| *** ianw has joined #openstack-shade | 00:00 | |
| *** noama has joined #openstack-shade | 07:40 | |
| *** abregman has joined #openstack-shade | 09:30 | |
| *** dootniz is now known as kragniz | 09:59 | |
| *** abregman has quit IRC | 12:30 | |
| mordred | Shrews, SpamapS: https://review.openstack.org/392120 Add unit test to show herd protection in action ... look, I wrote a patch which is _only_ a test | 12:38 |
|---|---|---|
| *** abregman has joined #openstack-shade | 12:51 | |
| *** abregman_ has joined #openstack-shade | 13:10 | |
| *** abregman has quit IRC | 13:14 | |
| Shrews | inconceivable | 13:16 |
| Shrews | it really chaps my hide when folks submit PRs that obviously haven't been tested even a little bit | 13:57 |
| * Shrews might need more coffee | 13:58 | |
| *** edtubill has joined #openstack-shade | 14:24 | |
| Shrews | mordred: I'm still having issues with your intentions on 388184 | 14:44 |
| Shrews | mordred: so, for example, let's consider the flavor field OS-FLV-DISABLED:disabled. This goes into the root object by default. Turn strict mode ON and it disappears completely. Is that the intent? Or did you mean for that to move into 'properties'? | 14:47 |
| Shrews | If you meant for it to disappear, then I believe I finally get it (I just may have confused your comments from yesterday) | 14:48 |
| mordred | Shrews: it shows up in properties in strict mode for me | 14:56 |
| mordred | wait | 14:56 |
| mordred | maybe I'm stupid | 14:56 |
| mordred | Shrews: ok. I'm wrong. and I am stupid. and I now finally understand your question | 14:57 |
| Shrews | \o/ | 15:06 |
| Shrews | i thought maybe we just had different meanings for "passthru values" | 15:06 |
| mordred | yah | 15:08 |
| mordred | Shrews: so there are 2 different philosophies happening atm ... | 15:09 |
| mordred | Shrews: server was passing through all of the things, regardless of whether we interpreted it and put an interpreted value in the root | 15:09 |
| mordred | Shrews: flavor was only passing through completely unknown values | 15:09 |
| mordred | Shrews: I think I like the flavor approach better ... basically "we have an API and you should use it - however, if there is a value we've never heard of yet, you can look for it in properties" | 15:10 |
| mordred | but if it's part of our API - like disabled - I can't see much value in a user looking for properties['OS-FLV-DISABLED:disabled'] ... right? | 15:10 |
| Shrews | mordred: i agree with those things. the flavor way is the more intuitive to me | 15:12 |
| mordred | cool. I will make the patch actually behave that way | 15:12 |
| mordred | and in fact, I think adding unittests for each of the normalizes is likely the best thing too | 15:12 |
| Shrews | i actually assumed the server stuff was doing it the same way (didn't get past the flavor part). | 15:13 |
| mordred | Shrews: I believe this confusion definitely points to the value in getting this all locked down, documented and tested | 15:16 |
| Shrews | mordred: agreed. glad my stupidity was useful | 15:18 |
| Shrews | also, dict.iteritems() might be more efficient than items() since it doesn't make a copy of things | 15:22 |
| Shrews | but that's just a nitpick | 15:22 |
| mordred | Shrews: unfortunately, iteritems is not a thing in python3 | 15:23 |
| mordred | Shrews: so you have to do like six.iteritems(dict) I think ... and at that point I think I stop caring unless it's important | 15:24 |
| Shrews | oh. hah | 15:26 |
| * Shrews makes note to update nodepool changes | 15:29 | |
| mordred | Shrews: ok. I updated the current change to match what we discussed. I have not added unittests for all of the normalize types - mostly wanted to make sure this version made sense to you | 16:38 |
| mordred | Shrews: (I also added a line to the docs) | 16:38 |
| Shrews | mordred: yes. this is MUCH clearererer to me (i especially like the pop_or_get utility for some odd reason) | 17:06 |
| *** edtubill has quit IRC | 17:24 | |
| *** edtubill has joined #openstack-shade | 18:04 | |
| *** abregman_ has quit IRC | 18:24 | |
| *** edtubill has quit IRC | 21:07 | |
| *** abregman has joined #openstack-shade | 21:29 | |
| *** abregman has quit IRC | 21:49 | |
| *** thingee has joined #openstack-shade | 22:11 | |
| * mordred waves at thingee | 22:12 | |
| thingee | thought I could use a little shade | 22:12 |
| * mordred throws some shade to thingee | 22:13 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!