19:11:19 <dtroyer> #startmeeting openstackclient
19:11:20 <openstack> Meeting started Thu Feb 18 19:11:19 2016 UTC and is due to finish in 60 minutes.  The chair is dtroyer. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:11:22 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:11:25 <openstack> The meeting name has been set to 'openstackclient'
19:12:06 <dtroyer> late ping: dhellmann, stevemar, terrylhowe, lhcheng, dstanek, MeganR
19:12:19 <dtroyer> apologies for my tardiness, anyone still around?
19:12:21 <dstanek> dtroyer: !
19:12:46 <dstanek> dtroyer: i'm addicted to glowing screens. so i'm always here :-)
19:13:19 <rtheis> still here :)
19:13:25 <dtroyer> dstanek: that's gotta mess with sleep patterns ;)
19:14:09 <dstanek> dtroyer: it's easier if you just don't do it
19:14:31 <dtroyer> tell my wife that when I'm dozing at the dinner table
19:15:04 <dtroyer> ok, we're light today, I didn't have anything specific
19:15:45 <dhellmann> o/
19:16:32 <dtroyer> it looks like we've got a deep queue atm
19:16:51 <dtroyer> just in time for our new core members ;)
19:17:01 <rtheis> :)
19:17:43 <rtheis> I'll start chipping away at the queue
19:18:07 <dtroyer> I'm geting started myself this afternoon
19:18:18 <dtroyer> is there anything specific that warrants priority attention?
19:18:56 <rtheis> I didn't see any at first glance
19:19:03 <rtheis> I did have a interface question on https://review.openstack.org/#/c/276041/
19:19:32 <rtheis> Should we validate the keys in a key1=value1,key2=value2 list?
19:19:46 <rtheis> or pass along to the API call
19:20:14 <rtheis> Example usage in https://review.openstack.org/#/c/276042/
19:20:21 <dtroyer> thats a good question.  I don't think we always know the valid keys
19:20:45 <dtroyer> and I'm not sure if that is the sort of thing that an API using JSONschema would be validating
19:21:50 <stevemar> oops
19:22:07 <dtroyer> IIRc we have some usages like that where the API specifically does not validate the dict, like for vendor-specific values
19:22:16 <dtroyer> so in general we can't assume validation
19:22:22 <dtroyer> by the api
19:22:23 <stevemar> i'm here now
19:22:40 <rtheis> okay
19:22:43 <dtroyer> welcome stevemar.  you're #ff6633 now, just so you know ;)
19:22:57 <stevemar> i *always* forget about this meeting
19:23:18 <dtroyer> in cases like 276042 I really wish we could do that without resortin to AVPs as an option argument
19:23:31 <stevemar> maybe we can talk about the unset/set commands using cliff.Command?
19:23:31 <stevemar> i think tangchen was looking for guidance there
19:23:38 <dtroyer> since those are required and well known.  in that case we probably _should_ validate them
19:24:57 <dtroyer> stevemar: I think we hashed that out a while back and decided to remove the output on set/unset.  I'm not a fan but apparently those with output were in the minority
19:25:14 <stevemar> o/
19:25:16 <stevemar> bueller?
19:25:18 <stevemar> rgr
19:25:56 <stevemar> dtroyer: oh i think i found a bug with osc... need time to reproduce it
19:26:05 <dtroyer> what?  a bug?
19:27:10 <stevemar> :)
19:27:10 <stevemar> https://github.com/openstack/python-openstackclient/blob/master/openstackclient/api/auth.py#L149-L158
19:27:19 <rtheis> dtroyer: so on 276042, do you have an alternative to AVPs?
19:27:29 <stevemar> this check here is breaking some of the non-traditional auth plugins we have in keystoneauth
19:27:43 <dtroyer> rtheis: not a better one when multiple routes can be set at once.
19:27:56 <rtheis> ok
19:28:12 <dtroyer> but I fear that becomes a precedent in general
19:28:28 <dtroyer> and hope it does not
19:28:51 <rtheis> ok
19:29:09 <dtroyer> stevemar: put it to occ, it is the auth arbiter now, right?
19:29:52 <dtroyer> we're going to wind up with some breaking changes to do all of that…that'll trigger 3.0.
19:29:55 <stevemar> dtroyer: basically
19:30:01 <stevemar> yep
19:30:03 <dtroyer> and I hope we can stuff the rename in there then too
19:30:10 <stevemar> oh?
19:30:13 <dtroyer> so, planning ahead a bit
19:30:16 <stevemar> i'm hoping to do this soon
19:30:29 <dtroyer> I thought we decided reanming the package was best doen at a major rev
19:30:46 <stevemar> folks internally are finally starting to use federation and are complaining about the broken CLI :)
19:30:54 <dtroyer> soon like a week or two?
19:30:59 <stevemar> probably not
19:31:00 <stevemar> :(
19:31:14 <dtroyer> well, major is a lot farther out than that…
19:31:43 <dtroyer> you can probably fix it without 3.0 but we need to rip out all of the OSC auth for 3.0
19:33:17 <stevemar> dtroyer: that's a beast to fix
19:34:11 <dtroyer> can we just rip out that check and let the plugins fail it later?
19:37:12 <dtroyer> ok, anything else?
19:37:56 <rtheis> I don't have anything else
19:38:34 <dtroyer> stevemar: ?  You are slowly turning #33cc33
19:38:56 <stevemar> dtroyer: not as good of a color
19:39:12 <dtroyer> unless you are a system status indicator
19:39:18 <stevemar> if anyone else wants to rip and replace the auth stuff, go ahead :)
19:40:02 <dtroyer> I started a while back, but that was before occ took over the world.  it should actually be easier now
19:40:48 <dtroyer> ok, I've got a tight schedule this afternoon, if nothing else is pressing I'd like to get going.
19:41:16 <rtheis> nothing else, thank you dtroyer
19:42:51 <dtroyer> ok, thanks everyone!
19:42:55 <dtroyer> #endmeeting