19:01:27 #startmeeting OpenStackClient 19:01:28 Meeting started Thu Oct 1 19:01:27 2015 UTC and is due to finish in 60 minutes. The chair is dtroyer. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:01:29 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:01:31 The meeting name has been set to 'openstackclient' 19:01:34 anyone here? 19:01:50 Hi, I am new, but here 19:02:01 alarm ping: dhellmann, stevemar, briancurtin, terrylhowe, lhcheng, sigmavirus24, dstanek 19:02:11 Hi MeganR, welcome 19:02:13 o/ 19:02:20 kind of half here 19:03:45 ok, might be a quick meeting then... 19:04:03 o/ kinda half here too :( 19:04:20 so we have one FTE there ;) 19:04:30 lol 19:04:49 o/ 19:04:50 I think one topic stevemar is interested in is: https://review.openstack.org/#/c/222046/ 19:05:28 * stevemar sneaks in late 19:05:30 dtroyer: o/ 19:05:40 ok, lets get going now that stevemar is here 19:05:59 #topic release 1.7.1 19:06:21 Let's start with the pending 1.7.1 release to fix an Image v2 issue 19:06:35 #link https://review.openstack.org/229694 19:07:01 it's based on stevemar's fix for find_resource() to handle resources/managers that don't have resource_class attribute 19:07:16 we're looking at you glancelclient v2 19:07:41 that combined with a novaclient change totally broke boot from volume 19:08:04 the novaclient fix was released a few hours ago, our fix is merged and release pending 19:08:46 I don't know if anyone has anything to add, I just wanted to get that out in case there were any questions. we quick-merged the reviews to do all of that 19:10:37 ok… 19:10:42 thanks stevemar for the quick fix! 19:10:54 #topic image v2 set and create 19:11:20 those were part of a chain of reviews addressing a set of image v2 changes, mostly to image set 19:12:01 https://review.openstack.org/#/c/226749/ is at the root of that chain 19:12:27 I've gone back and forth about the —property alias options that are there, I landed on leaving them in because we already have a release with them 19:13:48 I think we're going to go forward with this set and address further issues as follow-up, unless more objections surface 19:15:00 ok… 19:15:26 #topic Reviews 19:16:07 lhcheng mentioned https://review.openstack.org/#/c/222046/, stevemars set of object store additions 19:16:35 I've registered reservations about the object name 'account' but don't have a better alternative to propose 19:17:11 gah, i am back now 19:17:20 "object account", "object-store account"? 19:17:33 lhcheng: suggested 'object account' which I culd live with 19:17:40 if I culd spel 19:17:56 `object account` hmmm 19:18:08 i'd almost prefer object-store account 19:18:16 it's specific enough to remove my ambiguity concerns, if a little awkward 19:18:17 but i ain't no fancy UX guru 19:18:41 that is probably better from a naing perspective 19:18:51 specific and precise 19:18:55 and longer 19:19:08 not an UX expert, but just didn't want to type longer 19:19:25 I am fine either way 19:19:25 can we have a dash? 19:19:35 we've thrown out 'tab completion' before to answer that, but its not a great answer 19:20:01 stevemar: do we use 'object-store' anywhere now in the CLI? in another command or option? 19:20:23 dtroyer: right! forgot about that awesome tab feature :) 19:20:26 I think the dash is consistent with out internal use 19:21:43 dtroyer: actually we don't use a dash 19:21:47 http://docs.openstack.org/developer/python-openstackclient/command-list.html 19:21:51 we opt for a space 19:22:33 we have `object crud` and `container crud` 19:22:43 ?? we use 'object store' somewhere externally visible? 19:22:47 that's not the same thing 19:23:08 'object-store account' vs 'object store account' 19:23:15 we use this: http://docs.openstack.org/developer/python-openstackclient/command-objects/object.html 19:23:29 ugh store is a verb too 19:24:00 ?? it shouldn't be 19:24:05 and a noun :) 19:24:39 dtroyer: well, it's an 'object-store' it's a noun in this case 19:25:06 'Object Store' is used and the name of the subsystem/API in documentation 19:25:48 so the proposal is to use 'object store account' as a three word noun 19:25:57 we don't use dashes in nouns anywhere else 19:26:48 and I am OK with the precedent of using the API/subsystem name as a disambiguation. plugins are doing it (except for the ones using project code names,err) 19:27:35 ok, lets follow up in the review? 19:28:10 dtroyer can i request that we rename in a later patch 19:28:41 do I have to change the method/class signatures or just the entry points? :) 19:28:43 stevemar: sure. 19:29:27 we should do a 1.8.0 release when the swifty bits are in, i can then update devstack to remove some silly swift commands 19:29:28 change the classes if you think it would be confusing not to, otherwise we do have other places they're not a perfect match 19:29:46 dtroyer: the classnames should be obvious enough 19:30:04 ok 19:30:41 Any other reviews to mention? 19:30:45 dtroyer: good, review the code then :D 19:31:14 not from me 19:31:19 i'm swamped with keystoney bits 19:31:31 so congrats PTL-man! 19:31:52 yeh, grats stevemar ! 19:32:08 dtroyer: terrylhowe oh so much fun 19:32:20 i hate stable-maint 19:32:39 I wanted to mention that I've taken the initial SDK review further down the road of getting the auth right, even though KSA will change the SDK. 19:33:00 I am primarily building an example that steps through the whole thing to get it in one place and easy to understand 19:33:31 and considering doing KSA directly 19:34:38 we're using it indirectly now with occ 19:34:57 There are a couple other reviews that have sat for a bit between the efforts for image v2 and yesterday's fun 19:35:16 stevemar: occ is getting out of control 19:35:31 maybe it should configure logging too 19:35:35 19:36:45 ok, moving along then 19:36:52 #topic open discussion 19:37:35 The preliminary schedule for Tokyo is out, we have one fishbowl, one workroom (Wednesday) and a Friday morning meetup slot 19:37:44 sounds good 19:37:45 all subject to change yet 19:37:55 we realllllllly need a session on doing 2.0 and the rename 19:38:06 can we hack on that on friday :) 19:38:21 or maybe we should wait til we go to the sdk (whenever that happens) 19:38:46 SDK is going to be added as a dependency in a minor rev 19:39:25 after the last couple of weeks I'd personally like to be rid of glanceclient first 19:39:49 but yeah, we'll talk about that process there and hopefully gett he sequence worked out 19:40:37 we also need to address some significant testing holes 19:41:35 anything else? 19:43:18 well alrighty then… 19:43:56 Thanks everyone 19:43:59 #endmeeting