[12:13] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=stevea]  Compressing buildlogs before storing in Librarian. (patch-2651: celso.providelo@canonical.com)
[12:14] <kiko> rock on cprov
[12:14] <kiko> matsubara, bug 3101 confuses me. sabdfl, you have a link in the ticket page, +editpriority, that has no trace of implementation
[12:15] <Ubugtu> Error: I cannot access this bug
[12:15] <kiko> can I just nuke this link?
[12:15] <kiko> Ubugtu,  bug 3101?
[12:15] <Ubugtu> (bug <abbreviation> <number>) -- Look up bug <number> in the bugzilla associated with <abbreviation>.
[12:15] <Ubugtu> Error: I cannot access this bug
[12:15] <kiko> Seveas, that bug is public, isn't it?
[12:15] <Kinnison> it forces me to log in
[12:15] <kiko> it works for me logged out
[12:16] <kiko> does not force me
[12:16] <Kinnison> and then hands me a forbidden
[12:16] <kiko> bug 3103, sorry
[12:16] <Ubugtu> Malone bug #3103: Cannot edit the assignee and priority of a ticket Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Diogo Matsubara, Status: Accepted http://launchpad.net/malone/bugs/3103
[12:16] <kiko> doh
[12:16] <sabdfl> kiko: hmm... no idea
[12:16] <kiko> bug 3103, sorry
[12:16] <Ubugtu> Malone bug #3103: Cannot edit the assignee and priority of a ticket Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Diogo Matsubara, Status: Accepted http://launchpad.net/malone/bugs/3103
[12:16] <kiko> sabdfl, but that code is yours..
[12:16] <sabdfl> sec
[12:16] <kiko> I mean
[12:16] <kiko> nobody else will know if priority is supposed to be edited in a different page :)
[12:17] <sabdfl> kiko: looks like +priority
[12:17] <sabdfl> try that?
[12:17] <sabdfl> zcml/ticket.zcml suggests so, anyhow
[12:17] <sabdfl> am mid-commit and cannot fire up LP
[12:19] <Kinnison> dear gods, I've found the worlds worst docstring
[12:20] <kiko> sabdfl, hmrmph
[12:21] <kiko> welcome back salgado-with-new-power-supply
[12:21] <kiko> man it is hot today
[12:22] <salgado> and it's going to be even worst
[12:22] <sabdfl> kiko: you grumbled, sir?
[12:22] <salgado> you better buy 2 new power supplies for each one that burns
[12:23] <sabdfl> kiko: did you find it ther?
[12:24] <sabdfl> i suspect it got lost in the translation from actions portlet to menu
[12:27] <ddaa> sabdfl: thanks, at least I was able to assign the ticket to myself
[12:28] <ddaa> this started from an attempt to workflow import requests with the ticket tracker.
[12:29] <ddaa> Looks like I'm like the first person to actually try using it in a while :P
[12:29] <sabdfl> ddaa: bug reports welcome
[12:29] <ddaa> well, just what I did :)
[12:29] <lifeless> ddaa: malone
[12:29] <ddaa> lifeless: what malone?
[12:30] <ddaa> I mean, "malone what?"
[12:30] <lifeless> if its not in malone, it aint a bug report :)
[12:30] <niemeyer_> ddaa!
[12:30] <lifeless> niemeyer: !
[12:30] <niemeyer> lifeless!
[12:30] <niemeyer> :)
[12:30] <lifeless> how are you ?
[12:30] <ddaa> lifeless: I'm the one who filed bug 3103
[12:30] <Ubugtu> Malone bug #3103: Cannot edit the assignee and priority of a ticket Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Diogo Matsubara, Status: Accepted http://launchpad.net/malone/bugs/3103
[12:30] <lifeless> ddaa: sweet
[12:30] <niemeyer> lifeless: Fine, thanks!
[12:31] <niemeyer> lifeless: Travelling on LP land.. :)
[12:31] <ddaa> hey niemeyer, got my my sqlobject patch?
[12:31] <niemeyer> ddaa: Yep!
[12:31] <ddaa> Making any progress?
[12:31] <niemeyer> ddaa: Got a few questions for you, have some time?
[12:31] <lifeless> niemeyer: hows it feeling ?
[12:32] <ddaa> niemeyer: well, it will be off hours, since I already sent my activity report for the day, so let's make it quick :)
[12:32] <niemeyer> lifeless: It's great to learn something new.. a huge amount of information in no time.
[12:33] <Kinnison> Is it bad that I have the pending reviews edit link in my bookmarks?
[12:33] <ddaa> lifeless: I basically told him "hey, I'd like to write taxi"
[12:33] <niemeyer> ddaa: Nevermind then.. I'll bother you tomorrow.
[12:33] <ddaa> niemeyer: grrr
[12:33] <lifeless> niemeyer: watch out, he'll be taunting next
[12:34] <niemeyer> ddaa: There are several points, so I don't think I can manage to make it quick.. :)
[12:34] <lifeless> :)
[12:35] <ddaa> Well, just ask so I can think about it overnight and have a _really_ comprehensive anser ready tomorrow :D
[12:35] <niemeyer> ddaa: Only if you promise not to fix anything without me.. :)
[12:36] <niemeyer> ddaa: I'm curious about RevisionNumber, for instance..
[12:36] <gneuman> why my pagetest doesnt fail like my aplication???
[12:36] <ddaa> niemeyer: let's talk about it now while lifeless is around
[12:36] <ddaa> we braintorming that up toghether in london
[12:36] <ddaa> * brainstormed
[12:36] <niemeyer> ddaa: It doesn't seems to make sense, from my distant point of view, to have it being a many-to-many relationship table between revisions and branches.
[12:37] <ddaa> That's because of how baz pull works.
[12:37] <ddaa> hu bzr pull
[12:37] <niemeyer> s/seems/seem/
[12:37] <ddaa> phone...
[12:37] <ddaa> lifeless, please take over
[12:38] <sabdfl> ok
[12:38] <sabdfl> in bzr, the individual revisions are revisions, not changesets
[12:38] <sabdfl> in bzr, they are changeset
[12:38] <sabdfl> s
[12:38] <sabdfl> sorry
[12:38] <sabdfl> in baz, they are changesets
[12:38] <niemeyer> Yes
[12:39] <sabdfl> if two branches contain exactly the same bits, at some point, then they should have the same revision identifier
[12:39] <sabdfl> in fact, they are identical
[12:39] <sabdfl> even if they got there by different roads
[12:39] <kiko> lifeless, it is in malone
[12:39] <sabdfl> if you have two related branches
[12:39] <lifeless> except they may have a different path .. bingo
[12:39] <sabdfl> and you do a bit of mutual merging
[12:39] <sabdfl> they are then the same
[12:39] <sabdfl> so they will have the same revision identifier
[12:40] <sabdfl> now, to represent that in the db, it is not enough to have Revision.branch
[12:40] <sabdfl> you need to be able to have the same revision in multiple branches
[12:40] <sabdfl> presto
[12:40] <Kinnison> ciao all
[12:40] <sabdfl> how'd i do, lifeless?
[12:40] <lifeless> sabdfl: good
[12:40] <sabdfl> hercules sword, dude, that's wots good for knots
[12:40] <lifeless> sabdfl: gold star in fact
[12:40] <sabdfl> traditionally
[12:41] <lifeless> that would be a slip knot, yes ?
[12:41] <niemeyer> sabdfl: Given the fields in these tables (revision, branch, revisionnumber), that logic doesn't seem to make sense to me.
[12:41] <niemeyer> But then, I'm just arriving.. :)
[12:41] <sabdfl> there may still be some leftover cruft, i don't know if those are clean
[12:42] <sabdfl> should be Branch, Revision, and BranchRevision. ddaa?
[12:42] <niemeyer> sabdfl: For instance, revision has the following fields: gpgkey, revision_author, committed_against, log_body, and more.
[12:43] <sabdfl> hmmm.... some of those are not going to make sense
[12:43] <lifeless> niemeyer: still being cleaned up
[12:43] <sabdfl> they sound more changelogish
[12:43] <niemeyer> lifeless: Ah, understood..
[12:44] <niemeyer> It's a bit hard to fix code which was half-changed, even more when you don't actually understand the original code, nor the way it's being changed.
[12:44] <niemeyer> That's an exciting challenge for me :)
[12:45] <lifeless> niemeyer: the db is a roughly direct model of bzr
[12:45] <lifeless> niemeyer: having the same revision in two places is simply 'branch convergence'
[12:45] <lifeless> niemeyer: or in fact 'cp -r'
[12:46] <niemeyer> I was completely lost for hours looking at random bits of information until some of these bits started to make sense.
[12:46] <niemeyer> lifeless: Humm... I see
[12:47] <ddaa> Hu, RevisionNumber is something I wrote today and yesterday from what we talked about in London.
[12:48] <ddaa> I like RevisionNumber, because it says exactly what it is: something that assigne a number to a revision in a branch.
[12:48] <niemeyer> ddaa: I know.. but from what we're talking, we have a long road ahead to correctly populate tables with imported information.
[12:48] <ddaa> niemeyer you do not need to populate the table fully right now
[12:49] <ddaa> Though I think you can get most of the data easily with pybaz.
[12:49] <ddaa> The gpgkey would be the difficult bit, but I'm not even sure it's right.
[12:49] <ddaa> committed_against is a relatively tricky bit too
[12:50] <niemeyer> ddaa: Yes, but the model being used in taxi.py seems to be out of date, almost completely.
[12:50] <ddaa> Well... yes.
[12:50] <ddaa> Actually, no.
[12:50] <ddaa> Right now, we are doing baz on top of a bzr model.
[12:50] <niemeyer> ddaa: For instance, we don't have an Archives anymore, right? We have branches, which in the baz model map to archive+category+...
[12:51] <ddaa> niemeyer: that's correct, the db needs no archive anymore.
[12:51] <niemeyer> ddaa: Right now we're doing a big mess.. hehehe :)
[12:51] <lifeless> niemeyer: you and ddaa have the privilege of NUKING CRUFT
[12:51] <lifeless> :)
[12:52] <ddaa> lifeless: actually, I already nuked much cruft using a very powerful cruft-nuking tool: rm.
[12:52] <kiko> :)
[12:52] <jblack> speaking of crut..
[12:52] <kiko> nobody said anything about crut
[12:53] <ddaa> that was about the quickest way to fixed that code.
[12:53] <lifeless> I like croutons
[12:53] <niemeyer> ddaa: And what's an url, for a pybaz archive? It looks like right now we're using the "name" bit information with a mangled archive+category+... to identify the baz archive. Is that right?
[12:53] <lifeless> niemeyer: the url for a pybaz branch is the url of the archive + the c--b--v section
[12:53] <ddaa> niemeyer: look at importd.archivemanager.ArchiveManager
[12:53] <jblack> I need to spec out the covering of bzr and bazaar branches at the same time.
[12:53] <niemeyer> I know baz works with url, but the archive_manager and other bits of infrastructure doesn't seem to be aware about this, are they?
[12:53] <jblack> Though I still think the better answer is to baz2bzr everything and be done with it.
[12:54] <kiko> the code included? :)
[12:54] <ddaa> In that case the url will be archive_manager._mirror().url
[12:54] <ddaa> except without the brackets
[12:54] <ddaa> archive_manager._mirror.url
[12:54] <ddaa> ArchiveManager works with pybaz.ArchiveLocation
[12:55] <ddaa> which is something I wrote to (minimally) support baz urls.
[12:55] <ddaa> Which is basically an url associated to an archive.
[12:55] <niemeyer> ddaa: About quickest and easiest way to fix it, I'm not criticising (or at least that was not my intention). I'm just lost in a changing code which was half-implemented, and tyring to understand what you think to be the way out to push together. :)
[12:56] <ddaa> niemeyer: I was merely making a tentatively witty and quite jubilatory statement
[12:56] <ddaa> we all hated this cruft with a passion
[12:56] <niemeyer> ddaa: Ah, ok.. that was another bit I was going to ask: mirrors. So there are no mirrors anymore.. only branches. Is it right?
[12:56] <ddaa> Yes, i mean No.
[12:57] <ddaa> We are still using baz.
[12:57] <ddaa> So there are still mirrors.
[12:57] <ddaa> But bzr there are no mirrors.
[12:57] <niemeyer> And what's a branch mirror?
[12:57] <ddaa> A branch.
[12:57] <niemeyer> Ahh, I understand now.
[12:57] <niemeyer> Humm.. no, I don't. :)
[12:57] <ddaa> That happens to be used in such a way that it always have the same revision history as its master.
[12:58] <niemeyer> At least not enough to reimplement taxi correctly :)
[12:58] <ddaa> That's fine.
[12:59] <niemeyer> Is it? I'd really like to reimplement it.
[12:59] <ddaa> So, those branch objects you will be creating bubble up to the launchpad webapp. Look at e.g. https://launchpad.net/products/samba
[01:00] <ddaa> Currently branches are showed as hyperlinks, whose text is the branch title (I think) and whose href is the branch url.
[01:00] <ddaa> The only url that's going to be useful to the user is the url at bazaar.ubuntu.com
[01:00] <ddaa> that is the mirror
[01:01] <ddaa> Right now, a mirror is an Arch mirror, so you need not worry much about bzr.
[01:01] <ddaa> Most of your work will be modeling Arch into the this bzr-based schema.
[01:02] <niemeyer> I see.. but what I don't understand is: we have archive_manager, and archive_manager has (for baz) an arch namespace to identify the branch. But this won't help us to fill the branch information, so we get the mirror url. Now, the mirror url was used for the branch itself. Where are the baz mirrors?
[01:02] <ddaa> Basically, revno(base-0)=1, revno(patch-1)=2, etc.
[01:02] <ddaa> mh
[01:02] <lifeless> ddaa: mmm
[01:03] <ddaa> Good point.
[01:03] <niemeyer> Ok.. that's for revisionnumber. That's the easiest part indeed.
[01:03] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial]  Fix for bug 3110: Typo in template admin page. Fix another host of similar typos by using grep. Also cleared up whitespace and reduced the number of false positives for simple grep tests (patch-2652: christian.reis@canonical.com)
[01:03] <lifeless> ddaa: iter_ancestors per favour
[01:03] <lifeless> ddaa: I think what we want, to prep for bzr, is to invert the master and mirror relationship
[01:03] <ddaa> lifeless: YAGNI
[01:04] <lifeless> ddaa: by which I mean, is that the 'official' location we pass around is the output location, and we 'pull from there' to start work, and 'push when finished'
[01:04] <niemeyer> I think that what we *really* want is to rip out the baz code from launchpad and yield RIP!
[01:04] <ddaa> mh... not YAGNI...
[01:04] <niemeyer> But we probably can't do that, so let's find a way to clean up that mess. :)
[01:04] <lifeless> ddaa: I was thinking this as I reviewed the locations patch, but it was out of scope for that
[01:05] <lifeless> and I thought we should really land that bastardo patch
[01:05] <sabdfl> anybody know if stub has announced a tag point?
[01:05] <lifeless> not offhand
[01:05] <ddaa> lifeless: improtd-archivelocation is merged all around
[01:05] <lifeless> yay
[01:05] <ddaa> it's not rolled out, but I see no compelling reason to do so
[01:06] <ddaa> niemeyer: so, there are actually two mirror locations of interest to you. The one you use to mirror to (sftp://...) and the one you want to register into the database (http://bazaar.ubuntu.com/...)
[01:09] <ddaa> I think you would have to add knowledge about the latter as public_mirror_base='http://bazaar.ubuntu.com/' in config.py, stick it in the Job (in master.cfg in botmaster) and use it in Taxi.
[01:09] <ddaa> The rest of the branch url you can construct by appending archive_manager.version.fullname
[01:10] <ddaa> niemeyer: am I being helpful?
[01:10] <niemeyer> ddaa: No doubts
[01:10] <ddaa> Though, I might be wrong somewhere, maybe this knowledge is already there.
[01:11] <niemeyer> ddaa: You mentined that mirrors are branches as well, but we still have mirror* tables around.
[01:11] <niemeyer> ddaa: Are those leftovers as well?
[01:11] <ddaa> hu... really...
[01:11] <ddaa> right... maybe...
[01:11] <ddaa> This stuff is related to the SuperMirror
[01:12] <niemeyer> ddaa: And, do we have a relationship table to connect mirrors already?
[01:12] <ddaa> I think it would still make sense for the supermirror to know about multiple mirror location for a given branch. Even with bzr, you might want to check with jblack. When in doubt, nuke.
[01:13] <ddaa> niemeyer: we do not have anything about mirror that was actually used meaningfully outside of taxi.
[01:14] <niemeyer> ddaa: So mirrors are not yet modeled according to the new scheme. Understood.
[01:14] <ddaa> I think that for immediate taxi needs, you can entirely get rid of the mirror table (db patch!)
[01:14] <ddaa> though I think we would need input from jblack about it
[01:14] <ddaa> since he's the MirrorMaster
[01:16] <ddaa> niemeyer: do you have other questions?
[01:17] <niemeyer> ddaa: Probably, but I'll let you do something else besides explaining launchpad to me.. :)
[01:17] <niemeyer> At least tonight ;)
[01:17] <ddaa> The next thing on my agenda involves a mattress and a pillow.
[01:17] <sabdfl> ddaa: snap :-)
[01:17] <niemeyer> ddaa: Thanks for the explanations
[01:19] <ddaa> where's my magic wand?
[01:19] <ddaa> argh, too late!
[01:21] <ddaa> good night guys
[01:21] <sabdfl> night ddaa
[01:21] <sabdfl> night pumpkin
[01:21] <niemeyer> ddaa: Good night!
[01:21] <Nafallo> lol
[01:21] <Nafallo> gnight ddaa :-)
[01:34] <segfault> any lp dev around?
[01:35] <segfault> I'm trying to merge my personal account and a team's contact account, is there any constraint about it?
[01:45] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial]  Remove portlet-actions from templates and remove some dead *_legend template code. (patch-2653: christian.reis@canonical.com)
[02:34] <gneuman> night all
[02:46] <dilys> Merge to rocketfuel@canonical.com/launchpad--production--1.36: Cherry picks (patch-9: brad.bollenbach@canonical.com, stuart.bishop@canonical.com)
[03:19] <stub> lifeless: Can you please mirror rocketfuel@canonical.com/launchpad--production--1.37
[03:21] <lifeless> apparently not
[03:21] <lifeless> having trouble getting to the chinstrap
[03:22] <lifeless> you have access to pqm - just run 'baz archive-mirror rocketfuel@canonical.com' please
[03:23] <lifeless> ah, I got in
[03:24] <lifeless> done
[03:33] <stub> ta
[04:12] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial]  Make logging to the librarian more informative when dealing with minimal or broken exceptions (patch-2654: stuart.bishop@canonical.com)
[04:47] <dilys> Merge to rocketfuel@canonical.com/dists--devel--0: [trivial]  production-1.37 config (patch-120: stuart.bishop@canonical.com)
[05:26] <dilys> Merge to rocketfuel@canonical.com/launchpad--production--1.37: Cherry pick patch-2654 into production (patch-1: stuart.bishop@canonical.com, rocketfuel@canonical.com)
[09:05] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial]  Add a robots.txt to the librarian that denies everything. (patch-2655: andrew.bennetts@canonical.com)
[09:33] <sabdfl> moin moin
[09:33] <sabdfl> stubarooney around?
[09:45] <Keybuk> morning, boss
[09:46] <Keybuk> so, today I'm starting the big FIVE HUNDRED PACKAGE test with the bzr version of hct/sourcerer
[09:59] <Kinnison> Morning
[10:02] <sabdfl> Keybuk: cool, have you done any other tests on smaller numbres of packages?
[10:02] <sivang> Good morning
[10:02] <sabdfl> stubarooney!
[10:02] <sabdfl> question for you
[10:02] <sabdfl> how do i get a resultset ranked by ftq?
[10:02] <Keybuk> sabdfl: yeah, lots of tests on random packages I could think of
[10:03] <sabdfl> Keybuk: and what sort of results have you been getting?
[10:03] <Keybuk> it's verrrry fast :p
[10:03] <stub> Check out the end of launchpad/doc/textsearching.txt - there are examples there. You need to ORDER BY rank(...)
[10:03] <stub> sabdfl: ^^^
[10:03] <dilys> Merge to rocketfuel@canonical.com/sourcerer--devel--0: [trivial]  sync patch logs from stable branch (patch-31: scott@canonical.com)
[10:04] <sabdfl> stub: can i do that from inside of SQLObject?
[10:04] <stub>     >>> runsql(r"""
[10:04] <stub>     ...     SELECT
[10:04] <stub>     ...         name, rank(fti, ftq('gnome')) AS rank
[10:04] <stub>     ...     FROM product
[10:04] <stub>     ...     WHERE fti @@ ftq('gnome')
[10:04] <stub>     ...     ORDER BY rank DESC, name
[10:04] <stub>     ...     """)
[10:04] <stub> sabdfl: perhaps
[10:05] <Keybuk> sabdfl: also I'm quite pleased with how efficient it is comparitive to the tla stuff
[10:05] <stub> sabdfl: If you can get sqlobject to order by an arbitrary function, then yes. Worst case is you need to create a view, using a query similar to the one I just pasted, and query that using sqlobject
[10:05] <sabdfl> should be, it's been a rather expensive exercise
[10:05] <Keybuk> all of the "dead" changesets to do things like make branches and move patch logs vanish -- which gives you a very clean bzr log output
[10:05] <carlos> morning
[10:06] <Kinnison> stub: I need to know what gina is doing when she gets those exceptions
[10:06] <Kinnison> stub: can you run her with her chatter turned sky-high?
[10:07] <Kinnison> stub: Since she should *NEVER* try to insert duplicated records
[10:07] <sabdfl> stub: so, if i can get it to ORDER BY rank(fti, ftq(query)) then that would be fine... but it would do the ftq twice, would it not?
[10:08] <sabdfl> once for the select, and once for the ORDER BY?
[10:08] <stub> Kinnison: I had a look. I expect that the rules for determining the build differ in the 'does this package exist' method and in the 'ensureBuild' method
[10:08] <Kinnison> stub: how hard to make them match up properly?
[10:08] <Kinnison> stub: It's essential I know how much gina can't import
[10:08] <stub> sabdfl: It won't run ftq() twice because it is an IMMUTABLE function - it will only get run once (per unique argument list) per statement
[10:09] <stub> Kinnison: I got confused tracing it through
[10:09] <sabdfl> and does fti @@ ftq('foo') do the same as rank(fti, ftq('foo')) ?
[10:09] <stub> Kinnison: I'll do another run of warty (the main bit) with verbosity on
[10:09] <Kinnison> stub: thanks
[10:10] <Kinnison> stub: a complete and reliable gina import is one of the things we're relying on for opening dapper
[10:10] <stub> sabdfl: fti @@ ftq('foo') is the bit that returns results that match using the fti index. rank(fti, ftq('foo') examines the particular bit of the index and calculates ranking. You could do 'WHERE rank(fti, ftq('foo') > 0' or something but it would be really slow I think
[10:12] <sabdfl> ProgrammingError: ERROR: non-integer constant in ORDER BY SELECT DISTINCT DistributionSourcePackageCache.id, DistributionSourcePackageCache.name, DistributionSourcePackageCache.binpkgdescriptions, DistributionSourcePackageCache.binpkgnames, DistributionSourcePackageCache.sourcepackagename, DistributionSourcePackageCache.binpkgsummaries, DistributionSourcePackageCache.distribution FROM DistributionSourcePackageCache WHERE fti @@ ftq('browser')
[10:13] <sabdfl> WHERE fti @@ ftq('browser') ORDER BY 'rank(fti, ftq(\'browser\'))'                                                                                                                    
[10:15] <stub> sabdfl: it is turning the function call in the ORDER BY into a string. I assume this is SQLObject mangling it? spiv might know if it is possible, but my guess is that SQLObject doesn't support it so you are stuck with a view.
[10:15] <sabdfl> yuck yuck
[10:15] <sabdfl> spiv: ?
[10:16] <Kinnison> is it a case of __sqlrepr__ needing to be overridden?
[10:16] <Kinnison> sabdfl: try:
[10:16] <Kinnison> class Rank:
[10:16] <Kinnison>     def __sqlrepr__(self):
[10:16] <Kinnison>         return 'rank(....)'
[10:16] <Kinnison> and then orderBy=Rank()
[10:17] <sabdfl> launchpad_dev=# SELECT DISTINCT DistributionSourcePackageCache.id, DistributionSourcePackageCache.name FROM DistributionSourcePackageCache WHERE distribution = 2 AND fti @@ ftq('browser') ORDER BY rank(fti, ftq('browser')); ERROR:  for SELECT DISTINCT, ORDER BY expressions must appear in select list 
[10:17] <stub> Kinnison: That is disgusting, but you could be right
[10:18] <Kinnison> sabdfl: :-(
[10:18] <stub> sabdfl: DISTINCT is biting you there
[10:19] <stub> sabdfl: ORDER BY max(rank(...))) might work, or it might need to be refactored into a GROUP BY
[10:19] <sabdfl> is this something that SQLObject could handle as a general case?
[10:19] <Kinnison> There were 1 imports of names not appearing in the __all__.
[10:19] <Kinnison> You should not import __doc__ from _strptime: canonical.archivepublisher.nascentupload
[10:19] <Kinnison> Anyone know why that might happen?
[10:19] <Kinnison> TTOTD: I'm *NOT* doing that
[10:20] <stub> Kinnison: import facist vs. standard library I'm afraid. Need to ping SteveA about that
[10:20] <stub> (_strptime is most likely the C implementation, and a python module does import * from it)
[10:20] <Kinnison> stub: so long as I won't get beaten for it, I'm happy
[10:20] <sabdfl> sabdfl: lifeless, ddaa, niemeyer: http://www.kernel.org/git/
[10:21] <stub> oops... gotta go. back in an hour or two.
[10:21] <sabdfl> does that not look like the launchpad branch listings i've been haranguing about since pre-UDU?
[10:21] <stub> Kinnison: gina is running
[10:21] <sabdfl> the world is converging on the things we set out to create 18 months ago
[10:21] <sabdfl> it's entirely our fault if they don't converge on us
[10:21] <sabdfl> we had this vision way ahead of everyone else
[10:22] <sabdfl> now we need to deliver on it before other people invent it
[10:22] <sabdfl> that page is basically what we want to see at launchpad.net/products/linux-kernel/+branches
[10:22] <sabdfl> make sense?
[10:23] <sabdfl> then, see why i get so excited about branch activity visualisation?
[10:23] <sabdfl> the number of branches will explode, into the hundreds, thousands for large projects
[10:23] <sabdfl> we need to be the most effective index to it all
[10:24] <sabdfl> the race is on :-)
[10:25] <sabdfl> stub: if SQLObject had a selectAlso= option to select, then I could do:
[10:26] <Keybuk> sabdfl: randomly, I thought of another karmic event last night
[10:26] <Keybuk> we can tell now when someone takes a revision you've created and pulls or merges it into their branch
[10:26] <Keybuk> that should be karmic
[10:26] <sabdfl> Foo.select("fti @@ ftq('txt)", selectAlso="rank(fti, ftq('txt')) AS rank", orderBy='rank')
[10:26] <sabdfl> Keybuk: YES
[10:27] <sabdfl> very cool indeed
[10:27] <sabdfl> "my code spreads fast"
[10:27] <Kinnison> That'll seriously help us spot vestigal branches for pruning and also finished branches
[10:27] <sabdfl> and if the branches themselves had some sort of relevance rating, then more karma for getting your bits into branches with greater consequence would work too
[10:28] <Keybuk> "sabdfl's quick patch to the kernel has been merged into LINUS'S TREE", sabdfl.karma += 10000000
[10:28] <Keybuk> type thing
[10:28] <zyga> heh :)
[10:29] <Kinnison> Keybuk: only if we can then do "Security patch #XXX to linux-source had to fix sabdfl's patch", sabdfl.karma -= 10000000
[10:29] <sabdfl> Keybuk: ah, the joys of being root on the karma allocating system :-)
[10:29] <sabdfl> if Branch.revision.owner.name == 'sabdfl': ...
[10:30] <sabdfl> Ooohh... stub, this works a treat
[10:30] <sabdfl> SELECT DISTINCT DistributionSourcePackageCache.id, DistributionSourcePackageCache.name, rank(fti, ftq('browser')) AS rank FROM DistributionSourcePackageCache WHERE distribution = 2 AND fti @@ ftq('browser') ORDER BY rank DESC;
[10:30] <carlos> zyga, hi, around?
[10:30] <sabdfl> how hard would it be to add the selectAlso= to SQLObject?
[10:30] <sabdfl> spiv: ?
[10:31] <sabdfl> i need it, oh, in the next few hours would do ;-)
[10:31] <sabdfl> stub: did you set a tag point yet?
[10:32] <sabdfl> Kinnison: ping, real world
[10:32] <Keybuk> "Cache" ?
[10:34] <zyga> carlos: morning :)
[10:35] <carlos> zyga, morning
[10:35] <carlos> zyga, I forgot to tell you that now that breezy is released, the new language pack exports contain only updates since the release date
[10:36] <zyga> hmm so they are diffs basically?
[10:36] <zyga> could you md5sum the last tarball that was used for breezy
[10:36] <zyga> I'd like to be sure I've got the right one
[10:41] <zyga> carlos: ?
[10:42] <carlos> zyga, no, those are not diffs
[10:42] <carlos> zyga, the tarball include full .po files
[10:42] <carlos> zyga, but only the ones that were updated since the release date
[10:44] <zyga> I see - that's even better
[10:44] <lifeless> sabdfl: ack
[10:44] <zyga> I'll simply extract them over previous tree and voila
[10:45] <sabdfl> lifeless: cool, could you put that little commentary in a mail to your team, please?
[10:45] <carlos> zyga, right
[10:45] <zyga> 16869b7ac3fc95f08dd208d5f7738359  rosetta-breezy-2005-10-11.tar.gz
[10:45] <zyga> that's my final tarball
[10:53] <dilys> Merge to rocketfuel@canonical.com/hct--devel--1: [trivial]  fix test case on machines without /etc/mailname (patch-21: scott@canonical.com)
[10:54] <Keybuk> heh
[10:54] <Keybuk> I submitted that one during the London sprint!
[10:55] <Kinnison> your laptop only just learned how to send mail again?
[10:55] <Keybuk> no, finally recovered the last bit of broken disk on my mail server
[10:56] <Keybuk> it's been spending the last week processing mail
[11:00] <Kinnison> heh
[11:00] <Keybuk> complete with echo -e ... \r
[11:04] <Kinnison> Keybuk: yum
[11:04] <Kinnison> SteveA: ping
[11:05] <Keybuk> (because my mail server bitches if you don't end lines with \r\n <g>)
[11:06] <Kinnison> hehe
[11:12] <Kinnison> gah, no stub
[11:13] <Kinnison> oh well
[11:16] <zyga> carlos: can you confirm the md5 sum?
[11:18] <carlos> sure, just a second...
[11:19] <carlos> 16869b7ac3fc95f08dd208d5f7738359  rosetta-breezy-2005-10-11.tar.gz
[11:20] <carlos> zyga, same checksum
[11:28] <BjornT> spiv: ping
[11:33] <zyga> carlos: thanks
[11:46] <spiv> sabdfl: "selectAlso" is an interesting idea, I'm not sure it's straightforward to do, but should be possible...
[11:46] <spiv> BjornT: pong
[11:46] <sabdfl> spiv: https://chinstrap.ubuntu.com/~dsilvers/paste/fileqsLxE1.html
[11:46] <sabdfl> i've mailed spiv with a request to review, polish, and land
[11:47] <sabdfl> he seems to be offline, could you handle it unless you can reach him?
[11:47] <spiv> You mean stub?
[11:47] <sabdfl> i'm depending on it for a landing due this weekend
[11:47] <sabdfl> spiv: sigh. yes :-)
[11:47] <spiv> :)
[11:47] <spiv> I'll take a look, and see if I can add some tests...
[11:47] <sabdfl> the patch is tiny. only thing i' not sure of is elegance and completeness
[11:49] <BjornT> spiv: https://chinstrap.ubuntu.com/~dsilvers/paste/fileMlKHqk.html
[11:50] <BjornT> spiv: is that intended behaviour? (i'm thinking of the type difference between existing_ticket.id and new_ticket.id)
[11:51] <spiv> The approach seems right to me, although I don't see how you'd extract the results with that.
[11:51] <spiv> BjornT: Hmm, seems a bit surprising to me.
[11:52] <spiv> BjornT: But the distinction between ints and longs generally doesn't matter any more... is it an issue for you?
[11:54] <BjornT> spiv: it's not that big of a deal, but it's quite annoying that in tests, you have to append 'L' to some ids.
[11:55] <spiv> sabdfl: I can see how your patch modifies the query to request the extra columns, but I don't see how you'd get the results back.  AFAICT, the extra columns will just be discarded.
[11:55] <spiv> sabdfl: What did you have in mind?
[11:57] <spiv> BjornT: Yeah, that is annoying.  Hmm.
[11:58] <spiv> sabdfl: Oh, you don't *need* the extra results, because you're only using it for ORDER BY... I see.
[12:00] <Keybuk> what's trialsuite.py ?
[12:00] <spiv> Well, trial is twisted's test runner, so perhaps it's something to do with that?
[12:01] <Keybuk> dunno
[12:01] <Keybuk> it just randomly appeared in baz merge output
[12:01] <Keybuk> it was "deleted" apparently
[12:01] <Keybuk> when it wasn't in my tree, and baz status doesn't mention it
[12:03] <sabdfl> spiv: yes. it's just a way to get something into the query that orderBy can use. the WHERE could also have limits on it, for example "rank > 0.3"
[12:04] <spiv> I see.  You'd only need this with SELECT DISTINCT, I guess.
[12:04] <pitti> Hi
[12:04] <Kinnison> hey pitti
[12:04] <pitti> Does malone allow me to search for/display all bugs that I filed?
[12:05] <Keybuk> https://launchpad.net/people/$USER/+reportedbugs
[12:05] <Keybuk> if you to your homepage, click on "Bugs", it's in the list on the right
[12:05] <Keybuk> "Bugs Reported"
[12:05] <pitti> ah, thanks
[12:09] <spiv> sabdfl: One minor issue with your patch is that we shouldn't really have any keyword arguments on selectBy other than connection, because it potentially clashes with **kwargs.  If we need special options there, we ought to use .selectBy(foo=bar).distinct().orderBy('foo'), or maybe .selectBy(foo=bar).clone(distinct=True, orderBy='foo').  So if you don't mind I'll leave that part of your patch out.
[12:13] <sabdfl> spiv: ok
[12:13] <sabdfl> can you use selectBy with ftq currently?
[12:14] <Kinnison> I don't think so
[12:14] <Kinnison> Unless someone overloaded something in a bizarre way to allow it
[12:14] <sabdfl> right, so it's moot. spiv: sounds fine
[12:14] <sabdfl> spiv: could you mail stub to let him know you are dealing with the thing i maile dhim about?
[12:15] <spiv> sabdfl: I'll do that now.
[12:15] <sabdfl> spiv: thanks
[12:21] <Keybuk> hmm, why did I never think about adding "baz archive-mirror" to pqm-submit-merge before? :p
[12:22] <Kinnison> cat bin/pqm-chinstrap-merge
[12:22] <Kinnison> #!/bin/sh
[12:22] <Kinnison> baz-mirror-here
[12:22] <Kinnison> exec arch-submit-merge "$*" $PQM_ADDRESS
[12:23] <Kinnison> :-)
[12:23] <Kinnison> where baz-mirror-here is essentially: baz archive-mirror $(baz tree-version | cut -d/ -f1)
[12:24] <Kinnison> which lets me just do: pqm-chinstrap-merge Some message with stuff in it, r=someone
[12:26] <Keybuk> I did:  baz archive-mirror $(baz parse-package-name -a $(baz tree-version))
[12:26] <Keybuk> heh
[12:27] <Kinnison> cut -d/ -f1 was shorter :-)
[12:27] <Kinnison> there's probably some zsh pattern to do it too
[12:28] <Keybuk> ${$(baz tree-version)%/*}
[12:31] <Kinnison> there we go :-)
[12:32] <Kinnison> now all you need is a pattern which means "the contents of the {arch}/++default-version file somewhere here or upwards..."
[12:32] <Kinnison> and you can get it without invoking baz
[12:32] <Kinnison> :-)
[12:33] <Kinnison> What's the syntax to get a Person's default address?
[12:33] <Kinnison> aah, is it .preferredemail?
[12:35] <spiv> sabdfl: I think we can do better that this -- really, you just want orderBy='expression rather than a column' to Just Work, rather than having to tell SQLObject how to write SQL.
[12:40] <spiv> sabdfl: I'll stick with selectAlso for now, though.
[12:41] <sabdfl> spiv: ok. i'm happy as long as *something* lands in sqlobject today (our branch) that preserves the current semantics, if there is also a better way, then thats cool
[12:42] <sabdfl> if there are better semantics, i'm only using it in a few places
[12:42] <sabdfl> however, if i get a chance, i'm going to improve ALL the places we use FTQ to use ranking. stub will love me and give me brownie points if i do that, see
[12:42] <sabdfl> so let me know soonish if the semantics will change
[12:42] <Kinnison> sabdfl: hash brownie points?
[12:43] <sabdfl> Kinnison: something like that
[12:43] <spiv> sabdfl: Ok.  I've already written a test for selectAlso, and I want to think a little more before I add magic to orderBy, so I'm going to be pragmatic and land selectAlso.
[12:43] <sabdfl> spiv: i have a rather large review coming up. can i ask you to handle it tomorrow, your time?
[12:43] <sabdfl> would credit time elsewhere
[12:43] <sabdfl> so i can land over the weekend
[12:44] <spiv> How large is large? :)
[12:44] <sabdfl> few thousand lines, but relatively straightforward stuff
[12:44] <sabdfl> mainly i would be looking for cut-and-paste bpb's
[12:45] <sabdfl> basically, finally reworking all the package UI and supporting classes
[12:45] <Kinnison> Okay, how do we send mails in a zopeless situation?
[12:46] <spiv> "bqb's"?
[12:46] <spiv> Er, "bpb's", I mean ;)
[12:46] <sabdfl> brown paper bag bugs
[12:47] <Kinnison> sabdfl: is that where you request a review, then put a paper bag on your head and go "lalalala" ?
[12:47] <sabdfl> Kinnison: when this has landed, could i ask you to clean up the organisation of your classes?
[12:48] <sabdfl> each into its own file
[12:48] <Kinnison> sabdfl: Sure, I can put that on my gardening TODO list
[12:48] <sabdfl> so someone else knows where to find sourcepackagepublishing etc
[12:48] <sabdfl> thanks
[12:51] <spiv> sabdfl: Ok, I'll do it -- I assume it will ready for me by the time I wake up tomorrow?
[12:52] <sabdfl> spiv: how many hours till then?
[12:53] <sabdfl> does anybody know what stub's intended branch point is?
[12:53] <spiv> sabdfl: About 14 or so :)
[12:54] <spiv> I'm not usually an early riser.
[12:54] <sabdfl> spiv: no problem. might even fit in a release party between now and then too :-)
[12:55] <spiv> sabdfl: andrew.bennetts@canonical.com/sqlobject--selectAlso--0 is mirrored -- do you want to take a look, or should I just say it's [trivial] ? ;)
[12:56] <sabdfl> spiv: go for it
[12:56] <sabdfl> hmm
[12:56] <sabdfl> Kinnison suggested being careful only to add ', ' if there was a column already, or something
[12:56] <spiv> Ignoring the test, the diff is even smaller than yours, so [trivial]  is appropriate.
[12:57] <spiv> Hmm.
[12:57] <spiv> That's true, I guess -- if you ever get a table with no columns other than id.
[12:59] <sabdfl> otherwise, kinnison and i have both looked at that code. what had me worried was that there might be other places where a query is constructed that might also need to be told about it
[12:59] <sabdfl> it works for the current case though
[01:01] <spiv> Hah, that case is broken without selectAlso anyway.
[01:04] <spiv> Hmm, the only case that can happen I think is with an accumulation, e.g. Foo.select(..., selectAlso=...).count().  Hmm.
[01:05] <sabdfl> i guess what i'm saying is that there may be other ways to use SQLObject that might also need to know about it, like .count() or something, or the set operation stuff
[01:05] <sabdfl> .count() probably doesn't need it
[01:05] <sabdfl> because it does not affect WHICH rows are selected, and hence counted
[01:05] <spiv> It's the only other place in the code I can see that does it.
[01:06] <spiv> Yeah, doesn't make so much sense with count()
[01:06] <spiv> It could theoretically matter with .sum()
[01:06] <spiv> So I'll take the easy route and add an assert to that code path, seeing as we don't have a use-case for it.
[01:10] <sabdfl> what's the proper way to get a row to delete itself? foo.destroySelf()?
[01:11] <Kinnison> yup
[01:13] <SteveA> Kinnison: you have something in the std library that is causing the import fascist to complain?
[01:13] <Kinnison> SteveA: yes
[01:13] <Kinnison> There were 1 imports of names not appearing in the __all__.
[01:13] <Kinnison> You should not import __doc__ from _strptime: canonical.archivepublisher.nascentupload
[01:13] <SteveA> do you know what line of code produces the error?
[01:13] <Kinnison> I assume the call of time.strptime
[01:14] <spiv> sabdfl: So, is that r=sabdfl,Kinnison?
[01:14] <SteveA> okay.  don't worry about it, i'll tell the importfascist about it.
[01:14] <Kinnison> SteveA: ta
[01:14] <sabdfl> spiv: yes, thanks very much
[01:14] <Kinnison> erk, don't put my name on it
[01:14] <sabdfl> will you pass upstream for consideration too, please?
[01:14] <SteveA> Kinnison: it's just causing you a bit of spew, not errors, right?
[01:14] <spiv> Kinnison: Heh, ok.
[01:14] <Kinnison> SteveA: aye
[01:14] <SteveA> okay.  it's all good.
[01:17] <SteveA> sabdfl: earlier, you asked "is there a way to have a menu item in the overview menu of object A take you to a page for object B?"
[01:17] <sabdfl> yes?
[01:17] <SteveA> so, that would be a menu item in the application menu for object A that is a link outside of object A's context?
[01:17] <SteveA> if so, you just say  def whateveritem(self):
[01:17] <SteveA>   return Link('/abs/url/to/wherever', text)
[01:18] <SteveA> you can use canonical_url to get the link you want, or use an absolute link or whatever
[01:18] <sabdfl> ok, very cool. thanks SteveA
[01:19] <spiv> sabdfl: Merge sent.
[01:19] <Kinnison> spiv: thanks
[01:19] <Keybuk> which one?
[01:19] <SteveA> so, if you have a link target of 'foo', it is relative to the menu's context... '/foo' is relative to the launchpad root, and http://whatever/foo is that absolute url
[01:19] <spiv> Keybuk: "The Life Aquatic with Steve Zissou"
[01:19] <Keybuk> good choice
[01:23] <dilys> Merge to rocketfuel@canonical.com/hct--devel--1: [trivial]  sync patch logs from stable branch (patch-22: scott@canonical.com)
[01:23] <Keybuk> aha, there we go
[01:28] <SteveA> bug 3110
[01:28] <Ubugtu> Malone bug #3110: Typo in template admin page Fix req. for: rosetta (upstream), Severity: Normal, Assigned to: Christian Reis, Status: PendingUpload http://launchpad.net/malone/bugs/3110
[01:39] <carlos> Kinnison, Could you tell me under which circumstances SecureSourcePackagePublishingHistory.status will be PENDING ? and, is it possible that we have more than one row for the same sourcepackagename and distrorelease with the status field set as PENDING or more than one set to PUBLISHED?
[01:39] <Kinnison> when uploads have come in but not yet gone to the archive
[01:39] <Kinnison> yes
[01:40] <carlos> Kinnison, but when it's PENDING, the sourcepackage is already built by the build farm, right?
[01:41] <Kinnison> no
[01:41] <Kinnison> we don't build PENDING (for now)
[01:42] <carlos> Kinnison, hmm, I'm confused, what's the point behind pending if it's not a queue to mark the ones that need to be built or are being built...
[01:42] <Kinnison> they've come in but not gone to the archive on disk
[01:42] <Kinnison> putting them on disk is expensive (takes some number of minutes)
[01:42] <Kinnison> so that's a periodic operation
[01:42] <carlos> Kinnison, what happens when an Ubuntu developer uploads a new sourcepackage?
[01:43] <Kinnison> new?
[01:43] <Kinnison> Do you mean NEW or new?
[01:43] <Kinnison> I.E. is it never-before-seen, or is it a new version of an existing one?
[01:43] <carlos> a new release is done and I'm as the maintainer, want to upload it to the archive
[01:43] <carlos> s/I'm/I/
[01:44] <Kinnison> You upload it
[01:44] <Kinnison> It enters queue/accepted
[01:44] <Kinnison> at some point, it moves from queue/accepted to queue/done and the publishing records are made in PENDING
[01:44] <Kinnison> at some point, the publisher publishes your package and it's made PUBLISHED, having SUPERSEDED the old version
[01:44] <Kinnison> then it gets built
[01:45] <Kinnison> the exact time relationships between them are not defined and elmo and I will chat about that later
[01:45] <carlos> hmm
[01:45] <carlos> ok
[01:46] <carlos> then I think I will just kill the attach script completely to substitute it with your call to the API I wrote.. It's easier...
[01:47] <Kinnison> If the attach script is needed for breezy/hoary/warty updates then keep it
[01:47] <Kinnison> We won't be building the old stuff in launchpad for quite a while
[01:47] <dilys> Merge to rocketfuel@canonical.com/sqlobject--test--0.6: [r=sabdfl]  Add selectAlso to SQLObject to allow SELECT DISTINCT with ORDER BYs that aren't plain columns. (patch-37: andrew.bennetts@canonical.com)
[01:47] <sabdfl> ka-ching
[01:48] <Kinnison> heh
[01:49] <carlos> Kinnison, no, we don't need it once all tarballs are imported
[01:50] <Kinnison> whatever you say
[01:50] <Kinnison> I'd not delete the scripts or anything, just turn them off though
[01:50] <Kinnison> just in case
[01:53] <carlos> Kinnison, I'm changing the way we attach files to Rosetta so I need to migrate that script or just remove it
[01:53] <Kinnison> if it's not too hard to, I'd migrate the script
[01:53] <carlos> so I think I will migrate the script only if I see it's needed
[01:54] <Kinnison> okay
[01:56] <SteveA> Kinnison: daniel.silverstone@canonical.com--laptop/launchpad--upload-and-queue--patch-23  <-- does that still need a review from me?
[01:57] <Kinnison> if it's in the queue then ues
[01:58] <SteveA> well, the output on https://chinstrap.ubuntu.com/~jamesh/pending-reviews/ is odd
[01:58] <SteveA> maybe it is already merged
[01:58] <SteveA> or maybe there is some other problem
[01:58] <SteveA> ah -- missing version
[01:58] <SteveA> version 0 ?
[01:58] <Kinnison> 1
[01:59] <Kinnison> oops
[01:59] <Kinnison> :-)
[01:59] <SteveA> okay, i'll change the page
[01:59] <Kinnison> thanks
[01:59] <SteveA> do you need it super urgently, or can i wait for the script to give me a nice diff?
[02:00] <cprov> SteveA: didn't you forget my review from yesterday, did you ?
[02:00] <SteveA> cprov: nope.  doing it now.   it's the builddUI one isn't it
[02:01] <SteveA> celso.providelo@canonical.com/launchpad--builddUI--0
[02:01] <cprov> SteveA: yes, I wonder if you can do the just added builder-action today too
[02:02] <SteveA> okay
[02:02] <cprov> SteveA: you deserve some gift for Montreal, I'm accepting suggestions ;)
[02:05] <SteveA> cprov: get dapper open, and i'll sleep better ;-)
[02:05] <Kinnison> SteveA: mine is blocking me a bit, but I need stub for my next patch
[02:06] <SteveA> jamesh: is there a way to ask the pending-reviews script to run on demand, from chinstrap?
[02:06] <SteveA> Kinnison: i'll get to it soon, then
[02:06] <cprov> SteveA: sorry dude, good sleeping is very rare also at this side of the world ;) anything easier ;) 
[02:07] <matsubara> good morning all
[02:07] <SteveA> you know, the best things about brazil don't travel well: hot beach babes and fresh fresh fruit juice.
[02:08] <SteveA> cprov: what's going to happen about the gina output that stub put on chinstrap?
[02:08] <SteveA> matsubara: good morning
[02:09] <carlos> Kinnison, btw: https://chinstrap.ubuntu.com/~dsilvers/paste/fileavhAhj.html
[02:09] <Kinnison> carlos: "upstream" ?
[02:09] <cprov> SteveA: ohh man do you facinated by hot beach babes ?!
[02:10] <cprov> SteveA: honestly I don't know much about gina nowadays ... was I little bit far from the last patches, ask Kinnison
[02:10] <carlos> Kinnison, you should set ispublished=True always as your tarball comes from 'upstream', in this case, a published sourcepackage
[02:10] <sabdfl> carlos: how's it going today?
[02:10] <sabdfl> what's the focus du jour?
[02:10] <Kinnison> carlos: okay
[02:11] <SteveA> Kinnison: do you need any assistance looking through the gina issues later?
[02:11] <carlos> Kinnison, ignore the importer argument, I will set it to rosetta admins if it's none and it's the same we need with those imports
[02:11] <carlos> sabdfl, hi
[02:11] <Kinnison> SteveA: It'd be handy
[02:12] <SteveA> Kinnison: okay, i'll start looking at the output when i've done these reviews.
[02:12] <SteveA> hopefully i'll be able to give you some more succinct information to work with
[02:12] <Kinnison> SteveA: I'd love for someone else to do it, but it seems, like everything else critical right now, everyone stuck their fingers in their ears and went "lalalala" when I asked to not be the only person who knows how it works
[02:12] <carlos> sabdfl, I'm working on the new import queue we talked to prevent the review-* templates to happen again
[02:13] <carlos> sabdfl, and at the same time I'm preparing my branch to get the translation tarballs directly from the builders
[02:13] <SteveA> Kinnison: i think kiko has a fair idea how it works too
[02:15] <Kinnison> SteveA: if kiko and can do it and leave me free to get on with the uploader I'd appreciate it
[02:15] <SteveA> Kinnison: i'll look at it, and let you know what needs to happen.
[02:15] <Kinnison> thanks
[02:15] <jordi> carlos: I'll be working on the review-breezy templates over the weekend.
[02:15] <SteveA> if we can fix it without much of your input, all the better.
[02:16] <jordi> carlos: do you think you can give me a quick howto of what I need to do now that I have access?
[02:18] <carlos> jordi, do you have rights to reach the potemplatenames page, right?
[02:20] <jordi> yes
[02:25] <carlos> jordi, ok
[02:26] <carlos> jordi, then, it's just a matter of going thru all the review-* templates, look at the .pot filename (you can see it from the +admin page)
[02:26] <matsubara> BjornT: ping
[02:27] <BjornT> hi matsubara 
[02:27] <carlos> and rename its potemplatename to that filename unless you know the translationdomain is not that one or it's something like 'template.pot' in which case you need to investigate it a bit more to get the right translation domain
[02:27] <matsubara> Hello BjornT, do you have time to talk about bug 759?
[02:27] <Ubugtu> Malone bug #759: re-duping an already duped bug causes system error. Fix req. for: malone (upstream), Severity: Normal, Assigned to: Diogo Matsubara, Status: Accepted http://launchpad.net/malone/bugs/759
[02:28] <BjornT> matsubara: sure
[02:28] <carlos> jordi, sometimes, instead of rename, you need to change the potemplatename link because you have already that translation domain in our database
[02:32] <sabdfl> carlos: ok, sounds good
[02:32] <sabdfl> is the idea, if we don't know the name of the potemplate, to put it into a queue
[02:32] <sabdfl> then through the web be able to connect the template to the existing one, or a new one, and have it imported
[02:32] <sabdfl> ?
[02:32] <stub> Kinnison: There is now chinstrap:~stub/gina-logs/warty-verbose.out, which contains all the debug output. It doesn't seem to help much though.
[02:32] <carlos> sabdfl, yeah, more or less
[02:33] <sabdfl> carlos: run me through the plan?
[02:33] <Kinnison> stub: urgh
[02:33] <carlos> sabdfl, with this kind of queue we will be able to support too the zope layout or any package that have the .pot and .po files in different directories
[02:34] <carlos> sabdfl, we have a queue where all files (.po and .pot) will be 'attached'
[02:34] <sabdfl> all files? or just the ones we could not deal with?
[02:34] <Kinnison> SteveA: can you see what you can work out?
[02:34] <carlos> that's done by Kinnison scripts after the build is done
[02:34] <carlos> sabdfl, all
[02:35] <SteveA> Kinnison: yes, i will
[02:36] <carlos> sabdfl, then, we have a script that merges the files from the queue into the final POTemplate/POFiles if there is no pending imports (this solve the problem that one upload overwrites the previous one if the import is not done)
[02:36] <jordi> carlos: good.
[02:36] <jordi> I'll get started on this over the weekend.
[02:36] <jordi> I still have no internet at the new flat. This sucks.
[02:36] <SteveA> Kinnison: i'm planning to do another cprov review and your review first, unless you tell me to go look at gina first.
[02:37] <Kinnison> SteveA: so long as gina gets looked at before stub goes to bed so if a fix is proposed he can try it out, I don't mind
[02:37] <Kinnison> SteveA: but right now, gina is worrying me
[02:38] <SteveA> cprov: can your builder-action review wait a few hours?
[02:38] <cprov> SteveA:  yes, in benefit of gina, of course 
[02:38] <SteveA> Kinnison: okay, i'm switching tracks to go look at gina.
[02:39] <Kinnison> cool
[02:39] <sabdfl> carlos: ok, so instead of attaching it to POFile or POTemplate, it always goes into a queue?
[02:39] <SteveA> has gina been hacked on lately?
[02:39] <carlos> sabdfl, and then, we will have a web UI that will show that queue and we will be able to set some imports to 'ignore' (this solve the problem we have with gtk that has the same .pot file in many different directories depending on the kind of build we do) or to associate manually the .po/.pot files we were not able to automatically associate with already existing POTemplate/POFiles with already existing objects or create new ones
[02:39] <carlos> sabdfl, yes
[02:39] <sabdfl> the point of this, though, was to deal with files where we did not know which POTemplate or POFile to attache it to
[02:39] <stub> SteveA: Changes landed yesterday
[02:39] <carlos> that way we can also show to the user the queue status
[02:40] <SteveA> stub: where?  production or RF or both?
[02:40] <sabdfl> carlos: is there a spc i can look at, please?
[02:40] <stub> rf
[02:40] <SteveA> cool
[02:40] <sabdfl> i'm a little concerned it will not be usable
[02:40] <SteveA> stub: do you have any ideas about the constraint issues?
[02:41] <carlos> sabdfl, not yet, I'm doing both things at the same time. but I should have the spec done later today
[02:41] <stub> I think the code that says 'does this BinaryPackageRelease' exist is broken
[02:42] <sabdfl> stub: all of that old code is horribly busted
[02:42] <carlos> sabdfl, it will not be usable? why?
[02:42] <sabdfl> i'm writing chunks of new broken code to replace most of it now
[02:42] <SteveA> with tests this time?
[02:43] <sabdfl> carlos: too many things in the queue, too little information to go by i.t.o. fixing the issues
[02:43] <sabdfl> let me see the spec, and i'll comment, i'm sure its workable
[02:45] <carlos> sabdfl, ok, let me concentrate on the spec and stop the development and will send it back to you 
[02:45] <SteveA> mpt: hello
[02:45] <stub> I'm falling asleep here. If you guys need to trial Gina, can you get elmo to give you access to launchpad@asuka and run it against the staging database? The database is all setup to accept connections as the gina user and there are suitable launchpad.conf files there already.
[02:45] <SteveA> stub: can you mail me instructions or where to find instructions?
[02:45] <stub> ok
[02:46] <SteveA> stub: bear in mind, i've never run gina
[02:46] <carlos> sabdfl, in the mean time: https://chinstrap.ubuntu.com/~dsilvers/paste/fileX0JeF7.html <- That's the DB schema I'm using atm
[02:46] <mpt> hi SteveA
[02:46] <SteveA> stub: is it safe to run gina multiple times?
[02:46] <SteveA> mpt: so, the breadcrumbs stuff so far has landed
[02:46] <mpt> great
[02:46] <SteveA> mpt: i'd like you to familiarise yourself with how to add breadcrumbs to things
[02:46] <SteveA> it is very simple
[02:47] <stub> SteveA: It had better be since we already have been
[02:47] <SteveA> when i get a respite from code review / debugging with Kinnison and cprov, i'll be doing the trailing > and not-link-to-current-page features.
[02:47] <SteveA> mpt: also, how's the menu coverage stuff going?
[02:48] <SteveA> stub: is there a way to tell it to just try one particular thing?
[02:49] <mpt> SteveA: I'm going to start the menu checking when launchpad--design-fascism--0 lands, because otherwise the cleanup I'd be doing at the same time as checking the menus would be conflicting with the major cleanup I did in that branch
[02:49] <mpt> though yesterday I did tidy up the person menus a bit
[02:49] <SteveA> mpt: when will the design fascism land?
[02:49] <stub> SteveA: You can tell it to just import warty, or warty-updates, or warty-security etc.
[02:49] <mpt> SteveA: when kiko-zzz finishes reviewing it :-)
[02:49] <SteveA> mpt: UBZ is drawing close, and i want us to have full menu coverage by then
[02:50] <mpt> fair enough
[02:50] <SteveA> stub: okay, but not more granular than that?
[02:52] <stub> SteveA: Nope
[02:53] <SteveA> okay
[02:55] <stub> I'm off to bed. I can look at it in 8 hours if people want. But I won't be around Sunday so I should try and finish off whatever needs doing on Saturday.
[02:56] <SteveA> stub: i'll be around somewhat tomorrow
[02:56] <SteveA> so, i'll ping you then.
[02:56] <SteveA> or you ping me or whatever
[02:57] <stub> ok
[02:59] <sabdfl> stub: sec
[02:59] <sabdfl> what's the tag point?
[03:00] <sabdfl> hey mpt
[03:00] <sabdfl> what's in design-fascism?
[03:00] <mpt> hi sabdfl, congratulations on breezy
[03:00] <carlos> stub, please, could you remove again the cache that prevents to import some pofiles on production? we have 122 files that are not imported and I need to know if it's because that cache or because a bug in our import code
[03:00] <sabdfl> ah. thanks, credit to #-devel mostly :-)
[03:00] <mpt> sabdfl: lots of markup and text fixes
[03:00] <sabdfl> cool
[03:00] <sabdfl> landing soon?
[03:00] <mpt> Making explanations tighter and more explanatory
[03:00] <mpt> yes, I'm pestering kiko about it every day
[03:00] <stub> carlos: eh?
[03:01] <sabdfl> i want to do a big weekend push and land polish monday for stub to roll out tuesday
[03:01] <sabdfl> that's why i;m hoping stub will reveal his tag point plans :-)
[03:01] <stub> sabdfl: patch-2654
[03:01] <sabdfl> stub: marvellous, thanks
[03:01] <stub> sabdfl: It should be in your email ;)
[03:01] <sabdfl> ok. my mail was a little scary this morning so i've just been coding
[03:02] <carlos> stub, I'm talking about the code you added to ignore, until next day, the import of a file if it fails 
[03:02] <sabdfl> stub: i have duplicated the distribution source package cache with a distrorelease binary package cache. ok?
[03:02] <sabdfl> in -41-0
[03:02] <sabdfl> with comments
[03:02] <stub> carlos: oh. ok. It is automatically removed every day but I'll do it manually now.
[03:02] <stub> sabdfl: ok
[03:02] <carlos> stub, hmmm, ok
[03:02] <SteveA> salgado: any idea where kiko is today?
[03:03] <sabdfl> stub: thanks muchly
[03:03] <sabdfl> SteveA: brazil
[03:03] <salgado> SteveA, he must be coming shortly
[03:03] <SteveA> sabdfl: that's a relief
[03:04] <stub> carlos: removed
[03:04] <carlos> stub, thanks
[03:04] <SteveA> stub: for future reference, where is it?
[03:04] <stub> carlos: scheduled removal is 10 past midnight
[03:04] <carlos> stub, BST?
[03:04] <stub> SteveA: /var/tmp/rosetta-poimport-seen.pickle - it is in the crontab
[03:04] <SteveA> .pickle (!)
[03:05] <stub> carlos: yes. or whatever dodgy timezone elmo runs the servers in
[03:05] <niemeyer> tmp!
[03:05] <carlos> ok
[03:05] <niemeyer> :)
[03:05] <stub> SteveA: yeah - yeah. it is a hack.
[03:06] <SteveA> good night stub 
[03:06] <kiko-zzz> kiko
[03:06] <kiko-zzz> it 
[03:06] <kiko-zzz> hurt!
[03:06] <kiko> man it is hot today
[03:07] <sabdfl> carlos: what are ignore and ispublished?
[03:08] <mpt> ooh
[03:08] <mpt> PTRuntimeError
[03:08] <mpt> A server error occurred. 
[03:09] <mpt> why do I get that instead of a normal traceback, SteveA?
[03:09] <mpt> it was just an HTML error
[03:10] <SteveA> when did you get that error?
[03:10] <carlos> sabdfl, 'ispublished' is to know if the upload is or not published (like the pofile.published field). I'm planning to let users to do the initial .po and .pot upload for productseries too so we only need to handle the request from that queue and jordi does not needs to look for the po files, create the tarball and do the upload by hand
[03:10] <SteveA> when rendering the page, or when starting the server?
[03:10] <mpt> when rendering the page
[03:10] <carlos> sabdfl, that way we introduce that kind of uploads in our workflow and people don't need to send any email, the system will tell us when we have work to do
[03:10] <SteveA> probably, you made an error so that the page was so far outside of parseable html / xml that the page template machinery could no longer make sense of it
[03:11] <mpt> ah
[03:11] <SteveA> unbalanced <div> for example
[03:11] <mpt> All I did was leave out the end " for an attribute
[03:11] <SteveA> ah
[03:11] <SteveA> that's kinda gross
[03:11] <SteveA> from a parsing perspective
[03:11] <SteveA> do you use syntax highlighting?
[03:11] <mpt> unbalanced <div>s I get told off for explicitly
[03:11] <mpt> yes, I do
[03:13] <carlos> sabdfl, about 'ignore', there are sometimes when we get the same .pot file more than once inside a .deb's source package, it's just a way to mark it as something that we are not interested on so next time we see it, we don't need to expend time with it. That way we don't need to have POTemplate objects hidden or 'disabled', we just ignore them before the import happens
[03:13] <carlos> sabdfl, I'm explaining all things in the spec
[03:21] <SteveA> mpt: well, we could add a nice error page for it, but i don't really see the point
[03:22] <mpt> SteveA: No, no worries, I was just wondering if it was a symptom of something more serious
[03:22] <SteveA> ddaa: i wonder if it's possible to get a bzr tree of zope3?  just wondering, because their svn server just went down again...
[03:23] <ddaa> well, do not even have a baz import of it
[03:23] <kiko> that's a small tragedy ddaa 
[03:23] <kiko> but we do have zope3 imported in rocketfuel :)
[03:23] <trax> is there an easy way to view all the bugs you have reported in launchpad?
[03:23] <SteveA> hmm... it uses svn:externals a lot
[03:23] <SteveA> kiko: no history to speak of
[03:24] <SteveA> trax: yes
[03:24] <kiko> that is true
[03:24] <SteveA> trax: go to your own page on launchpad
[03:24] <trax> done
[03:24] <SteveA> go to the "bugs" tab
[03:24] <kiko> what is /up/ with my firefox fonts
[03:24] <trax> those are only for bugs assigned to you
[03:24] <trax> it says "No bugs assigned to ...."
[03:24] <SteveA> there's a link to see that
[03:24] <ddaa> SteveA: what's the magic to get action menus?
[03:24] <trax> aaah!
[03:24] <trax> thanks!
[03:24] <trax> I missed that
[03:25] <SteveA> ddaa: the magic to get them?
[03:25] <ddaa> I mean, to get them on the .pt
[03:25] <SteveA> ddaa: you need to write an action menu, and register it in zcml
[03:25] <ddaa> I removed the section about portlet-action
[03:25] <Kinnison> Why do we set nowrap on the 'target' of our bug listings?
[03:25] <ddaa> I thought that the @@+portlet-action should die...
[03:25] <SteveA> ddaa: yes
[03:25] <kiko> portlet-actions must die
[03:26] <kiko> alo gneuman 
[03:26] <SteveA> kiko: can you help ddaa out please?  i'm supposed to be scrutinizing gina output
[03:26] <kiko> sure
[03:26] <kiko> sure I can
[03:26] <SteveA> thanks man
[03:26] <kiko> how may I help you ddaa 
[03:26] <ddaa> probably... checking out the LaunchpadMenus spec again
[03:26] <SteveA> ddaa: that's a tad out of date
[03:27] <SteveA> sorry
[03:27] <mpt> ddaa: see LaunchpadMenusInProgress
[03:27] <mpt> at the bottom
[03:27] <gneuman> alo
[03:27] <gneuman> kiko
[03:27] <mpt> but yeah, the spec needs finishing up now
[03:27] <SteveA> yeah, my job, i guess
[03:29] <kiko> Duplicates of this bug:
[03:29] <kiko>     * Bug #1975
[03:29] <kiko>     * Bug #2185
[03:29] <kiko>     * Bug #2200
[03:29] <kiko>     * Bug #2414
[03:29] <kiko>     * Bug #2749
[03:29] <Ubugtu> Error: I don't have a bugzilla Bug.
[03:29] <kiko>     * Bug #2921
[03:29] <Ubugtu> Error: I don't have a bugzilla Bug.
[03:29] <kiko>     * Bug #3146
[03:29] <Ubugtu> Error: I don't have a bugzilla Bug.
[03:29] <kiko> maybe it's time we fixed bug 1419, or at least worked around it
[03:29] <Ubugtu> Malone bug #1419: Rosetta fails to locate second_lang_msgset message ID when changing alternate language.  Fix req. for: rosetta (upstream), Severity: Major, Assigned to: Nobody, Status: Accepted http://launchpad.net/malone/bugs/1419
[03:29] <kiko> carlos?
[03:29] <mpt> Seveas?
[03:30] <Seveas> meh, what broke this time
[03:30] <Seveas> (didn't get to complete the rewrite yet)
[03:30] <ddaa> kiko: product-index.pt overrides portlets_two, that's the cause of my confusion I think
[03:30] <ddaa> kiko: do you think that sholud be fixed?
[03:31] <Seveas> Bug 1419
[03:31] <Ubugtu> Error: I don't have a bugzilla Bug.
[03:31] <Seveas> bug 1419
[03:31] <SteveA> Seveas: maybe you want to try out bzr for ubugtu?
[03:31] <Ubugtu> Malone bug #1419: Rosetta fails to locate second_lang_msgset message ID when changing alternate language.  Fix req. for: rosetta (upstream), Severity: Major, Assigned to: Nobody, Status: Accepted http://launchpad.net/malone/bugs/1419
[03:31] <Seveas> right -- case sensitive crud
[03:31] <kiko> SteveA, is there a join-like pattern in TAL?
[03:31] <ddaa> mh.
[03:31] <SteveA> kiko: what does that mean?
[03:31] <ddaa> No... I'm confused
[03:31] <Seveas> SteveA, hehe, I'm just reading baz/tla/arch docs and have not yet setup an archive :)
[03:32] <kiko> ",".join("[foo", "bar"] )
[03:32] <carlos> kiko, ?
[03:32] <SteveA> Seveas: nooo... look up bzr.  jblack can help you
[03:32] <kiko> SteveA, to avoid the bug I'm seeing at https://launchpad.net/products/launchpad/+spec/shipit-workflow
[03:32] <kiko> SteveA, where the sprints end with a comma
[03:32] <ddaa> Well... I was confused but right.
[03:32] <kiko> carlos, can you hack a workaround for bug 1419 today?
[03:32] <Ubugtu> Malone bug #1419: Rosetta fails to locate second_lang_msgset message ID when changing alternate language.  Fix req. for: rosetta (upstream), Severity: Major, Assigned to: Nobody, Status: Accepted http://launchpad.net/malone/bugs/1419
[03:32] <carlos> kiko, today I'm a bit busy to have the new attachment infrastructure ready for dapper import next week
[03:32] <kiko> hmmm, right
[03:32] <kiko> ok
[03:33] <kiko> carlos, do you have  a  hint for a workaround at least?
[03:33] <kiko> the problem doesn't seem to be very difficult to solve
[03:33] <SteveA> kiko: there is, but this is better done in the view class.
[03:33] <kiko> SteveA, any hint?
[03:33] <Seveas> !reload Bugzilla
[03:33] <Ubugtu> The operation succeeded.
[03:33] <Seveas> Bug 1419
[03:33] <Ubugtu> Malone bug #1419: Rosetta fails to locate second_lang_msgset message ID when changing alternate language.  Fix req. for: rosetta (upstream), Severity: Major, Assigned to: Nobody, Status: Accepted http://launchpad.net/malone/bugs/1419
[03:33] <SteveA> kiko: mpt knows about the sequence/name/last and all that
[03:33] <kiko> SteveA, but then the view class would have to return <a href>s too?
[03:33] <SteveA> yes
[03:33] <SteveA> so?
[03:33] <Seveas> SteveA, what's so great about bzr compared to baz?
[03:34] <kiko> SteveA, it cramps my style
[03:34] <SteveA> faster, better ui, better merging, IT IS THE FUTURE
[03:34] <Seveas> hehe
[03:34] <kiko> and I quote "IT IS THE FUTURE"!
[03:34] <trax> is bzr bazaar-ng yet, or is that still going to happen?
[03:34] <SteveA> kiko: okay, you can do it in tales.  ask mpt about the 'sequence' thing
[03:34] <trax> i've tried bzr, but I prefer mercurial
[03:34] <Seveas> is bzr included in Ubuntu?
[03:34] <kiko> Seveas, batteries and nuclear warheads included
[03:34] <SteveA> trax: when did you last try it -- loads of work on it recently
[03:35] <carlos> kiko, hmm, I think I did something about that problem already...
[03:35] <SteveA> we're converting development of launchpad to use bzr in a few weeks
[03:35] <trax> SteveA: that was about 3 months or so ago, maybe I should take a look again
[03:35] <kiko> carlos, not in production you didn't :)
[03:36] <trax> SteveA: is this bazaar-ng or bzr we're talking about?
[03:36] <carlos> kiko, no, it's not on production
[03:36] <SteveA> trax: wow... loads has changed over the last three months.  bazaar-ng is bzr
[03:36] <mpt> trax: yes :-)
[03:36] <carlos> kiko, but the fix was not too hard
[03:36] <kiko> carlos, want to send me a diff? I'll handle the rest
[03:36] <carlos> let me check if it's inside my language pack branch
[03:36] <kiko> please
[03:36] <BjornT> sabdfl: who should be able to edit a support ticket? (at the moment it seems that not everyone can)
[03:36] <trax> SteveA: how does bzr compare to mercurial?
[03:37] <SteveA> trax: they're now using a neat 'weave' format at the back end, sftp support arriving soon... pop over to #bzr sometime perhaps
[03:38] <SteveA> jblack can compare bzr to hg.  i haven't used hg, just read about it.
[03:38] <trax> SteveA: weave?
[03:38] <kiko> weaves are also the future
[03:39] <ddaa> trax: roughly feature equivalent (no real push in bzr though), slower but the gap is closing
[03:39] <ddaa> recent assessment by jblack gave hg the speed edge
[03:40] <kiko> ddaa, are your problems sorted?
[03:40] <sabdfl> BjornT: i'm happy to let AnyPerson edit a ticket
[03:41] <ddaa> kiko: my problem is: how do I add a context menu to the product-index.pt as this page overrides the overrides portlets_two
[03:41] <ddaa> I'm working on adding a context menu to Branch, not blocked there yet
[03:42] <SteveA> ddaa: you just register context menus and application menus.  you don't need to explicitly include them in your page templates.
[03:42] <SteveA> ddaa: you can still use other portlets.
[03:42] <SteveA> ddaa: it just works.
[03:42] <BjornT> sabdfl: ok, i'll change the permissions then.
[03:42] <sabdfl> thanks!
[03:42] <kiko> ddaa, does what SteveA said make sense to you?
[03:42] <ddaa> SteveA: I must be confused, because my product page does not have the "add branch" link...
[03:43] <SteveA> ddaa: did you register the menu in zcml?
[03:44] <ddaa> I added ProductCodeMenu between ProductsBugsMenu and ProductSetContextMenu in the product.zcml
[03:44] <mpt> BjornT: mpt@canonical.com/launchpad--translation-form--0510 has just finished mirroring with conflicts resolved
[03:45] <BjornT> mpt: cool, i'll review it later on today
[03:45] <mpt> thanks
[03:48] <SteveA> ddaa: tell me a branch to look at perhaps
[03:49] <ddaa> okay, I'll commit the relevent changes
[03:49] <mpt> BjornT: https://launchpad.net/products/launchpad/+bug/3148 seems like your field of expertise
[03:49] <Ubugtu> Malone bug #3148: Strip hard-wrapping newlines from e-mail only if it's actually e-mail Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Nobody, Status: New http://launchpad.net/malone/bugs/3148
[03:49] <carlos> kiko, yeah, I have a patch
[03:49] <SteveA> mpt: thre
[03:50] <mpt> SteveA: fou?
[03:50] <kiko> carlos, send it to meex
[03:50] <kiko> or give me a patch id
[03:50] <SteveA> mpt: there's a zope3 discussion about making page templates use xhtml only (or other xml... but not non-xml html)
[03:50] <carlos> kiko, I'm on that, just waiting for this fucking slow connection...
[03:50] <SteveA> mpt: do you see any problem for us with that?
[03:50] <carlos> I cannot believe that we were using 56K modems some years ago...
[03:51] <mpt> SteveA: Apart from the fact that the browser used by ~90 percent of people doesn't support XHTML?
[03:51] <carlos> kiko, https://chinstrap.ubuntu.com/~dsilvers/paste/filedIRRHR.html
[03:51] <mpt> then again, that's not one of our target browsers ;-)
[03:51] <carlos> kiko, please, ignore the changes to the imports sections
[03:52] <carlos> kiko, in fact... I think I should be able to merge that branch directly
[03:52] <kiko> carlos, can you explain why the second language wouldn't have that message id? isn't the potfile the same?
[03:52] <carlos> kiko, it's reviewed but I had a weird error running tests and I was not sure if it was my fault or one of those random errors that we see from time to time
[03:53] <kiko> carlos, I can handle it for you so you don't loose any more time, really
[03:53] <carlos> kiko, because it's a pomsgset and we only create it if we have a translation
[03:53] <carlos> kiko, I can request a merge now 
[03:53] <carlos> and if it fails, ask you to merge that small patch
[03:54] <carlos> kiko, for instance, we import a pofile
[03:54] <carlos> and then a potemplate that adds a new potmsgset
[03:54] <carlos> that old pofile will not have a pomsgset related to that new potmsgset unless someone adds a translation
[03:55] <kiko> carlos, understood, I hadn't realized it was a pomsgset
[03:55] <SteveA> mpt: i think it supports xhtml 1 enough.  the point is, the document must be well-formed XML.
[03:55] <mpt> SteveA: XHTML (and XML in general) is much more brittle than HTML, such that minor escaping bugs could easily cause an unreadable front page
[03:55] <mpt> IE6 doesn't support XHTML at all
[03:55] <SteveA> that's what tests are for
[03:55] <ddaa> david.allouche@canonical.com--2004/launchpad--branchdatastorage--0--patch-5
[03:55] <ddaa> http://localhost:8085/products/thunderbird
[03:56] <SteveA> but, we don't have to tell the browser it is xhtml
[03:56] <mpt> SteveA: We don't, and probably never will have, perfect test coverage
[03:56] <carlos> kiko, it's normal, I'm renaming many variables to note if it's a potmsgset or a pomsgset to make it clear
[03:56] <SteveA> the point is, we need to write our pages in xhtml
[03:56] <SteveA> we can tell the browser anything we like
[03:56] <SteveA> the point is to have page template processing require xhtml as input
[03:56] <SteveA> and produce valid xml as output
[03:56] <SteveA> but we can put whatever doctype etc. we want on it
[03:57] <mpt> SteveA: ok, if you're sending XHTML-as-HTML, then that's faintly silly, but I have no objection if it makes parsing etc easier
[03:57] <kiko> carlos, cool, thanks
[03:57] <BjornT> mpt: hmm, that's a though one. i'll add some comments to the bug later. it will have some implications to the formatting of outgoing emails as well.
[03:58] <SteveA> Kinnison: i have a gina-debuggery question for you
[03:58] <Kinnison> go on.
[03:59] <SteveA> Kinnison: i want to check all the .dsc files in the archive to see if there is spurious stuff in them.  can i do this on some machine somewhere?
[03:59] <carlos> kiko, merge request done
[03:59] <SteveA> or do i need to go over http?
[04:00] <Kinnison> You want to look at the katie-produced-archive?
[04:00] <Kinnison> if so, look on mawson in /srv/archive.ubuntu.com/ubuntu/
[04:00] <SteveA> Kinnison:  i want to look at whatever gina is using as input data
[04:00] <Kinnison> check it out on mawson then
[04:01] <SteveA> ta
[04:04] <SteveA> Kinnison: is that just breezy on there?
[04:05] <Kinnison> it should have the full pool
[04:05] <Kinnison> it may only be source+i386 though
[04:05] <Kinnison> dunno
[04:06] <SteveA> where would i find hoary and breezy?
[04:07] <Kinnison> the dists?
[04:07] <Kinnison> in dists/
[04:08] <SteveA> argh... they're talking about making you add explicit xmlns declarations to all page templates and fragments thereof
[04:09] <Seveas> hmm
[04:09] <Seveas> bzr IS easier than baz....
[04:09] <Seveas> thanks SteveA and all other bzr cheerleaders :)
[04:11] <niemeyer> ddaa: Where do we get the owner for new objects in taxi?
[04:11] <niemeyer> ddaa: Hello, btw :)
[04:11] <ddaa> mh
[04:12] <ddaa> I think the owner should be the owner of the corresponding product.
[04:12] <Kinnison> what's the "now" constant we should be using
[04:12] <Kinnison> ?
[04:12] <ddaa> niemeyer: but TBH I do not have a good answer for that, maybe the owner should be the Lanuchpad Admin team...
[04:13] <niemeyer> ddaa: Ok, 1 sounds like a good value for now then.. :)
[04:13] <ddaa> since arguably, if a product is first assigned to Alice, then reassigned to Bob, there's no reason that the rcs-imported branches and revision still belong to Alice...
[04:14] <ddaa> niemeyer: please make it 25 (launchpad admins)
[04:14] <niemeyer> Done
[04:14] <niemeyer> ddaa: Indeed.. the whole concept of branch owner seems strange.
[04:15] <ddaa> niemeyer: it makes sense in the webapp stuff I'm doing. When you can register a third-party branch.
[04:15] <carlos> see you later
[04:15] <kiko> carlos, rock on!
[04:15] <kiko> thanks dude
[04:15] <carlos> see you later
[04:15] <kiko> I appreciate it
[04:15] <niemeyer> ddaa: If I register your branch, is it mine?
[04:15] <carlos> will be back in an hour or so
[04:15] <ddaa> But the launchpad owner concept is awkyard for stuff that is grabbed from out there... that's why branches have "owner" and "author".
[04:16] <carlos> kiko, np
[04:16] <niemeyer> ddaa: Understood
[04:16] <ddaa> If you register my branch, it's owner=you, then you can set author=me.
[04:17] <ddaa> then the page does various sorts of magic to collate ownership and authorship.
[04:18] <ddaa> Actually, I'm starting to think that revision.owner should be dropped.
[04:18] <niemeyer> :-)
[04:18] <ddaa> it's not a useful granurality
[04:18] <ddaa> lifeless: ping?
[04:18] <niemeyer> Having it always == 25 doesn't seem useful indeed. :)
[04:19] <niemeyer> Perhaps we should change to 42?
[04:19] <ddaa> With branches, it makes sense.
[04:19] <ddaa> As when you register a third-party branch, the owner will be set to you.
[04:19] <niemeyer> I'm hacking on the revision code now (not sure if I mentioned it)
[04:19] <ddaa> But the revisions will be populated by taxi anyway.
[04:20] <ddaa> niemeyer: bah, unless your turning it upside down, we can handle the potential conflicts.
[04:21] <niemeyer> ddaa: On taxi.py? Yes, turning it upside down.
[04:21] <ddaa> No, I mean revision.py
[04:21] <niemeyer> Not teaching it (uff)
[04:22] <ddaa> Please turn this taxi bitch upside-down, inside-out and teach it how to live.
[04:22] <niemeyer> Hehehe :)
[04:27] <bradb_> ddaa: Do you have a couple mins to help me with porting a critical fix to mainline (from a cherry-pick-compatible branch that got rolled out yesterday)?
[04:28] <ddaa> bradb: how can I help you?
[04:28] <bradb> ddaa: so, i made a branch of prod yesterday to make sure that my patch was cherry pickable...
[04:29] <bradb> (particularly because it changed sample data, at the same time as other recent patches have also done so, etc.)
[04:29] <bradb> bradb@oxygen:~ $ baz branch $rocketfuel launchpad--malone-priority-fix--0
[04:29] <bradb> bradb@oxygen:~ $ baz get --link launchpad--malone-priority-fix--0 malone-priority-fix-for-mainline
[04:29] <bradb> bradb@oxygen:~/malone-priority-fix-for-mainline $ baz merge launchpad--malone-priority-cherry-pickable-fix--0
[04:29] <ddaa> please give me the full name of your branch
[04:30] <bradb> ddaa: brad.bollenbach@canonical.com/launchpad--malone-priority-fix--0
[04:30] <bradb> when merging, it hangs:
[04:30] <bradb> ...
[04:30] <bradb> * Scanning for full-tree revision ............................................................................................................................................. done.
[04:30] <bradb> * from archive cached: christian.reis@canonical.com--lozenge/launchpad--devel--0--patch-86
[04:30] <bradb> ...
[04:30] <bradb> there
[04:31] <ddaa> mh, it's not hanging
[04:31] <ddaa> it's just being veeeeeeeery slow
[04:31] <niemeyer> As usual :))
[04:31] <bradb> ddaa: it appears to be haning
[04:31] <bradb> hanging even
[04:31] <bradb> no disk space being taken up or anything
[04:31] <niemeyer> bradb: Have you waited for a few hours already? :)
[04:32] <ddaa> bradb: apparently your branch is not up on chinstrap
[04:32] <bradb> ddaa: should i go with a reverse replay approach instead?
[04:32] <bradb> ddaa: oh, i guess it isn't
[04:32] <ddaa> cannot tell, I do not know what you are trying to merge...
[04:33] <bradb> ddaa: is it worth me mirroring the branch then for you to explore this further?
[04:33] <ddaa> if you want me to do anything more than vague hand-waving or hip-shooting, I'm afraid that is necessary
[04:33] <SteveA> Kinnison: found one problem with gina that looks easy to fix.
[04:33] <ddaa> wait
[04:34] <bradb> ?
[04:34] <ddaa> you probably want to "uncacherev base-0" from your cherrypick branch
[04:34] <Kinnison> SteveA: cool, can you and stub try it out?
[04:34] <bradb> ddaa: is that just to save diskspace?
[04:34] <SteveA> Kinnison: that is, the tagfiles parser misses out the 'files' if the .dsc file isn't signed, perhaps because it ends with no NL char.
[04:34] <ddaa> so you we will not wait for 15 mins or so why baz uploads a stupidly huge tarball using a stupidly slow network implementation
[04:35] <ddaa> * when baz
[04:35] <SteveA> Kinnison: i'm going to ensure that all dsc files can be parsed properly and then maybe do another run
[04:35] <Kinnison> SteveA: cool
[04:35] <bradb> ddaa: so you mean i should do that for my porting-to-mainline branch, right?
[04:35] <SteveA> Kinnison: is there any point doing another run while there are still DB constraints problems?
[04:35] <bradb> (the "cherry pick branch" is the branch i want to merge into this branch)
[04:35] <ddaa> bradb: from whatever new branch you created by branching from rocketfuel
[04:35] <bradb> right
[04:36] <Kinnison> stevea: anything it fails to insert can clearly not harm things in the future
[04:36] <Kinnison> SteveA: gina is additive
[04:36] <SteveA> Kinnison: is a .dsc file allowed not to be signed?
[04:36] <Kinnison> yes
[04:37] <SteveA> i see
[04:37] <SteveA> okay, i'll ping you when i have news
[04:37] <Kinnison> cool
[04:38] <SteveA> i can't do the gina run until i get sudo access on asuka, btw
[04:38] <SteveA> i have an account there now, but karl is awaiting confirmation that i'm allowed that access
[04:38] <SteveA> do you have access there?
[04:39] <Kinnison> I don't
[04:39] <Kinnison> elmo is in front of me
[04:40] <Kinnison> shall I ask him?
[04:40] <bradb> ddaa: mirrored now
[04:40] <SteveA> Kinnison: if he's going to remain around, i'd say don't disturb him now
[04:40] <Kinnison> okay
[04:40] <SteveA> Kinnison: the "usual channels" will take their course
[04:40] <SteveA> if i have things ready to run again, and still don't have access, i'll ping you
[04:41] <SteveA> it'll be a while debugging and fixing though
[04:41] <SteveA> oh and my indian takeaway has arrived
[04:42] <ddaa> bradb: okay, I see a branch freshly tagged from rocketfuel. I presume you want to cherrypick a patch from another branch.
[04:42] <ddaa> bradb: right?
[04:42] <gneuman> has anybody merged from rf today?
[04:43] <ddaa> bradb: generally speaking cherrypicking is done by "baz replay $REVISION"
[04:43] <bradb> ddaa: i want to port a cherry picked patch into the mainline so (i *think*) i want to merge that branch on which i created the cherry picked patch *into* the freshly tagged rf branch, if you see what i mean
[04:43] <bradb> ddaa: yeah, that's what i thought, seeing as the other alternative appears to hang :/
[04:43] <ddaa> please specify "a cherrypicked patch"
[04:44] <bradb> a patch that was cherry picked into production. made possible because it was a branch off production.
[04:44] <kiko> friggin bugs.kde.org
[04:44] <ddaa> bradb: okay, it makes sense... where is that patch?
[04:45] <bradb> ddaa: i called it brad.bollenbach@canonical.com/launchpad--malone-priority-cherry-pickable-fix--0
[04:46] <kiko> SteveA, is it trivial to add an inline doctest to be run when make check runs?
[04:46] <SteveA> fairly
[04:46] <kiko> SteveA, in particular malone/externalsystem.py
[04:46] <SteveA> look in webapp/tests
[04:47] <kiko> hmmm it's a bit unfortunate that it's dangerous to depend on it
[04:47] <ddaa> bradb: mh... okay, your fix was written on the malone-priority-cherry-pickable-fix branch, right?
[04:47] <SteveA> kiko: we shouldn't have stuff in malone/ any more really
[04:47] <kiko> SteveA, components/remotebugwatch.py ?
[04:47] <bradb> ddaa: yeah, and has been cherry picked, and now i want to get that same code merged into mainline
[04:47] <SteveA> is it adapters / utilities?
[04:47] <SteveA> is it a script?
[04:47] <SteveA> is it script support code?
[04:48] <kiko> it's a library
[04:48] <kiko> support code
[04:48] <SteveA> for use by scripts, or just in general?
[04:48] <ddaa> bradb: baz replay brad.bollenbach@canonical.com/launchpad--malone-priority-cherry-pickable-fix--0--patch-1 (or some equivalent) is the only way to get that patch into rocketfuel without dragging in launchpad--production--1.36--patch-1 to patch-4.
[04:49] <ddaa> merge will... merge... that is carry in all the changes in the ancestry of what you are merging from.
[04:49] <bradb> ddaa: i just did a get-changeset/apply-changeset here. should that be an ok alternative?
[04:49] <ddaa> that's exactly what replay REVISION does
[04:49] <bradb> cool
[04:50] <bradb> conflicts in the sampledata, unsurprisingly
[04:50] <bradb> ddaa: thanks for clarifying. i'm think i'm back on the golden path.
[04:50] <ddaa> yellow brick road?
[04:51] <bradb> more or less
[04:51] <Kinnison> SteveA: OOI what file is that?
[04:52] <SteveA> lib/canonical/archivepublisher/tagfiles.py
[04:52] <ddaa> this_variable_is_lying = 42
[04:52] <bradb> true = False
[04:52] <Kinnison> SteveA: right, for the most part that's taken direct from james' tagfile parser
[04:52] <SteveA> mr troup?
[04:52] <Kinnison> yes
[04:53] <Kinnison> I have a patch to allow unsigned files sat in my branch here
[04:53] <SteveA> that's what's currently causing breezy's gina run to die
[04:53] <SteveA> oh?
[04:53] <SteveA> well, maybe i can get it?
[04:53] <Kinnison> I'll do a file-diff and nopaste it, one sec
[04:53] <SteveA> ta
[04:53] <Kinnison> you can try it
[04:54] <Kinnison> If you think it's a lack of trailing newline then my patch won't fix it
[04:54] <SteveA> i'll still add comments and improve variable names i think
[04:55] <SteveA> or, just nopaste the whole file
[04:55] <SteveA> and i'll diff
[04:56] <Kinnison> https://chinstrap.ubuntu.com/~dsilvers/paste/fileOdmqnl.html
[04:56] <Kinnison> that's my patch
[04:56] <Kinnison> I.E. really teeny
[04:56] <SteveA> ta
[04:56] <mpt> carlos: ping
[04:57] <mpt> oh, he's not here
[04:57] <Kinnison> I'm not sure how the dsc being moosed is breaking the binary adder though
[04:57] <mpt> serves me right for not using tab-completion
[04:58] <bradb> ddaa: strangely, there appear to be no diff markers in the conflicted files. maybe that's a feature unique to {get,apply}-changeset. "feature".
[04:58] <SteveA> Kinnison: i hope you don't say 'moosed' at UBZ
[04:58] <SteveA> bradb: yes
[04:59] <ddaa> bradb: this works using diff-patch, you have rej file
[04:59] <SteveA> bradb: it works differently
[04:59] <ddaa> bradb: you can use apply-delta do do cherrypicks with diff3
[05:00] <ddaa> like "baz apply-delta --three-way $BRANCH--base-0 $BRANCH--patch-1"
[05:01] <SteveA> I read that as "moose"
[05:01] <SteveA> ewww
[05:01] <mpt> mmmmm, mousse
[05:01] <bradb> ouch
[05:01] <Kinnison> SteveA: you sick sick man
[05:04] <SteveA> Kinnison: that's the badger.
[05:04] <sivang> Kinnison: what does "moosed" mean ? :)
[05:04] <SteveA> all .dscs can be read now
[05:04] <Kinnison> SteveA: rock on!
[05:04] <Kinnison> sivang: broken
[05:05] <elmo> SteveA: done
[05:05] <sivang> Kinnison: ah
[05:06] <SteveA> thanks elmo !
[05:06] <SteveA> elmo: you can probably kick me off it after the weekend 
[05:16] <ddaa> I have something that reads
[05:16] <ddaa>       <tr tal:repeat="rev_no context/latest_revisions">
[05:16] <ddaa> 	<td tal:content="rev_no/rev_no">1</td>
[05:16] <ddaa> but I get NotFoundError: (<RevisionNumber at 0x42561eec>, 'rev_no')
[05:17] <ddaa> I sort of guess I forgot to setup some security-proxying stuff to allow accessing the attribute
[05:17] <ddaa> (it's properly there in the interface)
[05:17] <SteveA> ddaa: what kind of objects does latest_revisions give to you?
[05:18] <SteveA> ddaa: do you have a <content directive in zcml for RevisionNumber?
[05:18] <ddaa> Ha... that must be it...
[05:18] <ddaa> where should that go?
[05:18] <ddaa> I'm sure I knew that in London, but I have already forgotten
[05:22] <ddaa> SteveA: thanks
[05:23] <ddaa> Yay!
[05:24] <ddaa> SteveA: thanks for your patch (a bit buggy, it should have used stepto instead of stepthrough), I now have a (mostly) working branch-index page
[05:24] <SteveA> what patch?
[05:24] <SteveA> cool
[05:24] <ddaa> The traversal thing.
[05:28] <SteveA> cool
[05:33] <bradb> Did the u/p change for wiki.canonical.com?
[05:39] <bradb> n/m
[05:39] <jblack> spiv: ping
[05:47] <SteveA> jblack: hiya
[05:47] <SteveA> there were some bzr and hg questions on here earlier
[05:49] <gneuman> does anyone know why series change from a product to another?
[05:50] <kiko> ddaa must know
[05:50] <kiko> or jblack 
[05:50] <kiko> or lifeless the asleeper
[05:51] <ddaa> because different products have different series
[05:51] <ddaa> but maybe i do not understand your question
[05:52] <gneuman> ok,hold on
[05:53] <kiko> I think he's asking why the code allows for changing series from a product to another
[05:53] <ddaa> well, because that's useful...
[05:53] <ddaa> there's an historical use case, which was moving series out from the "unassigned" product
[05:53] <gneuman> when u go to /products/(productname)/+series/(seriename)/+review u change the serie form productname to another porduct
[05:54] <gneuman> wich you choose form the form
[05:54] <gneuman> unassigned?
[05:55] <kiko> ddaa, anything apart from the unassigned product that you can see?
[05:55] <ddaa> and a current use case which is when a user create multiple unrelated series in a product instead of multiple products, or when a user creates a product "silva" and a an admin creates a product "sylva" with a productseries in it. 
[05:56] <ddaa> it's rarely used, but it still proves very useful now and then
[05:56] <ddaa> or when a user create noise productseries, so we can reassign them to a garbage product and get them out of the way
[05:57] <gneuman> i got 
[05:57] <gneuman> good
[05:57] <gneuman> i will be back with more questions
[05:58] <gneuman> thx ddaa
[05:58] <ddaa> the "change project" feature of products is similarly useful
[05:58] <ddaa> For example when a single product is created, and then the use finds out he wanted a project.
[05:59] <ddaa> BTW, I remind everybody, when you want such an admin change to be done, ASK ME
[05:59] <ddaa> because it will break RCS imports
[06:00] <gneuman> ok, so when u change a serie from product1 to product2, should the pae be redirected to prod2/+series/serie? or go to product1/+series?
[06:00] <gneuman> pae=page
[06:00] <ddaa> mh
[06:01] <ddaa> I think the former is more useful
[06:01] <ddaa> the latter can be obtained with "back".
[06:01] <gneuman> ok
[06:01] <ddaa> but hey, it's a rarely used admin feature, it's no biggie if it ends in a 404
[06:01] <ddaa> (just annoying)
[06:03] <ddaa> gneuman: you might want to add some comments about that somewhere, so the use cases are documented
[06:03] <ddaa> because it's understandably surprising at first
[06:05] <ddaa> mpt: don't you think that the html source of launchpad pages look like shit?
[06:06] <SteveA> ddaa: who cares? ;-)  we should remove all unneeded whitespace to reduce the size of the pages ;-)
[06:06] <ddaa> I care, because that makes writing page tests painful
[06:07] <ddaa> also, there's indeed a lot of wasted bits
[06:07] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial]  port the fix for #3077 (null priority causing exceptions to be raised on two pages) to mainline (patch-2656: brad.bollenbach@canonical.com)
[06:08] <ddaa> congrats bradb
[06:08] <bradb> scary, but i survived (i think)
[06:09] <mpt> ddaa: Learn the wonder of "..."
[06:09] <ddaa> mpt: I know of it, but it's still painful
[06:09] <mpt> How would you improve it?
[06:10] <gneuman> ddaa i will spend a lot of time on it!!!
[06:10] <ddaa> mpt: get all those empty lines back in control, for starts
[06:11] <ddaa> have discipline with page templates so emptylines are introduced in the source only where they are significant
[06:11] <ddaa> that should address most of it, actually :)
[06:12] <ddaa> remove pointless <span>
[06:12] <ddaa> remove uneeded <b> and <br>
[06:12] <ddaa> etc.
[06:12] <ddaa> I'm sure you know all of them
[06:13] <ddaa> same, remove <i> (use strong and em instead of b and i)
[06:28] <voidz0r> hello
[06:44] <jblack> ddaa: You around? 
[06:51] <roy_> hello everyone
[06:55] <carlos> jordi, hi, around?
[06:59] <sabdfl> hey roy_
[06:59] <roy_> hi
[07:00] <roy_> what's up mate?
[07:02] <sabdfl> carlos: ok, good plans
[07:03] <carlos> sabdfl, ?
[07:03] <carlos> sabdfl, dude your lag is really big :-D
[07:03] <sabdfl> the queue, for people to attach files
[07:03] <sabdfl> sorry, bits flying in the office, and code to be written
[07:04] <carlos> don't worry
[07:04] <carlos> sabdfl, https://wiki.launchpad.canonical.com/TranslationUploads
[07:04] <carlos> sabdfl, there is the initial braindump I'm doing
[07:04] <roy_> i need a girl in here to model in the ubuntu thong for me :D
[07:05] <carlos> sabdfl, I'm now dumping the implementation details, but more or less I think you can get the idea of the problems I'm trying to solve
[07:07] <roy_> that would so turn me on
[07:07] <carlos> kiko, the merge failed due conflicts, I have fixed them and requested a new merge
[07:07] <mpt> ddaa: Agreed on the <span>s (there's too much <span ... tal:content="..."> that should be <span ... tal:replace="...">
[07:07] <mpt> ddaa: But more <i>s in Launchpad do not mean <em>, and most <b>s in Launchpad do not mean <strong>
[07:08] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial]  One more fix to externalsystem -- make it cope with bugs.kde.org's weirdness. Also add tests and improve PEP8ness (patch-2657: christian.reis@canonical.com)
[07:09] <SteveA> Kinnison, cprov: i'm getting back to code reviews.  which of your reviews is more urgent?
[07:09] <cprov> SteveA: daniel's one
[07:09] <SteveA> okay
[07:09] <cprov> SteveA: I'm happy to land builder-action during the weekend
[07:10] <SteveA> cprov: i'll do them both today
[07:10] <cprov> SteveA: okidoki
[07:10] <mpt> ddaa: Most of the blank lines will be from <tal:foo> that gets stripped out by the processor. I always wondered how HTML source of dynamic sites got so full of blank lines, and now I know.
[07:11] <carlos> SteveA, where should we register specifications about Rosetta? inside Rosetta or Launchpad products?
[07:12] <SteveA> can we say that a rosetta spec depends on a launchpad one?
[07:13] <SteveA> i think we should make them all launchpad specs
[07:13] <SteveA> it will be easier to manage them that way
[07:13] <SteveA> and track their progress at the conference etc.
[07:14] <SteveA> this split between launchpad / rosetta / malone / buildd / etc. products is sometimes useful, but often a bit of a pain
[07:16] <carlos> SteveA, well, my spec does not depends on any launchpad spec, not sure if the system allows you to do that...
[07:16] <carlos> anyway, I'm happy adding the specs directly to Launchpad
[07:16] <mpt> BjornT: When you reviewed launchpad--translation-form--0510, did it include any pagetest changes?
[07:17] <carlos> SteveA, perhaps we should create a 'Launchpad' project that has all products as part of it so people can see the relation easily
[07:17] <SteveA> that's a good idea
[07:18] <SteveA> do specs for products get aggregated under their project?
[07:19] <ddaa> jblack: pong
[07:19] <mpt> carlos: That already exists
[07:19] <jblack> ddaa: I'm going over your post quite heavily.
[07:19] <mpt> SteveA: ohhh, https://launchpad.net/projects/launchpad/+specs
[07:20] <ddaa> jblack: the one about BranchLineage?
[07:20] <jblack> Yeah.
[07:20] <ddaa> cool
[07:20] <carlos> mpt, Rosetta is not member of any project
[07:20] <jblack> What does a revision mean to you these days? How are you defining it? 
[07:20] <mpt> carlos: https://launchpad.net/projects/launchpad says it is
[07:20] <carlos> then we have a bug
[07:20] <carlos> mpt, because products/rosetta does not says anything
[07:20] <ddaa> jblack: more specifically?
[07:21] <ddaa> A revision is a row in the Revision table...
[07:21] <mpt> carlos: and so does https://launchpad.net/products/rosetta
[07:21] <jblack> Yeah, exactly, if possible.
[07:21] <jblack> I mean conceptually.
[07:21] <mpt> carlos: I know it's in a portlet so it's invisible, but it actually does say that :-)
[07:21] <ddaa> Conceptually, a revision is a revision is a revision...
[07:21] <jblack> Damn the torpedos, abentley is wrong? 
[07:21] <ddaa> I'm sorry, but I do not know what you want to know.
[07:21] <mpt> "the launchpad suite:"
[07:22] <ddaa> Well, put another way
[07:22] <ddaa> Revision form an append-only DAG.
[07:22] <Kinnison> SteveA: if I do foo = dict(bar)
[07:22] <BjornT> mpt: no, no pagetest changes. should there have been any?
[07:22] <Kinnison> SteveA: will that use __iter__ and __getitem__
[07:22] <Kinnison> SteveA: or will it attempt .iteritems() ?
[07:22] <carlos> mpt, oh!, I was not on the main page O:-)
[07:23] <mpt> BjornT: yes, I think I'll get carlos to review them
[07:23] <mpt> carlos: I've got a branch ready for landing that alters the Rosetta newline pagetest. Can you check it before it lands?
[07:23] <mpt> so that I don't break anything
[07:23] <carlos> mpt, sure
[07:23] <carlos> mpt, branch?
[07:24] <mpt> carlos: mpt@canonical.com/launchpad--translation-form--0510
[07:24] <mpt> thanks carlos
[07:24] <SteveA> Kinnison: iter
[07:24] <Kinnison> funktastic
[07:25] <SteveA> i think...
[07:25] <Kinnison> I have a thing which smells a bit like a dictionary y'see
[07:25] <SteveA> hmm
[07:25] <SteveA> class dict(object)
[07:25] <SteveA>  |  dict() -> new empty dictionary.
[07:25] <SteveA>  |  dict(mapping) -> new dictionary initialized from a mapping object's
[07:25] <SteveA>  |      (key, value) pairs.
[07:25] <SteveA>  |  dict(seq) -> new dictionary initialized as if via:
[07:25] <SteveA>  |      d = {}
[07:25] <SteveA>  |      for k, v in seq:
[07:25] <SteveA>  |          d[k]  = v
[07:25] <SteveA>  |  dict(**kwargs) -> new dictionary initialized with the name=value pairs
[07:25] <SteveA>  |      in the keyword argument list.  For example:  dict(one=1, two=2)
[07:25] <SteveA>  |
[07:25] <SteveA> 
[07:25] <SteveA> so, if you provide __getitem__
[07:25] <Kinnison> right
[07:25] <SteveA> you are a mapping
[07:26] <SteveA> but, maybe it does use keys() and __getitem__ underneath
[07:26] <SteveA> i'd have to peek at the source
[07:26] <Kinnison> s'okay I've just realised I have to do it explicitly anyway
[07:35] <sabdfl> mpt: i'm looking through the design-fascism patch
[07:35] <sabdfl> I prefer Welcome to The Bazaar
[07:35] <sabdfl> "bazaar" is the tool
[07:36] <sabdfl> "the bazaar" is the launchpad-hosted marketplace of branches
[07:36] <sabdfl> hmm.... want me to review this for you?
[07:36] <kiko> sabdfl, please do
[07:36] <sabdfl> kiko: have you started?
[07:38] <kiko> sabdfl, not yet
[07:39] <kiko> this is design-fascism, right?
[07:41] <mpt> sabdfl: ok
[07:48] <sabdfl> kiko: yes
[07:50] <kiko> ok
[07:55] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: r=mpt Fix for bug 3128: 'Homepage' information should be at least 96 pixels high. (patch-2658: christian.reis@canonical.com)
[08:00] <carlos> mpt, https://chinstrap.ubuntu.com/~jamesh/pending-reviews/mpt@canonical.com/launchpad--translation-form--0510/filtered-diff does not shows me anything that changes the textareas nor any test update and the full checkout is taking a lot of time here....
[08:01] <Kinnison> SteveA: thanks for that
[08:01] <Kinnison> ciao all
[08:01] <SteveA> Kinnison: r=SteveA with tests, or r=SteveAUnderProtest without adding those tests ;-)
[08:01] <SteveA> we need to do them, so it's like taking out a loan
[08:02] <SteveA> with high interest rate
[08:02] <carlos> mpt, if you didn't touch anything inside textareas, that's ok. If you touched something... I will not be able to check that until tomorrow morning because I'm leaving now
[08:02] <SteveA> if they're not done
[08:02] <carlos> sabdfl, I will try to finish the spec for you tomorrow too
[08:02] <carlos> see you, time to leave!
[08:02] <mpt> carlos: ok, that diff is out of date, thanks anyway
[08:02] <carlos> mpt, I will take a look to it tomorrow then
[08:02] <mpt> ok
[08:03] <sabdfl> oh, mpt, this could be painful
[08:03] <sabdfl> we are going to disagree on capitalisation
[08:03] <sabdfl> page headings need to use capitals
[08:03] <sabdfl> they give the page stature
[08:03] <sabdfl> it does need to be more consistency
[08:03] <sabdfl> but
[08:03] <sabdfl> Add a milestone
[08:04] <sabdfl> should be
[08:04] <sabdfl> Add a Milestone
[08:04] <sabdfl> i agree with some of your changes
[08:04] <sabdfl> Create a New DistributionRelease
[08:04] <sabdfl> should be
[08:04] <sabdfl> Create a new Distribution Release
[08:04] <SteveA> cprov: just want to check i'm goingn to be reviewing the right thing: celso.providelo@c.c/launchpad--builder-action--0
[08:04] <sabdfl> it's a subtle thing, but i'm not happy with turning everything into sentence case
[08:04] <sabdfl> title case is important
[08:06] <SteveA> sabdfl: are you saying that you want to capitalize "pillars" or rather "names of kinds of content in launchpad" ?
[08:06] <sabdfl> I Don't Want Everything Capitalised
[08:06] <sabdfl> but I do want more cpaitalisation in titles than mpt does
[08:06] <sabdfl> the fascism patch turns every title into sentence case
[08:06] <SteveA> ideally, we'd be able to write down Guidelines on the Hackers' FAQ
[08:06] <sabdfl> "Register a new product release"
[08:06] <sabdfl> it should be
[08:06] <sabdfl> Reigster a new Product Release
[08:07] <SteveA> so Reviewers can make it so for new Pages 
[08:07] <SteveA> there's a certain cadence to it
[08:07] <sabdfl> -      <h1>Welcome to Rosetta!</h1>
[08:07] <sabdfl> +      <h1>Welcome to Rosetta</h1>
[08:07] <sabdfl> Why drop the !
[08:07] <mpt> because it looks lame
[08:07] <sabdfl> it's a welcome page
[08:07] <SteveA> one way to look at it is this: if you keep only the capitalized words, do you understand what's going on?
[08:08] <voidz0r> haha no
[08:08] <voidz0r> then I'd say "I'm not rosetta"
[08:08] <mpt> SteveA: I prefer: Are there any more words you can remove and have it still make sense?
[08:08] <sabdfl> mpt: i know this has been a lot of work
[08:08] <sabdfl> and most of the other changes i see, i like
[08:09] <sabdfl> but since you were changing my capitalisation in tons of places, it might have been worthwhile checking on a policy before doing it
[08:09] <mpt> NYT capitalization is good for menu items, because they're almost always 3 words or less
[08:09] <sabdfl> bugger, and most of these are changed in both page and tests
[08:09] <sabdfl> so its going to be a lot of work to undo
[08:12] <voidz0r> sabdfl (or anyone), that logout button really looks just like an input field, I even accidentally logged out trying to change the text >.< =)
[08:12] <mpt> voidz0r: that's Plone styling
[08:13] <voidz0r> ah
[08:13] <voidz0r> i kinda dislike plonk :p
[08:13] <mpt> which I could quite happily get rid of right now
[08:13] <mpt> since I'm hacking the style sheet
[08:13] <voidz0r> yay :)
[08:14] <SteveA> the apple websites seem to use the same kind of "title case, but for important words only" thing
[08:15] <mpt> sure, they're not presenting huge amounts of data
[08:16] <mpt> imo "Bugs assigned to Steve Alexander" is easier to scan than "Bugs Assigned to Steve Alexander"
[08:16] <mpt> because Bugs Assigned isn't a person
[08:16] <voidz0r> mpt, have you put an example online?
[08:16] <mpt> voidz0r: An example of what?
[08:17] <voidz0r> of your stylesheet :)
[08:17] <voidz0r> a demo
[08:17] <mpt> no
[08:17] <voidz0r> :'( ok...
[08:18] <SteveA> I'd say "Bugs assigned to Steve Alexander" if we're reading it as a sentence.  If there's a concept "Bugs Assigned" that is used throughout then i'd capitalize both, because "Bugs Assigned" has become an important term in its own right.
[08:18] <mpt> voidz0r: And if you want buttons to look like buttons, you will need to help me persuade sabdfl :-)
[08:18] <SteveA> but, "Bugs" should be capitalized a lot, because it is an important thing.
[08:18] <voidz0r> no, I was just curious
[08:19] <voidz0r> i like creating nice stylesheets :)
[08:19] <voidz0r> with least amount of hacks possible
[08:19] <mpt> SteveA: it's just one of what should eventually be many Malone reports
[08:19] <mpt> "Bugs Steve Alexander is subscribed to", "Bugs reported by Steve Alexander", etc
[08:19] <mpt> "Translations made by Steve Alexander"
[08:21] <voidz0r> hey, i just got an idea
[08:21] <voidz0r> launchpad has 3 columns, and all of them are fluid
[08:22] <sabdfl> mpt: people don't read web pages
[08:22] <sabdfl> structured information is better
[08:22] <voidz0r> so the columns can get wider then they're supposed to be, and the middle column also gets kinda stretchy
[08:22] <sabdfl> let me see some button designs... agree the current one looks too form like
[08:22] <mpt> sabdfl: right, scannability is at a premium
[08:23] <voidz0r> so what if the left and right colums have em units for width, and they're position uses percentages, creating a fluid gutter between the side columns and the middle one
[08:23] <mpt> sabdfl: which means Each Word Capitalized for things that are about three words or fewer (e.g. menu items), and sentence case elsewhere
[08:23] <carlos> mpt, still downloading your branch... I will review it tomorrow
[08:23] <carlos> later!
[08:23] <sabdfl> mpt: not for titles and headings
[08:23] <voidz0r> or,
[08:23] <sabdfl> look, this is a style thing. we need to agree a policy, and then be consistent
[08:24] <sabdfl> there was clearly a policy in place, with some inconsistency
[08:24] <sabdfl> but instead of simply addressing the inconsistency and querying the policy, your patch sets a new policy
[08:24] <voidz0r> if you can get the middle column to have a max-width (which is hard to do), then it would be even easier, and it won't get really stretched
[08:25] <mpt> sabdfl: Sorry, I didn't see any policy -- I saw sentence case in some places, and even words like "to" capitalized in other places
[08:25] <sabdfl> -        <h3 metal:fill-slot="heading">Create a new Team</h3>
[08:25] <sabdfl> +        <h1 metal:fill-slot="heading">Create a new team</h1>
[08:25] <sabdfl> h1 is correct, but Team should be capitalised
[08:27] <kiko> salgado, you got mail
[08:28] <mpt> voidz0r: I'm not sure what you mean by "their position uses percentages", but I think a fluid gutter would look odd
[08:29] <mpt> the columns vary in width because we have some pages with primary stuff that is too wide to fit in the 3-column layout, and some pages with portlets that are too wide also
[08:30] <zyga> hi
[08:30] <sabdfl> mpt: what's the difference between class="actions" and class="controls" ?
[08:30] <voidz0r> hm, maybe that can be addressed with an id or class on the body tag
[08:30] <zyga> I'm trying to create a new calendar entry
[08:30] <zyga> but I cannot reassign the start date
[08:30] <zyga> for some obscure reason it keeps setting itself back
[08:30] <voidz0r> but a fluid gutter should look ok, or just have them on the far left and right sides
[08:30] <kiko> sabdfl, note that this "policy" stated should really be written down -- why don't you go over it with mpt so he can specify or LaunchpadHackingFaq it?
[08:30] <zyga> I've set it to 12:00 
[08:30] <zyga> and I keep seeing 12:36
[08:31] <zyga> any clues? is this a know inssue?
[08:31] <voidz0r> it's kindof a tradeoff...
[08:31] <kiko> zyga, hmmm, possibly
[08:31] <sabdfl> kiko, mpt: sure, at UBZ
[08:31] <mpt> sabdfl: Where's class="controls"? There's no such class in the style sheet in my current branch
[08:31] <zyga> kiko: now I've set 12:30 and I've got 12:06
[08:31] <voidz0r> imho it's a good idea to keep often used things close to the browser controls
[08:31] <zyga> it smells like some parse error + calculation error
[08:32] <mpt> oh, weird
[08:32] <mpt> errorservice-*
[08:32] <kiko> sabdfl, I thought it was a simple rule..
[08:32] <kiko> I mean, do we need a bof to discuss capitalization?
[08:32] <voidz0r> a page with like 420px fixed width, centered and scrollbars way further on the right edge of the screen is kinda lameish :p
[08:32] <sabdfl> kiko: no, it ends up being more subtle than that
[08:33] <sabdfl> in some cases, i agree with mpt's patch
[08:33] <voidz0r> (imho)
[08:33] <sabdfl> but mostly, on the capitalisation, not
[08:33] <sabdfl> Its Not All Title Caps
[08:36] <mpt> voidz0r: all Launchpad pages are 100% wide, so the only way you'll get a 420px-wide page is in a 432px-wide browser window
[08:37] <voidz0r> no, i mean in general
[08:37] <voidz0r> as an example
[08:39] <SteveA> cprov: you have a review
[08:41] <cprov> SteveA: thank you
[08:43] <voidz0r> one more thing, if i may...
[08:43] <cprov> dudes, PQM looks like stalled on salgado's patch ...
[08:44] <voidz0r> i dont want my email adress shown on my profile, because of spam
[08:44] <voidz0r> can i change it?
[08:44] <SteveA> cprov: yes
[08:44] <mpt> voidz0r: No, but there's a bug report asking for that
[08:44] <voidz0r> ah ok :)
[08:45] <SteveA> cprov: elmo can unhang it.  i'll send an RT request about it...
[08:45] <voidz0r> tnx
[08:45] <kiko> mpt: https://launchpad.net/malone/bugs/3163
[08:45] <Ubugtu> Malone bug #3163: Smeg - No Translatable Templates Available Fix req. for: rosetta (upstream), Severity: Normal, Assigned to: Nobody, Status: New http://launchpad.net/malone/bugs/3163
[08:45] <mpt> voidz0r: https://launchpad.net/products/launchpad/+bug/1360
[08:45] <Ubugtu> Malone bug #1360: Inappropriate display of personal data Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Nobody, Status: New http://launchpad.net/malone/bugs/1360
[08:47] <cprov> SteveA: RT for restart PQM, aren't we being extremmely  bureaucratic ?
[08:47] <SteveA> cprov: got a better suggestion?
[08:47] <mpt> kiko: hmmm?
[08:48] <voidz0r> ok lets get on with it then (working on an gcbspci :)
[08:48] <salgado> cprov, SteveA, It doesn't look like it's staled
[08:48] <elmo> I've got a better suggestion, how about we not bother with bogus requests that just waste time?
[08:48] <voidz0r> cya
[08:48] <kiko> mpt, what's this smeg thing? :)
[08:49] <elmo> because it's pretty obviously not stalled
[08:49] <kiko> Elmo The Friendly One
[08:49] <mpt> kiko: It's the horrible name for the menu editor in 5.10
[08:49] <SteveA> kiko: with some effort, you could make that recursive
[08:49] <kiko> mpt, I don't believe you
[08:49] <salgado> and you guys please leave my branch there. I've been trying to merge it the whole morning
[08:50] <jbailey> Is there a particular component name I need to use to look for bugs in the +poll thing?
[08:50] <kiko> jbailey, launchpad.
[08:50] <jbailey> tx
[08:50] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial]  Say in the /requests page if an order was already sent for the shipping company or not. (patch-2659: guilherme.salgado@canonical.com)
[08:50] <mpt> kiko: yeah, it really is that bad
[08:50] <SteveA> i think we should have just 'rosetta' and 'launchpad'.  it's rather confusing otherwise.
[08:50] <elmo> kiko: look to your own, your boy's the one who started with the "beureaucratic" stuff
[08:50] <mpt> My girlfriend was asking me about it yesterday
[08:50] <mpt> but at least it's called "Menu Editor" in the menus
[08:53] <kiko> at least he can spell!
[08:54] <kiko> mpt, I thought smeg was something offensive
[08:54] <mpt> No, smegma is something icky
[08:54] <kiko> salgado, sent /to/ the shipping company. english is tricky
[08:54] <kiko> I think stub referred to it as "dick cheese", cheese not being a surname in this case
[08:55] <mpt> kiko: Smeg is a TV replacement expletive used in Red Dwarf, widely believed to be inspired by smegma
[08:55] <SteveA> not fridges?
[08:55] <mpt> (for another example of TV replacement expletives, see "rack off" in Australian soap operas)
[08:55] <sabdfl> thank god
[08:56] <sabdfl> can you imagine being "Richard Cheese"
[08:56] <bradb> because 1.2G Just Ain't Enough for 80 some-odd revisions to be applied, it seems
[08:56] <mpt> http://www.iloverichardcheese.com/
[08:56] <bradb> (1.2G /memory/ that is)
[08:57] <jbailey> mpt: Tell me you had to google for that, rather than knowing it off by heart...
[08:57] <SteveA> aw poo... another problem with gina
[08:58] <SteveA> sabdfl: john cleese from monty python, fish called wanda, etc. was originally "john cheese"
[08:58] <SteveA> which is almost as bad
[08:58] <sabdfl> mpt: can you show me this branch running somewhere, please?
[08:58] <sabdfl> i want to look at the specification listing change you made
[08:59] <sabdfl> i deliberately wanted not to use a table, and put quite a bit of work into that listing format
[08:59] <sabdfl> i'd like to see how it looks after your patch
[08:59] <sabdfl> can you send me a screenshot?
[08:59] <mpt> ok
[09:01] <kiko> SteveA, he changed his name?
[09:01] <SteveA> kiko: apparently so
[09:04] <sabdfl> mpt: ok, done
[09:04] <sabdfl> in general, excellent work, i'm very pleased with 90% of your changes
[09:04] <sabdfl> crisp and clear
[09:04] <mpt> thanks
[09:05] <sabdfl> there are some things that i'd like to review by seeing the pages in action. please organise for me to be able to see it one a web server runing over there
[09:05] <sabdfl> and the capitalisation i'm afraid gets a veto
[09:05] <sabdfl> please put a spec in place with your ideas, and mine, as a starting point for a UBZ bof
[09:05] <sabdfl> will mail shortly
[09:11] <sabdfl> SteveA: could you look into something for me with gina?
[09:11] <sabdfl> i noticed in the dump of dogfood (kinnison's) that it imports multiple identical builds of a sourcepackage
[09:11] <sabdfl> so the model is: source package release -> build -> binaries
[09:12] <sabdfl> it imports a build for hoary, and a build for warty, and a build for breezy
[09:12] <sabdfl> when in fact, it could look and say "all the binaries, exactly, are already there in build X, i'll just use that instead of creating a new one"
[09:12] <sabdfl> would it be hard to fix that?
[09:13] <sabdfl> SteveA: btw, nice breadcrumbs!
[09:14] <SteveA> is this for a sourcepackage with architecture=ALL ?
[09:14] <SteveA> ta
[09:14] <SteveA> actually, mpt designed it, i was just the codejockey
[09:14] <SteveA> there's a couple of refinements still on the breadcrumbs
[09:14] <SteveA> to be done
[09:15] <SteveA> i don't see that the binaries would be the same, unless architecture=ALL.  and even then, i don't see that they would be the same.
[09:15] <SteveA> but then, i don't have a deep understanding of the building stuff
[09:15] <sabdfl> it imports warty
[09:15] <sabdfl> and creates build + binaries
[09:15] <sabdfl> then hoary
[09:16] <elmo>         ed |     0.2-20 |        breezy | source, amd64, hppa, i386, ia64, powerpc, sparc
[09:16] <sabdfl> and instead of saying "i've seen those before" it appears to create a new build + binaries
[09:16] <elmo>         ed |     0.2-20 |         hoary | source, amd64, i386, ia64, powerpc, sparc
[09:16] <elmo>         ed |     0.2-20 |         warty | source, amd64, i386, powerpc
[09:16] <elmo> SteveA: ^-- example of binaries being the same across all 3 suites
[09:16] <sabdfl> they are different by arch
[09:16] <sabdfl> but the same across releases
[09:16] <sabdfl> and gina seems to miss that
[09:16] <SteveA> i see.
[09:16] <kiko> ah
[09:17] <kiko> that's why we're getting dupe inserts!
[09:17] <kiko> I wonder how deep the short-circuit needs to be
[09:17] <SteveA> kiko: can i punt this to you?  you know gina way better than i do.
[09:17] <sabdfl> kiko: i don't know how gina collects info before it decides to put it in the db
[09:17] <sabdfl> but i'm guessing..
[09:17] <sabdfl> it finds all the bin packages for the arch, from the source package
[09:17] <sabdfl> then creates a build for those
[09:17] <kiko> SteveA, noooooo
[09:17] <sabdfl> it should first look to see if the bin package is there already
[09:18] <sabdfl> if it finds one, it should find them all
[09:18] <kiko> a classic case of LBYL
[09:18] <sabdfl> then just reuse their build
[09:18] <sabdfl> so the new "upload" is just an upload of the same build
[09:18] <kiko> SteveA, it should be an easy fix
[09:18] <sabdfl> does that make sense?
[09:18] <SteveA> ah - i see, the version is 0.2-20 in each case
[09:18] <SteveA> all makes sense now
[09:18] <sabdfl> SteveA: not just the version, the actual sha1's should match too
[09:19] <sabdfl> i have a suggestion
[09:19] <sabdfl> spiv will review my work tomorrow
[09:19] <sabdfl> and i'll keep polishing it
[09:19] <sabdfl> (his review will cover all the heavy lifting, the rest is just polish)
[09:19] <sabdfl> but my branch makes it much easier to see, through the web, what the database says is going on in the archive
[09:19] <sabdfl> accurately
[09:19] <sabdfl> so
[09:20] <sabdfl> monday morning, we can look do the gina run on staging, with my branch running on it, and see if it all looks right
[09:20] <sabdfl> then we can go if it looks good
[09:20] <kiko> wow, great plan
[09:20] <SteveA> okay, i'll mail that to stub and Kinnison 
[09:20] <sabdfl> just delay the actual production rollout till tuesday
[09:21] <bradb> ddaa: I have another good one for you
[09:21] <kiko> sabdfl, the production rollout happens on tuesday :)
[09:21] <kiko> (please don't hurt me)
[09:23] <bradb> ddaa: https://chinstrap.warthogs.hbd.com/~dsilvers/paste/fileZo7PGV.html
[09:23] <sabdfl> i'm really nervous about this gina run
[09:23] <Mez> evening mark, long time no speak :D
[09:24] <sabdfl> hey Me
[09:24] <sabdfl> z
[09:24] <sabdfl> ok
[09:24] <sabdfl> release party time!
[09:24] <sabdfl> night all
[09:24] <kiko> night mark
[09:25] <Mez> stupid 6000 emails to download
[09:25] <bradb> jblack: feel free to have a look at the link above too, if you have time
[09:26] <bradb> it's happened twice in a row, even after i went to the trouble of manually adding revs to my revlib to make sure i'd be able to fork for diffs and all
[09:27] <kiko> bradb, network failure?
[09:27] <bradb> i'm back to do that now, a few revs at a time, in the hopes of getting up-to-date with rf in my revlib
[09:27] <bradb> kiko: i thought of that too, but it seems unlikely. this machine is on a wired connection too (as opposed to all the interference i get with wireless here)
[09:28] <bradb> fuck, again
[09:28] <bradb> even applying just 9 revisions
[09:28] <mpt> oh, he's gone
[09:32] <bradb> kiko: i hope it's not dodgy memory?
[09:33] <kiko> bradb, almost surely not
[09:33] <bradb> kiko: https://chinstrap.ubuntu.com/~dsilvers/paste/file2co4i2.html
[09:34] <bradb> i still have 1,3G on that partition
[09:36] <bradb> ddaa, jblack?
[09:37] <bradb> lifeless?
[09:38] <Mez> hmm
[09:38] <Mez> anyone here wanna do me a favou
[09:39] <roy_> what is the favor
[09:39] <Mez> explain to someone how the bzr system works
[09:39] <Mez> I find it hard to explain
[09:39] <roy_> i cant do that favor
[09:39] <SteveA> jblack: ?
[09:39] <roy_> i dont even know what bzr is
[09:39] <Mez> roy_ - in #katapult
[09:39] <Mez> jblack is away
[09:40] <SteveA> Mez: #bzr is a good place to ask
[09:40] <Mez> I cant find the log I had with him either
[09:40] <Mez> SteveA, no response in there
[09:40] <SteveA> well, it is the weekend
[09:40] <SteveA> is there something specific you want to know?
[09:40] <kiko> at least in .au
[09:42] <Mez> SteveA, just a general overview :D and maybe a Q&A
[09:42] <Mez> It's to explain to someone else
[09:45] <kiko> Mez, doesn't the bzr homepage clear up most questions?
[09:46] <Mez> kiko - it should but... *shrugs*
[09:46] <kiko> I HATE +BUG
[09:46] <kiko> @#@!$!@#@!#@!
[09:49] <bradb> it was an amazingly strange decision
[09:49] <bradb> like, intentionally weird, it would seem
[09:50] <mpt> bradb: Where can I see the table view bug listing in action nowadays?
[09:50] <bradb> mpt: too bad sucker! you CAN'T!
[09:51] <bradb> "tables aren't working for us", as mark said in his email. of course, users will perish without a table layout, so we'll bring them back soon enough.
[09:51] <SteveA> Mez: try to catch up with jblack sometime
[09:51] <mpt> bradb: meh
[09:52] <bradb> mpt: there are still table layouts for other listings though, like in the FOAF namespace, but those aren't the "official" ones, AFAIK
[09:52] <mpt> ah, thanks
[09:52] <bradb> for now, i won't be landing *anything* until my SCM software works
[09:53] <Mez> SteveA, will do
[09:56] <kiko> matsubara, ping
[09:56] <kiko> oh, never mind
[09:57] <kiko> matsubara, ping
[09:57] <matsubara> kiko: pong
[09:57] <kiko> matsubara, you know your fix for bug 2478?
[09:57] <Ubugtu> Malone bug #2478: Un-duplicating a bug leaves no indication of the change Fix req. for: malone (upstream), Severity: Normal, Assigned to: Diogo Matsubara, Status: Accepted http://launchpad.net/malone/bugs/2478
[09:58] <kiko> well
[09:58] <kiko> matsubara, I think you forgot to update the mailnotification code
[09:59] <bradb> kiko: can you try this command: baz get-changeset "$rocketfuel--patch-2656" somerandomdir
[09:59] <bradb> assuming you have $rocketfuel set to the rf archive
[09:59] <bradb> i.e. rocketfuel@canonical.com/launchpad--devel--0
[09:59] <kiko> sure
[09:59] <bradb> i ran out of disk space during a commit, which might have corrupted the archive
[10:00] <bradb> (when i later mirrored it)
[10:00] <kiko> the revlib or archive?
[10:00] <bradb> the archive
[10:00] <bradb> rocketfuel
[10:00] <bradb> on chinstrap
[10:00] <bradb> if you succeed in grabbing that changeset though, then i'm probably wrong
[10:02] <kiko> I succeed
[10:02] <kiko> do you want it?
[10:02] <bradb> wow
[10:02] <kiko> matsubara? silence?
[10:03] <bradb> kiko: i just need to update my revlib properly
[10:04] <bradb> kiko: what versions and tar and gzip are you running?
[10:04] <kiko> are you seriously asking this question? this is stock hoary
[10:05] <matsubara> sorry, I was checking mailnotification code
[10:05] <bradb> kiko: yes i'm seriously asking this question. got a better idea for why it would fail with a "gzip: stdin: unexpected end of file ..." for me but not for you? :)
[10:05] <bradb> but, if you're using hoary, that pretty much says it
[10:06] <Seveas> Can people submit their projects to launchpad, like on sourceforge? 
[10:06] <kiko> Seveas, yes, and you don't even need to ask permission
[10:07] <Seveas> cool
[10:13] <Seveas> hmm, the 'Specifications' link does not work for new projects 
[10:13] <Seveas> I get a 404
[10:13] <SteveA> Seveas: file a bug please
[10:13] <dilys> Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=SteveA]  BuilddUI missing pages reusing code from BuildRecordsView. (patch-2660: celso.providelo@canonical.com)
[10:14] <SteveA> unless it is already filed
[10:14] <SteveA> i think mpt may have filed that earlier today
[10:14] <kiko> Seveas, what's an example link?
[10:14] <Seveas> https://launchpad.net/projects/dynamite/+specs
[10:15] <kiko> I suspect projects don't have specs per se, do they?
[10:15] <kiko> this bug is not filed
[10:16] <mpt> no, I didn't get around to reporting it, sorry
[10:16] <mpt> it should work
[10:16] <Seveas> Should I file it?
[10:16] <kiko> please
[10:18] <Seveas> filed as bug 3166
[10:18] <Ubugtu> Malone bug #3166: New projects cannot create specs Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Nobody, Status: New http://launchpad.net/malone/bugs/3166
[10:19] <kiko> thanks Seveas you're a rocker
[10:22] <Seveas> kiko, as a matter of fact I am ;)
[10:23] <bradb> ddaa: ping
[10:23] <ddaa> bradb: sorry, I'll have to leave in a couple of minutes
[10:23] <bradb> ok
[10:23] <ddaa> looks like your running out of memory or something
[10:24] <ddaa> I've never seen a corrupt tarball in all my use of tla or baz.
[10:24] <ddaa> mh... star-merge...
[10:24] <bradb> i've got 1.25G of memory and i'm only getting one changeset
[10:24] <ddaa> Applying 16 revisions
[10:24] <bradb> the second URL
[10:25] <bradb> ddaa: i.e. https://chinstrap.warthogs.hbd.com/~dsilvers/paste/file2co4i2.html
[10:26] <bradb> even baz get-changeset "$rocketfuel--patch-2656" alsdjfasd gives the same error
[10:26] <ddaa> you can get the tgz changeset and try by hand
[10:27] <ddaa> looks like you really have something botched
[10:27] <bradb> i'm slightly worried that my memory may be corrupt but, at the same time, that seems somewhat remote
[10:28] <bradb> this one you mean?
[10:28] <bradb> bradb@chinstrap /home/warthogs/archives/rocketfuel@canonical.com/launchpad--devel--0/patch-2656 $ ls
[10:28] <bradb> checksum  launchpad--devel--0--patch-2656.patches.tar.gz  log
[10:29] <jblack> SteveA: I'm here
[10:29] <bradb> fwiw, that one untars fine
[10:30] <jblack> mez: What's up?
[10:30] <SteveA> hi jblack 
[10:33] <jblack> SteveA: I'm talking to them now
[10:51] <bradb> kiko: btw, if i can't find a good mountain bike, pignon sur roues has a 53" miele gara road bike. would that be ok if i can't find a good mountain bike?
[10:53] <kiko> bradb, yes, that's perfect. what are the components, and how much?
[10:54] <bradb> kiko: no idea about the components; what bits are you most keen to know about?
[10:54] <kiko> well, usually a road bike is either shimano or campagnolo, and either one has a few options (shimano 105, ultegra, dura-ace; athena, record, ...)
[10:55] <kiko> I'm curious if it has STIs
[10:55] <kiko> and if it's a real (I mean, training-grade) road bike (I've never heard of miele gara)
[10:56] <kiko> ah
[10:57] <bradb> http://www.mielebicycles.com/2005/en/velo_adultes.asp?BIKE=GARA&COLOR=JAUNE,MINUIT -- *no* idea if that's the one they have though
[10:58] <kiko> any idea of cost?
[10:58] <kiko> nice bike
[10:59] <bradb> free dude, free ;)
[10:59] <kiko> STI and shimano tiagra
[10:59] <kiko> bradb, huh?
[10:59] <kiko> free?!
[10:59] <bradb> it's the interest accumulated on the network card + late bday present + etc.
[11:00] <bradb> as long as it's only you and bjorn, i'll pay for the bikes
[11:00] <kiko> !
[11:00] <kiko> but you paid for the card!
[11:01] <bradb> i did? when?
[11:01] <kiko> well, when you bought your new one I guess
[11:02] <bradb> heh
[11:02] <kiko> ok ok whatever
[11:02] <kiko> this rocks and rolls
[11:02] <kiko> do you have a road bike?
[11:02] <bradb> no. just the mountain bike.
[11:02] <bradb> no pedals yet; might not get any this year
[11:04] <zyga> bah, I've crashed launchpad again
[11:04] <zyga> I've been creating a pool option
[11:04] <zyga> and the name contained spaces and non-ascii characters
[11:04] <zyga> if anyone is interested in checking the logs 
[11:04] <zyga> https://launchpad.net/people/moturuby/+poll/elect-team-leader/+vote-simple
[11:05] <zyga> the pool is now totally borked
[11:06] <bradb> jblack: around?
[11:07] <jblack> Right here, buddy
[11:08] <bradb> sweet
[11:08] <bradb> jblack: i repeatedly get this error: https://chinstrap.ubuntu.com/~dsilvers/paste/fileNvATQg.html
[11:08] <bradb> why do i get that error and what can i do to apply that CS to my revlib?
[11:09] <kiko> zyga, can you file a bug so I can paste in the error?
[11:09] <zyga> kiko: sure
[11:10] <jblack> lookin
[11:10] <mantiena> Hi all
[11:11] <kiko> hey mantiena 
[11:11] <jblack> bradb: can you check your browser password list, and msg me the chinstrap user/pass? 
[11:11] <zyga> https://launchpad.net/products/launchpad/+bug/3168
[11:11] <Ubugtu> Malone bug #3168: Creating pool option crashes the pool Fix req. for: launchpad (upstream), Severity: Normal, Assigned to: Nobody, Status: New http://launchpad.net/malone/bugs/3168
[11:11] <Mez> when i try to do an upstream RCS import, it ust sends me bck to the same page,k any idea why
[11:12] <kiko> Mez, hmmm, what URL?
[11:12] <Mez> https://launchpad.net/products/katapult/+series/0.3/+source
[11:12] <Mez> or do you mean what I'm trying to import
[11:12] <bradb> jblack: are you in?
[11:13] <mantiena> Anyone could help me import sources of one small product (live-installer udeb) into Bazaar ?
[11:13] <mantiena> look at https://launchpad.net/products/live-installer/+series/stable
[11:13] <kiko> Mez, it appears that the form is successfully posting, but gives no feedback? or what are you changing there?
[11:14] <Mez> It just comes back up the same page
[11:14] <Mez> as if i ad first gone there, just with the svn url filled in
[11:14] <jblack> bradb: I was taping up a box. ;) 
[11:15] <jblack> Bradb. got it
[11:15] <jblack> tar failed.
[11:16] <jblack> is tar missing, did you run out of diskspace, or memory?
[11:16] <Mez> me?
[11:16] <jblack> Actually, the problem isn't tar itself.
[11:16] <jblack> Its a bad gzip.
[11:17] <bradb> jblack: lots of diskspace and memory
[11:17] <kiko> Mez, what data are you filling out there?
[11:17] <jblack> bradb: Could you library be corrupted? 
[11:17] <Mez> svn://svn.berlios.de/kubuntupult/trunk
[11:17] <bradb> jblack: maybe, but if that's what's causing it, my mental model is shattered
[11:17] <jblack> check out your revision library by running baz my-revision-library
[11:18] <bradb> jblack: check it out how/
[11:18] <jblack> Well, something was ungzipped, and that gzip faile.d
[11:18] <bradb> i can get other patches, just not that one
[11:18] <jblack> which in turn caused tar to fail, which in turn cause baz to failed.
[11:18] <jblack> Yeah. Its a corrupt library or corrupt arch cache.
[11:18] <jblack> Lets check the arch cache first. 
[11:19] <bradb> jblack: that's weird, because i've been step-adding revs to my lib to play nice with baz
[11:19] <jblack> cd .arch-caches/archives/rocketfuel@canonical.com/... the branch you're pointed at.
[11:20] <bradb> yup
[11:21] <jblack> presumably, ~/.arch-cache/archives/rocketfuel@canonical.com/launchpad--devel--0--patch-2256
[11:21] <jblack> Is that there?
[11:21] <bradb> 2656, yeah
[11:21] <jblack> Ok. Kill it
[11:21] <mantiena> people, please help me
[11:22] <bradb> jblack: killed
[11:22] <kiko> mantiena, jblack and ddaa are the people to help you, and ddaa is the better one
[11:23] <mantiena> kiko, thanks
[11:23] <jblack> bradb: Now, there will be another cache.. the revlib one.
[11:23] <jblack> mantiena: What sort of problem is it? 
[11:23] <jblack> bradb: I don't know where your library is, because its user defined.
[11:23] <mantiena> I need to import sources of one small product (live-installer udeb) into Bazaar ?
[11:23] <mantiena> look at https://launchpad.net/products/live-installer/+series/stable
[11:23] <jblack> mantiena: I can give you a hand with that as soon as I'm done with brad
[11:23] <mantiena> Bazaar Publishing is Testing for about 4 days :(
[11:24] <mantiena> jblack, ok, thank you very much
[11:24] <jblack> Oh, so you've already started the import, but its still running.
[11:24] <jblack> bradb: However, it should have the same path style as your .arch-cache, but based off of whatever baz my-revision-library tells you
[11:24] <bradb> jblack: in my revlib, i see stuff like launchpad--devel--0--patch-2655, but no 2656
[11:25] <jblack> Ok. it could be failing on 2655, to take that one out
[11:25] <bradb> ok, removed
[11:25] <jblack> Ok. try get-changeset again
[11:26] <kiko> jblack, I meant "more appropriate" when I meant better -- but only because IIRC ddaa has been doing work on the imports recently, sorry
[11:26] <jblack> mantiena: What product? (and is it affiliated with a project?) 
[11:26] <jblack> kiko: Oh, I'm not offended at all. :) 
[11:27] <jblack> and if he's already got an important thats running - but -isn't, then I may have to punt to ddaa anyways.
[11:27] <kiko> I miscommunicated
[11:27] <jblack> Regardless, I can triage the problem for him, so that he can spend less time on it.
[11:27] <bradb> jblack: mm
[11:27] <bradb> bradb@oxygen:~ $ baz get-changeset "$rocketfuel--patch-2656" 
[11:27] <bradb> unable to rename "launchpad--devel--0--patch-2656.patches" to "../launchpad--devel--0--patch-2656.patches" (Directory not empty)
[11:27] <bradb> did i miss something there?
[11:27] <mantiena> jblack, I don't understand what you are asking :( I told product name to you 3 minutes ago
[11:27] <jblack> Its trying to create the directory, but it exists from the last thing you ran.
[11:28] <jblack> mantiena: Ahh, I'll scrollback for it then. 
[11:28] <jblack> ah, live-installer.
[11:28] <mantiena> ;)
[11:29] <gneuman> how can i get the new context and url from a serie that has changed product?
[11:29] <jblack> spooky. Why did searching for "live-instaler" work for "live-installer"
[11:29] <bradb> jblack: ah, ok, yeah, that was me manually going onto chinstrap and grabbing it. removed that dir...and it works!!
[11:29] <jblack> woo!
[11:29] <kiko> jblack, because of the way fti works
[11:29] <bradb> jblack++
[11:29] <kiko> I'm all about wooing on friday nights
[11:30] <jblack> mantiena: Its a failed job.
[11:31] <kiko> hey stub 
[11:31] <jblack> It failed because the webserver gave a 301 Moved Permanantly (http://ftp.ak1.lt"
[11:31] <stub> yo
[11:31] <SteveA> hi stub 
[11:31] <SteveA> i think i filled up the librarian
[11:31] <mantiena> jblack, where you see this message ?
[11:31] <SteveA> on staging
[11:31] <jblack> SteveA: did you enjoy it? 
[11:31] <SteveA> jblack: it was an experience
[11:31] <jblack> mantiena: I saw it on a backend maintenance page.
[11:32] <cprov> guys, good night, I'll melt at home, got full of melting in office (still 32 degrees and it's already dark)
[11:33] <jblack> mantiena: If you find me the right location, then I can fix that for you and get the import started up.
[11:33] <mantiena> jblack, but I can go without problems to http://ftp.akl.lt/users/mantas/live-installer/live-installer-0.1
[11:33] <SteveA> cprov: come to lithuania.  it's like montreal.
[11:33] <bradb> jblack: in closing, i wouldn't be surprised if the corruption happened when i ran out of disk space right at the end of a commit
[11:33] <mantiena> SteveA, hehe, you are from pov.lt ?
[11:33] <jblack> bradb: Thats my suspicion.
[11:33] <cprov> SteveA: should I say "heaven";) 
[11:33] <SteveA> mantiena: i'm squatting in their offices
[11:34] <kiko> :)
[11:34] <mantiena> SteveA, it seems I know your wife ... :)
[11:34] <jblack> mantiena: Perhaps its a missing file that the webserver can't find (that pysvn needs, that your client doesn't), and it redirects.
[11:35] <jblack> mantiena: I'll forward an email to ddaa for you though, ok? 
[11:35] <mantiena> jblack, ok
[11:35] <jblack> mantiena: would you like a carbon?
[11:36] <mantiena> jblack, maybe there could be problem because http://ftp.akl.lt/users/mantas/live-installer/live-installer-0.1/ really isn't SVN, but simpy unpacked sources ?
[11:36] <mantiena> what is carbon ?
[11:36] <jblack> carbon copy. CC: 
[11:36] <mantiena> ok
[11:36] <jblack> Where to?
[11:36] <mantiena> mantas@akl.lt
[11:37] <jblack> oh, .lt. That's why you asked about the pov.lt ;)
[11:37] <mantiena> yes
[11:40] <jblack> mantiena: While writing ddaa, I gave that listed url a try.
[11:40] <jblack> It doesn't work for me either.
[11:40] <jblack> Are you _sure_ that svn co http://ftp.akl.lt/users/mantas/live-installer/live-installer-0.1 is right?
[11:42] <mantiena> jblack, http://ftp.akl.lt/users/mantas/live-installer/live-installer-0.1/ really isn't SVN, but simpy unpacked sources ;)
[11:42] <jblack> Oh, that's right. You mentioend that. Yeah. We need an actual svn server to talk to
[11:43] <mantiena> jblack, but live-installer project doesn't have nor SVN nor CVS, only tar.gz and unpacked sources
[11:43] <SteveA> sveikas mantai.  trbut tau rekia paisti viso pov.lt.
[11:44] <jblack> mantiena: That's cool. We have tarball import method too
[11:44] <mantiena> jblack, how ?
[11:44] <jblack> What directory has the tarballs?
[11:44] <mantiena> Seveas, nesupratau k norjai pasakyti :(
[11:44] <Seveas> mantiena, ?!?
[11:44] <jblack> Stop saying I smell bad!
[11:45] <mantiena> jblack, almost the same - http://ftp.akl.lt/users/mantas/live-installer/
[11:45] <kiko> Mez, care to restate your problem?
[11:45] <jblack> mantiena: There's only one tarball? 
[11:46] <Mez> when trying to import a svn branch, it just brings me back to the same page as if i hadnt clicked submit
[11:46] <SteveA> mantiena: tik "sveikas".
[11:47] <jblack> Mez: Give me a moment to finish up with mantiena, and I'll lend you a hand
[11:47] <mantiena> jbailey, yes - in tallball are the sources
[11:47] <jblack> But thats it. This is the only version. There's no 0.0.9, no 0.0.1, etc? 
[11:48] <jblack> mantiena: Ok. You've already got the ftp info there, but the svn stuff is in the way. I don't have the permissions to fix that, so I'll mail this to ddaa.
[11:49] <mantiena> only one last (and first public) version is there - noone needs not working live-installer ;)
[11:50] <jblack> mantiena: Mail sent. ddaa has gone to bed for the night though, so he may not see it until tomorrow or monday.
[11:51] <jblack> Mez: Ok. Which product, and is it associated with a project?
[11:51] <mantiena> jblack, ok, thanks for help
[11:51] <Mez> jblack: katapult
[11:51] <jblack> btw, thanks for the help
[11:51] <jblack> Mez: I should have guessed. :) 
[11:52] <jblack> mez: Give me a moment to look at it
[11:52] <Mez> jblack :P
[11:52] <jblack> Mez: Ok. The page isn't as clear as it could be.
[11:52] <Mez> ...?
[11:52] <jblack> You probably didn't notice, but on the bottom right of https://launchpad.net/products/katapult/+series/0.3, a new box showed up showing an import status of Testing
[11:53] <jblack> Mez: However, I can tell you that its in the import system now, and because you're one of those cool guys, I've forced it to the top of the list. ;)
[11:54] <Mez> aw :D
[11:54] <Mez> Bazaar Publishing:  	Test Failed
[11:54] <jblack> Yes. 
[11:55] <jblack> Give me a moment to look at it a little closer.
[11:55] <jblack> hmmm. I'm guessing you left the arch fields blank? 
[11:56] <Mez> arch fields?
[11:57] <jblack> It may have asked you for an archive, a branch, a category. Perhaps not.
[11:57] <jblack> But they're in there now, and its testing
[11:58] <jblack> Mez: importing. 
[12:00] <jblack> Mez: Unfortunately there's a bug in import that will need fixing. 
[12:00] <Mez> lol
[12:00] <Mez> trust me to find the bugs
[12:01] <jblack> Good job, man
[12:01] <Mez> lol
[12:01] <Mez> I always seem to do this
[12:01] <Mez> try something new
[12:01] <Mez> find a bug
[12:01] <Mez> like when I first tried to package something
[12:01] <Mez> and found a bug in findutils