[00:06] <cjohnston> Ursinha: if there is anything we can do to help please let us know
[00:07] <Ursinha> wgrant, cjohnston , that page is only used by summit?
[00:07] <cjohnston> AFAIK
[00:08] <wgrant> Ursinha: Yes
[00:09] <Ursinha> and who maintains summit?
[00:09]  * cjohnston points to mhall119
[00:09]  * cjohnston hopes wgrant doesn't point at cjohnston
[00:09] <Ursinha> lol
[00:10] <Ursinha> I ask because if we change launchpad and that causes summit to go crazy, that wouldnt be a launchpad problem :)
[00:10] <cjohnston> Ursinha: its me
[00:10] <cjohnston> and yes
[00:10] <cjohnston> :-)
[00:10] <Ursinha> given that summit people are aware and so
[00:10] <Ursinha> ah, good then
[01:27] <StevenK> wgrant: I might have to fix the garbo job to use memcache. Or we deploy it and then cowboy disable it until the next NDT
[01:32] <wgrant> StevenK: Why?
[01:33] <wgrant> Oh, you mean to stop it running more than once?
[01:35] <StevenK> wgrant: Yeah.
[01:36] <wgrant> StevenK: Depending on timing we'll want to run it at least twice anyway
[01:36] <StevenK> wgrant: Really?
[01:36] <wgrant> I'd personally have used a feature flag to just run it once when we wanted to
[01:36] <wgrant> But it doesn't really matter
[01:36] <wgrant> StevenK: Because it's possible that garbo will start before everything that creates previewdiffs has the new code.
[01:37] <StevenK> Ah, true
[01:37] <wgrant> eg. code deploys to ackee, garbo runs and updates MP 1234, taotie updates MP 1234's diff with the old code, then taotie gets the new code
[04:56] <StevenK> wgrant: https://code.launchpad.net/~stevenk/launchpad/deny-ddeb-deletion/+merge/163633 or shall I refactor out and have the same thing for changeOverride() ?
[04:59] <wgrant> StevenK: I'd consider explaining that you should delete the deb instead, but otherwise fine
[05:02] <StevenK> wgrant: "Cannot delete debug publications directly, please delete the publication" doesn't sound right either
[05:03] <wgrant> "Cannot delete ddeb publications directly; delete the corresponding deb instead."?
[05:05] <StevenK> wgrant: http://pastebin.ubuntu.com/5663470/
[05:07] <wgrant> StevenK: right
[05:12] <StevenK> wgrant: I'll land it with --incremental, since the bug also mentions changeOverride()
[05:12] <wgrant> StevenK: Yep
[05:12] <wgrant> StevenK: Or do changeOverride in the same branch
[05:12] <wgrant> Because it's like two lines
[05:12] <wgrant> + a test
[05:13] <StevenK> So 12
[05:13] <wgrant> Won't quite be over the limit :)
[05:13] <StevenK> And I have like 35k LOC credit
[05:18] <StevenK> Hmmm, how do you pass kwargs through assertRaisesWithContent
[05:18] <StevenK> TypeError: assertRaisesWithContent() got an unexpected keyword argument 'new_phased_update_percentage'
[05:36] <StevenK> wgrant: changeOverride() now complains too
[05:38] <wgrant> StevenK: 50	+ AssertionError, "Cannot override debug publications directly; "
[05:38] <wgrant> 51	+ "please override the publication."
[05:39] <StevenK> wgrant: Hmmmm?
[05:40] <wgrant> StevenK: That's the old message
[05:40] <wgrant> You used the old message in changeOverride
[05:44] <StevenK> wgrant: PopulatePreviewDiffMergeProposal is done on prod, landing the branch to drop it
[05:44] <wgrant> StevenK: Do run a query to confirm that it worked first, pls.
[05:51] <StevenK> wgrant: The ddeb MP has updated
[06:20] <StevenK> wgrant: Can haz +1?
[06:26] <wgrant> StevenK: Should that be OverrideError rather than AssertionError? Otherwise +1
[06:27] <StevenK> Oh, I didn't realize there was an OverrideError
[23:55] <StevenK> Bah, MPJs can't really run on qas either
[23:55] <StevenK> But they're run under cron