[12:30] <mpt> Gooooooooooooooooooooooooooood morning Launchpadders!
[12:33] <LarstiQ> heya mpt
[12:37] <mpt> It's a bright and chilly morning here in Dunedin
[12:37] <LarstiQ> I wish it was chilly here in Den Haag
[01:34] <abentley-gaim> Is there any lag between uploading an SSH key and being able to push using bzr?
[02:39] <stub> ??signal
[02:41] <abentley-gaim> Someone set us up the bomb.
[06:43] <stub> Launchpad will be going down in 15 mins for its regular code update
[06:44] <stub> Launchpad will be going down in 15 mins for its regular code update. Downtime should be around 15 mins.
[07:45] <jamesh> stub: did the sprint meeitng export fix I merged last night make it into this rollout?
[07:45] <stub> jamesh: Yes. My telepathy is back online temporarily.
[07:46] <jamesh> thanks
[08:01] <ajmitch> evening mpt_ :)
[08:19] <SteveA> morning
[08:19] <SteveA> mpt_: ping
[09:02] <sivang> morning
[09:13] <stub> What is the magic to generate a class on the fly, assembled from a list of subclasses?
[09:14] <spiv> stub: type('name', bases, attrs_dict)
[09:14] <stub> Ta
[09:52] <carlos> morning
[09:58] <carlos> SteveA, jordi: Occitan and Amharic mailing lists are created
[09:59] <SteveA> good
[10:10] <SteveA> jamesh: ping
[10:11] <jamesh> SteveA: pong
[11:11] <jamesh> lifeless: btw, the tickcount extension is ready to be packaged: sftp://bazaar.launchpad.net/~launchpad/tickcount/devel/
[11:12] <jamesh> lifeless: you can either add the debian rules directly to the branch or run "make dist" to get a tarball base.
[11:12] <lifeless> ok, thanks
[11:12] <lifeless> wont be this week - focused on the sprint - but will be shortly after
[12:46] <bradb> lifeless: Is the answer to this question correct?
[12:46] <bradb> https://launchpad.canonical.com/LaunchpadHackingFAQ#head-451fd4372e4a0e7e1305a403e98dc3b173140f5f
[12:46] <bradb> lifeless: BjornT says it should be "Yes."
[12:48] <spiv> bradb: it should be yes.
[12:53] <bradb> spiv: Thanks. We updated the wording of the question (e.g. lifeless thinks it's confusing that "standalone" was referred to as a "story", etc.) and corrected the answer
[01:00] <carlos> spiv: do we have a way to use librarian with sampledata?
[01:01] <Znarl> stub : Ping?
[01:02] <Znarl> stub : Launchpad logs have reached 1.8gigs and are slowing down chinstrap.
[01:10] <spiv> carlos: Yes, so long as the test that wants the sampledata first uploads it...
[01:10] <carlos> spiv: hmm that's a 'no' for me ;-)
[01:11] <spiv> carlos: So, no pre-canned sampledata, but the librarian is otherwise fully functional, so you can upload and then download from it in tests :)
[01:11] <carlos> that's not sampledata for me ;-)
[01:11] <carlos> at least it's not 'real' sampledata
[01:11] <carlos> spiv: would be possible to add that feature in the future?
[01:12] <spiv> Well, you could make a LibrarianTestSetup subclass that does this.
[01:12] <carlos> well, I was thinking more a way to do 'make schema'
[01:12] <spiv> The problem is that the librarian stores stuff outside of the database.
[01:12] <carlos> and be able to use launchpad with some librarian files imported
[01:13] <spiv> And, well, where would it put its files?
[01:13] <carlos> spiv: sure, I'm thinking on someway to generate a tarball where we could store a default librarian tree
[01:14] <carlos> and that with something like:  'make librarian' untars it and we get a tarball that matches our db sampledata
[01:14] <carlos> sorry, /tarball/librarian tree/
[01:15] <spiv> To answer my own question: I guess it could use config.librarian.server.root.
[01:15] <spiv> (I was mistakenly thinking the librarian test setup used a temporary directory)
[01:15] <carlos> spiv: right
[01:16] <spiv> There'd be no reason to use a tarball, there's nothing wrong with plain files.
[01:16] <stub> Znarl: ok
[01:16] <spiv> I'm not really sure that this offers any advantages over my suggestion to do it in your test or test setup, though.
[01:17] <carlos> spiv: well, that tree is not going to be updated because the librarian tree is used from /var/tmp/fatsam.test
[01:17] <spiv> Which requires less infrastructure and would be less error prone.
[01:18] <spiv> So you're worried about optimising it?
[01:18] <spiv> We have much bigger test suite performance issues to worry about :)
[01:18] <spiv> The starting and stopping the librarian process every pagetest, for instance.  stub is working on avoiding that cost.
[01:19] <carlos> spiv: the answer to the question about why would we want this: https://launchpad.net/products/rosetta/+bug/31725
[01:19] <Ubugtu> Malone bug 31725 in rosetta "Sampledata should include a translation queue entry" [Medium,Unconfirmed]  
[01:20] <spiv> Ah, so this is for playing with the webapp locally?
[01:20] <carlos> right
[01:20] <carlos> for pagetests, I agree 100% with you that we should do the uploads
[01:20] <spiv> Sorry, I thought you were talking about tests (which is why I kept talking about them) :)
[01:21] <spiv> Well, I don't 100% agree that we should do the way I suggest for tests, more like 80% or 90% ;)
[01:22] <carlos> spiv: well, for me is the best way to test exactly what you want without breaking many tests when the sampledata changes ;-)
[01:22] <spiv> But right, I'd be happy for there to be a 'librarian_data' or similar make target that is invoked as part of 'make schema'.
[01:23] <carlos> spiv: storing it as a subtree in our bzr tree or as a tarball?
[01:23] <spiv> As a plain tree.
[01:23] <spiv> A tarball doesn't offer any benefits I can see, and just makes it harder to update the sampledata.
[01:24] <carlos> spiv: ok
[01:24] <spiv> And harder to inspect the sampledata.
[01:24] <carlos> yeah, that's right
[01:25] <spiv> It should live in database/sampledata/librarian, I think.
[01:26] <carlos> yeah, makes sense
[01:28] <spiv> carlos: the problem is I rarely actually run the webapp locally, but I run tests a lot :)
[01:28] <carlos> it's useful for new pages or UI changes
[01:28] <carlos> for bug fixes, it's useless
[02:10] <carlos> later
[05:53] <flacoste> I got weird failures after submitting my branch for merging to PQM
[05:53] <flacoste> anybody can help?
[05:53] <spiv> flacoste: pastebin it?
[05:58] <flacoste> spiv: https://chinstrap.ubuntu.com/~dsilvers/paste/fileCLSqVH.html
[05:59] <flacoste> looked to me likes these should have passed
[06:05] <spiv> flacoste: they do, yeah.
[06:05] <flacoste> what should I do in case like these?
[06:06] <spiv> flacoste: And if I compare those strings with doctest directly, it passes.
[06:07] <spiv> There's no missing voodoo I know of.
[06:07] <spiv> Those are just plain weird errors.
[06:07] <spiv> Where's the branch?
[06:08] <flacoste> sftp://chinstrap/home/warthogs/archives/flacoste/launchpad/ticket-tracker-bug40084/
[06:08] <spiv> Well, one of them is a real error...
[06:08] <spiv> (the <div class="error message">+notificationtest3 error</div>
[06:08] <spiv> one)
[06:09] <flacoste> indeed, and that's caused because LaunchpadView now doesn't render on redirect
[06:10] <flacoste> so, I should change the test there for sure
[06:10] <flacoste> and maybe the other ones have the same problem
[06:11] <flacoste> i.e. empty body makes doesn't match the ...
[06:11] <spiv> Well, the other ones all have Content-Length: 0, but that shouldn't cause them all to fail.
[06:14] <BjornT> flacoste: that's correct, ... doesn't match <BLANKLINE>. so if you made the page have an empty body, you have to remove the last ...
[06:14] <flacoste> will do, thanks for the help!
[06:14] <spiv> BjornT: weird that I don't see that when using doctest directly.
[06:15] <BjornT> spiv: it depends on if you use NORMALIZE_WHITESPACE or not. if you don't specify that flag, they will match.
[06:15] <BjornT> NORMALIZE_WHITESPACE is turned on in our pagetests
[06:15] <spiv> Ah.
[06:16] <BjornT> https://launchpad.net/products/launchpad/+bug/267
[06:17] <spiv> BjornT: Hmm, not even then, actually.
[06:18] <spiv> BjornT: https://chinstrap.ubuntu.com/~dsilvers/paste/fileiLR7hA.html says they match, even with ELLIPSIS + NORMALIZE_WHITESPACE
[06:20] <BjornT> spiv: right. i said that if you don't specify it, they will match. i.e. specify only ELLIPSIS
[06:20] <BjornT> at least i meant to say it.
[06:20] <spiv> BjornT: They match without it too.
[06:22] <spiv> BjornT: If you could tweak that paste to reproduce the failure, I'd like to see it :)
[06:22] <spiv> But for now, it's bedtime...
[06:24] <BjornT> spiv: hmm. either something wrong with that test you pasted, or it got fixed in upstream python...
[06:29] <spiv> BjornT: I tried with both doctest.py files in the launchpad+zope tree, too :)
[06:29] <spiv> BjornT: So I presume it's something wrong with my attempt to replicate the conditions of flacoste's failure, but I don't know what.
[09:46] <djk_> how do you delete a launchpad-account?
[09:47] <mongolito404> Hoi. I'm in charge of writing a PHP software that will probably use Rosetta for translation. I'm totally new to Rosetta and to i18n and a bit lost. Where should I look for Rosetta usable i18n technology in PHP ?
[10:17] <flacoste> i'm entangled in a bazaar merge puzzle
[10:18] <kiko> flacoste, what's up?
[10:18] <kiko> hey bradb 
[10:18] <flacoste> i've got a new branch which I started from rocketfuel from yesterday
[10:18] <bradb> hey kiko 
[10:18] <kiko> okay so far
[10:18] <kiko> bradb, got my email?
[10:19] <flacoste> it's called ticket-makebug-fixes
[10:19] <bradb> kiko: i did indeed
[10:19] <kiko> bradb, looked okay to you?
[10:19] <flacoste> it's for working on bug #3849 and bug #50007
[10:19] <Ubugtu> Malone bug 3849 in launchpad-support-tracker "No navigation to bug report created from a support request" [Medium,In progress]  http://launchpad.net/bugs/3849
[10:19] <Ubugtu> Malone bug 50007 in launchpad-support-tracker "When filing a bug related to the support tracker, the summary and description should be editable" [Medium,In progress]  http://launchpad.net/bugs/50007
[10:19] <kiko> flacoste, cool
[10:20] <flacoste> since it build on the changes in ticket-tracker-bug-40084 which wasn't merged yet in rocketfuel
[10:20] <flacoste> i merged that branch in ticket-makebug-fixes
[10:20] <kiko> okay so far
[10:20] <bradb> kiko: Yup. Those have basically been the hot topics so far.
[10:21] <flacoste> now, I'm done with the fixes and ticket-tracker-bug-40084 was merged this afternoon in rocketfuel
[10:21] <kiko> okay great
[10:22] <kiko> bradb, thats very cool
[10:22] <flacoste> I synced up rocketfuel but now when I do a bzr diff -r ancestor:../rocketfueld/launchpad I see that it would want to apply changes that were merged from ticket-tracker-40084 to rocketfuel
[10:22] <bradb> e.g. xmlrpc, release bug management, bug-reporting-tool, drinking-from-the-firehouse (dealing with large volumes of bug reports), etc.
[10:22] <kiko> flacoste, and indeed it would
[10:23] <kiko> because once you merged those revisions into your tree
[10:23] <kiko> they are "yours"
[10:23] <flacoste> i though that bazaar was able to see that the same set of changes was merged in both tree!
[10:24] <flacoste> guess, i've credited bazaar with too much cleverness
[10:24] <kiko> flacoste, it's not really possible to calculate that delta
[10:24] <kiko> because, well, if you depend on the features in the branch
[10:24] <kiko> then how can you expect your merge to RF to work?
[10:24] <kiko> i.e. you require some method added in that branch
[10:25] <flacoste> because I won't merge until the other branch is also merged :-)
[10:25] <kiko> how would your branch merged against RF without the prerequisite branch know
[10:25] <kiko> oh
[10:26] <flacoste> ticket-makebug-fixes is really the continuation of ticket-tracker-bug-40084
[10:27] <flacoste> i'll just save between this two and apply it to a brand new branch from rocketfuel, that will work
[10:27] <flacoste> s/save/save the diff/
[10:27] <kiko> that would work indeed
[10:28] <matsubara> flacoste: did you try bzr diff -r branch:../rocketfuel/launchpad ? AIUI, ancestor compares you current branch against the last branch you have from ancestor.
[10:28] <flacoste> well, the doc says that ancestor: is used for the merge, so I thought I should use that one
[10:28] <flacoste> it's probable that bzr diff -r branch: would work
[10:29] <flacoste> well, it does work but also shows me all the new stuff in rocketfuel
[10:30] <flacoste> when I tried bzr merge ../rocketfuel/launchpad, I did get conflicts on the files related to ticket-tracker-bug-40084
[10:44] <kiko> does anyone know if a rollout is scheduled for this week?
[10:44] <kiko> SteveA?
[10:45] <SteveA> kiko: there was a rollout this morning.  
[10:45] <kiko> SteveA, oh, fabulous!
[10:45] <kiko> SteveA, do you know what revision level?
[10:46] <SteveA> look at the wiki page
[10:46] <SteveA> https://launchpad.canonical.com/LaunchpadProductionStatus
[10:46] <kiko> amazing! 
[10:48] <kiko> fab
[10:48] <SteveA> f.a.b.
[10:48] <kiko> matsubara, interesting to see what timeouts we get tomorrow -- hopefully my work has helped us!
[10:49] <kiko> if everything worked out as well as I hope it did I now need to get the +translations page finally under control
[10:49] <kiko> anyway
[10:49] <kiko> outta here!
[10:56] <djk_> how do you delete a launchpad-account?