13:05:36 #startmeeting openstackclient 13:05:37 Meeting started Thu Jul 14 13:05:36 2016 UTC and is due to finish in 60 minutes. The chair is dtroyer. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:05:38 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:05:41 The meeting name has been set to 'openstackclient' 13:06:01 #topic recent breakage 13:06:15 So the transition to osc-lib is having a bit of fallout 13:06:51 dtroyer: Hi Dean, it is the meeting now, right ? or it is 6hours later ? 13:07:02 I attempted to do the ClientManager and OpenStackShell object transitions in pieces and managed to get a few things out of order 13:07:12 uh, am I at the wrong time now? 13:07:28 oh, I missed last week... 13:12:41 both of my calendars have out meeting now... 13:13:08 dtroyer: I'm not quite sure... 13:13:46 Actually I don't know this week is odd or even... 13:14:09 Maybe it is wrong time 13:15:09 And the osc-lib, I sent some patches. Some project core team chose to wait for the stable version. So for now, it is not breaking anything, I think 13:16:05 thos patches came a bit early, osc-lib is not really ready. the base stuf like utils, logging, etc are safe, but the upper stuff is not 13:16:52 dtroyer: OK. I can pend the rest patches. 13:16:54 I got a couple of things out of order and not all of it got merged before I left on vacation 13:17:31 anyway, I think we unbroke the people running master everywhere yesterday 13:18:51 dtroyer: OK 13:18:58 there is still an issue in the auth argument handling that I need to sort out to move forward converting osc to use osc-lib 13:19:15 #topic reviews 13:19:34 tangchen: since it seems to be just you an I, do you have any reviews you want to talk about? 13:20:14 no 13:20:36 dtroyer: I have just one question about swift. 13:20:46 sure 13:20:59 dtroyer: Are we going to implement swift commands in osc, or make swiftclient a plugin ? 13:21:52 swiftclient will never be used by us, everything we do currently is internal, the future is to either expand that or use the SDK 13:22:29 FWIW, swiftclient didnt' exist when we built our own commands, which were basically forked out of the existing at that time swift comamnd binary 13:24:58 dtroyer: So, object commands will be implemented in OSC, using SDK, right? 13:25:08 of course, in the future. 13:25:14 yes, once SDK is past 1.0 13:25:19 OK 13:25:30 #topic open discussion 13:25:32 anything else? 13:25:42 nothing else for me today 13:26:42 I have just one bit… 13:27:01 is https://review.openstack.org/#/c/302687/ the last of the IP command changes? I think a bunch of them merged while I was gone 13:27:28 Yse it is. 13:27:45 we can close the bug 13:27:47 coolness, thanks for your patience on getting those in tangchen 13:27:48 after this one 13:27:57 and for doing most of the work 13:28:19 not at all. It is my duty. and I learned a lot from it. :) 13:28:49 one other question on https://review.openstack.org/#/c/340226/ 13:29:00 Is anyone actually using cloudpipe? 13:29:41 we are not bound to implement every last think in the project APIs or CLIs, especially if they are not needed or used 13:30:15 tangchen: I'm hoping you might know or find out since the author is from EasyStack 13:30:17 dtroyer: I have never used it before. And neither do the people around me. 13:30:26 Yes 13:30:40 I'll confirm it tomorrow 13:31:20 ok thanks. If that is not driven by an actual need for it I'd like to not do it. 13:31:31 I think he just tried to find out which command has not been implemented. 13:31:38 OK 13:31:40 That was all I had 13:31:53 nothing for me either. 13:32:12 Quick question - I filed this blkueprint, what is the next step for me to get it approved? https://blueprints.launchpad.net/python-openstackclient/+spec/neutron-client-qos 13:32:53 (sorry to jump in, it just looked like you were possibly ending the meeting) 13:33:15 one min 13:33:20 njohnston: we don't follow the strict blueprint process, we mostly use them for coordination 13:33:50 OK. Should I file a spec next, then, or proceed to implementation? It's just migrating commands that exist in python0neutronclient. 13:33:56 what I would suggest os that you add to that the actual resource names you want to use, that is usually the hardest part of implementing new commands 13:34:08 OK, will do 13:34:15 http://docs.openstack.org/developer/python-neutronclient/devref/transition_to_osc.html 13:34:24 Please have a look at this page 13:34:29 Will do, thanks! 13:34:40 we usually don't like using abbreviations, but really 'qos' is common enough and a total pain to spell out so I expect that to be OK here 13:34:43 I think you 'd better to talk to neutronclient guys 13:35:00 I'm not sure if qos commands will be implemented in OSC 13:35:06 maybe a plugin 13:35:17 rthies is on vacation this week, he has been working with them and would be helpful to touch base with when he returns next week (I think it is) 13:35:21 but, I think neutronclient team can confirm that 13:35:39 OK, I confirmed that this was good with rtheis late last week. 13:35:42 the base guideline is that 'core API' stuff is to go into OSC 13:35:48 ah, ok, thanks 13:35:58 :) 13:36:10 Yeah, QoS is not in the same category as like FWaaS or LBaaS 13:36:18 it's part of the core code 13:36:40 OK, thanks very much, I will follow the document! 13:36:42 njohnston: one more bit, in your BP you use the verb 'update'. OSC uses 'set' for that 13:36:49 ok 13:37:02 I will fix that 13:37:33 thanks 13:37:40 before I rush out to get breakfast, is there anything else? 13:37:47 nope, thanks! 13:38:11 nothing 13:38:19 thanks a lot. ;) 13:38:24 ok, thanks guys! 13:38:35 thanks! 13:38:38 and tangchen, thanks again for the work you and EasyStack are doing 13:39:04 #endmeeting