19:00:24 #startmeeting python-openstacksdk 19:00:25 Meeting started Tue Dec 15 19:00:24 2015 UTC and is due to finish in 60 minutes. The chair is briancurtin. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:00:26 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:00:29 The meeting name has been set to 'python_openstacksdk' 19:01:10 terrylhowe: anything specific to cover on your end? 19:07:03 fwiw, i’m going to move on to getting the blah_id stuff sorted out next, service-by-service 19:07:24 unfortunately block_store is a little more complicated and involves a circular reference, but the rest should be relatively straightforward 19:20:33 nothing in particular briancurtin 19:20:57 I’ve been a little focused on osc and working the network stuff in over there 19:21:04 and $day_job 19:24:57 terrylhowe: yeah i noticed that. besides that _id stuff, I’m mostly making sure we’re consistent on a bunch of stuff, so probably a few more of these types of audit reviews 19:25:21 the _id stuff is related to pathargs, right? 19:26:38 if so that’d be great, biggest thing in the way of a 1.0 to me 19:32:22 terrylhowe: first i’m going to knock out the _id ones that we can just do simply by setting type=x.X, which is a bunch of them. some of the more complex ones will need something for path_args and I’m going to tackle it separately 19:32:58 yeh, I think you’ll find it isn’t that simple briancurtin 19:33:32 that is what I found at least, but maybe there is an easier way 19:37:25 terrylhowe: yeah, i haven’t looked deeply into implementation, but needing something more for path_args only affected a smallish percentage of the overall things needing access to it. not to say it’s not an issue, but i’m going to tackle the easy case and get it out of the way first and then dive in once it’s more isolated 19:38:23 I was just kind of hoping we could add a method to each one that implements path_args and hard code the method in each one 19:38:33 that might be the simplier appoach 19:39:05 that is what I keep thinking at least and not getting to it. 19:39:27 etoews: added some path args resources and didn’t make them resource.props. That might be a better way to go 19:39:28 yeah, that could work, especially since i think only a few places are complicated enough to need a more advanced feature 19:40:25 there is only one resource that has a resource.prop that is a path arg *and* in the body. 19:40:52 that one could just stay a resource prop 19:43:33 terrylhowe: another thing i thought of earlier but haven’t had the chance to check out: i wonder how the move to ksa might have affected how we were passing through query params straight into requests 19:44:07 yeh, I’m not sure briancurtin I haven’t tried it 19:44:15 i’ll look later/tomorrow 19:45:20 I think there some little things like what Qiming found with the exceptions; 19:46:11 speaking of exceptions, it might be a better approach to catch all ksa exceptions and covert them to our own so users don’t have to know about ksa 19:47:03 absolutely 19:47:51 i’ll create an LP issue for that one so we can track it 19:48:19 https://bugs.launchpad.net/python-openstacksdk/+bug/1526516 19:48:19 Launchpad bug 1526516 in OpenStack SDK "Catch all ksa exceptions and map them to SDK exceptions" [Undecided,New] 19:48:23 briancurtin: ^ 19:48:32 cool, there we go 19:49:27 The other thing I’d like to do is put a hook into ksa so we don’t have to override the Session class 19:49:53 we just need to do some differnent endpoint filter url parsing, I’d think other ksa uses would want to do the same 19:50:36 osc overrides Session too to collect timing data, taht would be nice to have in ksa 19:50:49 yep, true 19:52:07 anyway nothing else here 19:55:06 terrylhowe: cool, thanks 19:55:09 #endmeeting