[01:34] <sabdfl> sivang: register the spec against the product or distro that is relevant, then assign them all to the ubz agenda
[01:35] <sivang> sabdfl: ok, thx.
[01:39] <sabdfl> np. i will write up the full process and post to the lists this week so that we get as many people drafting specs before UBZ as possible
[01:40] <sivang> cool
[01:40] <sivang> sabdfl: I've filed a wishlist bug for the spec tracker,
[01:40] <sivang> sabdfl: that it should also allow grouping acording to tracks from the main sprint overview 
[01:41] <sivang> sabdfl: I already have a handful of specs to register in it.
[01:41] <sabdfl> sivang: there aren't tracks
[01:41] <sabdfl> the specs should be grouped by distro or product
[01:41] <sivang> eh, right
[01:41] <sabdfl> the conf is just an aggregation of them
[01:42] <sivang> yes that's what I meant :)
[01:42] <sabdfl> we'll use some sort of autoscheduler to figure out who can meet at any time
[01:42] <sabdfl> based on priority and status of the specs
[01:42] <sivang> yeah, so I've heared here and there :)
[01:42] <sivang> would be cool putting it to the test
[01:43] <sabdfl> will be, yes
[01:43] <sivang> sabdfl: is there some sort of linear programming algo there? (like ones used for solving transportation and logistic problems)
[01:43] <sivang> ?
[01:44] <sabdfl> sivang: nothing too fancy just yet
[01:44] <sivang> sabdfl: k, I should get near a fax machine acutally and stop asking, just "Use the source, Luke" ;-)
[01:45] <sabdfl> fax machine?
[01:46] <sivang> sabdfl: yeah, to sort out the NDA thingy 
[01:47] <sabdfl> ok
[02:01] <sabdfl> Keybuk: i have a bunch of test failures in HCT with my new branch, will need your help
[02:01] <sabdfl> could you baz switch to mark.shuttleworth@canonical.com/launchpad--newpackageclasses--0 and run the HCT tests please?
[02:02] <sabdfl> in summary, the branch creates a ton of new classes, that are basically a "source package within a distro" or "source package release within a distro", same for binary packages, with distro and distrorelease and distroarchrelease
[02:02] <sabdfl> lots of permutations and combinations
[02:02] <sabdfl> anyhoo... HCT doesn't like it, i'll focus on the test failures in LP itself (13 left) if you could have a look at the HCT ones
[02:03] <sabdfl> i think you made up a "distro source package" object yourself, for HCT, right?
[02:03] <sabdfl> i think you'll find the new one quite nice
[02:04] <sabdfl> these are the bits that need to go live tuesday :-)
[02:06] <ajmitch> heh, no pressure there :)
[02:06] <sivang> ajmitch: how come not? this is soyuz that's approaching landing :)
[02:07] <ajmitch> a smooth touchdown is always impressive :)
[02:07] <Keybuk> sabdfl: yup, no problem
[02:07] <sivang> yes, even for trained pilots
[02:07] <Keybuk> you're up to date with the hct branch?
[02:07] <Keybuk> (rocketfuel)
[02:10] <ajmitch> lifeless: AUD?
[02:11] <lifeless> ajmitch: reais!
[02:12] <ajmitch> ah right
[02:12] <sivang> lifeless: some sort of .au currency ?
[02:12] <lifeless> sivang: brazillion
[02:27] <sabdfl> Keybuk: i'm up to the point where i think stub branched for production
[02:27] <sabdfl> it was rocketfuel launchpad patch 2654
[02:27] <sabdfl> not sure what the equivalent hct patch level is
[02:28] <Keybuk> baz tree-id lib/hct
[02:30] <Keybuk> (is just there's a test fix for "on pure breezy" in there, want to make sure you've got that one)
[02:30] <Keybuk> (though I see failures in hctapi I think too ... so there's at least some other problems here)
[02:37] <sabdfl> Keybuk: rocketfuel@canonical.com/hct--devel--1--patch-20
[02:37] <sabdfl> that's out of date
[02:37] <sabdfl> but i didn't want to update to a point that you were depending on launchpad past the branch date (friday)
[02:37] <Keybuk> right
[02:38] <sabdfl> if youre expecting stub to roll the latest stuff out tuesday then am happy to update now
[02:38] <sabdfl> sourcerer too
[02:38] <Keybuk> some schema changes went in recently, let me check you haven't got mismatched versions
[02:43] <Keybuk> sabdfl: yeah
[02:43] <Keybuk> patch-2662
[02:43] <Keybuk>     2005-10-15 05:39:24 GMT
[02:43] <Keybuk>     [r=stub]  replace ManifestEntry patch_on with parent that can apply to all types of entry
[02:43] <Keybuk> you have the Launchpad side of my changes
[02:43] <Keybuk> just update lib/hct and lib/sourcerer
[02:43] <sabdfl> no i don't
[02:43] <sabdfl> i have patch-2654 only
[02:44] <Keybuk> hmm
[02:44] <sabdfl> at least, i darned well hope so
[02:44] <Keybuk> either you've accidentally pulled them in
[02:44] <Keybuk> or baz just lied to me
[02:44] <Keybuk> hang on
[02:44] <sabdfl> bugger damn shoot
[02:44] <Keybuk> oh, wait, maybe I ran the wrong command
[02:44] <Keybuk> no, worry
[02:44] <Keybuk> ignore me
[02:44] <lifeless> sabdfl: baz cat-log rocketfuel@canonical.com/launchpad--devel--0--patch-2655
[02:44] <Keybuk> I ran the wrong baz command :p
[02:44] <sabdfl> cat-log: no log for revision (rocketfuel@canonical.com/launchpad--devel--0--patch-2655)
[02:44] <sabdfl>   tree: /home/mark/projects/ubuntu/launchpad
[02:45] <sabdfl> phew :-)
[02:45] <lifeless> sabdfl: good.
[02:45] <Keybuk> I ran baz missing and somehow managed to think you *had* that revision

[02:45] <Keybuk> sabdfl: what's your tree-id of sourcerer?
[02:45] <sabdfl> Keybuk: patch-30
[02:46] <Keybuk> ok, I'll back mine down to the same ids
[02:47] <sabdfl> oh, this is joy
[02:48] <Keybuk> hmm? :p
[02:48] <sabdfl> there's a simple bug in some code that causes the test suite to spit out PO Template Header parsing error messages
[02:48] <sabdfl> so instead of the bug getting fixed
[02:48] <sabdfl> the error messages are happily embedded in about 90 tests
[02:48] <sabdfl> and i fixed the bug
[02:48] <sabdfl> so all the tests fail :-)
[02:49] <lifeless> gahr
[03:06] <Keybuk> sabdfl: ok, no errors in hct or sourcerer -- checking hctapi
[03:06] <sabdfl> you can run tests with no errors?
[03:07] <Keybuk> depends where the errors are
[03:07] <Keybuk> I am getting errors in hctapi
[03:21] <Keybuk> sabdfl: ahhh, I understand
[03:22] <Keybuk> some of the database functions return new classes, and the url resolver needs to know about them
[03:24] <sabdfl> Keybuk: that makes sense, yes
[03:29] <Keybuk> does anything return SourcePackage anymore?
[03:41] <sabdfl> Keybuk: yes
[03:42] <sabdfl> SourcePackage is a "DistroReleaseSourcePackage" in the new scheme of things
[03:42] <sabdfl> you get:
[03:42] <sabdfl>  DistributionSourcePackage
[03:42] <sabdfl>  DistributionSourcePackageRelease
[03:42] <sabdfl>  SourcePackage
[03:42] <sabdfl>   DistroReleaseSourcePackageRelease
[03:42] <sabdfl>  DistroReleaseBinaryPackage
[03:43] <sabdfl>  DistroArchReleaseBinaryPackage
[03:43] <sabdfl>  DistroArchReleaseBinaryPackageRelease
[03:43] <sabdfl> i think thats it
[03:47] <Keybuk> which one has the magic mapping back to ProductSeries ?
[03:48] <Keybuk> is that still only on SourcePackage ?
[03:49] <Keybuk> (which you could rename to DistroReleaseSourcePackage to match your new names <g>)
[03:58] <sabdfl> Keybuk: yes on both counts
[04:18] <Keybuk> uh, this is weird
[04:18] <Keybuk> baz branch is just hanging
[04:21] <Keybuk> meh, it started as soon as I traced it ... weeeird
[04:26] <Lathiat> dont you love those 
[04:26] <Lathiat> heisenbugs or whatever
[04:28] <Keybuk> it's consistently doing it too
[04:46] <sabdfl> Keybuk: ok, no more test failures in LP itself
[04:46] <sabdfl> do you want help on the hct front?
[04:48] <Keybuk> already done
[04:48] <Keybuk> just waiting for bzr to get its act in gear and actually make the commit
[04:51] <sabdfl> Keybuk: nice.thanks!
[04:51] <sabdfl> we'll have to get lifeless to do a coordinated commit
[04:51] <sabdfl> my stuff is not against current RF, but rather designed to land cleanly on the branch destined shortly for production
[04:53] <Keybuk> I branched from you
[04:53] <Keybuk> you'll just need to pull the one changeset
[04:53] <Keybuk> it was the bits of hct in launchpad that broke, not outside
[04:56] <Keybuk> sabdfl: pull from scott@canonical.com--2005/launchpad--newpackageclasses--0
[04:56] <Keybuk> it just mirrored
[04:56] <Keybuk> is one patch that applies to hctapi and test_hctapi and makes it use your new classes
[05:11] <Keybuk> (just gonna reboot to try and rid myself of baz's strange behaviour)
[05:59] <Koven> night
[05:59] <Koven> I've one question about launchpad
[08:02] <SteveA> good morning launchpadders
[08:07] <GoRoDeK> hi SteveA
[08:08] <SteveA> hi GoRoDeK.  how's is going?
[08:10] <GoRoDeK> SteveA: to sum it up: very good. thx for answering :) but have to go to some math lectures now ... cu
[08:11] <SteveA> cool
[08:48] <\sh> is soyuz landed already? :)
[08:56] <SteveA> the new ui and such from mark? not yet.  hopefully it'll be polished up today, and landed tomorrow.
[08:57] <\sh> sounds good :)
[09:03] <fabbione> \sh: look at the positive side of all this
[09:03] <fabbione> we can blame launchpad if dapper is broken :)
[09:03] <fabbione> any time :)
[09:04] <fabbione> "hey the kernel OOPS..." "must be soyuz bug"
[09:05] <SteveA> we're gonna be rewriting the kernel in python
[09:05] <SteveA> easier to maintain
[09:05] <fabbione> ehhehe
[09:05] <fabbione> have fun :)
[09:06] <fabbione> you will also need a new python interpreter in the MBR
[09:06] <fabbione> otherwise you can't boot
[09:06] <fabbione> oh did i mention that you have approx 512 bytes for that?
[09:06] <\sh> fabbione: this will never happen...but if it's happening, then we have to rename soyuz to apollo 13 ,-)
[09:06] <fabbione> \sh: ehhehe
[09:07] <\sh> fabbione: good morning btw :) 
[09:07] <fabbione> it can't be a good morning
[09:07] <fabbione> it's monday dude..
[09:07] <fabbione> morning sucks on monday
[09:07] <fabbione> ;)
[09:07] <ajmitch> heh
[09:07] <ajmitch> morning, then :)
[09:08] <\sh> fabbione: well, u r right...i have 3 digital tv issues to track down...and it looks like the problem was in front of the tv or settop box
[09:08] <ajmitch> \sh: isn't that usually the case?
[09:08] <\sh> ajmitch: only in 99% of the cases
[09:08] <\sh> 1% is my fault ,-)
[09:10] <fabbione> \sh: wrong :)
[09:10] <fabbione> that 1% must be somebody else fault
[09:10] <fabbione> you just need to figure of who :)
[09:11] <ajmitch> that's why you need a system like soyuz?
[09:12] <fabbione> ajmitch: exactly :)
[09:12] <\sh> fabbione: oh yes...but u know who managing works...100% sure, that the management is at fault, but this is wrong, so team lead is wrong and team lead is always right, so at this point the guy from ops is the fault ,-)
[09:12] <\sh> s/who/how/
[10:02] <carlos> morning
[10:04] <SteveA> hello carlos
[10:04] <carlos> hi
[10:05] <SteveA> i'm going to do a bit of work on potemplate UI today
[10:05] <SteveA> to make its facet menus better, and to make it produce a proper NotFound 404 error
[10:05] <SteveA> when going to a not found language
[10:10] <carlos> ok
[10:10] <carlos> SteveA, cool, thanks
[10:18] <jordi> If I need to report some issue with my connectivity/ability to read email, is it sounder@ or allhands@?
[10:19] <jordi> My main mailserver has bad hardware trouble, it seems
[10:22] <SteveA> not sounder@
[10:22] <SteveA> mostly people use warthogs@
[10:22] <SteveA> morning david
[10:34] <jordi> err, I meant warthogs, not sounder of course
[10:34] <jordi> thanks steve
[10:35] <SteveA> jordi: do you need an email account to keep up with the rosetta list etc?
[10:38] <jordi> SteveA: I probably should, yes. I don't know how much time my normal emial account will be down
[10:40] <SteveA> how about a gmail account?
[10:40] <SteveA> give me an email address for you, and i'll send you an invitation
[10:41] <carlos> SteveA, I think elmo said that we can get a pop3/imap account if needed
[10:42] <SteveA> yeah
[10:42] <SteveA> but
[10:42] <carlos> instead of an email alias
[10:42] <SteveA> why bother?
[10:42] <carlos> :-)
[10:42] <carlos> just noting it
[10:42] <SteveA> i can give jordi a gmail account right now
[10:42] <SteveA> we can ask the admins for an imap account too
[10:42] <SteveA> but, that will certainly take a bit longer
[10:43] <carlos> right
[10:43] <SteveA> i guess the canonical imap account would show mail as coming from jordi's canonical email address
[10:43] <jordi> oh, I have a gmail account.
[10:43] <SteveA> cool
[10:44] <jordi> I could temporarily use that
[10:44] <jordi> I'll subscribe
[10:44] <SteveA> so, i guess you just need your canonical email address pointed at it ?
[10:44] <jordi> if it can be snet to both, that'd be best
[10:51] <SteveA> jordi: fabbione had a suggestion on #canonical
[10:52] <jordi> SteveA: inside the DC
[10:53] <jordi> db.ubuntu.com does not exist here
[10:53] <SteveA> yeah, so ssh to chinstrap
[10:53] <SteveA> and use it from there
[10:53] <jordi> oh. AFAIK, I have no account.
[10:53] <SteveA> okay, then an RT request is what it'll have to be.
[10:54] <Kinnison> Hmm, no stub
[10:55] <SteveA> he should be around shortly
[10:55] <SteveA> he's having some 'ISP blocks irc' issues
[10:55] <SteveA> i've given him an account somewhere to use to do ssh tunneling
[10:58] <jordi> how can isps be so fascist
[10:59] <jordi> probably some enlightened mind thought IRC is only used by worm bots or whatever
[10:59] <Lathiat> quite easily apparently
[11:00] <SteveA> my home ISP has blocked one irc port
[11:00] <SteveA> but only at the adsl router
[11:01] <SteveA> so, i'll reconfigure it when i get around to it
[11:01] <SteveA> they don't mind my doing so
[11:01] <SteveA> it's just the default configuration
[11:02] <Lathiat> thats less bad
[11:02] <Lathiat> i like the ISPs that block incoming ports but let you turn it off
[11:02] <Lathiat> smtp, etc
[11:03] <jamesh> isn't stub on iinet?
[11:03] <jamesh> I haven't had any troubles
[11:04] <Lathiat> yeh he is
[11:04] <Lathiat> doesn't sound like something iinet would do
[11:04] <Lathiat> might just be a b0rked route or something
[11:06] <SteveA> jamesh: he's relocated to thailand.
[11:06] <jamesh> ah
[11:07] <SteveA> jamesh: good morning.
[11:07] <SteveA> how's the bugzilla conversion doing?
[11:10] <SteveA> lifeless: does anyone use a 307 response ?
[11:11] <lifeless> ECONTEXT
[11:12] <lifeless> its pretty much guaranteed that someone does, as the ietf likes reference uses before making a standard
[11:12] <jamesh> SteveA: I've got the basics in place for going through the list of bugs, creating users as needed
[11:12] <jamesh> SteveA: I'm just working out the details of converting the comments
[11:13] <SteveA> lifeless: does anyone *really* use it?
[11:13] <SteveA> jamesh: cool.  any issues so far?  anything you need to tell brad or bjorn about?
[11:14] <jamesh> SteveA: nothing in particular.  The existing malone interfaces seem sufficient to implement the whole lot
[11:14] <jamesh> which is good
[11:14] <SteveA> that's good
[11:14] <SteveA> and, any news on the autoscheduler magic?
[11:14] <jamesh> SteveA: some web browsers handle 307 responses (temporary redirects) different to permanent redirects
[11:15] <jamesh> SteveA: they update bookmarks if they see a permanent redirect but not otherwise
[11:15] <SteveA> okay
[11:15] <SteveA> i'm adding some stuff to Navigation to allow permanent redirects as an option
[11:16] <SteveA> i need it for making the shipit user.cgi permanently redirect to the shipit homepage
[11:16] <lifeless> SteveA: I dont know. if you want to make a permanent redirect, use a permanent redirect
[11:17] <SteveA> yeah.  was just curious about 307.  haven't seen any software that specifically supports it.
[11:17] <SteveA> and i figured you'd know
[11:17] <jamesh> we should support "402 Payment Required"
[11:17] <lifeless> SteveA: I haven't seen any bug reports about it :)
[11:18] <SteveA> for bounties?
[11:18] <jamesh> dunno :)
[11:18] <SteveA> for +addkarma ;-)
[11:18] <ajmitch> heh
[11:54] <sabdfl> moin moin sports fans
[11:54] <Kinnison> sports?
[11:54] <Kinnison> Is this lesser-spotted-lunchpad-developer hunting?
[11:59] <Kinnison> carlos: gah, your patch pipped mine into pqm's queue by nine seconds
[11:59] <carlos> Kinnison, well, It's fourth time I try to merge that patchset and was rejected always because baz conflicts....
[12:00] <carlos> I think it's time to get it merged ;-)
[12:00] <Kinnison> :-)
[12:00] <SteveA> Kinnison: should i put aside time to do reviews for you today?
[12:02] <Kinnison> SteveA: if stub can turn up and check a db patch and give me a blessing then yes
[12:02] <Kinnison> SteveA: otherwise I'm not sure how we're gonna open 'cos I have a patch which *has* to hit production and I've failed to get stub every time I've tried
[12:03] <SteveA> carlos: do you know why LanguageNotFound is a ValueError, and is not a NotFoundError ?
[12:03] <carlos> phone....
[12:04] <SteveA> Kinnison: have you mailed stub about it?
[12:04] <Kinnison> SteveA: No because I'm useless
[12:04] <Kinnison> SteveA: I was just moving my stuff to my desktop
[12:04] <SteveA> DOIT
[12:04] <Kinnison> SteveA: I'll mail him in about 5 minutes
[12:04] <SteveA> ta
[12:06] <Kinnison> stub@canonical.com ?
[12:08] <SteveA> stuart.bishop@
[12:08] <SteveA> stub might work
[12:08] <Kinnison> ta
[12:11] <carlos> SteveA, because I think it was a ValueError and I think you asked me to create its own exception and seems like I choose the wrong parent....
[12:11] <SteveA> okay
[12:11] <SteveA> i'm changing it to derive from NotFoundError
[12:12] <carlos> ok, thank you
[12:17] <SteveA> sabdfl: bug 2151
[12:17] <Ubugtu> Malone bug #2151: No page to edit architecture details (404 error) Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Nobody, Status: New http://launchpad.net/malone/bugs/2151
[12:17] <SteveA> sabdfl: does your branch fix this?
[12:29] <SteveA> Kinnison: stu will be around very shortly
[12:29] <Kinnison> SteveA: yeah, I'm chatting to him on jabber since I found him there
[12:29] <SteveA> k
[12:30] <sabdfl> SteveA: no
[12:30] <SteveA> sabdfl: ta.  i'm doing a bunch of bugfixing, but i don't want to conflict with your landing.
[12:40] <stub> Weee
[12:41] <Kinnison> stubby!
[12:41] <sabdfl> hey stubarooney
[12:42] <stub> Morning
[12:42] <SteveA> stub: i'm preparing a branch with a bunch of fixes of the most common errors in launchpad
[12:42] <SteveA> it's for merging into RF, but would be neat to cherrypick it if there's no conflicts
[12:42] <stub> This is all sounding like we are rolling out HEAD tomorrow
[12:42] <SteveA> ooh, fun
[12:42] <SteveA> that would be simpler
[12:42] <SteveA> although i'd like to see a day on staging first
[12:43] <stub> Pretty much everything that has landed recently is a bug fix for cherry picking, or critical.
[12:43] <SteveA> it's the "critical" i worry about
[12:43] <SteveA> these things sometimes introduce bad regressions
[12:43] <stub> We can roll out staging at the press of a button (well I can since I have the keys all setup). Will Kiko's minions have time to abuse it?
[12:44] <SteveA> hmm... i need to get some lunch.  i have a few hours more of fixes to make.
[12:44] <SteveA> matsubara and gneuman will be able to give it some stick
[12:44] <SteveA> i think it is worth doing.
[12:44] <stub> ok. I'm going to try roll out Keybuk's and/or sabdfl's Gina spanking onto staging
[12:44] <SteveA> kiko can help them be systematic about it
[12:45] <SteveA> okay, so we plan staging today, hammer staging tonight, roll out tomorrow.
[12:45] <sabdfl> stub: i'm just merging keybuk's changes now
[12:45] <SteveA> sabdfl: what do you think of above plan?
[12:45] <sabdfl> with a bit of luck that will result in tests passing across the board
[12:45] <stub> sabdfl: Good. Saves me trying to land it ;)
[12:45] <sabdfl> SteveA: why roll out head tomorrow?
[12:46] <SteveA> compared to what?
[12:46] <stub> sabdfl: Ping me when you have it merged and I can roll it out to staging while PQM is chewing on it
[12:46] <sabdfl> SteveA: patch -2654 + fixes
 Pretty much everything that has landed recently is a bug fix for cherry picking, or critical.
[12:46] <sabdfl> stub: ok. i also have some decruftifying work to do
[12:46] <stub> Yer - if it isn't HEAD, it sure will smell like it
[12:47] <sabdfl> hmm..
[12:47] <sabdfl> one thing i am concerned about with my changes is cron scripts
[12:47] <sabdfl> i don't thnk those have good tests, in general
[12:47] <sabdfl> so if my changes break those...
[12:48] <SteveA> it's been a weak point in the past.
[12:48] <SteveA> but, staging isn't good at picking up such problems
[12:48] <SteveA> i think we can just monitor them closely
[12:48] <SteveA> and fix as needed over the next couple of days
[12:48] <stub> I can cherry pick what I can and just delay any other 'critical' stuff until we have had a chance to kick the tyres on staging then
[12:49] <SteveA> stub: has staging picked up cronscript issues before?
[12:49] <sabdfl> stub: has most of the "critical" stuff other than mine actually landed on HEAD now?
[12:49] <sabdfl> i could update to head
[12:49] <SteveA> Kinnison: i'm going to get some lunch shortly.  i'll be available for reviews when i get back.
[12:49] <sabdfl> only 6 conflicts :-)
[12:50] <Kinnison> SteveA: cool
[12:50] <stub> SteveA: No. The cronscripts that have been tested there worked fine on staging. They have had database locking issues, which we didn't pick up, but I think we are learning to deal with that in our code now (?)
[12:50] <SteveA> stub: what about a cronscript with broken code
[12:51] <SteveA> rather than just contention-prone code?
[12:51] <SteveA> go dilys !
[12:52] <SteveA> Kinnison: the pqm queue going down, so i can get my fixes in ;-)
[12:52] <SteveA> ah, sod it... i'll get them in after lunch.
[12:55] <stub> ok. production--1.37 (scheduled for tomorrow rollout) is currently at patch -2659. There are only three other commits in rocketfuel after that, of which only one (patch-2663, a small fix from salgado) should probably be cherry picked. 
[12:55] <stub> sabdfl: ^^^
[12:55] <sabdfl> my changes are huge (patch is up to 11000 lines, filtered, now, with more to come) and will have bugs. but not bugs that cause data loss
[12:55] <sabdfl> nothing in there writes
[12:55] <sabdfl> its all just better ways to visualise and present the existing data
[12:55] <sabdfl> i'm hoping finally to remove the SourcePackageInDistro and VsourcePackageReleasePublishing view cruft
[12:56] <stub> so I think we are good to go for a production rollout tomorrow. My comment was partially based on the assumption that sabdfl's branch would need to be landed as soon as we could confirm it worked on staging in order to get Dapper open
[12:58] <stub> So.... gina has already been run on production. Has anyone given any though to it we can leave cleanup until later or should we do that now? Or is is a simple matter of just deleting the contents of a few tables and letting Gina-NG rebuild them?
[12:59] <sabdfl> erk
[12:59] <sabdfl> we agreed NOT to run gina on production
[12:59] <sabdfl> it has known
[12:59] <sabdfl> issues
[12:59] <stub> sabdfl: That was agreed *after* it had been run
[12:59] <Kinnison> sabdfl: gina ran on production (to an extent) before you raised the issues
[12:59] <sabdfl> and we were going to look at the results of a run on staging AFTER my branch had landed
[12:59] <sabdfl> fuck
[12:59] <sabdfl> that's not good
[01:00] <Kinnison> it's very easy to clean out what we're worried about
[01:00] <stub> Kinnison: Cool.
[01:00] <Kinnison> we empty the following tables (in the following order): binarypackagefile securebinarypackagepublishinghistory binarypackagerelease build
[01:00] <sabdfl> ok
[01:00] <sabdfl> then fix gina
[01:00] <stub> ok. I'll do that on staging before testing Gina.
[01:01] <sabdfl> then run again?
[01:01] <Kinnison> yeah, gina is additive. take away what she's done and she'll do it again
[01:03] <sabdfl> great. thanks Kinnison, made my day there
[01:03] <sabdfl> did kiko make the gina fix?
[01:03] <sabdfl> stub: keybuk's hctapi fixes merged and mirrored. running tests now
[01:04] <sabdfl> stub: when you make a production config, does it include specific patch levels of all the other branches? like hct and sourcerer?
[01:05] <sabdfl> what happens if they move on post-thursday, compatible with launchpad--devel--0 but not the thrusday branch for production?
[01:05] <stub> sabdfl: Generally no, but I have had to do it on occasions
[01:05] <sabdfl> ok
[01:05] <sabdfl> i just want to know what versions to be testing with
[01:05] <stub> I'll freeze it now
[01:10] <stub> sabdfl: so - I thought your branch you are landing now had the Gina fixes. I think your changes and Steve's are the only ones that have been made, and I don't know if your branch includes SteveA's.
[01:10] <sabdfl> stub: no. kiko was going to make the gina fixes
[01:10] <stub> ok.
[01:17] <matsubara> good morning!
[01:18] <ajmitch> morning
[01:22] <stub> sabdfl: https://chinstrap.ubuntu.com/~dsilvers/paste/fileV27R74.html
[01:30] <sabdfl> ah. the gina error in my tests was the absence of dpkg-source on the test machine :-)
[01:31] <sabdfl> (not the db gina error, the weird gina test failure)
[01:31] <sabdfl> stub: gracias, will test against that
[01:40] <sabdfl> stub: can i add this to 25-41-0.sql?
[01:40] <sabdfl> -- rid ourselves of unused views
[01:40] <sabdfl> DROP VIEW vsourcepackageindistro ;
[01:40] <sabdfl> DROP VIEW vsourcepackagereleasepublishing ;
[01:40] <sabdfl> i'm just running tests now
[01:41] <sabdfl> want to get rid of that old cruft. Kinnison will be happy.
[01:43] <stub> sabdfl: Sure
[01:43] <sabdfl> thanks
[01:44] <stub> Dinner time.
[02:06] <mpt> Goooooood morning
[02:09] <sivang> hey mpt  :)
[02:16] <carlos> see you
[02:18] <sabdfl> hey kiko
[02:18] <sabdfl> stub: looking good here, tests all passing with the extra views removed
[02:19] <sabdfl> vsourcepackagereleasepublishgin and sourcepackageindistro are *toast*
[02:19] <sabdfl> lifeless, ddaa: question. if I'm at --patch-20 and swtich to --patch-22 why does it not show modified files?
[02:19] <sabdfl> talks about patch logs etc, but not modified files
[02:20] <sabdfl> baz status then shows the correct patch level
[02:20] <dilys> Merge to rocketfuel@canonical.com/dists--devel--0: [trivial]  production-1.37 freeze (patch-121: stuart.bishop@canonical.com)
[02:21] <kiko> morning vietnam
[02:22] <sabdfl> it may be napalm city today, indeed
[02:22] <sabdfl> kiko: did gina get that dup-entry-fixin' love?
[02:23] <sabdfl> we've nuked the production run results in order to run it again with the fix
[02:25] <kiko> let me see
[02:25] <kiko> nope, not that I'm aware of
[02:26] <kiko> the last commit that touched it was stub's 
[02:28] <sabdfl> gina?
[02:28] <sabdfl> kiko: ok, you said that was a faily straightforward fix, will you do it? i thought that's what we planned
[02:28] <kiko> yes, that touched gina
[02:29] <kiko> verifying that the binarypackagerelease had already been built and early-returning, right?
[02:29] <kiko> what's kind of bad is that I don't have any test rig for gina set up
[02:29] <kiko> cprov, do you have one?
[02:29] <kiko> sabdfl, do you have a test rig set up?
[02:29] <kiko> I can try mailing you a patch instead of you like
[02:30] <cprov> kiko: no 
[02:31] <SteveA> re
[02:31] <kiko> suck
[02:32] <sabdfl> kiko: no, i don't have one handy
[02:34] <kiko> sabdfl, I need a full archive and a katie dump, right? I wonder if it's easy to run it on staging
[02:34] <kiko> cprov, Kinnison: is it?
[02:35] <cprov> kiko: at least one full component, katie isn't mandatory anylonger
[02:35] <SteveA> sabdfl: nice post to the faq about __getitem__ style.  i'll correct the bug in the sample code.  i agree with it.  one thing i want to avoid though is having 'heavy' or ambiguous __getitem__ methods.
[02:35] <SteveA> that is, they should give access to just one or at most two types of things.
[02:35] <cprov> kiko: try to run on gwyddion or use my external HD in your lappy
[02:36] <SteveA> otherwise foo[bar]  is hard to understand
[02:36] <kiko> cprov, the issue is that I don't think you can reproduce the error with a single run with no katie dump
[02:37] <uws> Hmmm. Is the oops-i'm-suffering-from-brain-damage-so-please-mail-me-my-password functionality broken or what?
[02:37] <kiko> is it broken?
[02:37] <cprov> kiko: I see, you're right, anyway we have complete warty/hoary here if you think it'd be udeful for you.
[02:38] <uws> kiko: I don't get any mail from it, it seems
[02:38] <uws> kiko: Are you a project admin?
[02:38] <uws> kiko: if so, you can grep logs for   "uws+launchpad@xs4all.nl"
[02:38] <ddaa> sabdfl: probably because there were no modified files, I guess
[02:42] <uws> kiko: nevermind, it was just terribly slow :P
[02:43] <SteveA> kiko: hi!
[02:43] <kiko> how are you SteveA 
[02:43] <SteveA> i'm good.  how are you in brazil?  still hot like the she-devil?
[02:44] <kiko> yeah, pretty hot.
[02:44] <SteveA> so, i've fixed a bunch of the "top error" bugs
[02:44] <SteveA> on my tree
[02:44] <SteveA> i've got a few more to do
[02:44] <kiko> cool
[02:44] <SteveA> going to try to get it into the rollout that is happening over tonight --> tomorrow, along with mark's soyuz stuff
[02:45] <SteveA> has anyone talked with you about organising QA work tonight?
[02:45] <kiko> not yet
[02:45] <SteveA> okay
[02:45] <kiko> I'm at the moment concerned with fixing gina
[02:45] <SteveA> so, stub checked, and there isn't much added to HEAD other than what would need to be cherrypicked anyway
[02:45] <SteveA> so, we're going to do a full rollout
[02:45] <kiko> cool
[02:45] <SteveA> can you organise a few people in brazil to do QA on staging later today?
[02:46] <kiko> land your top-fixes
[02:46] <kiko> sure, what time do we roll out?
[02:46] <SteveA> basically, to make a list of typical pages, and go through them
[02:46] <SteveA> trying stuff out
[02:46] <kiko> or are we blocked on me fixing gina? :-)
[02:46] <SteveA> we need to clear out the old gina data, and then to do a gina run with a fixedup gina
[02:46] <elmo> dapper is blocked on gina
[02:46] <SteveA> and that's a precondition for getting Kinnison
[02:46] <SteveA>  's latest stuff
[02:47] <SteveA> which is needed, as elmo says, for opening dapper
[02:47] <kiko> okay
[02:47] <kiko> well
[02:47] <SteveA> how's the gina looking?
[02:47] <kiko> tbh I can can try fixing gina but it will be a blind-fix unless I set up a local run
[02:47] <SteveA> we can do runs on staging
[02:47] <SteveA> stu will be back from lunch soon
[02:47] <kiko> SteveA, lunch?
[02:47] <SteveA> lunch
[02:47] <kiko> wtf
[02:48] <kiko> anyway
[02:48] <SteveA> or dinner
[02:48] <Lathiat> its 9-11pm in australia :)
[02:48] <SteveA> what does 'dinner' mean anyway
[02:48] <SteveA> stub is in thailand
 Dinner time.
[02:48] <SteveA> that's what he said
[02:48] <kiko> SteveA, can we do runs on staging?!
[02:48] <Lathiat> dinner is ambiguous
[02:49] <SteveA> when i debugged some gina stuff last week
[02:49] <Lathiat> some people say it means lunch, some say it means tea time (night meal)
[02:49] <SteveA> it was on staging
[02:49] <SteveA> staging can't take a complete gina run
[02:49] <SteveA> but we can do main, for example
[02:49] <SteveA> we can do various sections
[02:50] <SteveA> hmm... are my gina changes in RF?
[02:50] <Nafallo> dinner doesn't mean "put foodish stuff into <body>"?
[02:50] <SteveA> kiko: apparently not
[02:51] <SteveA> kiko: i'll mail you a patch
[02:51] <kiko> SteveA, thanks 
[02:51] <SteveA> this gets gina running to completion in production / staging
[02:52] <SteveA> there was one issue with gina not handling unsigned .dsc files
[02:52] <SteveA> and another issue about a simple bug in some SQL
[02:52] <SteveA> there's also a bit of extra debugging output
[02:52] <kiko> why can't we do a complete gina run on staging?
[02:53] <SteveA> disk space
[02:53] <kiko> is there a way of triggering the gina error quickly?
[02:53] <SteveA> what error?
[02:53] <SteveA> with that patch, gina runs to completion on main, security, for breezy, hoary and warty, on staging
[02:54] <kiko> it raises errors though
[02:54] <SteveA> the remaining issue is what mark was talking about friday night
[02:54] <SteveA> what kind of errors?
[02:54] <SteveA> with the patch i sent you, it raises no errors
[02:55] <kiko> psycopg.IntegrityError: ERROR:  duplicate key violates unique constraint
[02:55] <kiko> +"binarypackagerelease_binarypackagename_key"
[02:55] <kiko> what is the remaining issue then?
[02:55] <SteveA> yes.  i fixed it
[02:55] <SteveA> with Kinnison's help
[02:55] <SteveA> on friday
[02:55] <SteveA> and ran it on staging
[02:55] <SteveA> to check
[02:55] <kiko> and?
[02:55] <SteveA> the remaining issue is that gina is putting duplicate data into the database
[02:56] <kiko> oh
[02:56] <kiko> into what table?
[02:56] <SteveA> see the launchpad list
[02:56] <SteveA> "Gina run in production on tuesday, staging with mark's patch monday"
[02:56] <SteveA> i sent it
[02:56] <SteveA> with a transcript from this channel
[02:56] <SteveA> you were there
[02:57] <SteveA> you're in the goddamn transcript ;-)
[02:57] <kiko> I believe I actually said something
[02:57] <Kinnison> sorry guys, I was at lunch
 kiko: i don't know how gina collects info before it decides to
[02:57] <SteveA> put it in the db
 but i'm guessing..
 it finds all the bin packages for the arch, from the source package
 then creates a build for those
[02:57] <Kinnison> anything needing my input?
 it should first look to see if the bin package is there already
 if it finds one, it should find them all
[02:57] <SteveA> 
[02:57] <SteveA> Kinnison: yes
[02:57] <SteveA> please read the above
[02:57] <SteveA> and think about it
[02:57] <SteveA> and talk with kiko about it
[02:57] <Kinnison> how much of "the above" ?
[02:57] <kiko> yeah yeah
[02:58] <SteveA> the <sabdfl> part
[02:58] <Kinnison> Right, okay
[02:58] <Kinnison> so I know what sabdfl's on about
[02:58] <Kinnison> kiko: what do you want to know?
[02:58] <kiko> let me clean out my tree, 1m
[03:00] <kiko> SteveA, sabdfl: what area of launchpad should we emphasize the test on?
[03:00] <SteveA> souyz pages primarily
[03:00] <SteveA> but make sure that the rest still works
[03:00] <SteveA> we really don't want a failure in shipit
[03:00] <SteveA> so that should be tested to make an order or two
[03:00] <SteveA> and to sign up
[03:00] <SteveA> with a new account
[03:01] <kiko> sure
[03:03] <SteveA> is salgado going to be around today?
[03:03] <kiko> yes, he has a class on monday morning or something, that's all
[03:03] <SteveA> bug 2714 probably needs some love
[03:03] <Ubugtu> Malone bug #2714: sort doesn't sort entire data set Fix req. for: shipit (upstream), Severity: Normal, Assigned to: Guilherme Salgado, Status: Accepted http://launchpad.net/malone/bugs/2714
[03:04] <kiko> SteveA, salgado has some poll fixes that I would rather we did first -- shipit is coming along acceptably for now
[03:04] <kiko> the sorting is nice-to-have but not critical IMO
[03:04] <SteveA> k
[03:05] <SteveA> i'm going to get on with fixing the remaining top bugs.  we'll sort out the staging situation when stub returns.
[03:05] <kiko> Kinnison, my tree is clean, let me start looking at the code
[03:06] <SteveA> did you get that patch i mailed?
[03:07] <SteveA> jordi: is your mail sorted out yet?
[03:07] <kiko> SteveA, yes, I did
[03:08] <SteveA> cool
[03:09] <jordi> SteveA: let me check
[03:09] <kiko> jordi, hospital?
[03:10] <jordi> kiko: I had my back checked after 6 months
[03:10] <jbailey> LP seems to not want me to log in.
[03:10] <jordi> 6 months ago I had big pain
[03:10] <jordi> jbailey!
[03:10] <jbailey> Is it known to be broken, or is it somehow just me?
[03:10] <SteveA> jbailey: could be the 'lack of Vary' header problem
[03:10] <jbailey> jordi!
[03:10] <SteveA> jbailey: if you're going via an aggressive proxy
[03:10] <SteveA> it's high on my list of things to fix, after the top 10 not working pages
[03:10] <jbailey> SteveA: I have no idea what my ISP has.
[03:11] <SteveA> let me try logging out and then in...
[03:11] <jbailey> Is there a suggested workaround?  And/or does just replying to bug mail get it into malone? =)
[03:11] <SteveA> jbailey: try going to https://launchpad.net/+login
[03:11] <SteveA> try shift+reload of the page you want to see
[03:11] <SteveA> hi stub 
[03:12] <SteveA> so, kiko will need to try new-gina out on staging
[03:12] <mdke> jordi, got some more fixed po files for you if you are up for it
[03:12] <SteveA> can we get staging ready, with gina data nuked ?
[03:12] <SteveA> and maybe the librarian GC-ed ?
[03:12] <elmo> umm
[03:12] <elmo> staging doesn't have enough data to test the bug that needs fixed
[03:13] <BjornT> jbailey: replying to bug mail works as well
[03:13] <kiko> that was my concern
[03:13] <kiko> elmo, what do you suggest?
[03:13] <SteveA> elmo: don't we have main from both breezy and hoary on there?
[03:14] <jbailey> SteveA: Lovely.  Logging in on that other page and then shift-reload worked.
[03:14] <elmo> oh, err, don't mind me - I seem to have gotten confused
[03:14] <jbailey> Shift-reload hadn't worked before that.
[03:14] <jbailey> BjornT: Nice, thanks. =)
[03:14] <elmo> asuka apparently has a full mirror archive.u.c mirror
[03:14] <stub> SteveA: Librarian garbage collected means deleting all the files btw ;)
[03:14] <Kinnison> SteveA: Most python equiv of "touch $foo" => open(foo,"w").close() yes?
[03:14] <SteveA> jbailey: okay.  i wonder... can you try logging out, and seeing if you can reproduce the issue?
[03:15] <SteveA> stub: is the LGC running in production?
[03:15] <jordi> SteveA: fixed, thanks
[03:16] <stub> SteveA: LibrarianGarbageCollection is still a work in progress (nearly ready to land the first half IIRC)
[03:16] <SteveA> Kinnison: perhaps... maybe open it append.
[03:17] <Kinnison> SteveA: is that w+ ?
[03:17] <SteveA> stub: i'm concerned we'll need it run in production before a gina run in production.
[03:17] <SteveA> elmo: what's the current free space on the librarian please?
[03:18] <elmo> /dev/sda3             537G  361G  149G  71% /
[03:18] <SteveA> Kinnison: os.utime
[03:18] <SteveA> Kinnison: do we have enough space in the librarian to do a full gina run
[03:18] <jbailey> SteveA: Sure, gimme a sec.
[03:18] <Kinnison> SteveA: that doesn't work if the path doesn't exist
[03:19] <Kinnison> SteveA: What I'm actually trying to do is; "test -e $foo || touch $foo"
[03:19] <jbailey> SteveA: No, I cannot reproduce it right now.
[03:19] <elmo> SteveA: FWIW, archive.u.c all 6 arches is 99Gb
[03:19] <SteveA> jbailey: thanks.  please let me know if it happens again.
[03:19] <elmo> I don't know how that translates in gina tho - hopefully it shouldn't grow any
[03:19] <jbailey> SteveA: Will do!
[03:20] <SteveA> hey brad
[03:20] <kiko> ahoy bradb 
[03:20] <bradb> hey dudes
[03:20] <kiko> duderino
[03:20] <SteveA> Kinnison:  you know os.path.exists ?
[03:21] <Kinnison> SteveA: yes, currently I have:
[03:21] <Kinnison>                     if not os.path.exists(comp_path):
[03:21] <Kinnison>                         # When I think about you, I touch myself...
[03:21] <Kinnison>                         open(comp_path, "w").close()
[03:21] <SteveA> looks good to me
[03:21] <kiko> in the future grep -v "\s+#\s$" please
[03:22] <kiko> +$ even
[03:22] <Kinnison> kiko: Pardon?
[03:24] <mpt> sivang: Tell me about the "sivang" person in Launchpad
[03:25] <kiko> @#@!*!9881@&(!
[03:25] <kiko> deleted my old inbox
[03:25] <kiko> that's a sad way of clearing out email :-(
[03:25] <sivang> mpt: what do you mean? :)
[03:25] <mpt> sivang: We're getting many errors from URLs like https://launchpad.net/people/sivang
[03:25] <SteveA> that's interesting
[03:25] <sivang> mpt: errors?
[03:25] <SteveA> when i go there now
[03:25] <SteveA> i get an error
[03:26] <Lathiat> i get a page not found
[03:26] <mpt> sivang: Is that your account?
[03:26] <mpt> Is that an account you merged into another one?
[03:26] <sivang> I didn't merge it :)
[03:26] <sivang> Just changed it
[03:26] <sivang> to https://launchpad.net/people/sivan
[03:26] <kiko> bradb, are you going to fix 933 or can I give it to matsubara?
[03:26] <mdke> mine is giving an error too
[03:27] <sivang> I figured is nicer to have my first name registered, that way have the same for the u.c email alias
[03:27] <mdke> people/mdke
[03:27] <mpt> sivang: Had you posted the old URL anywhere else?
[03:27] <mdke> oops, it is there now
[03:27] <mpt> Like a mailing list?
[03:27] <mpt> or a wiki page?
[03:27] <SteveA> hmm... something interesting is happening when traversing 'people'
[03:27] <SteveA> looks like it somehow returns 'None'
[03:27] <bradb> kiko: You could give it to matsubara if you want. That would *rock* *rock* *rock* to have that fixed.
[03:27] <sivang> mpt: not that I know of, that's quite interesting, I must say
[03:28] <mpt> kiko, can you get the Referer headers included in those logs? SteveA was asking for those last week too, iirc
[03:28] <SteveA> mpt: i think i need to hack stuff to get referer headers easily
[03:28] <kiko> SteveA, yeah, I was amused by your comments about asking me to add referrer information -- there is none :)
[03:28] <SteveA> mpt: but i can get them with a bit of research, matching up error logs with traffic logs
[03:28] <SteveA> kiko: you can match up the two logs
[03:29] <sivang> mpt: no, I didn't mention those anyware. But I saw google has already cached almost all of my page's properties.
[03:29] <sivang> (When I look for myself on google, that is)
[03:29] <mpt> Yes, perhaps it's googlebot
[03:29] <mpt> http://www.google.co.nz/search?hl=en&lr=&q=link:ifqC7ngMP6UJ:https://launchpad.net/people/sivang
[03:29] <kiko> SteveA, that's a bit of work and I'm loathe to do it unless, well, you say it's not includable in the regular log
[03:30] <mpt> so we need the User-Agent header as well :-)
[03:30] <SteveA> mpt: okay, i'll look these up later
[03:30] <SteveA> i need to fix some bugs and get a branch landed now
[03:30] <SteveA> mpt: how's the menus work coming on?
[03:31] <stub> So kiko - you got a branch for me?
[03:31] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: Bits and bobs on distro{arch,}release. r=stevea (patch-2664: daniel.silverstone@canonical.com)
[03:31] <SteveA> mpt: maybe we should spec out leaving a 'redirect' record when a name is changed in launchpad
[03:31] <SteveA> so that, until another name occupies it, the page can say 'you're probably looking for this'
[03:31] <kiko> stub, no, I don't
[03:32] <stub> kiko: ok. Bath time then.
[03:32] <kiko> at the moment I don't even know how to run a test for the fix I have
[03:32] <SteveA> stub: 
[03:32] <bradb> kiko: Do you have a few minutes to wrap up this sort order patch?
[03:32] <SteveA> if you make staging ready for testing gina on
[03:32] <SteveA> i can log in and do the stuff
[03:32] <stub> ok
[03:32] <mpt> SteveA: Like I said, I want to land the design fascism branch first so that I'm not redoing work -- I thought I'd get to land that last week, but the capitalization blocked it
[03:33] <mpt> so I need to put up the branch for sabdfl to look at
[03:33] <SteveA> mpt: okay.  how much left on that?
[03:33] <mpt> and then it should be landable
[03:33] <kiko>     "low": SourcePackageUrgency.LOW,
[03:33] <kiko>     "medium": SourcePackageUrgency.LOW,
[03:33] <kiko>     "high": SourcePackageUrgency.LOW,
[03:33] <kiko>     "emergency": SourcePackageUrgency.LOW,
[03:33] <kiko> Kinnison, why is everything ow?
[03:33] <kiko> low?
[03:33] <SteveA> we're do damn mellow...
[03:34] <SteveA> this is medeterranean urgency levels
[03:34] <SteveA> kiko: how long until you want to try a gina run?
[03:34] <sabdfl> urgency? "maybe"
[03:34] <Keybuk> SteveA: would a patch that changes just two import lines be considered [trivial]  ?
[03:35] <Keybuk> (it also deletes about 15,000 lines
[03:36] <SteveA> Keybuk: perhaps... you can always paste the change into https://chinstrap.ubuntu.com/~dsilvers/paste/, with a summary of the deletion to get an instant review from any reviewer listening.
[03:36] <mattl_> hi, i've been looking at launchpad over the last few days.. it's based on Plone, right? is the source code to Launchpad available anywhere yet?
[03:36] <kiko> based on Plone. hmmm.
[03:37] <kiko> I don't know how to answer that :)
[03:37] <SteveA> mattl_: we use parts of the plone style sheet.
[03:37] <Kinnison> kiko: urgh, what file?
[03:37] <kiko> there's this plone.css file I want to to delete though :)
[03:37] <mpt> I've already deleted it
[03:37] <mattl_> okay, does it run on Zope?
[03:37] <SteveA> mattl_: but it isn't based on plone at all.  the source code isn't available right now.
[03:37] <kiko> Kinnison, handlers.py, right before fuck in ear
[03:37] <SteveA> mattl_: it uses Zope 3, sqlobject, twisted and a bunch of other stuff.
[03:38] <kiko> Kinnison, and /don't/ touch that file, just send me diffs if you like
[03:38] <kiko> SteveA, not to mention string and wire
[03:38] <mattl_> SteveA: okay, thanks. is there any plan for making it available at some point?
[03:38] <SteveA> kiko: i wasn't going to mention the string and wire.
[03:38] <mattl_> what colour is the string?
[03:38] <sabdfl> chewing gum & duct tape
[03:38] <Keybuk> SteveA: meh, that involves "baz diff" taking less than a metric week
[03:38] <Kinnison> kiko: urgh that's so totally shit
[03:39] <Kinnison> kiko: yeah, just fix them to be LOW/MEDIUM/HIGH/EMERGENCY
[03:39] <kiko> Kinnison, you should focus more on code and less on cursing PEP-8 perhaps?
[03:39] <SteveA> mattl_: management have said that it will be made open eventually.  but there is no definite plan.  we're making some parts of it open as we go.
[03:39] <kiko> sure
[03:39] <SteveA> Keybuk: don't see why.  paste the two changed lines into the paste thing, along with a paragraph about the deletions.
[03:39] <Kinnison> kiko: I'm certain I wouldn't have done that
[03:39] <kiko> sabdfl, it's vegemite flavor 
[03:39] <Kinnison> kiko: but if I did then yeah, perhaps I should
[03:39] <kiko> :)
[03:39] <mattl_> SteveA: Thanks.
[03:39] <SteveA> there will come a time when waiting for bzr for launchpad will be quicker than waiting for baz diff.
[03:40] <Kinnison> mmmm mussels
[03:40] <mpt> though, mussel-flavored gum would be interesting
[03:40] <bradb> kiko: do you have a moment to wrap up this sortorder patch?
[03:40] <SteveA> kiko: so... when shall i plan to do some gina wrangling on staging with you?
[03:40] <kiko> bradb, sure
[03:40] <kiko> SteveA, in about 30 minutes
[03:40] <SteveA> stub: can you let me know when staging will be ready?
[03:41] <bradb> kiko: so, from reading your email...
[03:41] <kiko> hey Kinnison 
[03:41] <SteveA> kiko: fine.  just enough time for me to fix another bug or two.
[03:41] <bradb> kiko: to be sure, what do you mean exactly by wanting the backend "unified"? there's exactly one function you call to get the sort params.
[03:41] <Keybuk> https://chinstrap.ubuntu.com/~dsilvers/paste/filea6ekEM.html
[03:42] <kiko> bradb, you said "old-style" versus "new-style" sorting initially, and I wanted there to be only one backend, not two. is that fixed? I can look at the code anyway though.
[03:42] <bradb> kiko: Should I just resend the latest version of the patch? I guess it's more or less different enough to have another look, if you're interested. Maybe that'll clear things up.
[03:43] <SteveA> Keybuk: is this just a straight relocation to put sourcerer at the top level?
[03:44] <kiko> bradb, yeah
[03:44] <kiko> Kinnison:    "source": PackagePublishingPriority.EXTRA #Some binarypackages ended up
[03:44] <kiko>                                               #with priority source.
[03:44] <Keybuk> SteveA: Sourcerer's been in both places for the most part of the last year
[03:44] <bradb> kiko: ok, bazzing my way there now
[03:44] <kiko> any clue why they ended up with priority source?
[03:44] <Keybuk> the lib/canonical/sourcerer one is when lifeless randomly checked an ancient copy in
[03:44] <Keybuk> despite there being lib/sourcerer pulled in by the config
[03:44] <elmo> kiko: it's a dak thing
[03:44] <Kinnison> kiko: bug in dak
[03:45] <Keybuk> the change kills the ancient one, and just adjusts the code to use the new Version parser (which has bug fixes too)
[03:45] <SteveA> Keybuk: i see.  the 'canonical.sourcerer' makes it clearer that it is canonical code.  but anyway...  so the code was actually using both versions, as is apparent from the imports you changed.
[03:45] <Keybuk> SteveA: yup, it was confused code
[03:45] <stub> Bah. fti.py is starting to take silly amounts of time to run.
[03:45] <SteveA> Keybuk: okay, r=me
[03:45] <Keybuk> if we like canonical.sourcerer -- we should make a change to the sourcerer branch in rocketfuel and stuff
[03:46] <SteveA> Keybuk: i think we have better things to do that relocate sourcerer right now.
[03:46] <Keybuk> indeed
[03:48] <sabdfl> mpt: spiv made a good review comment
[03:49] <sabdfl> in all the listings i tend to use a smaller font for the summary-below-the-main-bit
[03:49] <sabdfl> can we have a class "lesser"?
[03:49] <sabdfl> also, i often have indents
[03:49] <sabdfl> style="font-size: 80%; margin-left: 30px;" is common in these
[03:50] <sabdfl> class="lesser indented" would be nicer, or something similar
[03:50] <sabdfl> thoughts?
[03:50] <ddaa> 30px?
[03:50] <ddaa> 3em, good, 30px bad
[03:52] <mpt> sabdfl: class="lesser" already exists and already does what you want :-)
[03:52] <Keybuk> yes, sabdfl: don't ever use "px" in stylesheets, m'kay
[03:52] <sabdfl> mpt: ok. and the indent?
[03:52] <Keybuk> use "em" or "%" so those of us which bad eye sight can scale the pages up
[03:53] <ddaa> keybuk: border-width=1px is fine, because explorer has stupid values for logical line widths.
[03:53] <mpt> sabdfl: Wanting an indent is a side-effect of not using the listing tables consistently with the rest of LP
[03:53] <Keybuk> yeah, that kind of thing, and for absolute graphics alignment and stuff
[03:53] <mpt> sabdfl: which I'll show you once I've got this branch landed
[03:53] <sabdfl> ok. i'm sceptical of the tables based approach though
[03:54] <ddaa> call the doctors, Keybuk and I agree on something!
[03:56] <stub> SteveA: staging is awaiting your pleasure
[03:56] <SteveA> thanks stub.  i've still got that US$50...
[03:59] <SteveA> carlos: ping
[04:07] <bradb> kiko: Got distracted in #canonical there for a second, but I just sent the current version of the patch your way.
[04:08] <kiko> ok
[04:14] <SteveA> kiko: it's 30 mins... how's gina looking?
[04:16] <kiko> SteveA, I'm still following the code. Kinnison, sabdfl, elmo: can you tell me which table we're duplicating data in? build or binarypackagerelease?
[04:17] <sabdfl> kiko: could be both
[04:17] <kiko> sabdfl, "could"? I need to know
[04:17] <sabdfl> sorry, i had an imported database with examples on my laptop, but i dumped it to do normal testing
[04:18] <sabdfl> does Kinnison's dogfood have the results of a gina run?
[04:19] <sabdfl> Kinnison: ping
[04:19] <stub> IIRC it was build
[04:19] <kiko> okay
[04:20] <kiko> sabdfl, could it be a matter of altering handlers.py:*._getBinary() to stop checking for the distrorelease?
[04:20] <stub> BinaryPackageRelease was the insert that was failing due to violated constraints
[04:21] <kiko> that's fixed
[04:21] <sabdfl> kiko: i dont know gina at all
[04:21] <kiko> ok
[04:21] <sabdfl> that ho never felt my lovin touch
[04:21] <stub> Night ppls
[04:22] <Kinnison> sabdfl: yes?
[04:22] <stub> SteveA will be taking Gina for a spin
[04:22] <kiko> Kinnison, gimme a hand
[04:22] <Kinnison> hmm?
[04:22] <sabdfl> Kinnison: kiko needs your help to debug gina
[04:22] <sabdfl> specifically, kniwing whats being duplicated
[04:23] <sabdfl> night stub
[04:23] <bradb> kiko: How does the patch look?
[04:24] <bradb> kiko: re: the priority icon, note: 1. it was mpt's idea, 2. whether it's a good thing or a bad thing, I don't know. Personally, I hate icons, unless they're extremely obvious and standard (like pencil == edit, red X == delete, etc.), but I did add title attribs to them in this patch, and also linkified them, which makes them slightly more usable.
[04:25] <kiko> bradb, you need to wait for gina, but I'll give you a hint: get_sortorder_from_request() needs to loose all reference to "new-style" and "old-style" -- IOW, you should share the form variables in the request.
[04:25] <kiko> +    if request.get("orderby"):
[04:25] <kiko> +        # The old-style sorting, which is based on clickable column headers.
[04:25] <kiko> +        return request.get("orderby").split(",")
[04:25] <kiko> +    elif request.get("sort") and request.get("sortorder"):
[04:25] <kiko> +        # The new-style sorting, with a list format and dropdown sorting
[04:25] <kiko> +        # widget.
[04:25] <kiko> +        return request.get("sortorder").split(",")
[04:25] <kiko> just call everything orderby or sortorder
[04:25] <kiko> it doesn't need to be special-cased to new versus old
[04:26] <bradb> kiko: ok, question for you (i don't know the "right" answer):
[04:26] <bradb> let's say you've got the search form and you type in the search "crashes"
[04:27] <bradb> you see a header "1 -> 15 of 15 results"
[04:27] <bradb> underneath that, the sort widget
[04:27] <bradb> you change the sort order, to be ordered by priority (let's say that wasn't the default) and click "Sort Again"
[04:28] <bradb> rerenders
[04:28] <bradb> then you change it *again*, but don't click "Sort Again"
[04:28] <kiko> ok
[04:28] <bradb> then you do a search
[04:28] <bradb> do you expect the current sort order to persist, or the new one that you chose but, for whatever reason, didn't submit?
[04:28] <kiko> the new one
[04:28] <kiko> what sort of question is that?
[04:28] <SteveA> bradb, kiko: we need to focus on gina
[04:28] <kiko> why is there a separate "sort again" button?
[04:29] <bradb> kiko: what sort of question is that?
[04:29] <SteveA> i have 1.5 hours or so to go until i'll very much want to go home for the day
[04:29] <bradb> kiko: what did you have in mind? evil jumping screen js submits? :P
[04:30] <kiko> bradb, can you put it up for me to look at in like 1h?
[04:30] <bradb> ok
[04:30] <kiko> SteveA, I'm already going over it with Kinnison, ok?
[04:30] <SteveA> okay, great
[04:33] <SteveA> jamesh: are you available for a quick review?
[04:35] <SteveA> Kinnison: do you have stuff for me to review?
[04:37] <Kinnison> SteveA: not yet, will have in about 20 minutes
[04:37] <SteveA> okay cool
[04:37] <SteveA> i just finished my bugfixing work
[04:37] <Kinnison> SteveA: I'm just doing some work tidying my security.cfg
[04:40] <Kinnison> IIRC someone made it possible to alter what db user was used for a doctest
[04:40] <Kinnison> what does one have to do to do that?
[04:40] <SteveA> see poimport.txt
[04:41] <SteveA> see also launchpad/ftests/test_system_documentation.py
[04:41] <Kinnison> ta
[04:41] <SteveA> in fact
[04:41] <SteveA> the latter is better
[04:41] <SteveA> the former looks wrong
[04:42] <Kinnison> Erm, so that has poExportSetUp
[04:42] <Kinnison> but what do I do?
[04:44] <SteveA> copy what it does
[04:44] <SteveA> but for what you need
[04:44] <Kinnison> right, so I just create uploaderSetUp and uploaderTearDown ?
[04:44] <SteveA> so, add a set up and tear down function 
[04:44] <SteveA> yes
[04:44] <Kinnison> okies
[04:44] <SteveA> and register it in special = {...}
[04:45] <kiko> Keybuk?
[04:45] <Keybuk> kiko: ?
[04:45] <kiko> U-TURNin?
[04:46] <Keybuk> #canonical
[04:52] <jamesh> SteveA: sure.
[04:52] <SteveA> jamesh: actually, i reviewed the diff, and decided it was [trivial] .  but there is one part i'd like you to review even so
[04:52] <SteveA> that is a change to the query timeout stuff
[04:53] <jamesh> okay
[04:53] <SteveA> https://chinstrap.ubuntu.com/~dsilvers/paste/fileaFiLk0.html
[04:53] <SteveA> it concerns me because it is fragile to changes in the db connection code / database stuff
[04:54] <SteveA> i was thinking of lower()ing and normalizing whitespace before comparing the string, to make it more robust.
[04:55] <SteveA> maybe you can think of how to write a test for it?  maybe making a direct DB connection, using the adapter, and issuing a query that makes that ProgrammingError occur, and then issuing another bogus query to check that only that error is handled with a RequestQueryTimedOut exception.
[04:56] <jamesh> SteveA: I had some ideas about how to test for it
[04:56] <jamesh> SteveA: we already load some custom Python functions into postgres via plpython
[04:56] <SteveA> jamesh: okay, well this code is on its way to pqm.  but i'd appreciate if you can write a decent test for it :-)
[04:56] <SteveA> i'm afraid i need to go test gina with kiko shortly
[04:56] <jamesh> SteveA: so I wonder if we could add a custom function that just does a time.sleep() of a given length
[04:57] <SteveA> i see
[04:57] <SteveA> also, i guess my error class should be RequestStatementTimedOut...
[04:58] <SteveA> yes, i think that may be the only way to make it deterministic
[05:00] <SteveA> i think a good test is a much better idea than trying to make the string-matching robust
[05:01] <SteveA> tthe consequences of not matching properly aren't too bad though -- just a SystemError rather than a nice "request timeout" error
[05:09] <kiko> SteveA, sabdfl: ping
[05:09] <SteveA> yep
[05:09] <sabdfl> kiko: pong
[05:10] <kiko> so the issue seems to be that we are looking for binarypackagereleases that were built in the current distribution release
[05:10] <sabdfl> SteveA: ok, mirrored, so mark.shuttleworth@canonical.com/launchpad--newpackageclasses--0--patch-36 passes all tests with the rest of what's in the production branches, should merge cleanly, and has all reviewer feedback
[05:10] <sabdfl> kiko: don't do that
[05:10] <carlos> SteveA, pong
[05:11] <kiko> I've fixed this by changing the query to look for releases with that version and architecture in the current /distribution/
[05:11] <SteveA> hi carlos.  i had a question about some potemplate stuff, but then found out what i needed for now.
[05:11] <carlos> ok
[05:11] <carlos> SteveA, feel free to ask anything you need
[05:12] <kiko> sabdfl, that would avoid creating the release and the build for every distribution release after the one its source was uploaded to
[05:12] <kiko> sabdfl, don't do.. what?
[05:13] <kiko> bradb, https://launchpad.net/malone/bugs/3280
[05:13] <Ubugtu> Malone bug #3280: bug listing uses images to show priority, they lack titles Fix req. for: malone (upstream), Severity: Normal, Assigned to: Nobody, Status: New http://launchpad.net/malone/bugs/3280
[05:13] <bradb> oh, cool, thanks
[05:15] <SteveA> sabdfl: i'll try merging mark.shuttleworth@canonical.com/launchpad--newpackageclasses--0--patch-36 into staging
[05:15] <kiko> SteveA, sabdfl doesn't want to talk to me 
[05:16] <SteveA> sabdfl: are there any important database changes in there (other than removing some unused tables) ?
[05:17] <SteveA> sabdfl: i'm not sure how to apply such changes to staging
[05:17] <elmo> he's on the phone
[05:20] <SteveA> elmo: what's the best way to get baz to get archives from chinstrap to asuka?
[05:20] <SteveA> use sftp as my user and use my password?
[05:22] <elmo> SteveA: in the short term, I guess yes
[05:22] <SteveA> ok
[05:23] <kiko> SteveA, I believe I have a patch for you
[05:24] <SteveA> kiko: cool.  but we still don't know what mark's objection was
[05:24] <kiko> SteveA, note that he didn't even hear my solution before saying it
[05:24] <kiko> elmo, if he ever gets off the phone, we're waiting
[05:28] <salgado> SteveA, do you have anything against the fix stub suggested in https://launchpad.net/products/launchpad/+bug/1612? (if not I'm going to merge it)
[05:28] <Ubugtu> Malone bug #1612: Logging in from front page doesn't return to front page Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Nobody, Status: New http://launchpad.net/malone/bugs/1612
[05:28] <kiko> SPEAK NOW OR SILENT FOREVER
[05:31] <SteveA> okay... on staging i merged into rocketfuel@canonical.com/launchpad--devel--0--patch-2664 the branch mark.shuttleworth@canonical.com/launchpad--newpackageclasses--0--patch-36 by star-merge
[05:31] <SteveA> there were eight conflicts
[05:31] <SteveA> C   lib/canonical/launchpad/browser/build.py
[05:31] <SteveA> C   lib/canonical/launchpad/browser/distrorelease.py
[05:31] <SteveA> C   lib/canonical/launchpad/browser/sourcepackage.py
[05:31] <SteveA> C   lib/canonical/launchpad/database/distrorelease.py
[05:31] <SteveA> C   lib/canonical/launchpad/interfaces/sourcepackagerelease.py
[05:31] <SteveA> C   lib/canonical/launchpad/zcml/configure.zcml
[05:31] <SteveA> CA  lib/canonical/launchpad/zcml/.arch-ids/binarypackagepublishing.zcml.id
[05:31] <SteveA> CA  lib/canonical/launchpad/zcml/binarypackagepublishing.zcml
[05:31] <SteveA> 
[05:32] <SteveA> salgado: i think this should wait until the 'vary' header fix
[05:32] <kiko> pain
[05:32] <SteveA> so, jbailey reported a problem logging in today
[05:32] <salgado> SteveA, what's the 'vary' header fix?
[05:32] <SteveA> and going to that page was a part of the workaround
[05:33] <SteveA> if you're connecting to launchpad via a cache, or have your browser set up in a particular way, then logging in can appear not to work
[05:33] <SteveA> because the cache doesn't know that the page you were on has now changed
[05:33] <SteveA> the solution involves setting the Vary header to say about cookies, and making the cookie vary under certain circumstances
[05:36] <SteveA> kiko: i'm confused by these conflicts
[05:36] <SteveA> i thought mark said he'd synced up with RF
[05:36] <SteveA> ah... maybe just with production.
[05:37] <Kinnison> is PQM stuck?
[05:37] <SteveA> Kinnison: possibly.  celso's job has been there for a while
[05:37] <salgado> Kinnison, I don't think so
[05:38] <SteveA> nope... things are happening
[05:38] <Kinnison> naah, looks like chinstrap is busy
[05:38] <Kinnison> that's all
[05:38] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=SteveA]  landing Slave Manual Mode flag, it allows us to observe/select build job for a suspicious slave w/o stressing the whole environment. (patch-2665: celso.providelo@canonical.com)
[05:39] <kiko> lol
[05:40] <SteveA> kiko: i'm undoing the merge on staging
[05:41] <SteveA> where's this patch i can look at?
[05:43] <kiko> SteveA, I can commit it or I can send it to you, what's preferred?
[05:43] <kiko> it contains some changes I wouldn't call conservative
[05:44] <SteveA> can i take a look at it in Kinnison's pastebin?
[05:44] <kiko> no
[05:44] <kiko> I can send you a full diff though
[05:44] <kiko> I can also isolate parts of it for you
[05:44] <SteveA> sure
[05:44] <kiko> let me baz diffify and commit then
[05:45] <SteveA> ok
[05:45] <kiko> I suggest, SteveA, using baz replay
[05:45] <sabdfl> hi guys
[05:45] <kiko> it's the sab!
[05:46] <kiko> man am I hungry
[05:46] <SteveA> sabdfl: rocketfuel@canonical.com/launchpad--devel--0--patch-2664 is what i have on staging
[05:46] <kiko> DST makes me hungry
[05:46] <SteveA> and i get 8 conflicts merging your branch into that
[05:47] <sabdfl> kiko: that will work for now. but make sure you are using real files, not names + versions (because the ubuntu one and the debian one will be different, and gina will in due course be importing both).
[05:47] <sabdfl> SteveA: python ./upgrade.py -d datebasename i think
[05:47] <kiko> Kinnison, do you grasp what sabdfl is telling you?
[05:47] <kiko> or rather, telling me? :)
[05:47] <sabdfl> SteveA: hmm... why don't you have --production-1.37 on staging?
[05:48] <SteveA> sabdfl: i guess stu put rocketfuel@canonical.com/launchpad--devel--0--patch-2664 on there because we're going to be rolling out HEAD
[05:48] <sabdfl> we are? shit
[05:48] <Kinnison> kiko: he's being confused because he doesn't realise you're restricting name+version to "within this distro"
[05:48] <SteveA> we discussed that this morning
[05:49] <sabdfl> Kinnison: nonetheless, gina should be using bit-for-bit comparisons, via the librarian, not relying on name guessing magic
[05:49] <sabdfl> you have no idea how broken most of soyuz was until this weekend
[05:49] <sabdfl> its still borken, i'm sure, but in new and exciting ways
[05:49] <Keybuk> random question ... is pqm ok?
[05:49] <Kinnison> sabdfl: if name+version+within-distro isn't enough then we have larger and more scary problems
[05:49] <Keybuk> it looks stuck
[05:49] <Kinnison> Keybuk: yeah, chinstrap was busy busy busy
[05:49] <sabdfl> most of the old brokenness was because the people who wrote it, and they are the same people behind gina, don't have a deep understanding of the data model and archive structures
[05:50] <sabdfl> here's what gina needs to be doing
[05:50] <Kinnison> sabdfl: the archive mandates that name+version+distro is unique
[05:50] <Kinnison> sabdfl: full stop
[05:50] <sabdfl> Kinnison: yes
[05:50] <sabdfl> but
[05:50] <sabdfl> debian + foo + 2.1-1 
[05:50] <sabdfl> ubuntu + foo + 2.1-1
[05:50] <sabdfl> could be EXACTLY THE SAME SOURCEPACKAGE RELEASE
[05:51] <sabdfl> and gina will import it as two entries if it makes that kind of assumption
[05:51] <Keybuk> not just that
[05:51] <sabdfl> that's exactly the kind of assumptin that created this problem in the first place
[05:51] <Keybuk> ubuntu warty, hoary and breezy foo 2.1-1 are the same
[05:51] <Kinnison> sabdfl: Oh Christ
[05:51] <Keybuk> and only appear once in the pool
[05:51] <Kinnison> Keybuk: yes, don't confuse the issue please
[05:51] <Keybuk> but gina tries to import it treetimes
[05:51] <Kinnison> Keybuk: We're already on that, please don't confuse the issue now
[05:52] <Kinnison> sabdfl: I thought we were simply trying to get *Ubuntu* imported cleanly for now
[05:52] <Kinnison> sabdfl: If you want gina fixed to work for when we import Debian *too* then we're gonna be here for longer
[05:52] <sabdfl> Kinnison: if there are fixes to be made, i'm trying to argue for basing the import loging on librarian file based intelligence, not on "hey this is how the thing is supposed to work" handwaving
[05:52] <sabdfl> now just listen please
[05:52] <kiko> Keybuk, I've fixed that now.
[05:52] <kiko> however
[05:52] <kiko> sabdfl is right
[05:53] <kiko> I knew this was fragile between distributions
[05:53] <sabdfl> when importing a source package release, first step is to find out if the librarian knows about that file, and then if there is a SourcePackageRelease for it
[05:53] <sabdfl> next
[05:53] <sabdfl> when importing a BinaryPackageRelease, do the same thing
[05:53] <sabdfl> now
[05:53] <sabdfl> binaries come in sets
[05:53] <sabdfl> we call that a Build
[05:54] <sabdfl> so do an assertion check that the previously-imported-binary came from a Build that also has all the binaries you currently want to import
[05:54] <sabdfl> if not HOLLER
[05:54] <sabdfl> i can't think of a way to deal with that, it should never happen, so HOLLER
[05:54] <sabdfl> moving on
[05:54] <kiko> sabdfl, that fix is more intrusive than what I have now
[05:55] <sabdfl> now you know that the binary you are wanting to import is already imported, and the Build for it is good. So use that previously-imported one
[05:55] <sabdfl> that is all
[05:55] <sabdfl> in both cases, source and binary, its important to start from the librarian ID, which we KNOW is checked bit for bit
[05:56] <sabdfl> the extra trick that is required on the binary front is to deal with a case where for some reason something screwed up, and the previous Build was created without all the binaries, and now we are importing and are seeing extra binaries. i don't know how to deal with that case.
[05:56] <kiko> yeah, that's tricky but reasonable too
[05:56] <sabdfl> i don't know for example if we ever throw binaries away
[05:56] <kiko> sabdfl, do you want to review the emergency fix? I'm not suggesting to do it your way today
[05:56] <kiko> (because, if anything, SteveA needs to go to bed at some point)
[05:57] <sabdfl> i don't think we should be going live with stuff constructed in an emergency fix
[05:57] <sabdfl> i thought this was being worked on over the weekend
[05:57] <sabdfl> so lets delay the rollout a day
[05:57] <sabdfl> get it done properly
[05:57] <sabdfl> test it
[05:57] <sabdfl> and go
[05:57] <sabdfl> SteveA: do you want me to merge launchpad--0--patch-2664?
[05:58] <sabdfl> or to land my stuff on HEAD?
[05:58] <kiko> I am not the best candidate for this fix because my knowledge of the librarian isn't top-notch but I can do it -- sabdfl?
[05:58] <SteveA> kiko: do it, and give spiv to review
[05:58] <sabdfl> kiko: hmm... basically, i think there is an api to say to the librarian "here is a file"
[05:58] <sabdfl> it then gives you a unique pointer to that
[05:58] <SteveA> sabdfl: HEAD, because we'll be rolling out some other changes that haven't quite landed yet
[05:59] <sabdfl> you can check to see if there is already a source pacakge release that points at that pointer
[05:59] <sabdfl> etc
[05:59] <sabdfl> SteveA: ok
[05:59] <kiko> yeah, I assume that, but I'll need to cargo-cult
[05:59] <SteveA> the conflicts seem fairly minor
[05:59] <SteveA> but i don't feel confident dealing with them
[05:59] <sabdfl> lord knows what baz will do with this...
[06:00] <SteveA> Kinnison: got something for me to review?
[06:00] <sabdfl> SteveA: do you really think its a good idea to roll out HEAD? I thought stub meant that he was updating the target to TODAY's head, not HEAD-at-the-instant-he-does-it
[06:00] <sabdfl> sorry guys
[06:00] <sabdfl> another
[06:00] <sabdfl> call
[06:01] <Kinnison> SteveA: Just committing now
[06:01] <kiko> sheesh
[06:02] <SteveA> the HEAD stub will be rolling out is meant to include some of today's landings.  for example, i have some important bugfixes at #6 in the pqm queue
[06:02] <SteveA> at #7, we have a trivial from salgado to make voting work properly
[06:02] <SteveA> these are things that would be cherrypicked
[06:03] <SteveA> there have been no big landings that aren't bugfixes recently
[06:03] <SteveA> so, i'm not saying 'keep your branch up to date with HEAD until it is rolled out'
[06:03] <SteveA> i am saying that the patch level of RF that will be chosen will be now + a few hours still.
[06:04] <SteveA> right after salgado's voting fixes land is a good bet
[06:04] <kiko> SteveA, I have some fixes from gneuman for voting too, so..
[06:04] <Kinnison> to do with queue stuff
[06:04] <Kinnison> stub has blessed the db bits
[06:05] <SteveA> kiko: we need to be a bit careful.  if someone lands a big new feature / refactor, then we probably don't want to roll that out
[06:05] <kiko> it's not big
[06:05] <kiko> adds tests and fixes traversals
[06:05] <SteveA> sure.  i mean, don't let some big thing land between now and your landing that
[06:05] <SteveA> otherwise, it will have to be a cherrypick later
[06:06] <kiko> sure thing
[06:07] <Kinnison> SteveA: I've added a branch to your queue
[06:07] <Kinnison> SteveA: it's mirroring now
[06:07] <SteveA> Kinnison: do you happen to have a diff for it to hand?
[06:07] <Kinnison> SteveA: mirrored
[06:07] <Kinnison> SteveA: No, but I guess I can make one
[06:07] <Kinnison> SteveA: but I can't easily filter it
[06:07] <SteveA> don't worry about that
[06:08] <SteveA> but a diff would help me a lot
[06:08] <Kinnison> Can we still not trigger a run of jamesh's scripts?
[06:08] <SteveA> i don't know how to do so
[06:08] <SteveA> perhaps only james can
[06:08] <Kinnison> right
[06:08] <Kinnison> do we know when they run?
[06:09] <SteveA> i don't.  looking in jamesh's crontab would probably say...
[06:09] <Kinnison> elmo: Can you tell us when jamesh's pending-reviews stuff will next update?
[06:10] <elmo> 42 2,6,10,14,18,22 * * *  $HOME/pending-reviews/update-pending-reviews.sh
[06:10] <kiko> sabdfl, note that I wasn't suggesting rolling out without testing -- that's what SteveA and I were going to do
[06:11] <kiko> sabdfl, the difference between my fix and your suggestion is coping with multiple distributions properly, which could be done later because we are only importing a single distro today.
[06:11] <Kinnison> so 18:42
[06:12] <SteveA> Kinnison: i need to be getting home soon.  i can, however, do the review from my laptop at home later on, if there's a diff waiting on the pending-reviews page
[06:13] <Kinnison> SteveA: well, I imagine it'll make a diff later
[06:13] <Kinnison> SteveA: so go home and look later when you've rested a bit
[06:13] <Kinnison> SteveA: btw, the waitress who most often gave me coffee at the KK over the past two weeks... was called Aiste
[06:14] <SteveA> cool.  say "labas vakaras" to her. (if it is the evening)
[06:14] <Kinnison> well, I'm at home again now *phew*
[06:14] <Kinnison> two weeks in london was too much
[06:15] <elmo> london's a lot more dealable when you're living, and not in a hotel
[06:16] <Kinnison> aye
[06:16] <Kinnison> I managed to live there for a year
[06:18] <bradb_> Does it make sense for zopeless code to have its MAIL FROM be our standard bounce address?
[06:19] <Kinnison> Dunno, but I'm gonna need to know how to send mail zopeless fairly soon
[06:20] <SteveA> giveittomenow
[06:20] <bradb> Kinnison: I'll make it be our standard bounce address (I can't see that being any worse than using whatever email is being used in the From header) and if anybody whines, I can change it.
[06:21] <sabdfl> lifeless_, ddaa: help
[06:21] <sabdfl> slinky% baz status --conflicted                     ~/projects/ubuntu/launchpad
[06:21] <sabdfl>  C  database/schema/.arch-ids/patch-25-40-0.sql.id
[06:21] <sabdfl>  C  database/schema/patch-25-40-0.sql
[06:21] <sabdfl> i did a --star-merge where i should not have
[06:21] <kiko> sabdfl, undo is not a possibility?
[06:21] <sabdfl> kiko: yes
[06:21] <Kinnison> bradb: okay, can you then mail me with info on how to send mail when zopeless?
[06:21] <ddaa> undo barfs because of the duplicate ids
[06:21] <sabdfl> those files have just conflicted with themselves
[06:22] <kiko> sabdfl, just delete the id files?
[06:22] <sabdfl> can i just move the .orig's back into position
[06:22] <kiko> and baz undo
[06:22] <bradb> Kinnison: I can email you with how I /think/ it's done, if you want. :) I only know that this bit I'm looking at is zopeless-specific, because of the "if isZopeless():" block. :)
[06:23] <ddaa> do as kiko says, but remove anything that looks vaguely suspect in the database/schema/.arch-ids
[06:23] <kiko> that's how I do it
[06:23] <sabdfl> ddaa: here's the thing. there is NO file called database/schema/.arch-ids/patch-25-40-0.sql.id or  database/schema/patch-25-40-0.sql
[06:23] <sabdfl> each of those just has a .orig and a .rej
[06:24] <ddaa> funny
[06:24] <kiko> odd indeed
[06:24] <kiko> and baz undo still fails, sabdfl?
[06:24] <ddaa> also "baz resolved --all"
[06:25] <ddaa> sabdfl: just remove all the weird stuff and "baz resolved --all", then baz undo to get right back you were before.
[06:25] <ddaa> it's not even funny anymore to see it demonstrated again and a again that conflating metadata and user data is  a bad idea...
[06:26] <bradb> Emacs++ # multi-line search-and-replace that Just Works
[06:26] <sabdfl> ddaa: and then retry the merge? not with --star-merge? i'm *afraid* at what might conflict then
[06:26] <Kinnison> bradb: If you can mail me that then yes, that'll be good
[06:26] <bradb> Kinnison: sure, I'll do that shortly
[06:26] <Kinnison> bradb: star
[06:27] <ddaa> sabdfl: maybe if you have a patch-25-40-0.sql already, rename it before
[06:27] <ddaa> your problem sounds like a conflicting file addition
[06:28] <ddaa> but without some lengthy diagnostic, I cannot tell more...
[06:28] <sabdfl> ah, bugger it, i'll just rebuild the working directory
[06:29] <kiko> sabdfl, I'm mailing you a diff with what I have right now
[06:29] <kiko> hoping you understand the caveats
[06:29] <sabdfl> kiko: i've never looked at gina, not sure i'll understand whatever you send me
[06:29] <kiko> I suspect you will
[06:32] <looksaus> I had posted a bounty on the ubuntu wiki a few months ago
[06:32] <looksaus> I see aigarius has migrated it to Launchpad, which is fine on its own
[06:33] <looksaus> but it's lacking some info, and formulated somewhat suboptimal...
[06:33] <looksaus> only I have no rights to edit it
[06:33] <looksaus> even if my name is mentioned inside
[06:33] <looksaus> can anyone help?
[06:34] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: r=salgado Seriously whack the team membership edit view, improving the way validation  is done, preserving form values when submitting, and reducing the wordiness everywhere. De-indent some main page silliness. Simplify the template by using macros. Add a test that checks the form's behaviour (patch-2666: christian.reis@canonical.com)
[06:36] <looksaus> kiko, that was a fast reply to #3285, thx!
[06:37] <SteveA> kiko, sabdfl: i'm going out for some food.  i'll be back using the laptop at home a bit later to do Kinnison's code review and any staging work that's needed then.
[06:37] <SteveA> where "a bit later" is in a few hours.
[06:39] <sabdfl> oh baz baz baz
[06:40] <sabdfl> looksaus: hmm... file a bug on the bounty system, assign to me
[06:40] <sabdfl> there should be a "issuer",  who can be a person, who then gets to edit it
[06:40] <looksaus> k, will do
[06:40] <Nafallo> what's decided now? dapper opens ~wednesday? :-)
[06:42] <sabdfl> Nafallo: looks that way
[06:42] <Nafallo> I've read through some of the specs yesterday night. looks awesome :-).
[07:04] <ddaa> SteveA: is there any way to make creating POST pagetests easy?
[07:04] <sabdfl> ah.
[07:04] <sabdfl> stub gave out two patch-40's
[07:06] <ddaa> that's the time where the git fans giggle
[07:13] <kiko> ddaa, makepagetest.py?
[07:13] <ddaa> kiko: thanks
[07:14] <ddaa> in case it's not absolutely clear yet, I'm entirely clueless about launchpad devel practises
[07:15] <kiko> ddaa, if I can help you in whatever way, please ask
[07:15] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial]  Create empty override and file lists for apt-ftparchive (patch-2667: daniel.silverstone@canonical.com)
[07:15] <kiko> ddaa, remember that functional doctests are preferred to pagetests
[07:15] <ddaa> kiko: you know that learning machine in Matrix... could come in handy
[07:16] <ddaa> kiko: where can I learn more about what you just said?
[07:17] <kiko> ddaa, we have doctests and pagetests, right?
[07:18] <ddaa> Tell me more about doctests.
[07:18] <kiko> pagetests operate almost exclusively on the http level
[07:18] <kiko> doctests import modules and try doing things to them
[07:18] <kiko> take a look at doc/person.txt
[07:19] <kiko> and pagetests/foaf/40-*.py
[07:19] <ddaa> you mean, that's the only difference? Just do normal doctests in a patgetest .txt file?
[07:19] <kiko> yes, that's the only difference
[07:19] <kiko> (what you described is a pagetest)
[07:19] <ddaa> I guess it's going to take some practice to know where to draw the line.
[07:19] <kiko> pagetests test web pages using an http interface
[07:20] <kiko> they are what tests the templates, for instance
[07:20] <ddaa> yup, of course
[07:20] <kiko> nothing else can test templates (beyond trivial syntax verification)
[07:20] <ddaa> ack that much, that were you test that some specific bit is there. For example that there's a "add branch" link on the product index page.
[07:21] <ddaa> or that a branch is displaying in the branch listing for a product
[07:21] <kiko> exactly
[07:22] <kiko> you could do the latter using a functional doctest normally, though
[07:22] <ddaa> what's not clear to me is where to draw the line for "I do not want to test this bit of code using a pagetest" when said bit of code is used solely for generating pages
[07:22] <kiko> because that listing comes from some method in the view or content class, right ddaa?
[07:22] <kiko> ddaa, maybe we can discuss a concrete situation?
[07:23] <ddaa> kiko: well, right, but on still needs to test that the template for displaying the listing is correct
[07:23] <ddaa> Right now I'm reading docs.
[07:23] <looksaus> where's the right place and time to discuss the bounty system in Launchpad (apart from Montral in the beginning of november :)
[07:24] <ddaa> I'm thinking about getting rid of the annoying-to-maintain sampledata I introduced and just fill the db from the tests
[07:24] <looksaus> URL maybe?
[07:24] <kiko> ddaa, I guess I normally assume that trivial page templates don't need to be tested.
[07:24] <ddaa> because, asides from being annoying to maintain, it's also annoyingly nonsensical
[07:24] <kiko> ddaa, but they should be pagetested using xx-notfound-traversals
[07:24] <kiko> possibly the right thing to do, yes
[07:24] <kiko> but sampledata helps us web QA people
[07:25] <ddaa> :) tell me more about web QA
[07:25] <kiko> we click click click till it hurts
[07:25] <ddaa> with what?
[07:26] <kiko> so if there's nothing in the database we need to do extra clicking
[07:26] <ddaa> good, better test coverage
[07:26] <kiko> that has short and long-term effects, not all of which are beneficial :)
[07:26] <ddaa> hu... I guess that was the wrong answer :)
[07:26] <kiko> there's no such thing as a wrong answer
[07:26] <kiko> just bad context
[07:27] <kiko> remember that
[07:27] <ddaa> I'm thinking, maybe I could use the pagetest to create the sampledata
[07:27] <ddaa> ...
[07:27] <ddaa> nah... that's too circular...
[07:28] <kiko> well
[07:28] <kiko> you can create sampledata using the web, yes
[07:28] <kiko> just make newsampledata
[07:28] <kiko> in database/schema
[07:28] <ddaa> yup
[07:28] <ddaa> It's just that I need to put my sample data on different product and person than those the pagetests use
[07:28] <ddaa> Is there some policy about that?
[07:29] <ddaa> but then, this sampledata will end up bitrotten
[07:30] <ddaa> this starts to really feel like a no-win situation :(
[07:30] <kiko> that sounds fine
[07:30] <kiko> not the rotten bitdata
[07:31] <kiko> the other part
[07:31] <uws> Please look at the bottom of https://launchpad.net/people/uws/+karma   Where do these "Translation Suggestion Approved" numbers come from? I've never been involved in translations via launchpad/rosetta
[07:31] <Nafallo> uws: but your translations have been imported from upstream, right? :-)
[07:32] <uws> Nafallo: I do translate some GNOME modules, yes. But how does it find out it's me?  email?
[07:32] <uws> Another question:
[07:32] <Nafallo> uws: I don't know. I've just seen the same thing happens to Christian Ros ;-)
[07:33] <uws> if you're looking at a people/teamname page
[07:33] <uws> there's _no_ way to view which produccts this team manages
[07:34] <kiko> uws, known bug, filed
[07:36] <ddaa> kiko: so is there a way to make newsampledata out of a db modified by a pagetest?
[07:36] <kiko> no
[07:36] <kiko> that's a caveat
[07:36] <kiko> well
[07:36] <kiko> not directly
[07:36] <ddaa> so, there's a way :)
[07:37] <kiko> you could grep for the http() calls in the pagetests and issue them directly
[07:37] <kiko> the test machinery has one important caveat
[07:37] <kiko> it resets the database after every story or standalone test
[07:38] <ddaa> right, that's why I asked
[07:38] <ddaa> a standalone test is one that starts with xx-, right?
[07:38] <kiko> hmm
[07:38] <kiko> I guess :)
[07:40] <bradb> ddaa: pagetests/README.txt should help clarify
[07:40] <ddaa> kiko: so would it be right to have a doctest that contains the code used to create the sampledata, inactive, and a few functional tests to check the sampledata is indeed there?
[07:41] <kiko> ddaa, that's unusual -- you'd ned to check with SteveA 
[07:41] <ddaa> kiko: I have it before my eye, but it does not seem to say anything about standalone tests
[07:41] <ddaa> and when the db is reset
[07:42] <kiko> what I said is true though
[07:42] <ddaa> sure, it's just missing a definition for standalone :)
[07:43] <ddaa> bah... in the short term I'll just get rid of the sampledata, I'll look for a better solution later
[07:51] <uws> EHm
[07:51] <uws> https://launchpad.net/products/bazaar/+series/ng/+edit  I can edit this page
[07:51] <uws> Even though I'm NOT in any way connect to this project
[07:51] <kiko> talk to ddaa 
[07:52] <ddaa> uws: can you actually commit changes?
[07:53] <ddaa> I cannot test that, since I'm myself part of the owning team :)
[07:53] <uws> ddaa: I can
[07:53] <uws> I just changed the . to ...
[07:54] <ddaa> Arguably, that's bad...
[07:55] <ddaa> kiko: can you assign one of your minions to look into the issue?
[07:55] <kiko> ddaa, yes
[07:55] <ddaa> that should be a private bug, but I sort of gathered that private bugs were problematic now...
[07:55] <kiko> uws, is it too much to ask you to file a bug?
[07:55] <uws> ddaa: That's really bad
[07:56] <kiko> ddaa, problematic?
[07:56] <uws> Btw, what's the way to add a URL to the tarball of a release?
[07:56] <uws> editing a release doesn't work :(
[07:56] <ddaa> kiko: like they were not visible to those who need to see it...
[07:56] <kiko> ddaa, that was mostly a red herring
[07:58] <uws> ddaa: Okay, so you take care of this issue for now?
[08:22] <auth> Good evening
[08:23] <auth> I want to know who I should talk to about deleting an account in launchpad. I managed to create 2 accounts..
[08:24] <salgado> auth, you don't need to delete one. you can merge them
[08:25] <auth> ok, that seems nice. Where in the menus should I look for that option?
[08:25] <salgado> auth, login with the account you want to keep and then go to https://launchpad.net/people/+requestmerge
[08:26] <sabdfl> evening sports fans
[08:31] <auth> sabdfl: Thanks!
[08:31] <kiko> evening sabdfl 
[08:31] <auth> sorry, salgado: Thanks!
[08:32] <kiko> is there a way to do slices in TAL?
[08:32] <sabdfl> auth: i was about to say you're extremely, defiintely, welcome
[08:32] <sabdfl> kiko: nup
[08:32] <kiko> snarl snarl
[08:33] <sabdfl> be nice to do a bit more stuff in tal, like..
[08:33] <sabdfl> <tal:foo repeat="bar context/bars">
[08:33] <bradb> kiko: they should work in a python: expression, shouldn't they?
[08:33] <sabdfl>   <div tal:content="context/xxx.bar">
[08:33] <sabdfl> or something
[08:34] <sabdfl> bradb: python: in tal is a bad bad sign
[08:34] <sabdfl> alarm bells all over the show
[08:34] <kiko> it depends
[08:34] <bradb> not always, IMHO. practicality has been known to beat purity in many cases.
[08:41] <bradb> Who wrote sending-mail.txt?
[08:42] <bradb> I want to give Kinnison a good example of how to send mail in LP, Zopeless. AIUI, it's identical to Zopeful, but just wanted to be sure.
[08:43] <sabdfl> (19:35:13) bradb: not always, IMHO. practicality has been known to beat purity in many cases.
[08:43] <sabdfl> errr.. that's pretty funny, Mr IUpstreamEditableBugTask
[08:43] <sabdfl> ;-)
[08:44] <bradb> I knew that was crack when I wrote it. I just couldn't figure out a better way to do it at that point in team and wanted to get something working.
[08:44] <bradb> s/in team/in time/
[08:46] <BjornT> bradb: i wrote sending-mail.txt. and yes, sending mail in zopeless is exactly the same as in lp. the only thing is that if you use zopeless in tests, you can't test that the mail has been sent.
[08:47] <sabdfl> kiko: ok, i'll take a look at that gina patch now
[08:47] <kiko> thanks
[08:47] <sabdfl> is there a sane testing strategy now for whatever we come up with?
[08:48] <bradb> BjornT: presumably that's a bug (that zopeless email sending code can't be tested, if I understand you correctly)?
[08:49] <kiko> sabdfl, I think there is
[08:49] <kiko> sabdfl, privmsg?
[08:51] <BjornT> bradb: yeah, you could call it a bug. you can work around it, though. with the latest cronscript i wrote sending mail, i test the mail sending in a normal test environment, then i run the scripts and look for log statements that say that mails have been sent.
[08:52] <kiko> mpt, bradb already has a fix for that, are you trying to conflict with him?
[08:53] <bradb> kiko: has a fix for what?
[08:54] <kiko> bradb, look at your mail
[08:58] <bradb> Kinnison: so there you have it. doc/sending-mail.txt should contain everything you need to know about sending mail Zopeless (which appears to be identical to sending mail Zopefully)
[09:02] <ddaa> kiko: https://launchpad.net/products/launchpad/+bug/3292
[09:02] <Ubugtu> Error: I cannot access this bug
[09:04] <kiko> ddaa, I can't see it. can you add me and launchpad as a CC:
[09:04] <ddaa> yeah... that's the problem with private bugs
[09:05] <kiko> only if you forget to CC: people :)
[09:05] <ddaa> kiko: try again
[09:05] <bradb> kiko: BTW, have you had a chance to look at the sortorder patch again? I'm dying to merge this one. ;)
[09:06] <kiko> bradb, I told you my comments
[09:06] <kiko> did you address them?
[09:06] <kiko> thanks ddaa 
[09:06] <SteveA_> kiko: time for some asuka action?
[09:07] <SteveA_> yeah
[09:08] <bradb> kiko: Well, I could address them by simply deleting the code that looks for an "orderby" parameter, because the code that was using that should all be inactive now. That would mean that the table sorting would put the table code out of sync (i.e. break it), but unbreaking would mean actually making it visible again in the UI to begin with, which is a different patch not yet written.
[09:08] <bradb> "table code", i.e. the bi-directional column-click sorting
[09:08] <kiko> bradb, why don't you just use the same name instead of two different ones that do the same thing?
[09:09] <kiko> I am missing some part of the puzzle I am sue
[09:09] <kiko> sure
[09:09] <kiko> SteveA_, always!
[09:10] <bradb> kiko: because of the workflow that i outline in the earlier email. but you appear to say that that workflow sucks. *shrug*.
[09:10] <kiko> bradb, you said you were going to post a site for me...
[09:11] <bradb> kiko: sure, i can do that now.
[09:11] <bradb> that should make it clearer
[09:11] <kiko> thanks
[09:13] <kiko> elmo?
[09:14] <SteveA_> kiko: okay.  i'm reviewing a patch for Kinnison too.  do you have some gina stuff for me to merge into staging?
[09:16] <kiko> SteveA_, yepperulas
[09:16] <kiko> christian.reis@canonical.com--lozenge/launchpad--devel--0--patch-215
[09:16] <kiko> SteveA_, you may replay up to that patch
[09:17] <kiko> you can start with patch-206
[09:17] <bradb> kiko: http://69.70.209.33:8086/products/firefox/+bugs
[09:18] <bradb> Hm, I wonder if Ubugtu just hit me there
[09:18] <bradb> Either that or kiko's fast on the draw
[09:19] <SteveA_> kiko: replay up to that patch?
[09:19] <SteveA_> can't i just merge it in?
[09:20] <fabbione> hey guys
[09:20] <fabbione> how do we look?
[09:29] <sabdfl> is PQM borked?
[09:29] <kiko> SteveA_, sure you can
[09:29] <kiko> bradb, I conferred with mpt
[09:30] <kiko> bradb, we both agree the sort again button should go, and the optionmenu should be offered as an extra option under the search box, above the search button
[09:30] <kiko> bradb, both simplifying UI and backend code. whee!
[09:30] <sabdfl> SteveA: i'm ready to go, tests are all passing, and I even fixed the most embarrassing almost-bug
[09:30] <sabdfl> but... there are like NINE items in the pqm queue
[09:32] <SteveA_> Kinnison: reviewed!
[09:32] <bradb> kiko: how can it be under the search box while still being above the search button? are you suggesting some kind of three row layout?
[09:33] <SteveA_> sabdfl: pqm has been really slow today.  i think chinstrap is busy doing other stuff, and this slows pqm down.
[09:33] <SteveA_> maybe we need our own integration box...
[09:33] <SteveA_> kiko: okay, i'll try merging that new gina stuff
[09:34] <SteveA_> sabdfl: are you particularly keen to see your stuff running on staging before gina has run there?
[09:36] <sabdfl> SteveA_: no, i can't see my new pages without the results of a gina run
[09:36] <kiko> bradb, three rows, yes
[09:36] <bradb> ouch
[09:37] <sabdfl> an integration box would be good. elmo, can we have one please? dedicated box for pqm? little disk, but lots of ram and CPU?
[09:37] <bradb> kiko: ok, so if i change it to look like that, and change the parameter to be named "orderby", will you pre-approve this/
[09:37] <kiko> bradb, you'll notice the code gets a lot simpler!
[09:38] <SteveA_> sabdfl: i'll mail rt with that request
[09:39] <bradb> kiko: dude, it gets 2 lines of Python code simpler :)
[09:39] <sabdfl> ok SteveA_
[09:39] <sabdfl> with bzr, pqm will get much faster
[09:39] <bradb> kiko: which is more than made up for by the hopefully-works-everywhere .js addition you suggest :)
[09:39] <sabdfl> i think it spends quite a lot of time doing the build-config, the merge, and then the commit
[09:40] <SteveA_> yes, it does
[09:40] <kiko> bradb, there's no JS
[09:40] <kiko> bradb, I think you are confused
[09:40] <SteveA_> and running baz-related tests
[09:40] <sabdfl> with bzr that should be down to a minute or two, total
[09:40] <bradb> kiko: oh, you don't want js, ok
[09:40] <sabdfl> and a faster machine will blaze through the tests
[09:40] <SteveA_> the test suite still takes a while, but it would cut the time to 1/3 what it is, i reckon
[09:40] <sabdfl> be nice to get the whole cycle below 5 minutes
[09:40] <kiko> bradb, dead-simple: optionmenu that applies to searches
[09:40] <bradb> either way, it's exactly 2 lines of Python code simpler, FTR :)
[09:41] <bradb> (changing it now)
[09:41] <kiko> minus 4 lines of comments
[09:41] <kiko> and it may become even simpler
[09:41] <kiko> I need to look at the code to see
[09:41] <SteveA_> strange... moz thunderbird doesn't have enigmail included by default in breezy
[09:42] <fabbione> SteveA: no, enigmail is a plugin.. you nees to install it separately
[09:43] <sabdfl> i think its wedged
[09:43] <SteveA_> stub, elmo and lifeless can unwedge it
[09:43] <SteveA_> and karl
[09:44] <mpt> sabdfl: Do you see anything at http://async.com.br:8081/ ? It's the design-fascism branch
[09:44] <sabdfl> Znarl, elmo, lifeless: help unwedge pqm please, if it is in fact wedged?
[09:44] <sabdfl> mpt: ok, where do i look for the listing stuff?
[09:45] <mpt> http://async.com.br:8081/ http://async.com.br:8081/malone http://async.com.br:8081/support
[09:45] <mpt> they're much more consistent now
[09:45] <SteveA_> kiko: okay, gina improvements merged with no conflicts
[09:45] <mpt> though we still have too many macros for "show me a bug listing"
[09:46] <kiko> SteveA_, cool as juice
[09:46] <SteveA_> kiko: do you have a database query you want me to run before i run gina?
[09:46] <SteveA_> or to look for one of the examples of packages?
[09:48] <sabdfl> mpt: ok, +1, looks good
[09:48] <SteveA_> hurrah
[09:49] <kiko> hmmm
[09:49] <sabdfl> mpt: one thing - the line spacing seems to have widened a bit
[09:49] <sabdfl> could you tighten that up again plese?
[09:49] <sabdfl> please
[09:49] <sabdfl> seems to be across the board
[09:49] <mpt> sure
[09:49] <SteveA_> fabbione: you still around
[09:49] <SteveA_> ?
[09:50] <fabbione> SteveA: yeps..
[09:50] <mpt> p {margin: 0.5em 0em 1em 0em; line-height: 1.5em;}
[09:50] <mpt> that's the badger
[09:50] <SteveA_> can you suggest any packages in main that didn't change between hoary and breezy?
[09:50] <Lathiat> ~[1~[4~mpt: /malone/distros/ubuntu -> has a list heading, but the listing isnt organized in a list like that (is that intentional?)
[09:51] <fabbione> SteveA: probably... none?
[09:51] <SteveA_> if not, in universe?
[09:51] <mpt> Lathiat: What does "~[1~[4~" mean? :-)
[09:51] <fabbione> SteveA: i can try to check.. gimme a few minutes
[09:51] <SteveA_> thanks
[09:51] <mpt> Lathiat: bradb is fixing that bug right now
[09:52] <sabdfl> mpt: nudge the badger in a more compact direction?
[09:52] <Lathiat> mpt: its the result of ssh lag and me hitting a few arrow keys and it coming into irssi in one packet and it using paste mode to write them out properly rather than interpreting them as control codes
[09:52] <mpt> sabdfl: done, overridden in launchpad.css
[09:53] <sabdfl> mpt: thanks. did that change in plone.css recently?
[09:53] <fabbione> SteveA: libpam-radius-auth in universe
[09:53] <SteveA_> fabbione: awesome, thanks
[09:53] <fabbione> SteveA: both hoary and breezy have Version: 1.3.16-3
[09:54] <Lathiat> mpt: ok 
[09:54] <fabbione> SteveA: no problem.. i knew that slacking on a package i maintain would have been somewhat useful in the future :)
[09:54] <mpt> sabdfl: I don't know, I don't change plone.css
[09:54] <mpt> but the line-height was noticable in bug descriptions/comments
[09:55] <sabdfl> mpt: did you change root-portlet-rosetta.pt?
[09:55] <sabdfl> in that branch?
[09:56] <sabdfl> it needs a <br /> after each drl barchart
[09:56] <sabdfl> which is in my branch
[09:56] <sabdfl> i think i fixed it in my branch
[09:57] <mpt> sabdfl: I don't think I changed it in this branch, except that I may have added alt="" to the <img> tags
[09:58] <sabdfl> ok. it looks like i added a <br />
[09:58] <sabdfl> if you touched those lines, its going to conflict shortly, please don't drop the <br />
[10:00] <sabdfl> ok, we are still jammed on pqm
[10:05] <bradb> kiko, mpt: just to be sure, how much more do you want that page to be tweaked from how I've changed it to look now?
[10:06] <SteveA_> sabdfl: i've talked with kiko about a plan for testing gina on staging.  i can't do it right now for two reasons: 1. the old gina data is still on staging, and it needs to be gone, and i don't know how.  2. it is late here and i want to sleep.
[10:06] <kiko> bradb, I think mpt would like the fields aligned, but apart from that, that's what I wanted
[10:06] <SteveA_> so i'm mailing stu with the plan of what we want to do to test this, and your code, out on staging
[10:07] <bradb> kiko: how is the search button aligned?
[10:07] <SteveA_> what is the branch / patchlevel you have to try out?
[10:07] <SteveA_> (just in case it doesn't make it in to pqm shortly)
[10:07] <sabdfl> SteveA_: no problem. i will review kiko's proposed gina diff tonight
[10:07] <sabdfl> i don't have a katie handy
[10:07] <sabdfl> will ask elmo for a dump or something to make gina run locally, where i can test it
[10:07] <sabdfl> i have pleny of space and speed on the desktop
[10:07] <mpt> bradb: in the second column of a <table>, said second column containing the other controls while the first column contains their labels in <th><label>...</label></th> fashion
[10:08] <SteveA_> okay.  however, stu can straightforwardly run this on staging tomorrow
[10:08] <sabdfl> i will land my code on HEAD tonight, if we can unwedge pqm
[10:08] <SteveA_> the only information i'm missing is your branch and patchlevel
[10:08] <sabdfl> SteveA_ please get yourself and kiko fabulous pqm unwedging powerz
[10:08] <SteveA_> i'll ask lifeless about it
[10:09] <SteveA_> so... you're going to work on gina tomorrow, locally?
[10:09] <SteveA_> no need for running on staging by stu first thing?
[10:10] <kiko> hey gneuman 
[10:10] <gneuman> hey
[10:10] <bradb> mpt: mm, isn't it kind of evil to use tables to do this alignment?
[10:10] <kiko> gneuman, I have a new bug or two for you
[10:11] <gneuman> ok
[10:11] <kiko> I tried to merge your fixes today, but snif, overburdened by other chores, will try tonight
[10:11] <gneuman> np
[10:11] <mpt> bradb: no, it's evil to use tables that don't make sense when linearized
[10:11] <mpt> this one does
[10:13] <bradb> i was of the impression that text-to-speech agents make special noises for tables, and for when a new row starts in a table, which makes table-for-layout-rather-than-for-presenting-tabular-data seem all the more evil
[10:14] <bradb> but i'll go with what you said
[10:17] <mpt> bradb: http://www.joeclark.org/book/sashay/serialization/Chapter10.html
[10:17] <SteveA_> i'm going to ask stu to do the gina / mark's branch stuff on staging anyway.  i think it's worth it, given how important this is
[10:18] <kiko> agreed
[10:19] <kiko-afk> sabdfl, what's your branch name?
[10:20] <SteveA_> mark.shuttleworth@c.c/launchpad--newpackageclasses--0
[10:20] <SteveA_> i guess
[10:20] <SteveA_> that's the one that is in needs-reply on the reviews page
[10:21] <bradb> mpt: my reference is Chapter 1 of http://tinyurl.com/8qyed. Jakob Nielsen's preferred CSS book, written by the two guys who invented CSS. *shrug*. :)
[10:25] <mpt> bradb: If it contains a non-<table> way of achieving the same layout in real-world browsers, I'd be very interested in knowing what it is
[10:25] <kiko-afk> so would I
[10:28] <bradb> mpt: The way it looks now, is that way you had in mind?
[10:29] <bradb> s/way/what/
[10:30] <mpt> no, the way it looks now doesn't have multiple rows
[10:30] <Znarl> sabdfl, SteveA,  unwedged pqm.
[10:31] <bradb> mpt: are we looking at the same /products/firefox/+bugs page? That's got multiple rows.
[10:31] <kiko-afk> mpt?
[10:31] <bradb> mpt: i.e. the page off my server you're looking at, right?
[10:32] <kiko-afk> mpt, bradb: shouldn't there be a ":" after the labels, btw?
[10:32] <mpt> bradb: No, I was looking at rocketfuel
[10:32] <bradb> oh
[10:32] <mpt> bradb: kiko showed me your layout, I didn't notice the URL
[10:32] <kiko-afk> http://69.70.209.33:8086/products/firefox/+bugs?field.searchtext=&current_sortorder=datecreated&sortorder=datecreated&search=Search&advanced=&status=10&status=20&assignee=all
[10:33] <bradb> (i think)
[10:33] <mpt> that's good, apart from the colons
[10:33] <kiko-afk> bradb, hit by ubugtu?
[10:33] <bradb> kiko-afk: it swam all over that URL, it appears
[10:33] <bradb> again
[10:34] <kiko-afk> ubugtu? weird
[10:34] <lifeless> moin
[10:36] <bradb> mpt: when will this "Search" button not look like a text entry?
[10:37] <bradb> i'd say that bugfix just became that much more important with this layout change :)
[10:37] <mpt> Don't know
[10:37] <mpt> I've still got a fair bit of dead CSS to cut out
[10:37] <bradb> ok
[10:43] <bradb> kiko-afk: can i land this after i finish making the UI change suggested?
[10:43] <bradb> the only thing that really changes is 1. remove 2 lines of Python code and 2. shift around some HTML.
[10:51] <ajmitch> morning
[10:55] <fabbione> good night guys
[10:56] <ajmitch> bye fabbione 
[11:05] <zyga> hey guys
[11:05] <zyga> can anyone check the logs
[11:05] <zyga> https://launchpad.net/products/launchpad/+bug/3168
[11:05] <Ubugtu> Malone bug #3168: Public polls are broken Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Guilherme Salgado, Status: New http://launchpad.net/malone/bugs/3168
[11:07] <salgado> zyga, would you try adding an option other than "Frdric Logier", but with non-ascii characters?
[11:07] <zyga> salgado: sure, no problem
[11:07] <zyga> salgado: strange - check this out
[11:08] <zyga> https://launchpad.net/people/moturuby/+poll/elect-team-leader/+vote-simple
[11:08] <zyga> I hope you can see the same stuff as I can
[11:08] <zyga> I see 'Frederic' twice
[11:08] <zyga> once with all accents and once without
[11:08] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=SteveA]  landing Slave Manual Mode flag, it allows us to observe/select build job for a suspicious slave w/o stressing the whole environment. (patch-2668)
[11:08] <zyga> salgado: it worked
[11:08] <zyga> salgado: I've added anothe option and there was no problem
[11:09] <salgado> zyga, okay, I guess I know what happened
[11:09] <salgado> zyga, first you added an option with the accents ("Frdric Logier")?
[11:09] <zyga> yes
[11:09] <salgado> and then it crashed?
[11:09] <zyga> actually...
[11:09] <salgado> then you added one without the accents?
[11:09] <zyga> I've always added it with all accents
[11:09] <zyga> or maybe, no ... I did add the one without accents later
[11:10] <zyga> so: fist with accents, crash, then wihout accents, crashed
[11:10] <salgado> okay, in both cases, it didn't crash while adding the option, it crashed later
[11:10] <salgado> so, both options were added
[11:10] <zyga> salgado: I didn't actually notice that those options were already present
[11:11] <salgado> that's why I changed the description of the bug report. although it looked like the crash was when adding the option, the crash was actually after
[11:11] <zyga> salgado: OTOH: can you remove all options later on? I'd like to make an usefull poll in the end
[11:11] <zyga> ah
[11:11] <salgado> so, that problem is, in fact, fixed
[11:11] <salgado> the problem you saw now, when trying to add another "Frdric Logier" is because there can't have two options with the same name
[11:12] <salgado> it should tell you that instead of giving you a system error, though
[11:12] <zyga> hmm but I didn't get any message saying 'you cannot add this'
[11:12] <zyga> right
[11:12] <salgado> and gneuman is working on that
[11:12] <zyga> great
[11:12] <zyga> okay I get it now
[11:12] <zyga> can you please remove all options?
[11:12] <salgado> now, what to do with all the options
[11:13] <salgado> I can't delete them
[11:13] <salgado> but you can mark them as inactive
[11:13] <salgado> I mean, the ones you want
[11:13] <salgado> an inactive option won't show up for people to vote
[11:13] <zyga> hmm
[11:13] <zyga> I cannot
[11:13] <zyga> I get system errro on each 
[11:13] <zyga> https://launchpad.net/people/moturuby/+poll/elect-team-leader/+option/4
[11:13] <zyga> and /2 and /6
[11:13] <zyga> sorry /3 not /2
[11:14] <zyga> still something not right
[11:14] <salgado> dammit. kiko was going to merge a fix for this today
[11:15] <salgado> but it looks like he left without merging it. :-(
[11:15] <Lathiat> wow
[11:16] <Lathiat> i just realised how those buttons up the top
[11:16] <Lathiat> spread onto two lines right aligned
[11:16] <Lathiat> thats cool!
[11:16] <zyga> Lathiat: ?
[11:16] <Lathiat> zyga: resize yoru window horizontaly
[11:16] <Lathiat> to about half 1024
[11:16] <Lathiat> on launchpad
[11:17] <zyga> Lathiat: ah, nice indeed
[11:17] <zyga> Lathiat: no div, span :)
[11:17] <ajmitch> Lathiat: and it gets messy as you make it smaller
[11:17] <ajmitch> I know everyone here just loves the plone.css though :)
[11:18] <zyga> plone.css?
[11:18] <ajmitch> yes
[11:18] <zyga> what is it?
[11:18] <Lathiat> heh
[11:18] <ajmitch> the css style
[11:18] <Lathiat> anyway, im going back to bed, laters all
[11:18] <salgado> zyga, I'll find a way to either merge that fix so you can mark the option inactive or ask someone to delete the option
[11:19] <salgado> zyga, please poke me if it's not fixed by tomorrow morning. and sorry for the trouble
[11:19] <ajmitch> bye Lathiat 
[11:19] <zyga> salgado: thanks :-)
[11:19] <zyga> np
[11:25] <sabdfl> thanks Znarl
[11:32] <sabdfl> hmm... there are 4 patches from salgado in the pqm queue, from 2 different branches. i wonder if it would be possible to delete the earlier from each, snce they will then both go in together with the second in each branch. make sense? znarl, do you have the ability to do that?
[11:32] <sabdfl> we really need a web interface to the pqm queue
[11:32] <sivang> zyga: how do you start a poll for something on launchapd ?
[11:33] <sabdfl> sivang: on a team, you should have a poll creation page
[11:33] <sabdfl> everyone in the team can vote
[11:33] <zyga> re
[11:33] <sivang> sabdfl: ah cool , I'll check that 
[11:33] <zyga> sivang: answered already :)
[11:35] <sivang> zyga: yes 
[11:38] <ddaa> Are the passwords for the sampledata users documented somewhere?