whatever reason there is a lot of red in the queue but the eventlet and pbr chnages are pretty important
10:06:24 <tonyb> they both mean worhking with other teams to try and get the dependant patches merged
10:06:39 <tonyb> Oh and webob :/
10:06:45 <prometheanfire> for eventlet 1/3 have merged when I last looked
10:06:53 <tonyb> prometheanfire: cool.
10:06:57 <dirk> o/
10:07:04 <tonyb> hey dirk
10:07:28 <dirk> yeah, the pbr bump to 2.0 is pretty surprising to me
10:07:42 <dirk> and has a lot of fallout, since everything on the planet has a pbr  < 2.0 in its requirements
10:07:59 <dirk> can someone summarize the plan or did I miss that email?
10:09:00 <tonyb> dirk: Well PBR's git was broken and if it was released as is it would have blocked the (docs) gate in a way that we (requirements team) couldn't control/fix
10:09:15 <prometheanfire> dirk: actually, when I updated ocata's packages I didn't notice much <2.0 for pbr
10:09:24 <tonyb> dirk: so the we fixed pbr to $do_the_right_thing
10:09:54 <tonyb> dirk: that got released today and it was decided that it was an API break so we bumped the major version
10:10:23 <tonyb> that decision was preety quick and we (release team) didn't reliase that people had defensive caps in place
10:10:44 <tonyb> dirk: so far I've only seen 5 things in g-r that have the cap
10:10:58 <tonyb> so the fallout was a bit or a surprise
10:11:19 <tonyb> dirk: but the main aim of the release was to unblock/fix the sphinx mess we have ATM
10:11:53 <tonyb> ironically so far all the projects with defensive caps are unaffected by the API break
10:12:32 <tonyb> dirk: the plan to fix it is basically land the fixes in the identified projects and len use pbr
10:13:05 <tonyb> projects that don't use constraints *and* use one of the 5 projects will need a local patch to get by until then
10:15:00 <tonyb> dirk: is that better? tl;dr: you didn't miss anything
10:18:22 <tonyb> next topic?
10:20:15 <tonyb> #topic Community goals
10:20:25 <tonyb> So there are 2 community wide goals
10:20:29 <tonyb> py3 and wsgi
10:21:04 <tonyb> I'm pretty sure the latter doesn't matter for us, as it's my understanding that it's limted to API services
10:21:19 <tonyb> but py3 (well py35) does matter to us
10:22:13 <tonyb> I'd like to get us to the point this cycle where we're testing equally on py3 and pu2 and we set the basepython = python3 in the tox 'venv'
10:22:28 <tonyb> Opinions on how doable that is?
10:23:08 <tonyb> (keeping in mind the tooling rewrite we have to do)
10:23:13 <prometheanfire> It'll take some work I'm sure, but it's more of a known quantity
10:24:08 <tonyb> Should we focus on testing with the venv thing as a stretch goal?
10:24:34 <prometheanfire> we are using system libs?
10:24:37 <dirk> tonyb: thanks, sorry am not fully focused on the meeting
10:24:48 <tonyb> dirk: it's all good
10:24:57 <tonyb> prometheanfire: I don't follow?
10:25:12 <prometheanfire> what do you mean venvs?
10:25:42 <tonyb> If you run tox venv -- $some_command have it run under py3
10:25:57 <tonyb> most of our docs say that's the way to consume our tools
10:26:24 <dirk> tonyb: I understand the part of unblocking sphinx, I was the one writing the patch for sphinx enablement and the one unbreaking pbr :)
10:26:35 <dirk> it is just surprising that we bumped the major version, that is kind of unexpected
10:26:40 <tonyb> ... Oh no they don't strange
10:26:51 <dirk> but its fine, after such a long time its ok to bump the version
10:27:01 * asettle hears docs
10:27:02 <asettle> *looks around*
10:27:14 <prometheanfire> tonyb: ah
10:27:20 <tonyb> dirk: Yeah it was a bit of a surprise but people with more pbr history made that call
10:27:39 <tonyb> asettle: not real docs you can ignore us
10:27:59 <asettle> tonyb: yeah I saw.
10:28:02 <asettle> *ignoring*
10:28:38 <prometheanfire> :D
10:29:20 <tonyb> So python3 ... we need to do some planning and then pick someone to work on it
10:29:36 <tonyb> Who would like to be that person?
10:30:27 <prometheanfire> my next 2 weeks are more or less booked out
10:30:36 <prometheanfire> beyond that I may be able to
10:31:09 <tonyb> prometheanfire: undertood
10:32:39 <tonyb> I'll chat to the TC about what best aligns with the goal and then we can decide
10:32:54 <tonyb> #topic Open Discussion
10:33:11 <tonyb> did y'all recover from the PTG?
10:35:07 <tonyb> Anything?
10:35:29 <dirk> sort of, I'll go on vacation in order to recover :)
10:35:44 <dirk> did we sort out on whether we go first with eventlet or with webob?
10:35:47 <tonyb> dirk: Oh? when?
10:35:49 <dirk> I think webob might be less pain
10:35:59 <dirk> next week
10:36:11 <tonyb> dirk: nice.  Have fun
10:36:34 <prometheanfire> I'm dealing with nova fails right now :(
10:36:49 <tonyb> dirk: I don't think we had an order in mind.  I promised to write "all the test webob" changes
10:37:03 <tonyb> so I'll do that tomorrow to get data on how close we are
10:37:37 <tonyb> dirk: or if you have time I can shoot you the script to run and you can do it ...
10:38:38 <prometheanfire> I think it'll end up being whichver one has the deps merge first
10:38:45 <tonyb> dirk: but we're still blocked on the dependant patches anyway :(
10:39:20 <tonyb> prometheanfire: yeah.
10:39:52 <tonyb> dirk: but yes I agree webob is likley to less challenging
10:41:32 <tonyb> anything else?
10:43:19 <tonyb> If we think of anything else there is always #openstack-requirements
10:43:26 <tonyb> Thanks everyone
10:43:30 <tonyb> #endmeeting