19:00:02 <briancurtin> #startmeeting python-openstacksdk
19:00:02 <openstack> Meeting started Tue Nov 17 19:00:02 2015 UTC and is due to finish in 60 minutes.  The chair is briancurtin. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:00:03 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:00:06 <openstack> The meeting name has been set to 'python_openstacksdk'
19:00:31 <etoews> \o
19:00:46 <briancurtin> #topic KSA
19:01:22 <terrylhowe> https://review.openstack.org/#/c/186140/
19:01:34 <briancurtin> terrylhowe: did you see jamie’s comment about taking an existing session, e.g., from OSC? i think that plus two comments i had on the previous changeset and then it’s probably good, right?
19:02:18 <briancurtin> actually, is that something that would be another factory off of Connection?
19:02:22 <terrylhowe> didn’t notice that
19:02:31 <terrylhowe> probably
19:02:46 <briancurtin> so maybe we do that in another change rather than hold this off?
19:03:48 <terrylhowe> I guess the connection class should have a constructor that takes a session in case you have one already
19:04:13 <briancurtin> yeah, that’s probably the easiest way, to just add a session=None
19:04:19 <terrylhowe> yeh
19:04:52 <briancurtin> terrylhowe: can you take a look at the comments on https://review.openstack.org/#/c/186140/30/openstack/profile.py and https://review.openstack.org/#/c/186140/30/openstack/service_filter.py
19:07:42 <terrylhowe> fine points briancurtin, I’d need to look at the exception some more to see if there is some other reason for that
19:08:12 <briancurtin> terrylhowe: cool, whenever you get around to it, just wanted to make sure they didn’t get lost since the other bunch of comments were mostly just the same thing over and over
19:08:58 <briancurtin> #topic https://bugs.launchpad.net/python-openstacksdk/+bug/1517149
19:08:59 <openstack> Launchpad bug 1517149 in OpenStack SDK "find method needed for BaseProxy" [Critical,Confirmed] - Assigned to Brian Curtin (brian.curtin)
19:09:13 <terrylhowe> as to the ServiceFilter isa dict, it is just convenient to say endpoint_filter=cls.service rather than endpoint_filter=cls.service.get_filter()
19:10:13 <briancurtin> terrylhowe: can we get that if it was a MutableMapping? i don’t have a super strong case here, i’d have to dig in a bit more than i did
19:10:35 <terrylhowe> probably
19:11:03 <briancurtin> can adjust that after anyway, it’s fine for now
19:12:23 <briancurtin> on that find method bug, i’m going to try and hack that out quick this afternoon. i noticed it’s the only remaining proxy method that doesn’t run through BaseProxy but goes through the resources. in all of the fury around changing find a bunch of times, we never hooked it up through that, and i really only noticed it when someone contributed to the
19:12:23 <briancurtin> Rackspace extension
19:13:18 <terrylhowe> is there much to be gained by having a method in BaseProxy?
19:14:39 <terrylhowe> Aren’t most of them one liners like:         return profile.Profile.find(self.session, value,
19:14:40 <terrylhowe> ignore_missing=ignore_missing)
19:14:41 <briancurtin> terrylhowe: consistency is a big one since this now sticks out like a sore thumb. it’s probably a relatively simple one to add
19:15:45 <briancurtin> i guess it’s possible that i don’t end up adding it, but i’d like to at least look at it to see if it’s worth it. i’m not going to like documenting this in the contributor guide as “you do these 5 things this way, but this other one you do it a little different"
19:16:24 <terrylhowe> k
19:17:02 <etoews> agreed. i'd prefer to see it in BaseProxy.
19:17:46 <etoews> if it's not there, it leaves contributors pondering the reason why it wasn't included.
19:18:11 <etoews> (since they're not going to read the contrib guide anyway :P)
19:18:43 <stevelle> s/read/find/ ?
19:18:54 <etoews> that too
19:20:12 <briancurtin> other than that stuff, i don’t have anything pressing that requires much of a discussion, just working through LP and gerrit. anyone have anything to cover?
19:20:27 <etoews> yep
19:20:36 <etoews> https://bugs.launchpad.net/python-openstacksdk/+bug/1479977
19:20:36 <openstack> Launchpad bug 1479977 in OpenStack SDK "Get sdk docs on http://developer.openstack.org/python-openstacksdk/" [Undecided,New] - Assigned to Everett Toews (everett-toews)
19:20:56 <etoews> is that the url there the final resting place for our docs?
19:21:01 <briancurtin> #topic docs on openstack.org
19:21:30 <etoews> i've got a patch set ready to go but would rather confirm the url now.
19:22:01 <etoews> just start throwing out urls for us to consider if you have an opinion
19:22:18 <briancurtin> etoews: i guess so? as much as i’d like it to just be openstacksdk right there, as we’ve moved to that name in other places, it maybe makes sense to keep the python prefix there in the event that other language SDKs become official to the point that they also end up in the same spot as us
19:22:32 <etoews> d.o.o/sdk/python
19:22:41 <etoews> d.o.o/python-sdk
19:23:01 <etoews> (having openstack in the later part of the url seems redundant to me)
19:23:12 <terrylhowe> Anne question?
19:23:22 <etoews> not really
19:23:29 <etoews> but maybe?
19:24:06 <terrylhowe> python-openstacksdk is fine with me, then the structure is flat
19:25:01 <briancurtin> i’d probably just go with python-openstacksdk as the bug states. if we have to change in the future I’m sure someone can figure out some redirects or something
19:25:19 <etoews> annegentle: ^
19:25:42 <stevelle> would your patch do anything for the current state of http://developer.openstack.org/#sdk etoews?
19:25:53 <etoews> nope
19:26:25 <etoews> i know the application ecosystem working group is looking at revamping the landing page of d.o.o
19:26:31 <annegentle> stevelle: we have plans for that though
19:26:34 <annegentle> right etoews
19:26:37 <etoews> but that doesn't really impact us at this point
19:26:39 <stevelle> sounds lovely
19:26:47 <etoews> it really needs it
19:26:49 <stevelle> everything in it's own time
19:26:55 * annegentle goes to eavesdrop to catch up
19:27:40 <etoews> in my defense, when i first helped create d.o.o it looked the same as the rest of the crappy design of openstack.org :P
19:28:01 <etoews> then everything else got a facelift but d.o.o
19:28:29 <annegentle> etoews: heh
19:28:31 <stevelle> It looks a lot better than it did when I joined openstack, fwiw
19:29:49 <annegentle> sorry etoews I dont' know your question and can't find it on eavesdrop
19:30:17 <terrylhowe> https://bugs.launchpad.net/python-openstacksdk/+bug/1479977
19:30:17 <openstack> Launchpad bug 1479977 in OpenStack SDK "Get sdk docs on http://developer.openstack.org/python-openstacksdk/" [Undecided,New] - Assigned to Everett Toews (everett-toews)
19:30:20 <etoews> do we want  http://developer.openstack.org/python-openstacksdk/ to be the final url for the sdk docs
19:30:25 <etoews> ?
19:31:03 <annegentle> etoews: ah. So the pattern we're establishing even last week is this so far: http://developer.openstack.org/api-guide/compute
19:31:14 <annegentle> and http://developer.openstack.org/api-guide/quick-start/
19:31:21 <annegentle> those are the only two we have besides the large api ref page.
19:31:35 <etoews> briancurtin and terrylhowe are for it. i'm wobbly on the redundant "openstack" in "python-openstacksdk".
19:31:36 <annegentle> so, what about http://developer.openstack.org/sdks/python-openstacksdk?
19:31:46 <annegentle> gives us a starting pattern?
19:32:00 <briancurtin> if we’re going /sdks then i’d probably say /sdks/python
19:32:01 <etoews> previously i had suggested d.o.o/sdk/python
19:32:03 <annegentle> and potential for a better landing page for sdks :)
19:32:16 <annegentle> etoews: there may be more than one python sdk for openstack, how would we handle?
19:32:29 <briancurtin> that would be a bad idea, so don’t do that
19:32:29 <etoews> ooooooh...i like the opportunity for an sdks landing page
19:33:52 <annegentle> there's already libcloud tho briancurtin
19:34:01 <etoews> agreed that more than
19:34:06 <annegentle> not that they're gonna publish their docs, but they may have a landing page with links
19:34:15 <etoews> one official python sdk is bad news
19:34:22 <briancurtin> libcloud is libcloud though, not an openstack sdk?
19:34:56 <annegentle> briancurtin: agreed, but this audience needs info. At the summit for example I fielded questions about the 2-3 Go SDKs. So would it be better that there was a landing page listing possibillities
19:35:10 <annegentle> dev.o.o/sdks/go/, dev.o.o/sdks/python, etc?
19:35:15 <annegentle> really just talking outloud
19:35:16 <briancurtin> I’m really just worried about the docs for my project
19:35:22 <annegentle> briancurtin: sure
19:35:33 <annegentle> dev.o.o/sdks/python/openstacksdk
19:35:54 <terrylhowe> sounds good to me
19:36:51 <etoews> works for me
19:37:59 <etoews> briancurtin: any thoughts on that one?
19:38:06 <briancurtin> its fine
19:38:35 <etoews> #agreed sdk docs to be published to developer.openstack.org/sdks/python/openstacksdk
19:38:54 <annegentle> ok, I've triaged that bug with that info as well as a link to a patchset that does the api-guide/compute publish configuration
19:39:40 <etoews> cool. i've got a patch set to infra ready to go. just need to update it with the new url.
19:40:05 <etoews> one last thing from me
19:40:09 <etoews> https://bugs.launchpad.net/python-openstacksdk/+bug/1515074
19:40:09 <openstack> Launchpad bug 1515074 in OpenStack SDK "Add support to stack update broke create server" [Undecided,New]
19:40:20 <etoews> serious facepalm here
19:40:35 <etoews> create server got broken.
19:40:36 <briancurtin> oh yeah, there’s that
19:40:51 <etoews> that's bad
19:41:10 <etoews> i'd be freaking out a bit more if ksa wasn't rolling over the change that broke it
19:41:31 <terrylhowe> I had to fix that for ksa
19:41:38 <etoews> i'm sure
19:42:00 <etoews> terrylhowe: did you wind up getting to the bottom of the func test failures in the ksa patch set?
19:42:20 <terrylhowe> they are all passing now etoews
19:42:30 <etoews> awesome
19:43:11 <etoews> we can close that bug when ksa lands
19:43:19 <terrylhowe> yeh
19:43:22 <etoews> i suppose that's all i have to say about that
19:45:36 <etoews> nothing else from me
19:45:46 <briancurtin> terrylhowe: anything else?
19:46:06 <terrylhowe> nothing from me
19:46:16 <briancurtin> alrighty then
19:46:18 <briancurtin> #endmeeting