[00:31] aaaaa [00:31] bbbbb [00:31] want to make /bin/py start up in 0.07 seconds (ish) rather than 0.5 ? [00:32] touch lib/site.py [00:32] in other news: setuptools is bad [00:32] orly [00:32] does lib/site.py exist normally? [00:33] it has some behaviour that is quadratic in the length of sys.path [00:33] lifeless: no, touching it prevents setuptools' site.py from loading [00:33] hah [00:41] The other reason setuptools is bad? I can't find any way to make its namespace package support compatible with having system installed lazr.* and buildoutified lazr.* [00:42] well [00:42] eggs are bad [00:42] the rest follows [00:51] maxb: "the" other? :) [00:52] someone should write a rant [00:52] well, I was thinking more *immediate* reasons :-) [00:54] the overriding problem with setuptools is that people use it [00:54] which is not the correct thing to do with software from pje [00:54] LOL [00:54] quotes page.... [00:54] the correct thing is to be inspired by it [00:55] heh [06:38] Good morning! [08:45] morning [08:46] yo [08:46] spm: how goes the puller? [08:46] thumper: as expected. problem free now we have sufficient logging to detect the actual problem. [08:47] spm: excellent [08:48] mwhudson: hi [08:48] thumper: hello [08:49] mwhudson: were you able to glean any extra info from that puller log? [08:50] thumper: no [08:50] thumper: it looks totally impossible :/ [08:50] :( [08:50] did you cowboy any more logging? [08:50] no, no more yet [08:51] sun spots? [08:53] I hear that Dallas has the right kind of hats for us [08:59] spm: what is the branch puller delay? [09:00] thumper: I think it's this one: select greatest(0.0, extract(epoch from (CURRENT_TIMESTAMP AT TIME ZONE 'UTC' - min(next_mirror_time)))) from branch where branch_type <> 4; or a variant of. [09:58] mrevell: why does staging *still* say it's 2.2.7? To confuse the enemy? [09:58] Same for edge, actually [10:00] jtv: I think danilos or beuno have a branch to fix that [10:00] * beuno has no suck branch [10:00] *such [10:01] wow [10:01] I have filed a bug though [10:02] We could file one automatically for every cycle. [10:02] (No, _not_ serious) === dpm__ is now known as dpm === mthaddon changed the topic of #launchpad-dev to: Code hosting offline 10.00-10.40 UTC 1st Oct -- http://is.gd/3MyHY | This is Launchpad Development Channel | Week 00 of 3.1.10 | PQM is OPEN | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting | On-call review in #launchpad-reviews | Use http://paste.ubuntu.com/ for pastes | This channel is logged: http://irclogs.ubuntu.com/ === mthaddon changed the topic of #launchpad-dev to: This is Launchpad Development Channel | Week 00 of 3.1.10 | PQM is OPEN | https://dev.launchpad.net/ | Get the code: https://dev.launchpad.net/Getting | On-call review in #launchpad-reviews | Use http://paste.ubuntu.com/ for pastes | This channel is logged: http://irclogs.ubuntu.com/ [11:57] rockstar: *ping* [11:58] I managed to pinpoint the error back to Python that doesn't like to be inside a Vmware Server 2.0.1 [11:59] The people from VMware were shocked to see that this could happen and are thinking about ways to make sure this keeps working === jtv is now known as jtv-afk [12:19] jtv: or we could just not /have/ a version number in the footer [12:19] Fly-Man-: wow. what did vmware say was the problem? [12:20] Python version 2.6 from Jaunty [12:20] It seems that over the period of time [12:20] with the rocketfuel setup [12:20] it sets up 2.4, 2.5 and 2.6 [12:20] and those seems not to be happy with each other and the tools that VMware released [12:21] huh. interesting. [12:21] at least, not with the machine i'm running it on :p === Ursinha-afk is now known as Ursinha [13:14] select milestone.name [13:14] from milestone,product [13:14] where product.name='launchpad' [13:14] and product.id = milestone.product [13:14] and dateexpected < current_timestamp at time zone 'utc' [13:14] order by dateexpected [13:14] limit 1; [13:14] which would work if we bothered to keep our data up to date ;) [13:15] oh - we do. I just forgot the desc [13:15] select milestone.name [13:15] from milestone,product [13:15] where product.name='launchpad-foundations' [13:15] and product.id = milestone.product [13:15] and dateexpected < current_timestamp at time zone 'utc' [13:15] order by dateexpected desc === matsubara-afk is now known as matsubara [13:41] kfogel: i am an idjit :) see my last follow up on that thread. i think there's no bug there [13:44] barry: so, uh, what was the bug Henning saw, though? [13:49] kfogel: i don't know [13:57] abentley: hello [13:57] ? [14:04] thumper: Hi. [14:05] abentley: hey, how's it going? [14:05] It's going fine. Today just started, but I'll be trying to get an improved bzr cherrypicked. [14:06] thumper: And then back to the inline-commenting stuff. [14:06] abentley: how much faster is the preview diff generation with your patch? [14:06] abentley: and has that patch gone/going into bzr.dev? [14:07] abentley: it is bzrlib isn't it? [14:07] thumper: On my example test, it is 22x faster, but that will vary a lot because it changes O(n^2) scaling to O(n) scaling. [14:07] excellent [14:07] example test was what? [14:07] not the mysql test? [14:07] I have the mysql branches locally [14:08] if you want me to compare [14:08] thumper: It has landed in the 2.0 branch, and will propagate to bzr.dev and appear in the next release. [14:08] * thumper nods [14:08] has an official tarball available, or are we going to add our own? [14:09] thumper: I don't have the patience to run the sql merge without the patch, but with the patch, it completes in about a minute. [14:09] awesome [14:09] without the patch it is over an hour [14:09] (on my machine) [14:09] thumper: Just under an hour on loganberry. [14:10] that's awesome [14:10] so... [14:10] adding a new bzr tarball to the download cache? [14:11] how is the inline editing widget work coming along? [14:11] code folks! How do I create a stacked branch in a test? Is passing stacked_on to factory.makeBranch enough? [14:11] thumper: Yes, and setting up a corresponding branch. [14:11] jtv: yes, if you just want a stacked test branch [14:11] jtv: it won't create the branch on disk [14:12] thumper: I've got it working just for comments, and it puts the comment in the wrong place, and it doesn't support replies. [14:12] oh kay... [14:12] thumper: I want to reproduce bug 375013 in a test, basically [14:12] Bug #375013: Commit to a stacked branch does not work [14:12] jtv: ah [14:13] jtv: you want to create real branches [14:13] yes [14:13] * thumper tries to remember test examples [14:14] jtv: lib/codehosting/scanner/tests/test_bzrsyncd.py:537 [14:14] jtv: is an example of creating a real stacked branch [14:15] thumper: just what I was hoping for, thanks. [14:15] separate call to set_stacked_on_url, I see. [14:15] That'd do it, yes :) [14:16] thumper: stacked_on_branch is kinda sucky as a variable name... why not call your branches "pea" and "mattress"? [14:17] * thumper shoots jtv with a pea shooter# [14:17] (Of course since I'm testing a stacking limitation, my test would be called Princess) [14:17] jtv: i'd have to reject that branch [14:17] On *no* account look up "It's Princess!" (by the makers of South Park) on the intertubes [14:17] thumper: but you'd be grinning while doing so [14:24] Anyone got any idea why I'd be seeing timeouts in googletestservice.py during a test run? [14:26] ah netsplit [14:30] gosh isn't bigjools annoying [14:31] heh [14:46] gmb: ping [14:47] thumper: Hi [14:47] gmb: https://code.edge.launchpad.net/launchpad/+activereviews [14:47] gmb: you have two old approved branches [14:47] gmb: have they landed? [14:47] * gmb looks [14:48] gmb: they may be not identified correctly as we switched formats around then [14:48] thumper: No, they're branches that are waiting on subsequent branches before they can be landed. They're still unmerged. [14:48] * gmb wishes there was a way to express that that +activereviews could understand. [14:50] gmb: why can't they be landed? [14:50] gmb: what would happen now if you land them? [14:50] gmb: will it break something, or just be unused code? [14:52] thumper: Things would break. [14:52] * thumper wonders how to show this [14:53] thumper: They're large branches, or rather they're logical blocks of large branches. [14:53] Which I had to leave because of more pressing work. [14:53] hmm.. [14:54] ideally we'd want to have a way to say "this is approved but blocked" [14:54] so you won't get people like me bitching about it [14:54] Heh. [14:54] Right :) [14:55] thumper: Want me to file a bug for this so that you can get on with team leadery things? [14:55] sure [14:56] Will do. [15:10] thumper: https://bugs.edge.launchpad.net/launchpad-code/+bug/440011 [15:10] Bug #440011: There should be a way to say that a merge proposal is approved but blocked [15:21] ta [15:23] maybe I should ask here ... [15:23] I want to use a TAL formatter in a view and I found out I can do it this way: https://pastebin.canonical.com/22831/ [15:23] Is that the right way to do it? Or should I go through the registration somehow? I remember seeing an example once that was different ... === joey` is now known as joey === thekorn_ is now known as thekorn [16:27] Ursinha, sorry about missing the meeting. Drop off at the vet took longer than expected. [16:28] no problem rockstar [16:28] do you know why is update_review_diffs script failing? [16:29] Ursinha, yes, and abentley had a hack to fix it, and I think a correct fix as well. [16:30] rockstar: The "correct" fix stops it failing in this case, but not all cases. [16:30] abentley, yeah, I think you mentioned that last night. [16:34] Ursinha: It's been discussed on the list. See the threads under Scripts failed to run: loganberry:update_preview_diffs [16:46] thanks abentley and rockstar [17:01] my first MP for launchpad! :-) https://code.edge.launchpad.net/~mbp/launchpad/mbp-trivial/+merge/12732 === matsubara is now known as matsubara-lunch [18:15] beuno: https://code.edge.launchpad.net/~loggerhead-team/loggerhead/trunk-rich [18:15] thumper, I LOVE YOU [18:43] gmb: /me is here now. :) [18:43] * kees waves to bryce__ [18:44] whew [18:45] debug output and test script posted to https://bugs.edge.launchpad.net/malone/+bug/329917 [18:45] Bug #329917: Changing a task's target using the API OOPSes as NotImplementedError

[18:48] kees: I think we might have discovered a race condition here. [18:48] kees: Normally, that NotImplementedError will only get called if you try to access the bug using a target on which it isn't reported. [18:49] kees: e.g. https://bugs.launchpad.net/evolution/+bug/42 when bug 42 isn't reported on evolution. [18:49] Bug #42: Bug description listed in task is not the correct description [18:49] Bug #42: Bug description listed in task is not the correct description [18:50] kees: So it's almost as if the new bugtask is being accessed before it actually exists, which seems strange. [18:52] gmb: I haven't caught any OOPSes myself (how can I find an OOPS via the API?) but I always see the HTTP-500 error. [18:52] Hrm. [18:52] kees: OOPSes aren't available via the API because they're not part of LP, as such; that's a separate system that Launchpad links to. [18:53] ah-ha. do they show up in the html test from the API failure? [18:53] barry: have you run ec2 lately? [18:53] kees: I'm not sure; I'll find out for you. [18:54] kees: So, I don't know what's causing the problem but I at least understand the root cause of the problem. I'll update the bug and we'll get someone on it tomorrow. [18:54] (since we're all in the UK atm) === Ursinha is now known as Ursinha-afk [18:54] gmb: ok, great, thanks! === matsubara-lunch is now known as matsubara [19:05] bzr 2.0.0-lp-1? [19:24] BjornT: ping [19:32] * kees hugs matsubara for r9611 :) [19:32] kees, it's probably not me who you want to hug :-) [19:32] matsubara: oh, er, well, you touched the bug. :) [19:32] kees, I just run a script that parses commit logs for bug and mark them as fix committed [19:33] * kees looks for Muharem [19:47] good morning [19:52] intellectronica: uhm... bug 329917 is certainly not invalid. [19:52] Bug #329917: Changing a task's target using the API OOPSes as NotImplementedError

[19:53] kees: read my explanation? it's a different bug from the one described originally and marked as duplicate, and needs to be fixed client-side [19:53] kees: i'm looking at the real bug as we speak [19:53] intellectronica: which should be the master for the issue pitti, me, bdmurray, and bryyce are seeing? [19:54] intellectronica: aH, 342355 [19:55] kees: i think bug #342355 is the real bug. the one Ursinha-afk filed is a red herring. it would be nice to fix it (in launchpadlib, not launchpad) at some point, but it's not a show stopper [19:55] Bug #342355: transitionToTarget for an Ubuntu package yields HTTP 500 error [19:55] i also described a workaround [19:56] intellectronica: ok, cool. I'm interested in 342355 then. :) can you milestone/confirm/high that one instead? [19:56] maxb: you there? [19:57] kees: i'm investigating that one now and will update as soon as i know what's involved. i suspect it's an easy one and can get a fix out quickly [19:57] intellectronica: rockin'! thanks :) [19:58] kees: do you happen to have the foo to make me part of the ubuntu-bugs team (on staging)? [19:58] intellectronica: let me see [20:00] intellectronica: doesn't look like it, let me poke bdmurray [20:01] rockstar: Have you tried "make sync_branches" recently? [20:01] intellectronica: bdmurray is afk for lunch, it seems... let me see if I can find jcastro [20:02] abentley, no, but I remember that thumper and I had an issue with it when we were messing with the branch rewrite stuff. [20:02] matsubara: can you? ^^^^^ i think you've got admin foo powers? [20:02] abentley, I think there was a configuration issue. [20:02] abentley, although it looks like mwhudson is awake and alert, so he might have some insight. [20:03] mwhudson: Have you tried "make sync_branches" recently? (It's not scanning the branches for me) [20:03] * mwhudson tries to work out what's in trunk [20:03] intellectronica, I don't have lp admin powers [20:03] abentley: it's possible the "pull_branches" rule isn't right [20:03] abentley: does 'make pull_branches' seem happy? [20:04] intellectronica: hrmpf, all the admins are idle. :P is there some action you need to take that I can help with instead? [20:04] abentley: another problem can be that /var/tmp/bazaar.launchpad.dev/mirrors/ ends up 0700 for some reason [20:04] kees: mbarnett should be here [20:04] mwhudson: Yes. Earlier runs of sync_branches reported successful mirroring. [20:04] abentley: odd [20:04] kees: no big deal. i'll need to reproduce it locally anyway, just wanted to see what actually happens one you run your script and get a stack trace, but i can get that from the oops report [20:05] intellectronica: I take it back, ogasawara fixed it. you should be in now [20:05] * mbarnett looks around [20:05] mwhudson, mbarnett: sorry, I meant ubuntu-bugcontrol admins. :) [20:05] kees: awesome, thanks [20:05] mbarnett: ah, sorry [20:06] mwhudson: This looks like 0755 [20:06] * mbarnett wanders off again [20:06] mwhudson: np [20:07] abentley: when you say "not scanning" do you mean the scanner isn't even looking at the branches you think it should? [20:08] mwhudson: the branch scanner script reports 2009-10-01 18:54:29 INFO OOPS-1370BS1: Not a branch: "lp-mirrored:///~mark/bzr/bzrtools-silly". (~mark/bzr/bzrtools-silly) [20:08] mwhudson: I think the puller is putting the branches in the wrong place... [20:09] abentley: or cocking up the stacked_on_url [20:09] mwhudson: No stacking enabled for this project. [20:10] ok, that's one thing down :) [20:10] mwhudson: I looked at /var/tmp/bazaar.launchpad.dev/mirrors/ and the branches weren't there. (there were two presumably very old branches) [20:10] ok that's screwy [20:14] huh, the puller isn't working at all for me for some reason... [20:16] abentley: need to have coffee before looking at this more, brb [20:16] mwhudson: cool [20:26] mwhudson: I have an idea-- I bet you apache's mad at me. [20:34] abentley, you could say that all the time and 60% of the time you'd be right. [20:41] kees: i have a fix for your bug. review and test run, and if everything goes ok it should be fixed on edge after the next rollout [20:45] abentley: working now [20:45] ? [20:46] kinda working. It's got the wrong branch data, but that might be my fault. [20:46] * kees hugs intellectronica [20:46] mwhudson: It may be because I didn't have an /etc/hosts entry for bazaar-internal.launchpad.net [20:47] yes, that would certainly prevent the scanner from working [20:47] (i presume you mean bazaar-internal.launchpad.DEV ?) [20:47] mwhudson: Yes. === Ursinha-afk is now known as Ursinha [21:14] mwhudson: Up for a chat about jobs and code import? [21:15] abentley: ok, one sec [21:21] intellectronica, I understand what you mean in that bug, I think we should create one bugtask for launchpadlib, and mark the bug as dupe in case we find the lplib bug you mentioned [21:21] Ursinha: i think that's a good idea [21:22] intellectronica, done, then :) [21:23] Ursinha: great, thanks [21:34] * rockstar lunches [21:34] james_w, version 1.5.2 of launchpadlib has been released for your packaging enjoyment [21:40] cprov-afk, when you're around could you take care of https://answers.edge.launchpad.net/launchpad/+question/81974 please? [21:42] matsubara: uhm ... sure [21:43] cprov-afk, thanks! [21:43] matsubara: actually, I can't [21:43] matsubara: I'm not a soyuz-member anymore :`( [21:43] cprov-afk, oh! [21:44] well, since the user is OK with the team being deleted, I think I'm going to suggest that. [21:45] cprov-afk, do you know if merging the team away be blocked by the ppa linkage? [21:46] matsubara: merging is allowed [21:46] cool. I'll ask the losas then [21:46] thanks cprov-afk [21:46] matsubara: np [21:58] abentley: lib/canonical/twistedsupport/processmonitor.py === Ursinha is now known as Ursinha-afk === joey is now known as Guest45878 === Guest45878 is now known as joey [23:16] Oh, cprov's leaving? :( [23:16] left, even :) [23:16] :( rather [23:17] i think [23:17] Yea, his last day was yesterday. [23:17] Wow. [23:18] however what with launchpad now being open source, i'm sure he won't escape that easily :) [23:22] Heh. === matsubara is now known as matsubara-afk [23:29] rockstar: ping [23:30] EdwinGrubbs: ping [23:36] barry, yo [23:37] rockstar: hi. do you have a few minutes to look at some screenshots? [23:37] barry, yes. I still plan on being around for a bit. [23:38] rockstar: cool, thanks. i'm playing around with some fixes for bug 440220 [23:38] Bug #440220: registering slot interferes with watermark [23:38] rockstar: there are some screen shots attached to that bug report. the first one is a snap of the current problem, then three snaps of some fixes [23:39] rockstar: i'm very interested to know what you think. note that this approach is keeping the registering slot where it currently is and just playing with layout and nowrap [23:39] rockstar: but another possibility of course is to move the registering slot to a different location [23:39] rockstar: anyway EOT