[13:55] <openstackgerrit> Scott Moser proposed stackforge/cloud-init: tox: disable proxies when running nosetests  https://review.openstack.org/195294
[13:55] <smoser> claudiupopa, are you there ?
[13:55] <smoser> can you tell me if https://review.openstack.org/#/c/195294/1 works on windows ?
[14:05] <claudiupopa> yeah, seems to work.
[14:07] <smoser> can you +1 ? or +2  if you want.
[14:07] <claudiupopa> done.
[14:44] <smoser> claudiupopa, so why did that not merge ?
[14:45] <smoser> does it need a jenins run, and if so why didnt it run
[14:46] <claudiupopa> http://status.openstack.org/zuul/ seems to be a pretty big merge queue.
[14:59] <openstackgerrit> Merged stackforge/cloud-init: HACKING.rst: mention signing the contribors agreement  https://review.openstack.org/194340
[15:14] <openstackgerrit> Scott Moser proposed stackforge/cloud-init: tools/tox-venv: support running other than ./tools/tox-venv  https://review.openstack.org/195631
[15:19] <openstackgerrit> Merged stackforge/cloud-init: tox: disable proxies when running nosetests  https://review.openstack.org/195294
[15:20] <openstackgerrit> Scott Moser proposed stackforge/cloud-init: tools/tox-venv: support running other than ./tools/tox-venv  https://review.openstack.org/195631
[15:22] <openstackgerrit> Scott Moser proposed stackforge/cloud-init: Clean up stale auto-generated autodoc files.  https://review.openstack.org/191139
[15:47] <Odd_Bloke> smoser is on a rampage.
[16:47] <smoser> harlowja, around ?
[16:54] <openstackgerrit> Scott Moser proposed stackforge/cloud-init: Bring over the 'safeyaml' from bzr  https://review.openstack.org/170252
[16:58] <smoser> claudiupopa, if you want, you requeseted test there, i added some ^
[17:02] <smoser> or Odd_Bloke if you want to review that.
[17:06] <openstackgerrit> Merged stackforge/cloud-init: Clean up stale auto-generated autodoc files.  https://review.openstack.org/191139
[17:24] <harlowja> smoser suppp
[17:33] <smoser> harlowja, i was gonna ask about how to add tests to your existing change id
[17:34] <smoser> its really wierd.. ijust --amend
[17:34] <smoser> and git review
[17:34] <smoser> and my patchset rplaced yours
[17:34] <harlowja> hmmm, lol
[17:34] <smoser> which is wierd in my opinion
[17:34] <harlowja> :)
[17:34] <harlowja> touching my patchesets, how dare u
[17:34] <harlowja> lol
[17:35] <harlowja> so what u did is one way, the other way is to make a dependent review
[17:35] <harlowja> where u make a new patchset (using the initial one as the base) and then do git-review -R
[17:35] <harlowja> *which makes that new review have a dependency on the first
[17:36] <harlowja> otherwise ya, the patchsets sorta just get added ontop
[17:40] <smoser> yeah. i think the way i did it makes more sense here.
[17:40] <smoser> as the dependent review would then hav both of them in the git log
[17:40] <smoser> but yours wasnt acceptable without the tests
[17:42] <harlowja> :-P
[17:44] <smoser> hm..
[17:44] <smoser> what to do here.
[17:44] <smoser> looking at the templeter
[17:44] <smoser> and tests for it.
[17:45] <harlowja> up to u, how much of it u want to keep :-P
[17:50] <harlowja> smoser one option, remove the basic_rendering support, leave jinja2
[17:50] <harlowja> that might help reduce the amount of stuff to test (and meh, do we need that other way anymore?)
[17:51] <harlowja> otherwise, include a bunch of test-templates.txt, ensure they get expanded to expected-out-test-templates.txt or whatever
[17:51] <harlowja> and blow-up if not
[17:51] <smoser> i think just if its not declared, then we use builtin
[17:52] <smoser> its fine for some things.
[17:52] <harlowja> k
[17:52]  * harlowja doesn't have any attachment to it :-P
[17:52] <harlowja> my precious!
[17:52] <harlowja> lol
[17:54] <smoser> harlowja, i have this right now:
[17:54] <smoser>  http://paste.ubuntu.com/11774661/
[17:54] <harlowja> seems fair to me
[17:54] <smoser> the value in it is that we have *some* builtin renderer
[17:54] <harlowja> agreed, which probably suits 90% of the cases anyway
[17:55] <harlowja> except for weird i need to have for loops in my tempaltes and such
[17:55] <smoser> well. yeah, i think as we start offering $datasource.hostname
[17:55] <smoser> and such, the builtin will probably fall appart
[17:56] <Odd_Bloke> Why do we need a bultin renderer at all?
[17:56] <smoser> i'm wondering if you were right though, an djinja should be the default
[17:56] <harlowja> as long as jinja2 doesn't die like cheetah, lol
[17:56] <smoser> i dont know, Odd_Bloke
[17:56] <Odd_Bloke> It's more stuff for our users to learn, and more code for us to maintain.
[17:56] <smoser> its only really for backwards compat.
[17:56] <Odd_Bloke> Backwards compatibility is probably an argument for it though.
[17:57] <Odd_Bloke> ^_^
[17:57] <smoser> so lets say things..
[17:57] <harlowja> should we support cheetah to then :-P
[17:57] <harlowja> for backwards compat ;)
[17:57] <smoser> the thing it comes down to is selecting a default
[17:57] <Odd_Bloke> /kick harlowja NO
[17:57] <harlowja> lol
[17:57] <harlowja> but but but
[17:58] <harlowja> ha
[17:58] <smoser> any template can declare what it needs to be rendered with
[17:58] <smoser> but if there is no header ('##template: jinja2')
[17:58] <smoser> then what do we do?
[17:58] <smoser> what i did just right there was fall back to the builtin
[17:58] <smoser> if we say instead "undefined is jinja2"
[17:58] <smoser> then in the far future world of jinja2-is-dead
[17:58] <harlowja> jinja3 ftw
[17:59] <smoser> then we are in the position of basically *having* to break someone
[17:59] <smoser> as we wont be be able to render their undeclared thing
[17:59] <Odd_Bloke> Jinja2 is very unlikely to go anywhere.
[18:00] <harlowja> jinja3 ftw
[18:00] <harlowja> lol
[18:00] <Odd_Bloke> And if it does, we just vendor it in and warn people that it's deprecated.
[18:00] <smoser> meh
[18:00] <smoser> i think i like basicaly requiring the declaration
[18:01] <harlowja> ya, we can do that to
[18:01] <smoser> you either declare what you want, where we can then definitively render it or not
[18:01] <smoser> or you dont
[18:01] <smoser> and you get harlowja's crappy builtin renderer
[18:01] <harlowja> da best ever!
[18:01] <harlowja> lol
[18:01] <smoser> i think i'll add a alias for 'basic' from 'harlowja-crap'
[18:01] <Odd_Bloke> I think it would be good to warn people against using harlowja's crappy builtin renderer.
[18:01] <harlowja> u have to rename it 'crappy_built_in_renderer'
[18:01] <Odd_Bloke> Something in the log or somesuch.
[18:01] <smoser> Odd_Bloke, why? we can maintain that.
[18:01] <smoser> its not hard.
[18:02] <Odd_Bloke> I guess.
[18:03]  * harlowja surely will miss cheetah though, lol
[18:04] <harlowja> we could also default to https://github.com/Yelp/yelp_cheetah
[18:04] <harlowja> lol
[18:04] <smoser> harlowja, we can bring it back.
[18:04] <smoser> just for old times sake
[18:04] <harlowja> :-P
[18:08] <smoser> We really recommend that you don't choose cheetah for new projects, and certainly not this hacked fork
[18:09] <harlowja> lol
[18:09] <smoser> i think i'm going to add things like that to all my code from now on
[18:09] <harlowja> don't listen to them, jeez
[18:09] <harlowja> lol
[18:09] <harlowja> :)
[18:09] <smoser> "i really recommend you dont use this."
[18:09] <harlowja> :)
[18:09] <smoser> instead, maybe go outside and get some exercise!
[18:09] <harlowja> isn't that what all those licenses are for, lol
[18:10] <harlowja> thought they did the same thing, use at your own risk, its just opensource mannnn
[18:10] <harlowja> lol
[18:10] <harlowja> *aka u can't sue me ...
[18:11] <harlowja>  https://github.com/stackforge/cloud-init/blob/master/LICENSE-Apache2.0#L144 (that section, lol)
[19:32] <smoser> alright. i almost have tests for templter whoo hoo
[19:33] <harlowja> woooo
[19:33] <harlowja> u da man
[19:43] <smoser> the key to improving test coverage is reducing lines of code
[19:44] <openstackgerrit> Scott Moser proposed stackforge/cloud-init: Bring over the 'templater' from bzr  https://review.openstack.org/170257
[19:50] <harlowja> :)
[19:50] <harlowja> i still see lines of code there
[19:50] <harlowja> its not 0
[19:50] <harlowja> lol
[19:56] <smoser> what i didn't say was one of the things i did was join lines.
[19:56] <smoser> because 2 lines of 30 characters is 2 lines.
[19:56] <harlowja> :)
[19:56] <smoser> but one line of 60 chars is only 1
[19:56] <smoser> :)
[19:58] <openstackgerrit> Scott Moser proposed stackforge/cloud-init: Bring over the 'templater' from bzr  https://review.openstack.org/170257
[19:59] <smoser> harlowja, you want to review your code there + my tests ?
[19:59] <harlowja> :)
[19:59] <harlowja> sureeeee
[19:59] <harlowja> oh man, whoever wrote that code is super
[19:59] <harlowja> lol
[20:06] <harlowja> smoser ok added a comment
[20:06] <harlowja> i think the fixtures test library can replace that custom tempdir
[20:07] <harlowja> commonly used in other parts of openstack (since the author works there)
[20:14] <smoser> harlowja, i dont want to just grab dependencies.
[20:14] <smoser> i dont like writing code just because either
[20:14] <harlowja> this would only be a test dep
[20:14] <smoser> yeah, thats not so bad.
[20:14] <harlowja> :-P
[20:20] <smoser> harlowja, if you want to add fixtures useage thats ok.
[20:20] <harlowja> k
[20:20] <harlowja> i can after that goes in
[20:20] <harlowja> ok with u
[20:21] <harlowja> i'm cool with that
[20:21] <harlowja> bb
[20:21] <smoser> but can we make sure it works iwth trusty
[20:24] <harlowja> sure,
[20:24] <harlowja> afaik the package exists there
[20:31] <smoser> its 0.3 in trusty
[21:01] <harlowja> ok, thats pretty old, but should be ok for this i think