[02:12] <tmh__> can I install launchpad tools such as blueprint on my own private server? where is the source code?
[02:14] <Fujitsu> tmh__: No, it's not currently open source.
[02:31] <mpt> Goooooooooooooooooooood afternoon Launchpadders!
[02:31] <ajmitch> hi mpt 
[02:36] <Nafallo> mpt: morning :-)
[03:10] <Ubugtu> New bug: #68111 in malone "Message warning of lack of bug watch is too wordy" [Low,Unconfirmed]  http://launchpad.net/bugs/68111
[08:41] <mpt> spiv, have you noticed WARNINGs during make schema?
[08:47] <stub> There are two I think at the moment which remind us that we have only partially completed some schema modification.
[08:47] <stub> (but I can't recall just what, so it isn't doing a brilliant job...0
[08:47] <stub> )
[08:52] <mpt> right
[08:54] <carlos> morning!
[08:55] <jamesh> stub: did you end up running product-release-finder in production?
[08:56] <stub> jamesh: Yes. lifeless even gave me a blow job.
[08:58] <jamesh> great.  Do you have a log of the run somewhere?
[08:58] <jamesh> or would it be on the launchpad-errors list?
[09:00] <Ubugtu> New bug: #68126 in launchpad "Mirror prober go boom on staging" [Undecided,Unconfirmed]  http://launchpad.net/bugs/68126
[09:03] <carlos> do we have a launchpad-dependencies package for Edgy?
[09:04] <jamesh> carlos: if we do, then it is in a third party repo I don't know about
[09:05] <carlos> I know we were supposed to have it
[09:05] <carlos> but I saw nothing about it in our mailing list
[09:22] <jamesh> carlos: http://people.ubuntu.com/~robertc/lp-depends/ <- looks promising, but doesn't seem to have been updated since all the Python packages dropped the "2.4" suffixes
[09:22] <carlos> I see, ok. Thanks
[09:24] <stub> jamesh: devpad.canonical.com:~stub/product-release-finder.log
[09:24] <jamesh> stub: cheers
[09:25] <stub> (yes I restarted it - it was running for a bit before I remembered I should be running it under screen)
[09:31] <SteveA> morning
[09:39] <jml> good evening
[09:44] <SteveA> mpt: got anything for me to merge to brilliant?
[09:46] <mpt> SteveA, pushing the front page right now, I'll let you know when it's done
[09:46] <mpt> plus pagetest fixes
[09:46] <mpt> Other pagetest fixes are in PQM right now to land on mainline
[09:46] <mpt> so you'll need to merge from mainline to get all the fixes
[09:47] <SteveA> ok.  if they're just test fixes, then I can still merge and push out changes to brilliant
[09:47] <mpt> right
[09:47] <SteveA> then merge in devel later today and merge the ui branch to pqm
[10:02] <mpt> SteveA, 2006-08-ui has finished pushing
[10:12] <mpt> SteveA, it looks like my front page work has caused more test failures though
[10:15] <carlos> stub: hi, could you give me a patch number and db review for https://devpad.canonical.com/~jamesh/pending-reviews/carlos/launchpad/bug-2322/full-diff ?
[10:15] <carlos>  stub: jamesh did already the code review 
[10:15] <carlos>  so I just need the db number to merge it
[10:32] <stub> carlos: https://devpad.canonical.com/~andrew/paste/filerToPwd.html
[10:32] <carlos> stub: ok, thanks!
[10:33] <stub> carlos: PostgreSQL will optimize sequential JOIN's normally, but there are so many that it thinks you are doing it deliberately to optimize the join order. Rewriting it like that lets it do its thing.
[10:34] <carlos> ok
[10:34] <carlos> I just copied what main db schema has and removed the field from it
[10:39] <jamesh> stub: carlos's version uses left joins.  Doesn't your version do inner joins?
[10:39] <stub> carlos: Urgh... scratch that. I rewrite the LEFT OUTER joings to INNERs
[10:39] <stub> yer
[10:39] <stub> (I always say LEFT OUTER instead of LEFT so I don't get confused)
[10:40] <stub> carlos: Use your original, same patch number
[10:40] <jamesh> well, there isn't such thing as a left inner join
[10:40] <stub> jamesh: There is at the end of a sprint ay
[10:40] <jamesh> so I usually just say join, left join, right join and outer join
[10:40] <stub> day
[10:41] <carlos> jamesh: good catch!
[10:41] <mpt> SteveA, test fixes have landed on mainline
[10:42] <SteveA> ok
[10:44] <SteveA> jamesh: I'd think "join" would mean "cross product"
[10:44] <SteveA> but then, that's not useful mostly
[10:44] <jamesh> SteveA: in terms of SQL, it is "foo join bar on condition"
[10:45] <jamesh> SteveA: it is a cross product
[10:46] <jamesh> an outer join is also a cross product, but also returns the rows where the condition evaluates to NULL (effectively)
[10:47] <SteveA> for those reading along at home, here's a nice summary: http://en.wikipedia.org/wiki/Join_%28SQL%29
[10:49] <SteveA> mpt: I merged from rf/devel and resolved a conflict in the Makefile
[10:49] <jamesh> yeah.  My description of outer joins is wrong
[10:50] <SteveA> it's wikipedia.  you can make your description right be editing that page.
[10:50] <SteveA> mpt: sftp://devpad.canonical.com/code/stevea/launchpad/ui
[11:34] <lifeless> malcc ping
[11:34] <malcc> lifeless: pong
[11:34] <lifeless> malcc: trying it this way
[11:34] <lifeless> hi
[11:35] <lifeless> so, PPA - we've been talking about a thing called SourcePackageBranch
[11:35] <lifeless> which will be used to ty bzr branches and soyuz otgether
[11:35] <lifeless> s/ty/tie
[11:35] <malcc> Ok
[11:41] <jamesh> spiv: I assigned brad's branch to you for review since he said you'd handled it before.  You might want to ask him for an incremental diff since it sounds like you've reviewed part of it before.
[11:41] <spiv> jamesh: I always note revision numbers in reviews subject lines, so it shouldn't be too hard to calculate the diff myself...
[11:42] <jamesh> spiv: brad gave the last reviewed revision in the description
[11:43] <spiv> Even better.
[11:44] <lifeless> sorry, link glitch
[11:44] <lifeless> malcc: so, when PPA comes in, what information will I need to know to be able to use a buildd to do a bzr checkout and then run dpkg-buildpackage -S in there, and from outside the xen instance gpg sign it
[11:44] <lifeless> and finally submit the result to the upload queue
[11:46] <lifeless> malcc: and is there already a spec on this ? :)
[11:46] <malcc> lifeless: I don't think there's a spec on this, but I'm still trying to catch up with what "this" is :)
[11:46] <malcc> lifeless: You're talking about building for PPAs in a NoMoreSourcePackages kinda way?
[11:46] <lifeless> malcc: yes, and for real distros too
[11:47] <lifeless> I'm knida getting low food wise now, so will try to catch up more seriously on this next week
[11:47] <malcc> lifeless: Since we need a source package for publication anyway, I assumed we'd create a procedure to make a source package from the branch
[11:47] <malcc> lifeless: And then let everything else proceed as it does today
[11:47] <lifeless> but for now, has the soyuz side of NMSP been well specced ?
[11:48] <lifeless> malcc: making a SP from the branch involves the checkout and dpkg-buildpackage -S
[11:48] <malcc> lifeless: No, NMSP is barely on our radar as of yet
[11:48] <lifeless> malcc: that runs user code
[11:48] <lifeless> malcc: so it needs to run under xen
[11:48] <malcc> lifeless: Ah, right, now I follow
[11:48] <malcc> lifeless: So the step you're talking about is precisely that sourcepackage building step
[11:48] <lifeless> thats the step that I know of, but not about, the machinery in soyuz to make happen
[11:49] <malcc> Well we'll need to set up Xen-based buildds to do the PPA building, so I'd hope there's not much extra machinery needed to use them for dpkg-buildpackage -S
[11:49] <malcc> But as of yet we have no machinery
[11:50] <lifeless> can I feed requirements in ?
[11:50] <malcc> Yes
[11:50] <lifeless> I need to run:
[11:50] <lifeless> 'bzr checkout' from either inside or outside of xen. This will require http access to the supermirror (which is accessible on an internal http side)
[11:50] <lifeless> 'dpkg-buildpackage -S' from inside xen.
[11:51] <lifeless> 'dpkg-signpackage' (IIR the command right) from outside xen with a local gpg key.
[11:51] <lifeless> in that order
[11:51] <malcc> lifeless: Makes sense
[11:51] <lifeless> and then I will have the result to hand over to your upload queue as files on disk.
[11:55] <lifeless> well, see you tomorrow.
[12:07] <SteveA> jamesh, spiv: can we talk about launchpad-dependencies for a few minutes?
[12:07] <jamesh> sure.
[12:07] <SteveA> I asked stuart to take this on, but he's currently very busy with sprints and such
[12:08] <SteveA> so, here's what I suggest: we use bugs to track what we want in the package
[12:08] <SteveA> etienne is doing the actual packaging, and is just waiting for our requirements
[12:09] <SteveA> so to sort it out, how about filing a bug listing the dependencies we know about for dapper, and one for edgy
[12:09] <jamesh> lifeless asked for bug reports to be filed at https://launchpad.net/products/launchpad-development-infrastructure
[12:09] <SteveA> I'm not sure if we need a separate package for developers and one for deployment on servers
[12:09] <SteveA> probably do
[12:09] <SteveA> that place is fine to file bugs
[12:10] <SteveA> so, if you'd like to get things moving, file a bug describing what we need, and ensure that stu is subscribed to it and etienne is assigned to it
[12:10] <SteveA> then mention it to etienne by email/irc
[12:10] <SteveA> what do you think?
[12:10] <jamesh> sounds good
[12:12] <jamesh> I've noticed that a few of the changes in edgy cause problems for running launchpad
[12:13] <jamesh> for instance, we try to run a "twistd2.4" script, and there now appears to only be a "twistd" script.
[12:20] <SteveA> right
[12:20] <SteveA> I think we should all be using these packages.  It's also helpful when new people join the team.
[12:22] <jamesh> perhaps in the case of twisted we should be invoking the versions inside our tree
[12:22] <jamesh> sourcecode/twisted/bin/twistd and trial
[12:27] <SteveA> that's a good point
[12:28] <SteveA> it's probably just a coincidence that using the OS versions works
[12:35] <jml> which version of twisted is in your tree?
[12:57] <SteveA> jordi: ping
[01:05] <Ubugtu> New bug: #68161 in launchpad "Launchpad should use in-tree versions of twistd and trial" [Undecided,Unconfirmed]  http://launchpad.net/bugs/68161
[01:13] <carlos> later!
[01:14] <lucasvo> jamesh: wasn't every 2.4 package renamed?
[01:14] <jamesh> lucasvo: yeah.  Above I was talking about executable names though.
[01:14] <lucasvo> ah, ok
[01:14] <jamesh> lucasvo: in dapper, a /usr/bin/twistd2.4 script was installed, and in edgy there isn't.
[01:31] <malcc> SteveA: Ping
[01:38] <jordi> SteveA: pong
[01:43] <SteveA> malcc: hi
[01:54] <spiv> jamesh: yeah, it would be better to use the twistd in our tree.  It'll be at sourcecode/twisted/bin/twistd.
[01:55] <spiv> jml: some random version of trunk that has a sufficiently new version of twisted.vfs, it doesn't correspond exactly to a particular release.
[01:56] <spiv> jml: Ah, it's a snapshot of SVN r17186 in fact.
[01:56] <spiv> Hmm, I wonder if the version in edgy is new enough?
[01:57] <spiv> Although, I have a suspicion the supermirror sftp server is likely to keep needing the bleeding edge as it develops.
[02:25] <salgado> hey spiv.  how about that bug 54791?
[02:25] <Ubugtu> Malone bug 54791 in launchpad "The mirror prober should check a few files from each mirror in paralel instead of a lot of files from a single mirror" [Medium,Confirmed]  http://launchpad.net/bugs/54791
[02:40] <Ubugtu> New bug: #68175 in launchpad-bazaar "vcs import should report cvs config and svn externals information" [Medium,Confirmed]  http://launchpad.net/bugs/68175
[04:14] <carlos> danilo[out] : ping
[04:15] <Ubugtu> New bug: #68189 in launchpad "execute_zcml_for_scripts should support different setup when run under tests" [Undecided,Unconfirmed]  http://launchpad.net/bugs/68189
[04:42] <salgado> spiv, no love from you, eh?
[05:06] <danilos> carlos: pong
[05:06] <carlos> danilos: hi, do you got your VISA ?
[05:07] <danilos> carlos: yeah, I did, but Americans are real silly: I got it for 3 years, and my passport is valid until 2008 :)
[05:07] <carlos> X-)
[05:08] <carlos> well, they gave you 3 years, but it's your fault you can only use it for next 2 years... :-P
[05:08] <carlos> danilos: I wanted to ask you about the status of your Firefox branch
[05:08] <danilos> hah, yeah, sure :) the common visa issuance requirement is to have passport that is valid at least 6 months after the end of your visa
[05:08] <danilos> or trip, or something like that
[05:09] <danilos> carlos: well, I think I uploaded it
[05:09] <danilos> unless I messed something up, let me recheck
[05:10] <Ubugtu> New bug: #68200 in launchpad-support-tracker "dateanswered really tracks confirmation date" [Undecided,Unconfirmed]  http://launchpad.net/bugs/68200
[05:10] <spiv> jamesh: it turns out the diff for bradb's malone-release-management on pending-reviews is the one I need to review, because the last reviewed revision has been merged to rf.
[05:11] <spiv> jamesh: (and in fact so have some later ones that got merged as [trivial] )
[05:11] <carlos> danilos: well, from what I see at https://devpad.canonical.com/~jamesh/pending-reviews/danilo/launchpad/firefox-import/full-diff (it was generated one hour ago) nothing changed since yesterday
[05:15] <danilos> carlos: (bzr push is taking quite some time; apparently it hasn't been pushed through last time)
[05:15] <danilos> btw carlos, did you manage to track the "reverting to old translations" bug people have started to report?
[05:16] <BjornT> spiv: fwiw, the revision of malone-release-management was indeed merged to rf, but then later it was pulled out. so you shouldn't have to review everything again, although it's hard to produce a good diff of the changes made, so maybe it's easier to do a complete re-review.
[05:16] <spiv> danilos: have you tried bzr+ssh://?  It should be ~20% faster.
[05:16] <spiv> BjornT: Oh, I see.  That wasn't clear from the notes in the review queue.
[05:16] <danilos> spiv: as a matter of fact, no, let me try that
[05:16] <spiv> I'll look closer at the commit logs.
[05:17] <carlos> danilos: no clue
[05:18] <carlos> danilos: I asked for concrete messages that were reverted to investigate at db level, but I'm not sure whether would be too helpful... we should log translation changes to be able to debug this kind of actions
[05:18] <flacoste> BjornT: ping
[05:18] <BjornT> spiv: rev 3682 is a bit special; it reapplies the changes reverted in the previews 'merge rf'.
[05:18] <jamesh> spiv: given the reversion, it'll probably be a bit difficult to get an incremental diff that filters out the RF merges
[05:19] <BjornT> hi flacoste 
[05:19] <flacoste> hi BjornT
[05:19] <spiv> jamesh: Yeah.
[05:19] <danilos> carlos: logging translations would significantly slow everything down, and we already have performance issues, imho :(
[05:19] <flacoste> BjornT: regarding using comment index in the confirmation URL, where is this used in Malone?
[05:20] <flacoste> i'm looking for an implementation example
[05:21] <BjornT> flacoste: it's used to display the individual comments. it's mostly browser code, i think.
[05:21] <carlos> danilos: well, I was thinking in some extra metadata information more than using an specific table to log the events
[05:22] <flacoste> BjornT: ok, and I'm going to go forward with your suggsetion of tightening the security with permissions
[05:23] <flacoste> BjornT: I'll need a new permission for the reject one though, I was about to call it launchpad.Moderate, what do you think of this name? that way, it's not too specific
[05:24] <kiko> aieee
[05:24] <kiko> salgado, is marilize happy? I saw no email
[05:25] <flacoste> BjornT: other thing, since i'll be using an index in the URL, should I rename the parameter to answer_idx?
[05:25] <salgado> kiko, yes, she said everything looks okay
[05:25] <jamesh> spiv: I guess a diff from 3682 to head would be fairly easy to produce (using the double merge algorithm I mentioned on the list)
[05:25] <kiko> salgado, you are the man
[05:25] <jamesh> spiv: then see if 3675..3681 is easy to review on its own
[05:25] <Ubugtu> New bug: #68203 in malone "Bug description with only whitespaces should return a nice error message instead of crashing" [Medium,Confirmed]  http://launchpad.net/bugs/68203
[05:26] <spiv> jamesh: yeah, that's basically what I'm going to try
[05:29] <BjornT> flacoste: launchpad.Moderate would work for me. it'd be good to send mail to SteveA, though, to see what he thinks.
[05:29] <BjornT> flacoste: i think answer_id looks better than answer_idx. the index will still be an id, unique within the ticket.
[05:30] <flacoste> BjornT: fine with both
[05:35] <Ubugtu> New bug: #68206 in launchpad "custom mail headers for specifications " [Undecided,Unconfirmed]  http://launchpad.net/bugs/68206
[06:15] <doko_> carlos: ping
[06:16] <carlos> doko_: pong
[06:17] <doko_> please have a look at http://people.ubuntu.com/~doko/ooconv/
[06:18] <doko_> currently po files like test_de_merge.po are created
[06:19] <doko_> can rosetta handle files like  test_de_msgid_comment.po as well?
[06:20] <doko_> and if yes, what happens with the old strings if I upload these?
[06:20] <doko_> carlos: ^^^
[06:20] <carlos> doko_: I guess the change is the '_: ...' part, isn't it?
[06:21] <doko_> yes, problem is, that with the current approach, some strings are marked as fuzzy (and maybe not imported in rosetta?)
[06:23] <doko_> at least we do loose the string "Initialienzeichen" with out current approach.
[06:23] <carlos> doko_: no, we don't yet support that in Rosetta
[06:24] <carlos> we will show the msgid as it's atm
[06:24] <carlos> that's with '_: ...' part
[06:24] <carlos> which sucks
[06:24] <carlos> yes, we plan to support that
[06:24] <carlos> no, I don't think you should use it for OO.org
[06:24] <carlos> is better to move to use native GSI support
[06:25] <carlos> doko_: danilos told me that we should have it ready to review next week
[06:25] <doko_> ok, then the only thing I can do with the ku translation is to ignore rosetta at the moment and use erdals GSI file directly.
[06:26] <doko_> carlos: nice
[06:26] <carlos> doko_: yes, I think that's the best and fast option right now
[06:27] <salgado> kiko, do you think it'd be fine to create the missing mirrors with an SQL/python script?
[06:31] <BaKKaR> hi
[06:31] <BaKKaR> can someone tell me pls what's gnome-orca?
[06:32] <BaKKaR> as far as i know it's acc application
[06:32] <BaKKaR> but is it the same Gnopernicus?
[06:53] <fabbione> stupid question...
[06:53] <fabbione> https://launchpad.net/people/ubuntu-toolchain/+branch/ubuntu-toolchain/glibc-2.5-package <-
[06:53] <fabbione> how often are these pages regenerated?
[06:53] <fabbione> i committed 36 like one hour ago
[07:28] <carlos> see you!
[08:03] <salgado> carlos, already left?
[10:00] <Ubugtu> New bug: #68261 in rosetta "OOPS exporting an empty POTemplate." [Medium,Confirmed]  http://launchpad.net/bugs/68261
[10:42] <kiko> matsubara, I'm confused by your patch
[10:42] <kiko> oh, I see
[10:42] <kiko> it crashes in unpatched trees
[10:42] <kiko> but not in a patched tree
[10:42] <kiko> matsubara, can you add something to check that the desired effect was obtained?
[10:44] <matsubara> kiko: thinking about it, it doesn't make much sense let the user export an empty POTemplate.
[10:44] <kiko> (i.e. that the export worked)
[10:51] <kiko> matsubara, do you think we should try to stop him? :)
[10:56] <matsubara> kiko: well, I don't know. Maybe avoid uploading empty potemplates in the first place. 
[10:57] <matsubara> kiko: but I don't know if there's a real use for POTemplates that have messageCount == 0
[10:58] <kiko> matsubara, until they are imported they do
[11:02] <matsubara> kiko: https://sodium.ubuntu.com/~andrew/paste/fileeuTuLx.html
[11:04] <kiko> matsubara, good job :)
[11:04] <kiko> matsubara, hold on to that patch let's see what carlos says
[11:06] <carlos> kiko, matsubara: I'm moving from __getitem__ like methods to explicit ones as I'm touching code
[11:06] <carlos> so I'm fine removing __len__
[11:07] <kiko> carlos, what about exports of empty files
[11:07] <carlos> well, export empty files makes no sense
[11:08] <kiko> should we allow them?
[11:08] <carlos> and it's an UI bug that we show them
[11:08] <kiko> mmm
[11:29] <kiko> carlos, what should we do then?
[11:30] <carlos> well, current patch fixes the issue, right?
[11:33] <kiko> carlos, it does, but I want to know what you want us to do
[11:33] <carlos> well, the faster solution would be to go with that patch
[11:34] <carlos> the final solution, would be to remove from the navigation links any potemplate with zero messages
[11:34] <carlos> the url should be valid, but we shouldn't link to them from sourcepackages or productseries
[11:35] <kiko> carlos, but then export will still be a valid action item.
[11:36] <Ubugtu> New bug: #68277 in malone ""newer first" doesn't sort bugs correctly" [Undecided,Unconfirmed]  http://launchpad.net/bugs/68277
[11:36] <carlos> kiko: we could disable that action then
[11:36] <kiko> carlos, yeah
[11:55] <kiko> carlos, so for now allow people to generate empty exports?
[11:55] <kiko> and file a bug for the fact that we do?
[11:55] <carlos> yes, please
[11:55] <kiko> cool.
[11:55] <kiko> matsubara, I'll take care of that as well
[11:55] <carlos> good night guys
[11:55] <carlos> kiko, matsubara: Thanks for handling that
[11:55] <matsubara> thanks kiko