[10:43] <leonardr> abentley, can you take a look at https://code.edge.launchpad.net/~leonardr/launchpad/launchpadlib-pagetests-take-2/+merge/22444 ?
[10:53] <wgrant> leonardr: The catastrophic test failures did actually occur on EC2, buildbot and locally. But they were sufficiently catastrophic that they killed the test suite in such a way that the failure wasn't detected by the outer test process.
[10:54] <mwhudson> leonardr: did you see https://code.edge.launchpad.net/~mwhudson/launchpad/aiee-everything-is-broken/+merge/22185 ?
[10:55] <mwhudson> mind you i think your branch is mostly better
[11:00] <leonardr> mwhudson: i hadn't seen that. it sucks that we duplicated work
[11:02] <mwhudson> leonardr: well, if i'd actually run the tests after my last commit on friday...
[11:02] <leonardr> mwhudson: shall i go ahead with my branch? is there anything you'd like me to change?
[11:02] <mwhudson> leonardr: the work i did since friday took about 5 minutes, so i don't mind really
[11:02] <leonardr> ok
[11:02] <mwhudson> leonardr: i didn't see anything to change, let me have another look
[11:03] <leonardr> all right
[11:03] <leonardr> you can do the review if you want, i haven't heard from abentley
[11:06] <wgrant> Looks like he tried to unassign himself from the topic many hours ago, but instead spoke the new topic into the channel.
[11:07] <leonardr> ahh
[11:10] <mwhudson> leonardr: done
[11:11] <leonardr> mwhudson: fyi, the xx-wadl thing
[11:11] <leonardr> previously there was a pre-cached 'testrunner wadl file' which was used in xx-wadl.txt
[11:11] <leonardr> if you tried to use launchpadlib from a pagetest, it would pick up the testrunner wadl file and believe that launchpad had no capabilities at all, since the testrunner wadl file was basically empty
[11:11] <mwhudson> leonardr: ah
[11:11] <mwhudson> that makes sense
[11:12] <leonardr> so i removed the testrunner wadl file, but then i needed a bigger test to prove that launchpad's wadl caching worked
[11:15] <wgrant> The WADL's not generated for every test, right?
[11:24] <leonardr> the wadl is generated when it's requested and not present on disk
[11:25] <leonardr> and not present in the in-memory cache
[11:26] <wgrant> Great.
[11:26] <leonardr> that's one wadl generation per version requested per launchpad restart
[11:27] <wgrant> Excellent.
[13:22] <bac> morning everyone.  bring out yer dead.
[16:03] <sinzui> bac: I have a short branch when you have time. I will have spm test it to verify we want to land the trivial fix.
[16:24] <salgado> bac, can you review https://code.launchpad.net/~salgado/launchpad/bug-547054/+merge/22460 for me?
[16:53] <bac> sinzui:  are you seeking a r-c for your branch?
[16:53] <sinzui> no
[16:54] <bac> ok, i'll mark it approved then
[19:51] <jml> some brave soul might want to review https://code.edge.launchpad.net/~jml/launchpad/ssh-key-auth/+merge/21801 -- but please don't feel obliged
[19:51] <jml> I'll look at it later to try to break it down into smaller branches.
[19:52] <jml> mentioning in case someone would prefer to spare me the trouble
[19:52] <salgado> thanks for the review, bac. :)
[19:53] <bac> salgado: np
[19:54] <bac> jml: is it mostly mechanical?
[19:54] <jml> bac, sadly no.
[19:54] <bac> jml: ick
[19:54] <jml> bac, I mean, a lot of it is just moving stuff into another place and then tweaking it a little
[19:55] <jml> bac, but it deserves quite a thorough review, IMO.
[19:55] <bac> jml: right
[19:55] <bac> jml: sadly for you i was experimenting with using a smallest-first approach to +activereviews today
[19:56] <jml> bac, heh, that's ok :)
[19:56] <jml> bac, I'll leave it in the IRC backlog just in case someone is feeling really really generous.
[21:08] <sinzui> bac: https://code.launchpad.net/~sinzui/launchpad/auto-yui-tests/+merge/22485 could unblock you
[21:08] <bac> sinzui: and you move to the front of the queue
[21:08] <bac> well, you were already at the front
[21:09] <sinzui> I am also at the back of the que :(
[21:09] <sinzui> And I lost a "ue"
[21:09] <bac> potato/potato
[21:09] <sinzui> potatoe!
[21:09] <sinzui> thanks you happy campers
[21:11]  * sinzui ponders how two Bush presidents could pick such diametrically opposed VPs
[22:09] <bac> sinzui: your yui test runner only works for lp.registry, right?
[22:09] <sinzui> yes
[22:10] <bac> but my JS for product series was moved to lp.code.
[22:10] <bac> :(
[22:10] <sinzui> I can make a small change to include other tests
[22:10] <bac> yeah, we'll have to.  but it would've been nice to keep it isolated
[22:11] <sinzui> I need to set the  layer dynamically
[22:11] <sinzui> Subclass the test runner for code
[22:11]  * sinzui could move the test class to lp.testing
[22:15] <sinzui> bac: Do you want me to move the class into lp.testing then subclass it for registry?
[22:16] <bac> yes i guess that makes the most sense.
[22:16] <bac> then i'll do it for code
[22:16] <sinzui> or move into lp.testing and dynamically set the test name/layer part
[22:16] <bac> the former makes more sense to me
[22:17] <sinzui> okay
[22:49] <sinzui> bac ping
[22:49] <bac> hi
[22:50] <sinzui> bac: I moved the base test class to lp.testing...
[22:50] <sinzui> I got a circular import for canonical.launchpad.testing.pages...
[22:50] <bac> ok, can you paste a diff and put a link in the MP?
[22:51] <bac> doh
[22:51]  * bac hides
[22:52] <sinzui> https://pastebin.canonical.com/29881/
[22:54] <sinzui> bac: I think paste of the actual code is easier to read
[22:54] <sinzui> https://pastebin.canonical.com/29883/
[22:57] <bac> yeah, the diff is a mess since the file got gutted
[22:57] <bac> but the change looks great
[22:57] <sinzui> I think I know how to make timelines work...
[22:57] <sinzui> maybe I'll do that for edwin to review tomorrow