[00:27] sinzui: you were right. before ff bug count = 21961. after = 21984 [00:29] unless we had those new bugs filed in the few minutes between running the queries which is possible [01:27] wallyworld: Could you please review https://code.launchpad.net/~wgrant/launchpad/trim-dak-utils/+merge/77629? [01:27] sure [01:28] as soon as the diff updates [01:28] bring on longpoll [01:29] Hmm, it should already be up to date. [01:29] The warning is gone now. [01:42] wgrant: did you run lint? there's an unused import in sync-source.py [01:50] wgrant: the error handling sematics seem to be changed. dak_utils.fubar wrote to stderr and exits. now we throw an exception. will the caller of the script "do the right thing" when it receives an exception? do we need to change any tests to match the new behaviour? [01:51] wallyworld: I find it amusing that you think the scripts in question are tested. [01:51] StevenK: they're not? [01:52] I suspect they aren't. [01:53] oh well. i thought the question was worth asking [02:03] qa-tagger disabled [02:03] I'm playing with where it is published [02:16] wallyworld: Oh, I guess sync-source.py might possibly be worth linting. [02:16] wallyworld: The others I dare not try. [02:16] nigelb: Do you happen to be here? [02:16] Haha [02:17] wallyworld: But the only unused import I see is _pythonpath. [02:17] Which isn't unused at all. [02:17] wgrant: that's the one [02:17] wallyworld: All scripts need to import that. [02:17] To get the right python path. [02:17] Imports with sideeffects ftl. [02:17] oh, ok. it was showing up as unused for me [02:17] It is unused. [02:17] It just has side-effects. [02:17] All scripts do it :( [02:17] right. [02:17] Yay buildout. [02:18] Those two scripts are hopefully not long for this world, anyway. [02:18] the other questions about error handling? [02:18] As for the dak_utils.fubar replacement... it's a LaunchpadScript now (I ported it two weeks ago), so raising LaunchpadScriptFailure logs an error and exits. [02:18] And there are no tests. [02:18] Well, no tests for error handling. [02:19] qa-tagger re-enabled, please visit our new location at http://lpqateam.canonical.com/qa_reports/deployment-stable.html [02:19] Just a couple of trivial end-to-end functionality checks. [02:19] ok. thatnks for the explanation [02:19] StevenK: Excellent. [02:19] StevenK: Although it is tempting to s/qa_reports/qa-reports/ [02:19] And redirect in place [02:19] wgrant: I can if you wish [02:19] Underscores make me sick :) [02:19] LP forbids them in most places. [02:20] wgrant: So, do eet, or we don't care? [02:20] Might as well do it. [02:20] Pretty URLs are nice. [02:20] Thanks. [02:21] Everything changed. [02:21] Drafting announce message to -dev [02:22] Although I am glaring disapprovingly at your tag-nator-o-matic.sh diff :) [02:22] Why? [02:22] wgrant: r=me [02:22] wallyworld: Thanks. [02:22] StevenK: Hardcoding devpad-specific paths is evil. [02:22] The old way was slightly evil too. [02:22] But this is worse. [02:23] * StevenK handwaves [02:23] Patches welcome, etc, etc. [02:23] StevenK: You probably also need to tell the dashboard about the new paths. [02:23] Ah [02:23] Yes, since I moved them [02:24] carob doesn't have bzr grep. For shame. [02:26] wgrant: Dashboard's config fixed. [02:26] Thanks. [02:33] Oh, staging's DB actually managed to restore last weekend. [02:34] Gasp! [02:34] Or something. [02:40] wgrant: Did you see my kill-+viewstatus MP? [02:41] StevenK: I did. [02:41] 92 failures, 15 errors [02:41] lol [02:42] wgrant: However, I wanted to pick your brain -- due to removing +viewstatus, some TAL now fails due to data/task_link being gone, or something [02:43] wallyworld, The sql statement has still *not* completed [02:43] Haha [02:43] * StevenK points sinzui at the launchpad-dev mailing list [02:43] sinzui: bollocks. perhaps we just set them all the true [02:44] just to get some data there [02:44] hooray feature flags are still working on prod [02:44] wallyworld, Maybe we should consider restoring the db to qastaging more than twice a year [02:44] That would be nice. [02:44] sinzui: +1 on that [02:45] then this issue would not have happened [02:45] TBH, I think sourcherry can just restore qas's db after stagings [02:45] once a day would be nice [02:45] Sure, if you want qas down for six hours [02:46] six hours! jeez [02:46] StevenK, congrats on making things a bit more open [02:46] iwbni that page also showed recently deployed revisions [02:46] poolie: Thank you :-) [02:46] poolie: That's lpqateam.c.c, which is already open [02:46] rather than cutting off at the point of what is actually deployed now [02:47] Indeed. [02:48] wgrant: http://pastebin.ubuntu.com/699465/ [02:48] wallyworld: That's roughly how long staging is down over the weekend [02:49] wallyworld: A DB restore takes about 24 hours. [02:49] And the service has to be down for a little while (not quite 6 hours, though). [02:50] why can't we just use normal replication type processes? ie feed across archive logs as they occur (to use Oracle terminology) [02:50] qastaging and staging have writes. [02:50] And there is data we prune before restore [02:50] Syncing archive logs won't help unless we then use PITR to regress the DB. [02:51] so no easy naswer [02:51] Plus using backups means we have evidence that our backups work. [02:51] true [02:51] StevenK: THe problem there should be pretty obvious :) [02:51] StevenK: You no longer define task_link... [02:52] - task_link = edit_link if can_edit else view_link [02:53] wgrant: Right, but what the heck can I set it too [02:53] edit_link is probably fine [02:54] The number of people who have deliberately gone to +viewstatus in the last three years is probably vanishingly small. [02:54] So just task_link = edit_link = ... ? [02:54] Yes. [02:54] Right [02:55] wallyworld: If you want we can probably kick off a qastaging restore now. [02:56] wgrant: that would save sinzui some sql grief [02:56] Although just before the weekend might not be ideal, because it's not automatic. [02:56] But say we do it on Monday morning. [02:56] You can land your DB patch on Tuesday. [02:56] ok [02:57] it's only a column not null patch so no biggie [02:57] monday morning apac time is probably the best time for having a losa around but not disrupting people too much [02:57] mwhudson: not when it's a public holiday [02:57] yes [02:57] Is it really? [02:57] is it a public holiday on monday? [02:57] lol ACT [02:57] NSW/ACT [02:57] lol NSW too. [02:57] Which public holiday? [02:58] Labour Day [02:58] does it matter? :-) [02:58] We have nothing between like June and November. [02:58] Ahh. [02:58] So we celebrate by doing none. [02:58] ACT has our silly public hol a week later. [02:58] Family and Community Day, or something. from memory. [02:58] Oh great, AFL Grand Final this weekend. [02:58] So Vic is having ALL of next week off. [02:58] Heh [02:59] % bzr log | head -n 7 | tail -n 1 [03:00] did the fuss about someone claiming that the reason women don't get paid as much as men was down to women taking time off to have their periods make it over the ditch? [03:00] http://twitter.com/#!/Hilary_Barry/status/88679250365919232 in any case [03:00] Heh. [03:00] spm: it's the NRL final too, so i think NSW will be hung over for a weel too [03:00] *week [03:00] Both on one weekend? Is that normal? [03:00] no idea [03:01] i don't live in/come from victoria, why would i know anything about afl? [03:01] ^^ this [03:01] Heh, indeed. [03:01] and yes, I believe it is very normal. totally difference markets, so no real overlap. [03:02] different* [03:02] i only really know about the nrl final because the nz warriors are in it [03:02] Silly Freenode [03:03] still chuckle at the pic a mate had; the poster reading I support NZ and any team playing against Australia. [03:04] ... unless it's the springboks, i suspect === wgrant changed the topic of #launchpad-dev to: https://dev.launchpad.net/ | On call reviewer: wallyworld | Critical bugs: 254 - 0:[#######=]:256 [03:09] Gasp [03:10] StevenK: I am now [03:10] StevenK: \o/ Public deployment reports! [03:11] nigelb: I thought you'd be excited [03:11] nigelb: I've requested the deployment of your statusexplanation patch for tonight. [03:11] StevenK: First mail I opened today :) [03:11] Careful. Now nigelb might explode with happiness [03:11] wgrant: Thanks! I was planning on poking you about that. [03:11] hahaha. Entirely possible. [03:13] Mailing list archive still doesn't have my mail, though. [03:16] How does db-devel work for deployment? [03:17] nigelb: We push the desired revision to the DB master (wildcherry), run full-update.py to apply the patches, then merge that revision into devel. [03:18] ah, more manual process than for devel? [03:22] nigelb: Pretty similar. [03:22] huwshimi: Did you find a solution to your Unity oddness? [03:22] huwshimi: I just upgraded and got the same thing. [03:23] ah, I need to sort out an issue I had with my second db-patch. [03:23] wgrant: No, I was going to report a bug, but didn't know how so haven't yet [03:23] thumper: Is Nautilus' root window meant to be unhappy with sitting entirely on the screen now? [03:24] wgrant: ahhh... whut? [03:24] wgrant: It's a new Unity feature called The Void [03:24] thumper: Like this: http://imgur.com/GeKHl [03:25] thumper: The desktop is out of aligment, but everything works (except the void behind) [03:25] that looks like a fubar [03:26] thumper: windows, dash, launcher etc. all sit on top of the void with no problems [03:26] DnD in the shifted root window is also a bit odd. [03:27] It detects the drop at the location it would be if it wasn't shifted. [03:27] But normal clicks work fine. [03:27] This reminds me - is everything alright with Mark's dictionary? :) [03:27] nigelb: It is getting a bit late :( [03:27] wgrant: I feel weird with UDS scheduling started and no name yet. [03:29] thumper: people.canonical.com/~wgrant/what.png is mine [03:30] * wgrant tries it on different hardware. [03:31] Bug #862743 [03:31] <_mup_> Bug #862743: Desktop drawn with offset

< https://launchpad.net/bugs/862743 > [03:31] wgrant: Is that one dialog box that you dragged in "the void"? [03:31] nigelb: Yes. [03:32] There's no window ther. [03:32] Ew. [03:32] So it starts as what looks like uninitialised GPU memory. [03:32] 3 weeks to release? WIN. [03:32] 3 weeks? [03:32] 2 weeks - 1 day :) [03:32] Oh, right. Its friday. [03:33] Hm, Ubuntu Mono is pretty nice. [03:33] Screenshot plz? [03:33] Upgrade from Lucid :P [03:33] Maybe soon. [03:34] I don't want to setup about 5 to 6 build environments again. [03:34] Launchpad, firefox, ubuntu web projects, mozilla web projects, work env. [03:40] And now X crashes. [03:46] wgrant: Bwahaha! I haz Ubuntu Mono [03:47] I remembered that Ubuntu members have access to the private PPA :) [03:47] And yeah, its pretty neat [03:50] nigelb: You can actually download it from http://font.ubuntu.com/ too [03:50] huwshimi: yeah, but the PPA was easier :) [03:50] huwshimi, wgrant: please file bugs with pictures and graphics drivers [03:51] thumper: In addition to the existing 36-affected bug #862743? [03:51] <_mup_> Bug #862743: Desktop drawn with offset

< https://launchpad.net/bugs/862743 > [03:51] wgrant: nah [03:51] There's no proprietary driver listed in that bug, so it's probably not fglrx's fault. [03:54] huwshimi: what GPU? [03:55] wgrant: also, what's your GPU? [03:55] thumper: erm [03:55] thumper: How do I find out? [03:55] thumper: I used to know this stuff, but Ubuntu has been to easy for too long :) [03:56] huwshimi: glxinfo [03:56] thumper: Radeon 5700 of some variety. [03:56] With whatever fglrx oneiric has these days. [03:57] Radeon HD 5700, that is. [03:57] * thumper is passing information on [03:57] thumper: it's an AMD of some kind [03:57] My GM45 should be upgraded in a few minutes. [03:57] Will see if that is borked too. [04:00] wgrant: The fix in that thread fixed it for me [04:00] *thread/bug [04:00] what was the fix? [04:00] "pkill nautilus" [04:01] heh [04:01] thumper: https://bugs.launchpad.net/ubuntu/+source/unity/+bug/862743/comments/12 [04:01] <_mup_> Bug #862743: Desktop drawn with offset

< https://launchpad.net/bugs/862743 > [04:03] huwshimi, wgrant: have either of you had that happen after a reboot (when not upgrading)? [04:03] thumper: I've never seen it before today. [04:03] Ah, pkill nautilus worked for me this time. [04:04] thumper: Yeah I rebooted after the upgrade to see if that would fix it. It didn't. [04:04] I tried it half an hour ago and it just crashed unity. [04:04] And X crashed 30s later, before I could get to apport. [04:05] Which caused compiz to crash again, wiping out the original report. [04:08] * thumper snorts [04:08] StevenK: Its been an hour and I still don't see your email in the web archive :( [04:08] that was definitely badly timed :| [04:12] thumper: Affects my Intel GM45 too. [04:12] weird === wallyworld is now known as Guest64579 [04:15] I see wallyworld is eshewing underscores for esoteric guest nicks. [04:23] wgrant, huwshimi: looks like jason just committed a fix for that [04:23] thumper: :D [04:24] thumper: Great. [04:24] Ah, I see. [04:42] StevenK: How often does Launchpad mailing list archive get updaetd? [04:43] Heh. [04:43] Poor forster is a little overloaded at the moment. [04:43] It's meant to be ~instantaneous. [04:43] But ubuntu-x-swat has a very large archive. [04:43] Its ben ~2 hours :P [04:43] *been [04:44] A week ago it was around two days behind :/ [04:44] I fear we may need to disable ubuntu-x-swat's archive. [04:45] my theory was it hates StevenK. This explanation works as well :P [04:46] That works. === jtv1 is now known as jtv === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === jtv1 is now known as jtv [05:43] wgrant: Do we still use a DebugLayer for anything? [05:44] I'm still tempted to rip out config.canonical.show_tracebacks in this branch, too. [05:45] We still run a separate DebugLayer HTTP listener in dev. Not sure why. [05:45] if canonical.launchpad.layers.DebugLayer.providedBy(self.request): [05:45] self.debugging = True [05:55] hah, bug has connections! https://bugs.launchpad.net/ubuntu/+source/software-center/+bug/752790/comments/6 [05:55] <_mup_> Bug #752790: unbuntu software center

< https://launchpad.net/bugs/752790 > [05:58] Holy heck that diff generation was fast [05:58] wallyworld: https://code.launchpad.net/~stevenk/launchpad/kill-plaintext-traceback-rendering/+merge/77647 [05:59] wgrant: Do you notice that Ubuntu Mono is smaller than the default mono? [05:59] (well, the old monospace font) [05:59] nigelb: Significantly so :( [06:00] Yeah, I had to increase font size from 10 to 12 [06:05] How can I figure out which revision removed a file? [06:15] StevenK: I tend to "bzr log -v" then search through that. [06:15] There may be a better way. [06:16] StevenK: doing it now === wgrant changed the topic of #launchpad-dev to: https://dev.launchpad.net/ | On call reviewer: wallyworld | Critical bugs: 247 - 0:[#######=]:256 [06:16] StevenK: do you know the file? [06:17] you should be able to do bzr log /path/to/file [06:17] That says not versioned [06:17] Either way, bzr log -v is a good idea [06:17] That's intersting. I just tested it on a standalone repo on my machine and it worked. [06:17] And one more tech-debt bug closed. [06:17] \o/ [06:18] I have a tech-debt bug to close as well. [06:21] StevenK: i wonder if PageTestLayer itself is still used at all? [06:22] wallyworld: Story tests [06:22] ok [06:23] StevenK: r=me [06:35] wgrant: did you just assign an ancient "we should rewrite cron.publish-ftpmaster" bug to me just so I could get the karma for it being Fix Released? :) [06:44] jtv: Probably. [06:44] I am gardening. [06:44] Well thanks. [07:01] [16:44:06] I am gardening. <== with a chainsaw? at ~ 3-5 inches above the ground? [07:02] That's the best way to do it. [07:03] spm++ [07:04] my F-in-law is a former cane farmer. from tropical queensland. his idea of a light prune involves a chain saw at a similar "height" [07:05] With wgrant's history of deleting large parts of launchpad codebase, I'm guessing his idea of gardening is similar. [07:07] HEH [07:07] oops. sorry, caps fail === wallyworld changed the topic of #launchpad-dev to: https://dev.launchpad.net/ | On call reviewer: - | Critical bugs: 247 - 0:[#######=]:256 [08:01] morning [08:02] Morning bigjools. [08:03] Have a nice weekend everyone [08:03] hello huwshimi [08:03] thanks for the icons [08:03] \o huwshimi [08:04] bigjools: Hey. No problems :) [08:04] bigjools: hey [08:04] bigjools: so I may have been a bit cryptic on the bug - let me know if os [08:04] *so* [08:05] lifeless: not even in my email yet [08:05] lifeless: oneiric's network manager doesn't write resolve.conf properly for static interfaces .... screws me every day [08:05] but thanks :) [08:05] bigjools: win! [08:05] not so much! [08:05] sarcasm!! [08:06] irony! [08:06] !!! --- !!! [08:06] good morning [08:13] wallyworld: wgrant: win - OOPS-2099AR35 - fallout from updating subscriptions when retargeting [08:15] lifeless: you mean associating a bug with a different project/distro as opposed to changing privacy/security settings ? [08:15] yes [08:15] its correctly unsubscribing launchpad-security [08:15] but the redirect is to a 404 (and so oopses as its internally generated) [08:15] i'm not aware of what actions we take when retargetting. i just did the privacy/security stuff [08:16] when one retargets such that one can't see, we probably want to return the,m to the original context [08:16] that sounds reasonable [08:16] wallyworld: I know, its a corner case we didn't identify in advance [08:16] i assume this can wait till next week? [08:16] I'm filing a bug for it. I'll tag it disclosure (as disclosure made it possible to happen). [08:16] yes, it will be rare. [08:17] cool. i'll make sure it's fixed first up === Beret- is now known as Beret === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan [08:45] bigjools: hi, is bug 241646 still a bug ? [08:47] jcsackett_: thanks! (yes, the Ubuntu archive admins are to the best of my knowledge the only users of this) [08:54] lifeless: technically yes, but I've never seen it a problem in practice [09:05] wallyworld: hi === adeuring changed the topic of #launchpad-dev to: https://dev.launchpad.net/ | On call reviewer: adeuring | Critical bugs: 247 - 0:[#######=]:256 === al-maisan is now known as almaisan-away [11:44] wgrant: I'm adding rabbitmq-management to lp-messagequeue-deps now that you packaged oneiric. Current version is 0.99, might either head to 0.99.1 or bit the bullet and go 1.0 [11:46] bigjools: Or 0.100..? [11:46] But yes, good plan to add it. [11:46] I typed that and it looked horrible! [11:47] Better than 1.0 :) [11:47] what's this aversion to 1.0! [11:47] We're hardly stable :P [11:48] it's a f***ing dependency package! [11:48] We could also just drop the 0. and go straight to 100 [11:50] we really should do bug 288645 [11:50] * wgrant pokes _mup_ [11:50] Bug #288645 [11:50] grr [11:50] mup was a victim of the netsplit I think [11:51] Ah, that one. [11:51] Bug #288645: "Register a branch" on hosted branches is confusing [11:51] There's a lot of LP UI bits that could do with some fixing... [11:54] pls to be reviewing https://code.launchpad.net/~julian-edwards/meta-lp-deps/add-rabbit-management/+merge/77692 [11:54] gah, still "Updating diff...". If only we had some way of making that update automatically. [11:55] It often takes >30min at this time of night :/ [11:56] hmmm should I be targeting oneiric [11:57] does it matter with recipes? [11:57] Doesn't matter hugely. [11:57] Recipes will do it. [11:57] k [11:58] https://lpstats.canonical.com/graphs/BuildersActiveVirtual/20110924/20111001/ is somewhat upsetting. [11:58] Perhaps they will come back automatically. [11:59] StevenK: Whenever it is that I meet you, I'll buy you a beer :D [12:00] (Deployment report) [12:00] wgrant: no, we need to ping brendan [12:00] Most of them have been gone for slightly under a day. [12:00] But some have been gone for longer. [12:01] Not sure what's going on there. [12:01] they have a cron job to put them back each morning [12:01] Apparently not :/ [12:02] it's broken :) [12:02] I'm going to start the publisher on DF BTW [12:02] let it steam over the weekend [12:03] A -A or -P will crash. [12:03] Due to The 53. [12:03] grar [12:04] right, off for a liquid lunch [12:04] You could grab them into the pool manually and then run the rescue script. [12:04] There should only be 4 in oneiric that are problematic. [12:05]