[00:06] <wgrant> StevenK: Erm
[00:06] <wgrant> StevenK: Why does https://code.launchpad.net/~launchpad/+recipe/launchpad-convoy use nest-part rather than merge?
[03:43] <nigelb> Ah. It's a holiday today?
[03:54] <wgrant> nigelb: Yeah, ANZAC Day for AU/NZ.
[03:54] <StevenK> wgrant: :-(
[03:59] <nigelb> wgrant: Ah. I get worried when this channel has very little activity ;-)
[04:43] <G> nigelb: haha, especially since it's the middle of the week :)
[04:44] <nigelb> G: Exactly
[07:32] <adeuring> good morning
[08:02] <czajkowski> aloha
[11:55] <gary_poster> Hey StevenK.  The convoy packaging problem is an issue for parallel testing because our automated system to initialize ec2 instances for tests breaks.  It looks like you worked on it but didn't get it working, seemingly because of a trivial version-already-exists-in-ppa issue.  Should I just try making a trivial change to convoy and firing off the recipe to see if that helps?
[11:57] <wgrant> gary_poster: Perhaps we should revive the old known-good version and try fixing it in a staging PPA or something?
[11:58] <gary_poster> wgrant, yeah, that sounds reasonable.
[11:59] <wgrant> gary_poster: Do you know how to do that?
[11:59] <gary_poster> wgrant, no, but it didn't sound too daunting & I was going to give it a try.  hints much appeciated.
[11:59] <gary_poster> r
[12:01] <wgrant> gary_poster: Delete the current version from the PPA, then copy the last known-good versions back from the grave (search for convoy with "Any status" on +copy-packages, select the last good version for each series, copy with binaries)
[12:02] <gary_poster> wgrant ack.  Then I have good versions in the PPA.  Then we can temporarily change the current recipe to build into a different PPA while we work out the kinks?
[12:02] <wgrant> Looks like 0.2.1-0~21+r21 is the version to go for, I think.
[12:03] <wgrant> gary_poster: Right, you can manually ask the recipe to build into a different PPA
[12:03] <wgrant> So we don't break everyone.
[12:04] <rick_h> is there any way to get the old recipe? I broke things by adding a Makefile to help run tests and upload to pypi, I can change that back if I knew what the old working recipe was
[12:04] <rick_h> that's what I was trying to figure out yesterday
[12:05] <wgrant> The recipe probably hasn't changed. You might have changed the packaging, which is fully versioned in the bzr branch.
[12:05] <wgrant> It's likely that the makefile broke it
[12:05] <wgrant> Simply by existing
[12:05] <rick_h> yea, I know that broke it, but then he went in and 'fixed' it and I couldn't tell what his 'fix' was
[12:05] <wgrant> Since the debian/rules just uses dh $@, which tries to autodetect the build system.
[12:05] <rick_h> right
[12:07] <gary_poster> wgrant, rick_h, I believe I have done the right temporary fix in the PPA.  Verification appreciated.
[12:07] <wgrant> gary_poster: Looks good
[12:07] <wgrant> Heh
[12:07] <gary_poster> cool thank you
[12:07] <wgrant> Yeah, seems OK
[12:07] <gary_poster> :-)
[12:07] <rick_h> wgrant: so where would this branch be Stevek had going with the debian dir? https://code.launchpad.net/~launchpad/+recipe/launchpad-convoy
[12:07] <wgrant> Should be published in a couple of minutes.
[12:07] <gary_poster> cool
[12:08] <rick_h> ok, cool
[12:08] <rick_h> sorry gary_poster, forgot the holiday
[12:08] <wgrant> rick_h: The second branch referenced in the recipe
[12:08] <wgrant> https://code.launchpad.net/~launchpad/convoy/trunk
[12:08] <rick_h> ah, gotcha. Sorry, read that as convoy/trunk
[12:08] <wgrant> (neither StevenK nor I are here today, but we can look tomorrow)
[12:08]  * rick_h drinks more coffee
[12:08] <wgrant> Yeah, that's a bit confusing.
[12:09] <wgrant> Should probably s/trunk/packaging/
[12:12] <rick_h> gary_poster: so if I can bug you a sec, where did you make a change that's now building?
[12:13] <gary_poster> rick_h, I went to https://launchpad.net/~launchpad/+archive/ppa
[12:13] <gary_poster> clicked on the delete link on top right
[12:13] <gary_poster> deleted convoy
[12:13] <gary_poster> went back to https://launchpad.net/~launchpad/+archive/ppa
[12:13] <rick_h> ah ok, I thought you were making a 'trivial' change to it somewhere and didn't see an edit
[12:13] <gary_poster> clicked on copy link
[12:14] <rick_h> gotcha, thanks
[12:14] <gary_poster> searched for convoy with Any status and copied the one that seemed like it might not be brokacool np
[12:18] <StevenK> I did question rick_h adding a Makefile, I felt it was pointless, but I utterly missed that dh would try it first. :-/
[12:22] <wgrant> A Makefile is handy here
[12:24] <rick_h> I've gotten too used to having make test and make upload, make sdist for all my setup.py etc needs
[12:24] <wgrant> Yup
[12:24] <rick_h> didn't realize I was going to bork things, I need to find some time to get back into playing with packaging bits
[12:24] <wgrant> Lots of our stuff has Makefiles for that purpose.
[12:25] <rick_h> yea, since LP I've gone through the Make oreilly book and gone nuts
[12:25] <rick_h> https://github.com/mitechie/Bookie/blob/develop/Makefile bwuhahahaha
[12:26] <StevenK> rick_h: I tend to put .PHONY at the end with everything
[12:26] <StevenK> Since make parses it in one-shot
[12:26] <rick_h> StevenK: yea, I started going that, but trying to remember/find them in teh list got to be a pain
[12:26] <rick_h> so I've gone the python "explicit" route
[13:00] <deryck> Morning, all.
[13:14] <dobey> morning deryck
[13:15] <dobey> did some part of the LP DBs get rolled back at all last night or earlier this morning?
[13:15] <wgrant> No.
[13:16] <wgrant> What's the issue?
[13:16] <dobey> https://code.launchpad.net/~ralsina/ubuntuone-control-panel/unique_in_ubuntu/+merge/103337
[13:16] <dobey> this branch had 2 reviews last night, and even a couple of comments from tarmac failing to land it
[13:16] <dobey> but now it has nothing
[13:17] <dobey> oh weird
[13:17] <dobey> nevermind
[13:17] <dobey> i see the problem
[13:17] <wgrant> It was probably deleted and recreated or something.
[13:18] <wgrant> We don't just randomly revert bits of the database :)
[13:18] <wgrant> Hmm
[13:18] <wgrant> I indeed have emails from 11 hours ago about that MP
[13:18] <dobey> no, it looks like he made 2 proposals at the same time
[13:19] <wgrant> But it has a different ID
[13:19] <wgrant> So yeah
[13:19] <dobey> sorry :)
[13:19] <wgrant> 103336 is the one tarmac attacked
[13:19] <cjwatson> Some people get very confused by the MP interface and think they need to re-propose any time they make a change
[13:20] <wgrant> In this case it looks like a swift double-click of the Propose Merge button
[13:20] <wgrant> The two were created seconds apart.
[13:21] <dobey> cjwatson: nah, this wasn't that case. more likely twitchy finger. i've made it very clear to our team to not do re-submit all the time as well :)
[14:20] <abentley> adeuring: I'm the OCR.  Could you please review https://code.launchpad.net/~abentley/launchpad/celery-everywhere-7/+merge/103483 ?
[14:21] <adeuring> abentley: sure
[14:39] <adeuring> abentley: r=me
[14:39] <abentley> adeuring: Thanks
[15:03] <gmb> deryck, I've bumped https://bugs.launchpad.net/launchpad/+bug/924378 up to Critical; it's now caused two build failures on db_lp in as many days (though not on consecutive runs, AFAICT).
[15:04] <_mup_> Bug #924378: buildbot spurious failure UncleanReactorError for TestPullerMasterIntegration <Launchpad itself:Triaged> < https://launchpad.net/bugs/924378 >
[15:04] <gmb> Just wanted to give you the heads up.
[15:04] <gmb> deryck, If you don't have resources enough to tackle it, gary_poster has suggested that ~yellow take it on.
[15:04] <gary_poster> Because we only have 8 other bugs in the queue now :-P
[15:04] <deryck> gmb, thanks. We can add it to our board, but it will likely be a bit before we can get to it.
[15:05] <gary_poster> thanks deryck.  We'll coordinate via the bug if we tackle it
[15:05] <deryck> gmb, but we could disable the test and wait on one of our squads to get bandwidth to fix it properly.
[15:05] <deryck> gary_poster, ok, cool.
[15:05] <gary_poster> deryck, +1 on disabling fwiw :-) although the bug makes reference to some reason why that's a bad idea
[15:07] <deryck> gary_poster, I think that was speculation from flacoste back when I filed it.  if no one recalls, we could disable and see the fallout to be sure ourselves. :)
[15:07] <gary_poster> :-) fair enough
[15:19] <mrevell> Apologies jcsackett, you weren't on the calendar entry. I've invited you now.
[15:30] <mrevell> sinzui, jcsackett, danhg, flacoste, matsubara, czajkowski: https://plus.google.com/hangouts/_/extras/talk.google.com/better-privacy
[15:33] <jcsackett> mrevell: having a bit of fun trying to get to the hangout. be there in a moment.
[15:34] <mrevell> jcsackett, no worries
[16:52] <sinzui> rick_h, do you know what I can do to convoy so that I can run Lp locally?
[16:54] <rick_h> sinzui: so gary_poster updated the ppa I thought with an older version that should work
[16:54] <rick_h> so update the packages?
[16:54] <sinzui> already done
[16:55] <sinzui> ImportError: No module named convoy.meta
[16:55] <gary_poster> rick_h, sinzui that will only work if you have not updated, or on a fresh install, IIUC
[16:55] <sinzui> ^ still present
[16:55] <gary_poster> Because all I did was remove the newer package and reinstate the old
[16:55] <rick_h> can you remove/reinstall?
[16:55] <sinzui> gary_poster, you mean visit lp and locate an older package, then lock the version?
[16:56] <gary_poster> Alternatively, sinzui, you could do (not exact spelling) sudo easy-install convoy and that would probably fix things up at the expense of a mess in your system python
[16:56] <gary_poster> sinzui, no I mean in the PPA, I deleted the new package
[16:56] <gary_poster> then copied the old one back into the PPA
[16:57] <sinzui> easy install is crack in that case
[16:57] <gary_poster> If I understand you, I agree that it would be nasty, but I think it would work
[16:57] <sinzui> I see precise is 4 version behind
[16:57] <gary_poster> yes
[16:59] <rick_h> I think r20 will still work for you
[16:59] <rick_h> the path stuff is only needed on production and I don't think we use it on dev, but not 100% sure
[17:34] <gary_poster> jml, just a heads up in case you are around: the test failure I emailed you about is also affecting our production usage, and therefore blocking our progress.  We will be working on it.  Any insights from you will of course be welcome.  benji, could you file a bug for this so we have a coordination point?
[17:35] <benji> gary_poster: sure
[17:35] <gary_poster> thanks benji.  benji, I doubt it affects us, but there are two subunit patches in new, recent bugs for subunit that I noticed.
[17:35] <gary_poster> bug 987490
[17:35] <_mup_> Bug #987490: Test failures with Python 2.6 <subunit:New> < https://launchpad.net/bugs/987490 >
[17:35] <gary_poster> bug 987938
[17:35] <_mup_> Bug #987938: subunit trunk packaging breaks subunit-* commands <subunit:New> < https://launchpad.net/bugs/987938 >
[17:35] <gary_poster> no not the last one
[17:36] <gary_poster> bug 987514
[17:36] <_mup_> Bug #987514: Regressions in support for Python 3 <subunit:New> < https://launchpad.net/bugs/987514 >
[17:36] <gary_poster> that one
[17:38] <jcsackett> is there a good/easy way to get +/- LoC for a branch out of bzr? like some magical incantation to pass to diff, maybe?
[17:39] <jelmer> jcsackett: you can feed the diff through diffstat
[17:39] <jelmer> "bzr send -o - | diffstat"
[17:39] <jcsackett> ah, awesome.
[17:39] <jcsackett> thanks, jelmer.
[17:42] <jelmer> we should add a lp-dev-tools script that does that :)
[19:09] <abentley> deryck: Could you please review https://code.launchpad.net/~abentley/launchpad/celery-everywhere-8/+merge/103545 ?
[19:09] <deryck> abentley, sure.
[19:09] <abentley> deryck: thanks!
[19:09] <deryck> abentley, np!
[19:13] <deryck> abentley, looks good to me. r=me.
[19:13] <abentley> deryck: Thanks.
[19:14] <deryck> abentley, np.  I've done enough of these now.  It's easy to spot the pattern. :)
[20:19] <lifeless> jcsackett: bug 988510 confuses me
[20:19] <_mup_> Bug #988510: Bug supervisor should not be subscribed on ubuntu bugs when transition from security to userdata <disclosure> <regression> <Launchpad itself:In Progress by jcsackett> < https://launchpad.net/bugs/988510 >
[20:19] <lifeless> jcsackett: I thought we'd checked that the vanilla policies around information types defined would work for ubuntu ?
[20:20] <jcsackett> the policies work, it's a matter of handling auto subscriptions.
[20:20] <lifeless> auto or implicit?
[20:21] <lifeless> is this perhaps a mixed-mode issue until we stop doing auto subscriptions entirely?
[20:41] <flacoste> sinzui: you did send a sketch on how to implement agreements in Launchpad on launchpad-dev a while back right?
[20:42] <flacoste> "Proposed team agreement" feature
[20:48] <sinzui> flacoste, I did
[20:51] <flacoste> gary_poster: who is chairing this week, you didn't update the wiki page after last week meeting
[20:51] <gary_poster> sorry flacoste, lemme find the email
[20:51] <gary_poster> flacoste, sinzui
[20:52] <sinzui> flacoste, That is the email thread I started. I cannot determine the real dates of when it happened because Shitehawk^hThunderbird does not honour my date formats. I cannot tell months and days apart
[20:52] <flacoste> sinzui: i found it in the archive, thx, it was 2011-11-08
[20:53] <sinzui> I see a sensible iso approved big-endian date just like I configured my desktop to use.
[20:58] <gary_poster> sinzui, you are still ok with hosting team lead call?
[20:58] <gary_poster> lifeless, you are unavailable so we do not need to try and invite you right?
[20:59] <bigjools> morning
[20:59] <sinzui> gary_poster, I am, I wish my wife would stop blocking me from doing it
[20:59] <gary_poster> ok
[21:39] <timrc> It feels like its taking excessively long to do ppa2ppa binary copies
[21:49] <timrc> still pending, wtf :(