[00:03] <JesperHansen> lifeless: and maybe I should give you a link to a bugreport that has RESDUP status
[00:09] <JesperHansen> https://bugzilla.mozilla.org/show_bug.cgi?id=454993 - this is the most recent one. The See Also modification was done 12 hours ago
[05:25]  * wgrant ponders adding "You are probably wrong." to the +nominate template.
[06:04] <lifeless> wgrant: hah
[06:04] <lifeless> wgrant: you saw that lovely little chat?
[06:05] <lifeless> wgrant: got anything you need reviewed?
[06:08] <lifeless> wgrant: also did you have any luck with https://bugs.edge.launchpad.net/malone/+bug/637854 ?
[06:50] <lifeless> wgrant: also https://bugs.edge.launchpad.net/soyuz/+bug/276950
[06:57] <lifeless> wgrant: you might like 'https://bugs.edge.launchpad.net/malone/+bug/366134' for an example of insane code.
[08:46] <wgrant> lifeless: I could really use a statement log for 276950.
[09:02] <lifeless> wgrant: you've got mail
[09:05] <wgrant> lifeless: Thanks.
[09:17] <wgrant> lifeless: That statement log is nauseatingly bad.
[09:17] <lifeless> yes
[09:18] <wgrant> I wonder if createMissingBuilds sucks in as-yet-unknown ways
[09:18] <lifeless> I like to think of it as 'plenty of room to improve'
[09:18]  * wgrant pokes.
[09:19] <wgrant> It does suck, but not the several-hundred-DAS-queries level of suckage.
[09:19] <lifeless> wgrant:     2397  OOPS-1721H1326  Archive:EntryResource:syncSource
[09:20] <lifeless> thats bad
[09:20] <lifeless> its a query count
[09:20] <lifeless> of course, worst is
[09:20] <lifeless>     2442  OOPS-1721L267   BugTask:+index
[09:20] <wgrant> lifeless: Same issue.
[09:20] <wgrant> Or similar.
[09:20] <wgrant> syncSource has copy checker + build creation
[09:20] <wgrant> == fail
[09:21] <lifeless> wgrant: If you can, writing a query scaling test would be a great idea
[09:21] <wgrant> Build creation is terribly naive, I knew that. But it seems to be far worse than I thought.
[09:22] <wgrant> Waitwhat.
[09:23] <wgrant> The P-a-s code is awful :(
[09:24] <wgrant> Oh, no, not too bad.
[09:24] <wgrant> Then what is it...
[09:25] <lifeless> man, I so want to turn memcache off and try https://bugs.launchpad.net/ubuntu/+bug/1/+index
[09:25] <lifeless> I just needs a losa
[09:28] <wgrant> Uggggh.
[09:30] <wgrant> Garrrr.
[09:30] <wgrant> SPR.getBuildByArch is the source of all our problems.
[09:30] <wgrant> It is shocking.
[09:30] <wgrant> It walks all the way up the series graph.
[09:30] <wgrant> Grabbing the corresponding DAS all the way up to Warty.
[09:31] <wgrant> Then *for each ancestor series*, it finds the distro archives.
[09:31] <wgrant> Despite the fact that the distro archives are for the entire distro.
[09:33] <wgrant> So that easily cuts out 1/3 of the queries.
[09:33] <wgrant> Ah, it attempts to rationalise that insanity.
[09:33] <wgrant> But it's still insanity.
[09:35] <persia> What's the value of P-a-s again?  Can it just be dropped?
[09:36] <wgrant> P-a-s is still wanted. And it's bad, but not as terrible as I first thought, so it can remain as it is.
[09:36] <wgrant> SPR.getBuildByArch is now the target of my fury... it has two callsites, one of which I was going to delete anyway, so it should be easy enough to fix.
[09:36] <persia> Well, I dis-want it, but I understand if others do :)
[09:37] <wgrant> It makes sense for some things...
[09:39] <persia> Go kill SPR.getBuildByArch : When I have time to get annoyed enough to really kill it, I'll benefit from from a long explanation.
[10:32] <wgrant> lifeless: <3 StormStatementTracer and HasQueryCount.
[10:32] <wgrant> Er, StormStatementRecorder
[13:01] <wgrant> lifeless: In the unlikely event that you're still around, could you see if there's any reason bug #65712 is still private?
[13:02] <wgrant> It's referenced in tests for the code I'm replacing, so it would be nice to see what it was about.
[13:03] <wgrant> I suspect I know its subject, but I would like to be sure.
[13:13] <jml> wgrant, I'll take a look
[13:14] <wgrant> jml: Thanks.
[13:14] <jml> wgrant, I've made it public.
[13:15] <wgrant> jml: Even better.
[13:15] <wgrant> Thanks.
[13:17] <jml> wgrant, np
[13:30] <wgrant> Bogus sampledata really does make for incredibly awful doctests.
[19:58] <lifeless> wgrant: I'm glad you like it
[19:58] <lifeless> jml: thanks
[20:03] <lifeless> jml: did you perhaps see my testtools branch?
[21:05] <jml> lifeless, I did
[21:05] <jml> lifeless, I have comments
[21:05] <jml> lifeless, but have yet to have an opportunity to write them down
[21:06] <lifeless> ah
[21:06] <lifeless> jml: I had hoped it would be a no brainer after our phone conversation
[21:13] <lifeless> I wonder if we need tags-per-task to make bugs like https://bugs.edge.launchpad.net/launchpad-registry/+bug/597738 work properly with QA processes
[21:19]  * beuno 's head hurts when thinking about the UI for such a thing
[21:21] <lifeless> beuno: indeed
[21:21] <lifeless> beuno: right now our QA process resets with each landing
[21:21] <lifeless> its ... bad
[21:22] <beuno> hrm
[21:22] <lifeless> we may need a separate thing to tags
[21:22] <lifeless> its all up in ze air atm
[21:23] <beuno> QA could be a thing on tis oqn on Launchpad
[21:23] <beuno> although I cringe at the idea of adding antoher knob
[21:23] <lifeless> beuno: EPARSE
[21:24] <lifeless> 'tis' 'oqn'
[21:24] <lifeless> your spanish words are confusing and infuriating me :P
[21:26] <beuno> it's still Sunday for me  :)
[21:27] <beuno> "on its own"
[21:27] <lifeless> rotfl
[21:27] <lifeless> ah I'd no chance of getting that did I.
[21:27] <lifeless> anyhow, yes, it could be.
[21:27] <lifeless> That would be an alternative to more bug statii
[21:34] <lifeless> grah
[21:34] <lifeless>     module = original_import(name, globals, locals, fromlist, level)
[21:34] <lifeless> ImportError: No module named mailman.monkeypatches.defaults
[22:03] <wallyworld> morning
[22:25] <lifeless> thumper: around?
[22:41] <thumper> lifeless: am now
[22:42] <lifeless> see -reviews
[23:24] <jml> lifeless, approved
[23:24] <lifeless> jml: thanks
[23:24] <jml> beuno, when are you visiting London next?
[23:24] <lifeless> jml: do you have anything else in the pipeline? perhaps we should release?
[23:25] <jml> lifeless, for testtools?
[23:25] <lifeless> yes
[23:25] <jml> lifeless, nothing in the pipeline. I'd like to fix the thing about matchers always including matchee in the stack trace
[23:25] <jml> lifeless, but I don't think that should block a release
[23:26] <lifeless> jml: I'm not sure what the right thing is there, though I too would like to meet james_w's needs
[23:26] <jml> lifeless, agreed
[23:27] <beuno> jml, not super soon, no. How about you buenos aires?  :)
[23:27] <jml> beuno, nothing planned yet.
[23:28] <lifeless> hmm
[23:29] <lifeless> there is definitely something funny with oopses happening
[23:30] <beuno> jml, had anything planned?
[23:31] <jml> beuno, no, just saw you online & thought it would be good to catch up
[23:34] <beuno> jml, yeah, this second part of the year I'm not getting on a plane  \o/
[23:34] <beuno> well, except for kiko's wedding
[23:34] <jml> beuno, zowie
[23:34] <jml> beuno, I'm toward the end of a 96 day stint in the UK :)
[23:35] <jml> beuno, my longest yet.
[23:35] <beuno> jml, congrats!
[23:36] <thumper> beuno: kiko's getting married?
[23:36] <beuno> thumper, I hope so, otherwise I'm going to be stranded somewhere in Brazil soon
[23:36] <thumper> beuno: when?
[23:37] <beuno> thumper, Oct 2nd
[23:37] <thumper> cool
[23:47]  * thumper lands the removal of branchrevision.id
[23:47] <jml> g'night folks.
[23:48] <lifeless> night jml
[23:53] <thumper> lifeless: is buildbot down?
[23:57] <wgrant> thumper: Is that going to make the next rollout take like two weeks?
[23:57] <thumper> wgrant: faster than making the id a bigserial
[23:58] <thumper> wgrant: I'm working on other branchrevision tweaks too
[23:58] <thumper> wgrant: so we store less
[23:58] <wgrant> Aha.
[23:58] <lifeless> thumper: fallout from the reboots
[23:58] <lifeless> spm will fix
[23:59] <thumper> lifeless: ok