[00:03] <bryceh> I'm getting an error about loom.branch when trying to run 'make schema': http://pastebin.com/d8i40sE1
[00:03] <bryceh> any ideas?  I tried installing bzr-loom but that made no difference.  Am I missing something else?
[00:10] <mwhudson> bryceh: what's in sourcecode/ ?
[00:10] <mwhudson> launchpad doesn't use the system plugins for bzr
[00:10] <bryceh> from bzrlib.plugins.loom.branch import (
[00:10] <bryceh>     BzrBranchLoomFormat1, BzrBranchLoomFormat6)
[00:10] <bryceh> from lp-branches/devel/lib/lp/code/bzr.py
[00:11] <mwhudson> sigh
[00:11] <mwhudson> bryceh: there's a directory called sourcecode in the launchpad tree
[00:11] <mwhudson> bryceh: what's in there?
[00:12] <bryceh> bryce@cardport:~/launchpad/lp-branches/devel$ ls sourcecode/
[00:12] <bryceh> bzr-git  dulwich               mailman   pygpgme  subvertpy
[00:12] <bryceh> cscvs    launchpad-loggerhead  Makefile  shipit   testresources
[00:12] <mwhudson> that's an odd mix
[00:12] <mwhudson> bryceh: run ./utilities/update-sourcecode
[00:13] <bryceh> mwhudson, ok will give that a shot
[00:14] <bryceh> hmm, that gives this error:  http://pastebin.com/nraERpsX
[00:15]  * bryceh tries deleting the .bzr dir
[00:18] <wgrant> You probably have to delete the entire shipit directory.
[00:27] <bryceh> wgrant, thanks, seems to be proceeding now
[00:52]  * thumper heads to lunch
[00:52] <thumper> bryceh: welcome
[07:00] <lifeless> can I get someone to land a patch for me ?
[07:02] <lifeless> https://code.edge.launchpad.net/~lifeless/launchpad/merge/+merge/23523
[07:03] <StevenK> lifeless: PQM is in RC
[07:04] <lifeless> oh
[07:04] <lifeless> what does that mean
[07:05] <StevenK> lifeless: release-critical fixes only
[07:11] <spm> lifeless: fwiw, unless it's like ZOMG STOP THE PRESS urgent; hasn't got a hope of getting in todays release.
[07:12] <lifeless> spm: oh clearly
[07:12] <lifeless> just keeping the inventory low
[07:12] <spm> heh
[07:34] <wgrant> lifeless: Option 1: violate separation of concerns in an already hackish script. Option 2: respect separation of concerns slightly more, and make your service look even less reliable and attractive than its already bad reputation.
[07:35] <lifeless> wgrant: doing the revisions at the same time would make the script take much much longer to run
[07:35] <lifeless> doing the scan staggered, so that new work happens fast but the new branches do get scanned, would allow the branches to be available more quickly.
[08:17] <adeuring> good morning
[09:00] <mrevell> Goodly morning
[09:02] <allenap> Good morning mrevell.
[09:03] <mrevell> yo allenap
[09:03] <allenap> mrevell: Today I am lanscallenape
[09:04] <mrevell> allenap, Oh! Man. Congrats :)
[11:36] <leonardr> i need to have a conversation with someone who knows a lot about how launchpad uses apache. i don't want to bother a losa about this since it's development work, but if a losa is free i'd love to talk
[11:36] <leonardr> jml, can you recommend someone?
[12:32] <jml> leonardr: flacoste or BjornT would be my two guesses
[12:37] <mwhudson> so...
[12:37] <mwhudson> code hosting hasn't imploded?
[12:38] <wgrant> Does this mean we're running hostedless?
[12:39] <jml> umm
[12:39] <jml> it probably hasn't imploded, but there have been some weird email behaviours: https://answers.edge.launchpad.net/launchpad-code/+question/109575
[12:40] <jelmer> jml: fwiw I received that email as well
[12:41] <james_w> jam apparently just proposed a moloolaba branch when I assume he is in bed, so presumably not an isolated incident
[12:41] <wgrant> I haven't got any.
[12:41] <james_w> https://code.edge.launchpad.net/~jameinel/bzr-fastimport/mooloolaba/+merge/14894
[12:42] <jelmer> jml: I've seen jam's fastimport proposal as well, but IIRC those are the only two.
[12:42] <james_w> Date: Tue, 04 May 2010 10:32:28 -0000
[12:42] <mwhudson> there's a new job script to send merge proposal email
[12:42] <mwhudson> maybe there were some odd old jobs in the db for some reason
[12:44] <mwhudson> jml: i notice https://code.edge.launchpad.net/~jml/zope.testing/subunit-error has a strange error message
[12:45]  * mwhudson goes to bed
[12:45] <jml> mwhudson: yeah, that is weird. sleep well.
[12:49] <james_w> jml: it seems that special permissions for source package branches have broken, but the code looks fine to be in lp:launchpad/devel at least
[12:49] <james_w> is there somewhere else I should be checking?
[12:50] <james_w> at least BranchMergeProposalEdit is no longer granted to uploaders, but that could be because they have lost EditBranch
[12:50] <jml> james_w: hmm.
[12:50] <jml> james_w: I'm dead certain that I added a unit test to check that uploaders have edit on official package branches.
[12:51] <jml> james_w: so, I guess I'd look for that test and make sure it was still valid
[12:51] <james_w> I'm dead certain I added a similar unit test to check that uploaders have edit on merge proposals too
[12:51] <jml> james_w: I don't know which branch most accurately represents the code that's on Launchpad.
[12:53] <jml> james_w: I'd recommend finding someone who is supposed to have upload permissions and getting them to try changing a branch's status and try pushing up to a branch.
[13:06] <gary_poster> mrevell: hi. :-)  did you notice during rollout if we can mark https://bugs.edge.launchpad.net/launchpad-foundations/+bug/553171 qa-ok?
[13:06] <mup> Bug #553171: "LP is read-only" and similar messages still point to old maintenance page <qa-needstesting> <trivial> <ui> <Launchpad Foundations:Fix Released by matthew.revell> <https://launchpad.net/bugs/553171>
[13:08] <mrevell> gary_poster, Looked good to me :)
[13:08] <james_w> jml: they have edit permissions, and the code looks good in all the lp:launchpad branches
[13:08] <mrevell> gary_poster, updated
[13:08] <gary_poster> thanks mrevell :-)
[13:09] <jml> james_w: so it's just merge proposals?
[13:09] <james_w> I think so
[13:09] <james_w> I'm going to check further
[13:12] <jtv> gary_poster: hi!  How do I run the doctest for lazr.batchnavigator?  I'd like to verify my fix for that oops I reported.
[13:13] <gary_poster> jtv: great, thank you! I hope ``bin/test`` will do the trick.
[13:13] <gary_poster> It is supposed to
[13:13] <jtv> gary_poster: I didn't see a bin in that branch, and the LP test runner doesn't seem to exercise it either.
[13:16] <gary_poster> jtv, acknowledging that this sucks and that we have a slow-burning process to make it better, did you see https://dev.launchpad.net/HackingLazrLibraries ?
[13:17] <jtv> gary_poster: just found it
[13:17] <gary_poster> great
[13:18] <jtv> running bootstrap... looks like that'll take a while
[13:18] <gary_poster> jtv, next time se the "Global cache" instructions
[13:18] <gary_poster> that makes subsequent efforts a lot faster
[13:18] <jtv> gary_poster: ok, thanks!
[13:20] <james_w> jml: not a regression, it's just that the check says that they can't upload to lucid now, but we have no way to propose the creation of a lucid-proposed branch with the changes. This will require some discussion.
[13:20] <leonardr> flacoste, bjornt, when you get a minute i'd appreciate your thoughts on bug 574697
[13:20] <jml> james_w, gnarr
[13:20] <mup> Bug #574697: Launchpad strips incoming TE header <Launchpad Foundations:Triaged> <https://launchpad.net/bugs/574697>
[13:21] <jml> james_w, you right to arrange that discussion?
[13:21] <james_w> yeah
[13:22] <jml> ta
[13:47] <jtv> gary_poster: all done & ready for review...  note: the hacking page says to merge your branches directly into trunk, but doesn't mention pushing.
[13:48] <jtv> Oh, that's further down.
[13:48] <jtv> No, it's not.
[14:00] <flacoste> leonardr: didn't we established that it was apache that removed those headers
[14:00] <flacoste> ?
[14:00] <flacoste> leonardr: if i recall, we had put a work-around in place for that behaviour?
[14:08] <leonardr> flacoste: i don't remember encountering this before
[14:08] <leonardr> we have a workaround for apache's treatment of Content-Encoding, but apache's behavior stops the workaround from ever happening
[14:10] <gary_poster> jtv: hacking page: it means bzr co the trunk and then merge and commit
[14:11] <gary_poster> jtv: forward me the review email when you get a chance
[14:44] <flacoste> leonardr: i'd suggest sending an email cc launchpad-dev to our local resident proxy expert: lifeless
[14:50] <leonardr> ok
[15:36] <mars> henninge, ping, looking for a CHR swap?
[15:36] <henninge> mars: Hi. Yes, I am.
[15:36] <henninge> mars: What can you offer? ;)
[15:37] <mars> henninge, how about Monday May 24th?  That's a National Holiday here in Canada.
[15:37] <mars> henninge, and I will be looking to swap it
[15:38] <henninge> mars: although it's a Monday, it should be fine ... ;)
[15:38] <mars> yeah, sorry about that :)
[15:38] <mars> henninge, great, I'll tell the list
[15:39] <henninge> last month was  a Monday too, but it wasn't too bad.
[15:39] <henninge> mars: yes, please do.
[17:05] <rockstar> abentley, should we allow recipes to be based of junk branches?
[17:06] <abentley> rockstar, I have no strong opinion.
[17:07] <abentley> rockstar, I guess so?
[17:10] <rockstar> abentley, I was thinking we shouldn't.  If they want to do that, they should create a project.
[17:11] <jml> rockstar, personally, I'd allow it.
[17:13] <rockstar> jml, rationale?
[17:14] <abentley> rockstar, or create the sourcepackage?
[17:15] <jml> rockstar, there's no compelling reason to forbid them that I know of. it requires work on our part to add the restriction. it requires work on their part to work around the restriction
[17:17] <maxb> It may be useful for people to experiment with the functionality in an obviously personally-sandboxed way
[17:22] <rockstar> jml, why not allow merge proposals on junk branches then?
[17:22] <jml> rockstar, partly because it makes the UI harder for us and harder for the user
[17:23] <jml> which branch do you merge into by default? how do you see the list of active reviews?
[17:24] <maxb> I think MPs ought to be allowed on +junk branches, if only to keep consistency, and allow people to play around with the feature when learning
[17:24] <maxb> default branch: none, they have to specify. list of reviews: only for a specific branch
[17:25] <jml> maxb, regarding junk branches, what I want before absolutely anything else is a dead simple way to turn them into the trunk branch of a new project
[17:25] <maxb> That would be nice. Bring back the ability to rename branches across projects
[17:26] <jml> in general, the reason why we make junk branches less awesome is to encourage people to use projects as the basis for collaboration
[17:26] <jml> (so bugs can be filed, translations done, related branches found etc)
[17:34] <rockstar> jml, so recipes would be an exception to that rule?
[17:35] <jml> rockstar, they'd buck the trend, certainly. I mean, you could also argue that "capable of being owned by teams" breaks the rule.
[17:35] <rockstar> jml, yes, I think it does to some extent.
[18:12] <mrevell> Night
[20:47] <leonardr> james_w, your wadllib code is present in the new version of wadllib
[20:57] <james_w> leonardr: great, thanks
[21:40] <kfogel> sinzui: if possible bug in Answers, should question assignee be ~launchpad-registry?  This is re https://answers.edge.launchpad.net/launchpad/+question/109538/  (that question is not an example of the bug, but it describes the bug's manifestation in 109518, which is a confusingly similar number)
[21:43] <sinzui> kfogel, I do not have an immediate answer. retarget it to launchpad-answers (not registry). One of us will get to it
[21:43] <kfogel> sinzui: no such team 'launchpad-answers' (that was my first try)
[21:44] <sinzui> kfogel, you do not need to assign teams or person in answers. We have an answer contact system
[21:44] <sinzui> kfogel, All the contacts get the email. I may not be the first person to answer the question
[21:45] <kfogel> sinzui: I am very confused by what just happened.
[21:45] <sinzui> kfogel, I get about 10 questions a week because I am a contact
[21:45] <kfogel> sinzui:  I first tried to retarget to launchpad-answers, and got an "invalid" sign underneath.  I wish I'd saved that screenshot.  So then I tried to re*assign* to a team of that name, but no such team.
[21:45] <kfogel> sinzui: now I re-blanked assignee and retargeted again, and this time it worked
[21:45] <sinzui> target not assign
[21:46] <kfogel> sinzui: right, that was the order I tried in.  And now it's wroking.
[21:46] <sinzui> assign is for actions that require a *specific* person or role. That is to say, assignee to not providing an answer but completing a task
[21:46] <kfogel> sinzui: so, nm, but that was very confusing.  No doubt I filled in a wrong field somewhere, but my memory has "corrected" that.
[21:47] <sinzui> I get 15 questions a weeks for the projects I am contact for and I answer within the hour
[22:03] <kfogel> sinzui: thanks for the quick response to that question (109538)
[22:04] <thumper> what do we tag things that can't be QAed?
[22:05] <thumper> like test fixes?
[22:05] <thumper> do we just remove the qa-needstesting tag?
[22:05] <salgado> qa-untestable
[22:53] <gary_poster> bac, ping
[22:53] <bac> hi gary_poster
[22:54] <gary_poster> hey bac.  OK, I have unit tests and changes for bug574493.  I only have theoretical proof that it should actually do what we want.  What else needs to be dine:
[22:54] <gary_poster> done
[22:55] <bac> gary_poster: skip lunch?  Freudian slip?
[22:55] <gary_poster> :-)
[22:55] <gary_poster> - actually verify that it does what we want.  Maybe you can do this on a dev machine, but I'd be more tempted to do it on staging, if possible
[22:55] <gary_poster> - get reviewed
[22:55] <gary_poster> - get landed
[22:56] <gary_poster> - get CP'd
[22:57] <gary_poster> bac, so, if you are willing, I need you to take over most of that.  What I can do is pass off the branch to you with reviewer notes.  We could do that verbally looking ata diff or something else like that