[00:01] ok, I'm wandering off again, available on the phone if it breaks more/again/harder [00:01] Night. [00:01] * wgrant files a bug. [00:05] elmo: thanks [00:05] wgrant: why would teams need a mail address ;P [00:06] Bug #614858 [00:06] <_mup_> Bug #614858: Recipe builds fail when requester does not have a preferred email address [00:06] lifeless: Well, they often do (preferredemail is the contact address). [00:06] wgrant: fvso often [00:08] lifeless: True. [00:11] I'd want to do a stats run on that before thinking about it [00:24] Ursinha-afk: / matsubara-afk: are either of you around ? [01:59] 'morning wgrant, lifeless [02:08] hai [02:14] jelmer: have you considered sleeping ? [02:15] lifeless: what's that, "sleep" ? [02:15] its the root for sleeping. [02:15] :P [07:18] elmo: When you wake up... the world has collapsed again. [07:18] I can't see what's done it, but it's not the same recipe. [07:30] wgrant, Lucid, lmsensors-trunk, team-iquik [07:32] cody-somerville: Sigh. [07:34] ........................... [07:35] What the *fuck*. [07:35] Seriously guys. [07:40] Unbelievable. [07:47] Anyway, everything's working again. [09:04] Bah, I think things are broken again. [09:07] Yes :( [09:09] details? [09:10] I can't tell. [09:10] Needs log access. [09:11] 2010-08-08 09:10:06+0100 [-] startBuild(Lucid, dailylucid, nova-core (on http://thorium.ppa:8221/)) [09:11] 2010-08-08 09:10:06+0100 [-] Scanning failed with: 'NoneType' object has no attribute 'email' [09:11] * wgrant kills it dead. [09:12] I wonder why this is only happening now. [09:12] Maybe dailies weren't being created until recently. [09:12] Oh! [09:13] They would have never reached the head of the queue until recently, because of the queue fuckup over the last couple of weeks. [09:13] elmo: can just disable the daily build script until we fix it? [09:13] thumper: if you give me some hint as to how to do so, sure? [09:13] * thumper loks [09:14] It's request_daily_build.py, but NFI where it runs. [09:14] request_daily_builds.py, sorry. [09:14] not cesium [09:14] so it must be loganberry [09:14] checking [09:14] Yeah. [09:14] request_daily_builds.py [09:14] loganberry [09:15] possibly as bzrsyncd [09:15] yeah, it is. ok, commenting out it and the check for it [09:15] elmo: ack, thanks [09:15] Thanks. [09:15] I'll take an urgent look tomorrow morning [09:15] wgrant: is there a bug? [09:16] thumper: For one of the issues: bug #614858. [09:16] <_mup_> Bug #614858: Recipe builds break the world when requester does not have a preferred email address [09:16] ta [09:16] For the others which I discovered a couple of hours ago... not yet. [09:16] what was the other issue? [09:17] ok, all done [09:17] (disabling) [09:18] 2010-08-08 09:18:30+0100 [-] startBuild(Maverick, rekonq-daily, rekonq (on http://lychee.ppa:8221/)) [09:18] I wonder how many there are left. [09:19] Probably lots. [09:19] there are quite a few daily recipies now with team owners [09:19] Shall we play suspend-all-builds-without-a-preferred-email-address? [09:19] I'm game if it involves SQL and not me tailing a log and pasting broken ones to the channel [09:20] That's what I was thinking. [09:20] (esp. since I need to get on a plane at some point today) [09:20] * wgrant SQLs. [09:20] Ah, IS sprint? [09:20] wgrant: ta [09:20] wgrant: yep [09:40] * wgrant waits for download-cache to update... [09:48] elmo: Ahem. One query for the wrong (10.08) schema later, http://paste.ubuntu.com/474867/ appears to work fine. [09:54] Ok. [09:54] #launchpad intrigues me. [09:54] How has this error come up so many times today? [10:14] interesting [10:14] possibly things to check [10:15] a change in sql for something in registry? [10:15] a cp of a soyuz change on friday? [10:15] something else? [10:15] like latency for backlog catchup [10:23] wgrant: thanks; there was only actually 3 [10:33] \o/ am up to actual new functional in qa-tagger refactoring. [10:33] OTOH mars and ursula are going to kill me Monday :) [12:05] \o/ success [12:07] jml: did I see you just come online ? [12:07] lifeless, maybe [12:07] lifeless, wassup [12:07] I want to sketch a fixtures thing [12:08] if you're up for a little voice - very little - on that that would be nice [12:08] sure. [12:08] but I need coffee first [12:08] ~!5 [12:09] ~15m [12:09] ok [12:09] well, ping me when you're back [12:09] if I'm up, we'll chat [12:09] if not, it'll brew eventually anyhow :) [12:16] hi [12:17] lifeless, ping [12:18] hi [16:09] is this actually monday for you guys or you're just having fun? :) === daker_ is now known as daker [20:59] echo hi [21:40] lifeless: you might be interested to look at https://code.edge.launchpad.net/~james-w/launchpad/more-matchers/+merge/32057 where I felt Matchers didn't support me too well in what I wanted to do. I filed a bug on one thing that would help. It may partly be that keeping the existing assert* API didn't help. [21:43] + except NotFoundError: [21:43] + return self.new(name) [21:43] james_w: is new an instance method ? [21:44] lifeless: I don't think I wrote that? [21:44] or maybe that's in the prerequisite branch, which got included in this diff to start with. [21:44] its in the diff for https://code.launchpad.net/~james-w/launchpad/more-matchers/+merge/32057 that I got mailed to me ;) [21:46] right, again only mailing the first diff wastes people's time [21:58] james_w: where was this bug you filed ? [22:04] lifeless: https://bugs.edge.launchpad.net/testtools/+bug/615108 [22:04] <_mup_> Bug #615108: Please provide a way to match an iterable of matchees [22:05] morning [22:07] james_w: MatchesMany ? [22:07] james_w: or 'IsIterableOf [22:38] bbiab -> allergy shot [22:43] I can't seem to access MPs on staging. [22:43] OOPS-1681S1581 is a timeout, and OOPS-1681S1580 is an actual OOPS. [22:46] wgrant: is it always timing out? [22:46] it could be a staging librarian issue [22:46] One times out, the others OOPS. [22:47] The OOPS would make sense if the diff was gone. [22:47] wgrant: is there an oops for bug 614858 ? [22:47] <_mup_> Bug #614858: Recipe builds break the world when requester does not have a preferred email address [22:47] But timing out as well on others is a little odd. [22:47] thumper: No OOPS. [22:47] Just stuff in buildd-manager logs. [22:48] And it doesn't do OOPSes (yet?) [22:49] * thumper sighs [22:49] Why? [22:49] The problem is obvious. [22:53] wgrant: I don't know the buildd code as well as you [22:53] wgrant: where do you think the exact issue is? [22:54] thumper: lib/lp/code/model/recipebuilder.py -- search for 'preferredemail' [22:55] heh [22:58] Ahem, oops. [23:04] ok, I'll admit to have not worked on any real recipe stuff [23:05] wgrant: I'm surprised it is possible for a user to not have a preferredemail address, when does that happen? [23:06] jelmer_: They're deactivated, or it's a team without a contact address. [23:06] (I vaguely remember writing that code in Wellington) [23:06] jelmer_: And a team 'requests' a recipe build if they are the owner of a daily recipe. [23:07] wgrant: Why would a deactivated account not have a preferredemail ? [23:07] is that removed when they are deactivateD? [23:07] I think a deactivated person still has a preferred email [23:07] the problem here is a recipe owned by a team [23:08] jelmer_, thumper: Deactivation resets the preferredemail to NEW. [23:08] So yes, the person ends up without a preferredemail. [23:08] This may be how mail to deactivated users is prevented. [23:09] (See the end of Person.deactivateAccount) [23:09] wgrant: Thanks === jelmer_ is now known as jelmer [23:17] thumper: bug #615144 [23:19] wgrant: ta [23:19] * thumper afk to get the car a warrent [23:30] * jelmer waves goodnight