[02:28] <StevenK> wgrant: https://code.launchpad.net/~stevenk/launchpad/stop-lying-please/+merge/165000
[03:42] <wgrant> StevenK: I hope that breaks some tests
[03:42] <StevenK> I doubt it, based on my grepping
[06:35] <lifeless> wgrant: / StevenK: hey, I would be very interested in knowning a perf comparison between ec2 and hp's test perf for lp parallel single-node test runs
[06:45] <wgrant> lifeless: At what sort of node sizes?
[06:46] <lifeless> wgrant: well, the ones you do ec2 test runs on .. xlarge isn't it?
[06:46] <wgrant> lifeless: ec2 isn't parallel today
[06:47] <lifeless> wgrant: oh!
[06:47] <lifeless> wgrant: why not :)
[06:47] <wgrant> lifeless: Because if we care about timeliness then we just throw it straight at buildbot
[06:47] <wgrant> ec2 test still uses a single test process on an m2.xlarge
[06:48] <lifeless> I would be curious if that ran any better on the same size in HPCloud, but less interested than in the actually-uses-the-cpus-case :)
[10:01] <wgrant> StevenK: I've shuffled gina around to be a bit more sensibly testable (eg. you now don't have to run scripts/gina.py to test anything at all...), so https://code.launchpad.net/~wgrant/launchpad/gina-skippage/+merge/165048 is now pretty trivial.
[10:10] <StevenK> wgrant: Your indentation for the test looks a little suspect
[10:12] <wgrant> StevenK: Howso?
[10:16] <StevenK> wgrant: You open a listcomp and don't indent another level
[10:17] <wgrant> StevenK: Ah, I indented one by a space, and one not
[10:17] <wgrant> Must have miscopied
[10:17] <wgrant> Or do you object to the second too?
[10:19] <StevenK> I'd prefer 4, but I can deal with one
[10:20] <wgrant> It's unambiguous and I think more readable than 4
[10:20] <wgrant> Though it is questionable
[10:20] <StevenK> wgrant: So fix the first one and land it, or refactor it to a function :-P
[10:44] <wgrant> StevenK: Oh yeah, that's done
[11:37] <StevenK> wgrant: It's been approved for a while
[11:50] <wgrant> StevenK: Thanks