[12:06] <bradb> mhz_design: ping me when you're ready to continue
[12:28] <mhz_design> bradb: i'm back but wife is calling me now :(
[12:28] <bradb> mhz_design: https://launchpad.net/people/edubuntu-chile/+editemail
[12:28] <bradb> I was going to ask you to try reconfirming it
[12:28] <bradb> To see if that helps
[12:28] <mhz_design> bradb: will you be around in 45?
[12:29] <bradb> probably not
[12:29] <mhz_design> bradb: I can loggin
[12:30] <mhz_design> I am asked to use the same email address or other.
[12:30] <mhz_design> bradb: which?
[12:30] <bradb> mhz_design: log in as your normal user
[12:30] <bradb> mhz.chile@gmail.com, if I'm not mistaken
[12:30] <mhz_design> bradb: I am logged in
[12:31] <mhz_design> An e-mail message was sent to 'mhz.chile@gmail.com'. Follow the instructions in that message to confirm the new contact address for this team.
[12:31] <bradb> yep
[12:31] <mhz_design> ok, bradb thx
[12:31] <bradb> did confirming work though?
[12:32] <mhz_design> not gotten any email yet
[12:32] <bradb> it might take a min or two
[12:32] <mhz_design> my wife will kick my butt
[12:32] <bradb> ok :)
[12:37] <kjcole> Hullo.  Any thoughts about deleting/undoing mistakes at launchpad.net?  I'm creating small bits of clutter on my way to understanding how things work.
[12:47] <bradb> kjcole: Sorry, Launchpad isn't very good at dealing with user mistakes yet. Don't stress too much about messing things up. When things get mucked up a bit because a new user was playing around and didn't have an undo facility, that's our fault, not theirs.
[12:49] <kjcole> bradb: Thanks. In that case, I'll be slightly less cautious, as I attempt to learn the brave new world (for me) of RCS by jumping right into bzr. ;-)
[12:50] <bradb> kjcole: have fun :)
[12:54] <mhz_dinner> bradb:  Oops
[12:54] <mhz_dinner> Sorry, something just went wrong in Launchpad. 
[12:54] <mhz_dinner> .9
[12:54] <mhz_dinner> :(
[12:55] <mhz_dinner> bradb: obviously, I got email, I visited url, I pressed ok, but error :(
[12:57] <bradb> mhz_dinner: Is there another contact email address you can try to confirm with?
[12:57] <mhz_dinner> bradb: yup mhz@tecnocimiento.cl
[12:58] <mhz_dinner> bradb: could it be that I am aproved ubuntu memeber but not updated yet ?
[01:00] <bradb> mhz_dinner: I don't think so, but I'm not 100% sure on that.
[01:00] <bradb> mhz_dinner: Can you try that other address?
[01:45] <mhz_dinner> re
[08:20] <carlos> morning
[08:21] <sivang> morning
[09:16] <SteveA> hi
[09:20] <carlos> SteveA, hi
[09:41] <carlos> jordi, ping
[09:43] <jordi> hey
[09:44] <jordi> carlos: I was about to ping you
[09:44] <jordi> I need the GNU team now
[09:44] <jordi> mailman has accepted
[09:44] <carlos> jordi, ok
[09:44] <carlos> jordi, I'm reading all pending emails from rosetta-users
[09:44] <carlos> and I think you missed "Translation of Django web framework"
[09:45] <carlos> jordi, at least I don't see any answer to that email
[09:45] <jordi> hmm
[09:45] <carlos> would you take a look, please?
[09:45] <carlos> in exchange ;-)
[09:45] <jordi> yes
[09:45] <jordi> carlos: should we create a GNU team and a TP team?
[09:46] <jordi> there's a difference
[09:46] <jordi> GNU team would be people with translation disclaimers
[09:46] <carlos> thank you
[09:46] <carlos> hmmm
[09:46] <jordi> TP team would be people doing non-GNU translations in the TP
[09:46] <jordi> we don't need the TP team now
[09:46] <jordi> but we want to make this clear in the team description
[09:46] <carlos> it's the same team
[09:46] <jordi> no
[09:46] <carlos> but with different status, right?
[09:47] <jordi> people without translation disclaimers in the FSF should not be allowed to translate mailman
[09:47] <jordi> oh, yes.
[09:47] <jordi> but we can't have statuses in Rosetta, right?
[09:47] <carlos> we don't support that difference, so perhaps the non-GNU translators should just joing Ubuntu's teams?
[09:48] <carlos> is more or less the same..
[09:48] <jordi> django, I had this message flagged, but unanswered due to the policy thing (which I wanted to finish writing now)
[09:48] <ajmitch_> evening
[09:48] <carlos> jordi, I need a name (used for the URL), title, and a summary
[09:49] <carlos> jordi, oh, I thought you already started the rejection/acceptance of the imports using the new policy
[09:49] <carlos> ok
[09:49] <carlos> ajmitch_, morning ;-)
[09:49] <jordi> carlos: I don't think so, because they are already established teams as GNOME or Debian
[09:49] <carlos> ok
[09:49] <jordi> ajmitch_: Iiiii waaaant a paackaaaageeee!!
[09:49] <ajmitch_> jordi: dude, check experimental
[09:49] <carlos> jordi, give me the information and you will have your team! :-P
[09:50] <ajmitch_> jordi: I just had to pull patches from CVS to get it building again
[09:50] <jordi> carlos: no, not yet, didn't finish it last week, went out for the weekend and didn't work yesterday
[09:50] <jordi> so today was The Day
[09:50] <jordi> ajmitch_: woops :)
[09:51] <carlos> jordi, ok
[09:52] <carlos> jordi, btw, I will be on holidays next week so please, prepare a list of things you need done as soon as possible from my side because I don't think I will have Internet access
[09:52] <carlos> jordi, anyway, I will send an email today to warthogs mailing list to notify that
[09:52] <carlos> jordi, I'm already handling the plural form requests
[09:53] <jordi> carlos: GNU Translators, gnu-translators, "The GNU translators are in charge of localising the official GNU projects. GNU translators need to send a signed Translation Disclaimer on paper to the FSF, for copyright reasons, before they can translate."
[09:53] <jordi> carlos: great
[09:54] <jordi> maybe translation Disclaimer can have a link to the place
[09:54] <carlos> jordi, sure, URL?
[09:54] <jordi> http://www.iro.umontreal.ca/translation/HTML/disclaim.html
[09:55] <carlos> thanks
[09:55] <carlos> hmm 
[09:55] <carlos> jordi, seems like I'm not able to add a link...
[09:57] <carlos> jordi, https://launchpad.net/rosetta/groups/gnu-translators/
[09:57] <SteveA> carlos: got any review for me to do?
[09:57] <carlos> SteveA, near there, I fixed all issues I got yesterday
[09:57] <carlos> fixing tests
[09:57] <carlos> well... running tests
[09:57] <carlos> ;-)
[09:57] <carlos> SteveA, btw, is the karma test fixed on rocketfuel?
[10:00] <jordi> carlos: file a bug :)
[10:00] <sivang> I am wondering, is the Karma on launchpad supposed to dwindle as time passes?
[10:01] <ajmitch_> yes
[10:06] <BjornT> carlos: stub fixed the karma test yesterday
[10:06] <carlos> BjornT, ok, thanks
[10:17] <carlos> jordi, which kind of permissions should mailman have?
[10:18] <carlos> Structured or Closed?
[10:18] <carlos> I guess Closed, right?
[10:22] <jordi> nuclear safe :)
[10:23] <lifeless> sivang: yes.
[10:23] <carlos> jordi, done
[10:24] <sivang> lifeless: ok, such that contribution should be maintained over time and not stick I suppose?
[10:26] <jordi> carlos: great
[10:33] <carlos> jordi, you will need to create the GNU translation teams
[10:33] <carlos> jordi, setting as the owner the official coordinator for GNU 
[10:33] <carlos> jordi, and setting the team as closed so only the owner can add new members
[10:34] <jordi> yes
[10:58] <SteveA> carlos: i'm sure there will be other changes based on the new review.  is there other stuff you can be doing while i review it and get it back to you, or would you rather that i did the review in chunks so that you can address one set of points while i continue reviewing?
[10:59] <carlos> SteveA, I have other things to do so I can wait for a full review
[10:59] <SteveA> ddaa: how's work going?
[10:59] <zyga> morning carlos 
[10:59] <carlos> zyga, morning
[10:59] <carlos> SteveA, does it mean you want the diffs now? 
[10:59] <ddaa> Little by little. Still reading morning email.
[10:59] <carlos> SteveA, the notification thing broke many tests and I'm still fixing some of them
[11:00] <ddaa> SteveA: once I'm done I think I'll start doing the importd2bzr tests that require a sandbox and sample archive (registration and checkout).
[11:01] <ddaa> SteveA: BTW, I hit again the problem of wanting to change sample data with permission of a user different from the user running the rest of a doctest.
[11:01] <ddaa> I hacked something with login(), but I think that's wrong.
[11:02] <ddaa> actually, not hacked... I do login('foo.bar@canonical.com'), make sample data changes, then login(ANONYMOUS), go on with the test.
[11:02] <ddaa> Is that supported?
[11:03] <SteveA> ddaa: yes, you can do that
[11:03] <ddaa> SteveA: I'm not too sure how that will interact with the fact that those tests shoud be run with a specific dbuser...
[11:04] <ddaa> but... well, I think I we can leave by running importd2bzr as the importd user, even if it does not need all the privs.
[11:04] <ddaa> * tha we can live
[11:33] <siretart> hi folks
[11:33] <ajmitch_> hey siretart 
[11:33] <siretart> can please anyone explain me why this bug: https://launchpad.net/distros/ubuntu/+source/pure-ftpd/+bug/3172 does not appear on this list: https://launchpad.net/people/motureviewers/+assignedbugs
[11:33] <Ubugtu> Malone bug #3172: ftpd (Ubuntu) - pure-ftpd does not install correctly. In: pure-ftpd (Ubuntu), Severity: Normal, Assigned to: MOTU Reviewers Team, Status: PendingUpload http://launchpad.net/bugs/3172
[11:34] <ajmitch_> siretart: it also doesn't show up on https://launchpad.net/distros/ubuntu/+source/pure-ftpd/+bugs
[11:34] <ajmitch_> (probably due to pendingupload)
[11:35] <ajmitch_> it's part of pending upload & fixed bugs not showing up, I think
[11:35] <siretart> I don't understand why 'pending upload' bugs should not appear on bug lists
[11:36] <ajmitch_> something to bug bradb about :)
[11:36] <siretart> hm. he is currently not here.. let's wait for him
[11:45] <carlos> jordi, will you answer the request for evolution's doc import?
[11:46] <SteveA> it all depends who the bug list is for
[11:47] <SteveA> if it is for a programmer or package maintainer then probably they're not immediately interested in "pending upload" bugs
[11:47] <SteveA> if it is for a person reporting a bug, looking for dupes, then they are interested in a "pending upload" bug, because the bug is not fixed in any usable version of the software.
[11:50] <siretart> SteveA: is there any possibility to get a list of all bugs assigned to a group which are 'pending upload' or 'new' or 'accepted'?
[11:51] <jordi> carlos: I wanted to talk about that with you
[11:51] <ajmitch_> probably just a matter of adding the advanced search to the lsit
[11:51] <jordi> carlos: do you plan to autoimport that as well?
[11:51] <carlos> jordi, if GNOME's CVS has it, yes
[11:51] <siretart> is there any 'advanced search' to bug lists at all?
[11:52] <ajmitch_> siretart: yes, for /distros/ubuntu/+bugs
[11:52] <jordi> carlos: GNOME CVS has xml
[11:52] <jordi> carlos: but danilo's webpage has templates
[11:52] <ajmitch_> siretart: it shows 92 currently pendingupload
[11:52] <jordi> http://kvota.net/doc-l10n/by-modules.html
[11:53] <jordi> thisi s what currently uses gnome-doc-utils
[11:53] <jordi> evo not included, as expected
[11:53] <carlos> jordi, hmmm, I suppose we could add them then
[11:53] <carlos> jordi, let's talk about it after my breakfast, please :-P
[11:53] <jordi> ok
[11:53] <carlos> see you in 15 minutes 
[11:53] <jordi> pretty late for breakfast :)
[11:54] <carlos> jordi, I was waiting for my gf to wake up ;-)
[11:54] <carlos> see you
[11:54] <siretart>  Sorry, Launchpad took too long to process your request. :(
[11:55] <SteveA> siretart: what URL?
[11:56] <jordi> laters
[11:56] <siretart> SteveA: I was trying to use https://launchpad.net/distros/ubuntu/+bugs-advanced. I want a list of all bugs assigned to motureviewers which are in state 'New' 'Needinfo' 'Accepted' or 'Pending Upload'
[11:56] <siretart> SteveA: would it be possible to add such a query to the motureviewers group?
[11:57] <SteveA> there are two things going on here, i expect
[11:57] <SteveA> the first is that the particular choice of query you're using on the +bugs-advanced page could do with better indexes
[11:57] <ajmitch_> siretart: it just worked for me, but fails sometimes
[11:57] <SteveA> so that the query doesn't time out.  please file a bug saying what query you chose, so that someone can reproduce it, and we'll look into what indexes are needed.
[11:58] <SteveA> the second is that you're asking to be able to use a preconfigured query, i think.  so, that's another bug, and one for brad to look at.
[11:59] <siretart> SteveA: should I ask brad here when he arrives or should I file a malone bug and assign it to him?
[11:59] <SteveA> file a bug anyway, but you can also talk with brad
[11:59] <SteveA> if we have the bug, then it's in the system
[12:00] <siretart> ok. filing bug now
[12:00] <SteveA> thanks
[12:04] <siretart> is bradb brad.bollenbach@gmail.com?
[12:05] <siretart> obviously yes. never mind
[12:05] <siretart> filed as malone bug #5163
[12:05] <Ubugtu> Malone bug #5163: Need advanced query on MOTUReviewers group In: malone (upstream), Severity: Normal, Assigned to: Brad Bollenbach, Status: New http://launchpad.net/bugs/5163
[12:08] <matsubara> good morning!
[12:09] <niemeyer> Morning!
[12:24] <cprov> morning guys
[12:24] <cprov> niemeyer:  could you please join ##soyuz1.0
[12:24] <niemeyer> Sure
[12:25] <carlos> jordi, ping
[12:55] <kiko> lifeless?
[12:55] <lifeless> kiko: ?
[12:55] <kiko> lifeless, any ETA on fixing the merge issue (crashing with a missing revision)?
[12:55] <kiko> it's basically blocked my work on the gina branch
[12:56] <lifeless> kiko: its essentially top priority
[12:56] <kiko> (yes I realize I can pull out patches and apply to a new tree but I am avoiding that because there are a ton of diffs and logs I don't want to replay)
[12:56] <lifeless> I've a new bzr I'm testing for it at the moment
[12:56] <lifeless> had some versioning issues getting it onto chinstrap w/out fucking everything
[12:57] <lifeless> wheres your branch published ?
[12:57] <kiko> SteveA and I had discussed having a way to know where status is on the various bzr work, do you have a suggestion?
[12:57] <kiko> on chinstrap, kiko/launchpad/gina-production IIRC
[12:57] <lifeless> I'm not sure what you mean about status on bzr work
[12:57] <kiko> chinstrap.ubuntu.com:/home/warthogs/archives/kiko/launchpad/gina-production/
[12:57] <kiko> well
[12:58] <kiko> there's a bunch of improvements and fixes to "bzr the platform" that are high-priority for the launchpad team.
[12:58] <lifeless> like ?
[12:58] <kiko> pqm move to balleny
[12:58] <lifeless> Sorry I spoke with steve yesterday
[12:58] <kiko> fixes to various bugs reported 
[12:58] <lifeless> and there was only one show stopper we discussed
[12:59] <lifeless> which is the one affecting you/jamesh/ a couple of other people
[12:59] <Kinnison> lifeless: feature request for pqm's web ui... some indication of when pqm started the currently active job would be really nice
[12:59] <kiko> okay, SteveA may have forgotten or chosen to talk about this at another time
[12:59] <lifeless> everything else is moving along as it moves along
[12:59] <lifeless> I don't mean that in an offhand manner
[12:59] <kiko> I would like a better plan or status report than that
[01:00] <lifeless> I mean that specific things like 'enable pqm on balleny' are in hand
[01:00] <lifeless> And other specific things like 'implement archives' are much more vague, and I don't know which things interest you and which don't.
[01:01] <SteveA> kiko: lifeless and I have discussed that the current serious can't merge bug was top priority for lifeless, and that the move of pqm to its own server was what was top priority before the merge bug appeared.
[01:01] <lifeless> so I suggest that you treat the me & bzr the way the sysadmins team is treated: have a wiki page on wiki.lp.canonical.com where the things that matter to you right now are listed
[01:01] <kiko> lifeless, we need WAY better response time than the sysadmins team
[01:01] <kiko> so I'm not sure what the best way to go about this is
[01:01] <SteveA> kiko: i don't think that's helpful
[01:01] <SteveA> we get really good response time from the admins
[01:02] <SteveA> when it matters, and when we communicate that it matters
[01:02] <kiko> perhaps once the major blockers (this bugfix, move to balleny) are done we'll feel less need for this.
[01:02] <kiko> I don't agree, SteveA, but we don't need to dicuss that now.
[01:03] <lifeless> so kiko, what is the issue - do you feel you are getting bad response times? or are you stressed that a bug has not been fixed (it takes 2 hours to do a single analysis run on it you know)
[01:03] <kiko> you're getting me sidetracked :-) so let me get back to my original question: is there an ETA on the fix for the bug.
[01:03] <lifeless> kiko: I wish there was - no.
[01:03] <kiko> okay. is it something that might take a week or a month?
[01:04] <lifeless> currently I have this new bzr I'm testing then once I have it passing with all the moving bits - pqm, cm, email, bzrtools - I will be will rolling it out
[01:04] <lifeless> next day or two I sincerely hope. I know its a huge impediment to have a branch you can't merge.
[01:04] <kiko> it's okay, I just want to know when to schedule a retry (because it takes 7 hours to run)
[01:04] <kiko> so I'll do that on thursday at the earliest
[01:04] <kiko> thanks
[01:04] <lifeless> it should only take that time once
[01:05] <kiko> unless it fails :)
[01:05] <lifeless> unless you are throwing away the dir afterwards
[01:05] <kiko> oh?
[01:05] <lifeless> even if it fails it should have updated the weaves
[01:05] <kiko> interesting. 
[01:05] <SteveA> how do we get the updated weaves onto chinstrap?  does that even make sense?
[01:05] <kiko> and as for the PQM move, ETA would be probably sometime next week, lifeless?
[01:05] <SteveA> i mean, it seems odd to have everyone doing tons of reweaving
[01:05] <lifeless> SteveA: we will be doing that in the big-upload-of-all-old branches
[01:05] <lifeless> SteveA: but we should not do that before then
[01:06] <SteveA> and, that's after the pqm move?
[01:06] <lifeless> SteveA: otherwise we will cause more reweaving than is needed
[01:06] <lifeless> thats after the pqm move
[01:06] <lifeless> pqm is running on balleny, its just not fully usable yet.
[01:06] <Kinnison> stub: ping?
[01:06] <lifeless> elmo fixed the issues a few hours ago in theory, so I'm testing it again shortly. Tomorrow morning I expect
[01:07] <kiko> okay, that's what I wanted to know for now. thanks.
[01:11] <jordi> carlos: pong
[01:11] <salgado> SteveA, I have a new dependency for launchpad development (python2.4-pylib, to run the sqlobject tests), but the package is available only in breezy. should I add it now to the RocketFuelSetup wikipage or wait until pqm moves to balleny and we can have that package installed on the pqm box?
[01:11] <ulinskie> hi!
[01:12] <ulinskie> anybody knows how I can digitally sign my UBUNTU code of conduct?
[01:12] <SteveA> salgado: does the package work on hoary?
[01:12] <salgado> SteveA, it should work, as that's where I built it the first time
[01:13] <SteveA> salgado: we should get mdz to make it a dependency of the launchpad dependencies package
[01:13] <SteveA> and then get that package installed on the pqm machine
[01:14] <SteveA> lifeless / elmo: does balleny run breezy?  are we using mdz's launchpad dependency package on there?
[01:14] <salgado> SteveA, yes, it does run breezy. I asked elmo already
[01:15] <niemeyer> salgado: pylib?
[01:15] <kiko> open question: has anyone had test failures in PQM in buildd-scoring ou person.txt?
[01:15] <kiko> we're seeing apparently random failures
[01:15] <kiko> lifeless, you apparently told niemeyer this had to do with postgresql?
[01:15] <niemeyer> salgado: Managed to convince them it makes sense? :)
[01:15] <SteveA> kiko: BjornT was looking at that i think
[01:16] <salgado> niemeyer, actually it was uploaded to breezy when I asked doko. it was rejected in debian only
[01:16] <niemeyer> salgado: Ah, I see.. :(
[01:16] <salgado> only in debian, even
[01:16] <niemeyer> salgado: Quite odd
[01:16] <salgado> I only noticed it last week
[01:17] <lifeless> kiko: I didn't, stevea said it might be a 8.0 vs 7.4 thing
[01:17] <kiko> ah.
[01:17] <kiko> sucks to be us.
[01:18] <SteveA> ulinskie: hello
[01:18] <SteveA> ulinskie: do you use GPG ?
[01:19] <Kinnison> SteveA: was my review response enough to get r=stevea on the buildd/uploader branch?
[01:19] <SteveA> Kinnison: if i didn't explicitly say so in my response, then no
[01:19] <ulinskie> I just created minde
[01:19] <ulinskie> I just created mine this morning
[01:19] <Kinnison> SteveA: okay
[01:19] <Kinnison> SteveA: thanks
[01:20] <ulinskie> and was trying to figure out how to digitally sign the ubuntu code of conduct with it
[01:21] <SteveA> Kinnison: did i reply to your last response?  i mean, who spoke last in the thread?
[01:21] <Kinnison> SteveA: I did
[01:22] <SteveA> ulinskie: don't go!
[01:22] <SteveA> Kinnison: then i owe you a reply
[01:22] <Kinnison> SteveA: okay, just thought I'd ask
[01:24] <BjornT> SteveA, kiko: i haven't looked at any buildd-scoring failures, i looked only slightly on the person.txt failure, but stub fixed it.
[01:24] <kiko> okay, 50% cool
[01:24] <BjornT> kiko: what is the error exactly?
[01:24] <SteveA> kiko: dude, you're at least 70% cool
[01:24] <kiko> I'll ask niemeyer to mail the mailing list
[01:25] <Kinnison> buildd scoring failures?
[01:25] <kiko> yeah
[01:27] <SteveA> jamesh: hello
[01:27] <SteveA> or spiv perhaps
[01:27] <Kinnison> kiko: erk
[01:29] <kiko> niemeyer is sending the email
[01:31] <carlos> jordi, so... evolution documentation...
[01:34] <spiv> SteveA: Hello
[01:34] <kiko> hey spiv
[01:35] <SteveA> hi spiv.  can you give some advice to Kinnison about testing a buildd related script?
[01:35] <spiv> Sure.
[01:35] <SteveA> he has sent email
[01:37] <spiv> Yep, got it.
[01:38] <SteveA> ta
[01:40] <lifeless> night
[01:46] <jordi> carlos: yeah
[01:46] <kiko> hey jordi my man
[01:47] <jordi> carlos: evo hasn't adopted gnome-doc-tools it seems
[01:47] <spiv> jamesh: Hmm, the pending-reviews is lacking diffs...
[01:47] <jordi> so there is no "official" pot
[01:47] <jordi> kiko: YOU are the man
[01:48] <kiko> I want some official pot
[01:48] <kiko> and if you can't get me any can I at least have some glue
[01:48] <jordi> I want some pot
[01:48] <jordi> hah
[01:49] <salgado> kiko, I know you're very good at this... can you suggest better names for this: https://chinstrap.ubuntu.com/~dsilvers/paste/fileDti2ia.html ?
[01:49] <carlos> jordi, then I don't think we should import it
[01:50] <carlos> jordi, ask him to bore upstream to move to use gnome-doc-tools
[01:51] <kiko> salgado, flattery could get you in trouble
[01:51] <spiv> Kinnison: Have you see canonical.librarian.ftests.test_gc?
[01:52] <Kinnison> spiv: No, I haven't
[01:52] <kiko> salgado, maybe getRequestedCDCount(self, country=None, approved=False) -- ?
[01:52] <spiv> Kinnison: I think the right approach to testing the buildd slave scanner would probably look quite similar.
[01:53] <Kinnison> spiv: I'm just merging into my buildd fixes branch and I'll take a look, thanks
[01:53] <spiv> Kinnison: test_gc is the tests for LibrarianGarbageCollection, which is a cronscript.  test_gc tests the functions directly, and also invokes the cronscript.
[01:53] <spiv> Kinnison: I think it's a similar, if slightly simpler, situation to yours.
[01:54] <Kinnison> spiv: thanks
[01:56] <jordi> carlos: ok
[02:01] <salgado> kiko, looks okay. I'll try that
[02:03] <kiko> salgado, the "or" there is what makes me uneasy about Requested, but if you think it's okay..
[02:03] <elmo> grr, why does the launchpad-dev pkg recommend the lunachpad-database-dev package?  it massively reduces it's usabilityt
[02:06] <Kinnison> it does?
[02:07] <elmo> % apt-cache show launchpad-dependencies | grep Recommends
[02:07] <elmo> Recommends: launchpad-database-dependencies
[02:07] <stub> Kinnison: pong
[02:07] <elmo> the ratio of machines that need database to that don't in the DC is something like 8:1
[02:07] <Kinnison> stub: Librarian GC fails when it has to merge > 2 aliases pointing at the same content
[02:08] <Kinnison> stub: as a stop-gap to get it running on dogfood I subbed the comma for a percent on the relevant cur.execute
[02:08] <Kinnison> stub: this is running from allegedly up-to-date rocketfuel
[02:11] <Kinnison> stub: basically the cursor is turning "123, 456" into a quoted string in the interpolation, rather than a list of numbers
[02:12] <Kinnison> stub: so it needs to be interpolated in first, or else you need to use an SQL function in the query to split it apart again
[02:12] <stub> Kinnison: Hmm... I could have sworn that was fixed and merged in.
[02:13] <stub> Kinnison: It is fixed on the version I'm running on staging anyway
[02:14] <Kinnison> stub: How did you fix it?
[02:14] <stub> Kinnison: Do you have BATCH_SIZE = 1 near the top of lib/canonical/librarian/librariangc.py ?
[02:14] <stub> Kinnison: The fix it to replate ', vars()' with '% vars()'
[02:16] <Kinnison> No I don't
[02:16] <Kinnison> and yes, I did that replace
[02:21] <stub> Kinnison: Hmm... then either my changes havn't landed or you aren't up to date enough
[02:23] <stub> I just pushed that branch to pqm again
[02:25] <jordi> carlos: there are more updates to do in the DB
[02:26] <carlos> jordi, which ones did I miss?
[02:26] <stub> Send 'em through - I was about to do the one Carlos emailed me
[02:26] <jordi> carlos: wait
[02:26] <jordi> ok
[02:27] <carlos> jordi, I saw another plural form request but I asked form more info. And the spokenin update is also pending for more info
[02:27] <carlos> s/form/for/
[02:28] <jordi> carlos: also pending: add ubuntu-l10n-mr to ubuntu translators
[02:29] <carlos> jordi, I can do that from the website
[02:29] <jordi> The other one with pending info, is it "Sng"?
[02:30] <carlos> don't think so, no
[02:30] <carlos> when was it send?
[02:30] <stub> So would the buildd's scoring system (whatever it is scoring) degrade scores over time by any chance?
[02:30] <jordi> Oct 23
[02:30] <carlos> jordi, btw, which language is mr ?
[02:30] <stub> cprov, Kinnison: ^^^
[02:31] <cprov> stub: the opposite, older builds in queue has higher scores. 
[02:32] <jordi> Marathi
[02:32] <jordi> carlos: the spoken in thing, is it the Sri Lanka request?
[02:32] <cprov> stub: this way we enforce they get processed ASAP. Did I missunderstand your question ?
[02:33] <jordi> that would be Sinhala (si) and Tamil (ta)
[02:33] <jordi> Ah, I see you replied
[02:33] <jordi> Sinhala = si, yes
[02:33] <carlos> jordi, yes
[02:33] <jordi> Sinhala and Sinhalese are equivalent
[02:33] <jordi> so you can add that to the DB
[02:33] <carlos> ok
[02:34] <kiko> stub, for soyuz testing, is it okay that we are testing using the development config and not the production config? 
[02:34] <stub> cprov: Oh - just thinking that  niemeyer 's problem with the buildd tests failing on him might have had the same cause as the karma tests failing on me the other day. If the score calculation changes, then the scores of the sampledata items will change over time. And tests that encode the scores will start failing.
[02:34] <jordi> did you find Sng?
[02:35] <stub> kiko: What bits of soyuz? The config you are using shouldn't matter (and you couldn't use production config anyway since that will only work for production systems)
[02:35] <cprov> stub: the score tests have relational interval times, NOW - X seconds 
[02:35] <kiko> stub, ftp, upload queue and publish bits
[02:36] <cprov> stub: I have the same karma errors and the wrong about that was the current karma datecreated not a 2025 pattern as the other, it's time sensible too
[02:36] <jordi> carlos: according to the guy, Sango
[02:36] <jordi> n == 1 ? 0 :  (n == 0 ? 0 : 1)
[02:36] <jordi> is this
[02:36] <jordi> but, doesn't that look wrong?
[02:36] <stub> kiko: Should all be fine.
[02:37] <kiko> thanks stub 
[02:37] <salgado> cprov, it's weird though, that no merges should be going through pqm if that was the cause of the karma test failing
[02:37] <stub> cprov: Fix for the karma has landed now, along with a test to ensure bad karma doesn't end up in the sample data again.
[02:38] <stub> Sounds like you haven't merged from rocketfuel recently
[02:38] <carlos> jordi, it's the same as 'n > 1'
[02:38] <carlos> jordi, I answered that email already
[02:38] <cprov> salgado: yes, it's weird 
[02:38] <carlos> preparing the spokenin patch so Stuart can apply all them at the same time
[02:39] <jordi> carlos: heh, right
[02:39] <jordi> ok
[02:39] <jordi> carlos: I haven't got your reply yet then
[02:40] <cprov> stub:  "Sounds like you haven't merged from rocketfuel recently" ... me ? why ?
[02:41] <carlos> stub, I have the other sql commands, do you want them by email ?
[02:42] <stub> carlos: whatever ;)
[02:43] <salgado> hi daf. are you around?
[02:43] <Kinnison> +s
[02:43] <daf> salgado: hi
[02:44] <salgado> daf, have you tried merging the accounts again, after you got the timeout?
[02:44] <daf> I'll do that now
[02:44] <carlos> jordi, stub https://chinstrap.ubuntu.com/~dsilvers/paste/fileK7R7hn.html
[02:44] <daf> I tried to merge two accounts at the same time
[02:44] <daf> one of them worked, one of them timed out
[02:45] <kiko> hey daf
[02:45] <kiko> do you have a traceback for the timeout?
[02:45] <salgado> oh, that could be the problem then
[02:45] <daf> could be
[02:46] <daf> it happened when Launchpad detected accounts from my GPG key (clever!) and suggested I merge them
[02:46] <daf> aha, it timed out again
[02:46] <salgado> daf, did you get a traceback this time?
[02:47] <daf> remind me -- how do I get a traceback?
[02:47] <daf> it just says "orry, Launchpad took too long to process your request. Blah blah blah."
[02:47] <daf> well, not literally
[02:48] <carlos> jordi, ubuntu-l10n-mr added. Anything else?
[02:48] <salgado> daf, all members of the launchpad team should see tracebacks in the error pages
[02:48] <daf> mm
[02:49] <spiv> salgado: https://launchpad.net/errors/showEntry.html?id=1133271955.030.903115280185 looks like it.
[02:49] <jordi> carlos: not right now, I think.
[02:49] <salgado> spiv, yeah, I got it too
[02:49] <daf> interesting -- I am logged in
[02:49] <spiv> Assuming you get the right appserver, I guess...
[02:49] <SteveA> daf needs adding to some team, i expect
[02:49] <SteveA> jamesh's error reporting stuff should land soon
[02:49] <daf> I'm a member of Launchpad Administrators and Launchpad Developers
[02:49] <SteveA> which will improve our QA a HUGE amount
[02:49] <daf> ooh
[02:49] <carlos> jordi, ok
[02:49] <daf> sounds good
[02:50] <SteveA> daf: you should see TBs.  try going to a 404 URL
[02:50] <jordi> carlos: quim unsubbed from rosetta :)
[02:50] <salgado> daf, stub, https://chinstrap.ubuntu.com/~dsilvers/paste/filexYywyj.html
[02:50] <carlos> jordi, yeah, I know
[02:50] <jordi> hey daf!
[02:50] <salgado> daf, the problem could be because you have too much stuff to merge or because the posubmision table is locked by another transaction
[02:50] <salgado> stub, what do you think?
[02:50] <daf> *cough* "Accept expired security information?
[02:50] <daf> The security information for "chinstrap.warthogs.hbd.com" expired on Sat 28 May 2005."
[02:51] <carlos> jordi, dude, you are completely blind!  daf was here for a while already! :-)
[02:51] <spiv> salgado: Hmm, there's an index on posubmission.person, so updates to that column are likely to be very expensive.
[02:51] <daf> hey jordi :)
[02:51] <jordi> carlos: I was not looking you know :)
[02:51] <jordi> there's this thing called lliurex
[02:51] <jordi> carlos: besides, he's in the Other Secret Channel :)
[02:52] <carlos> jordi, which one?
[02:52] <carlos> jordi, tell me its name here, no one else will see it, really...
[02:52] <carlos> ;-)
[02:53] <daf> spiv: aye, that must be it
[02:54] <jordi> carlos: mako's channel.
[02:54] <jordi> #yukidoke
[02:54] <stub> daf: Something else was holding open a lock on posubmission or person. Did this just happen?
[02:54] <spiv> I guess person merge is likely to be one of our most expensive operations, because it can write to a fairly large number of rows in many tables.
[02:54] <salgado> spiv, that's bad. we should have indexes on the foreign key to person on other tables that are updated when merging accounts
[02:54] <SteveA> we could do a queue for it
[02:55] <SteveA> timemachine!
[02:55] <salgado> they already did. :-(
[02:55] <kiko> salgado, SteveA even beat you to the message!
[02:55] <SteveA> let's do the timewarp again
[02:55] <spiv> Of course, guessing about performance is likely to be completely wrong...
[02:55] <kiko> dude I want salgado to READ MY EMAIL
[02:55] <spiv> So I should stop speculating :)
[02:55] <SteveA> adding indexes may be bogus
[02:55] <SteveA> because they need to be updated with the updates
[02:56] <SteveA> so, although you can find the information quicker, it will take longer to change
[02:56] <jordi> going home
[02:56] <daf> I think finding happens much more often than updating in this case
[02:56] <jordi> carlos: will work on the mail queue today
[02:56] <jordi> carlos: whe do you leave?
[02:56] <carlos> jordi, cool
[02:57] <carlos> jordi, around 18:00 local time
[02:57] <jordi> today, until when?
[02:57] <carlos> oh
[02:57] <carlos> holidays?
[02:57] <jordi> kiko: you're going to love my zillions of requestss to populate the GNU translators group :)
[02:57] <jordi> yeah
[02:58] <jordi> oh, so it's not today
[02:58] <jordi> :)
[02:58] <carlos> no
[02:58] <carlos> ;-)
[02:58] <jordi> ok
[02:58] <jordi> nm then
[02:58] <carlos> From Monday till Friday both included
[02:58] <jordi> k
[02:58] <jordi> woah man
[02:58] <jordi> mega-puente
[02:58] <carlos> jordi, but wasting holidays
[02:58] <carlos> ;-)
[02:58] <jordi> I'll probably be in Barcelona during the 2nd weekend.
[02:58] <jordi> "wasting"
[02:59] <carlos> jordi, ok
[02:59] <kiko> jordi, do you want to come to brazil for xmas?
[02:59] <carlos> :-D
[02:59] <jordi> not "wasting" if you have fun :)
[02:59] <jordi> kiko: will it be BORING?
[02:59] <kiko> YES and it will include NAKED ALIENS
[02:59] <jordi> kiko: actually, maybe yes.
[02:59] <carlos> jordi, "puente" is when you don't get holidays but just free days :-D
[02:59] <jordi> OH DUDE
[02:59] <jordi> carlos: I need to talk about the "puente" here :)
[02:59] <jordi> laters
[02:59] <kiko> jordi, look for tickets, week before and after xmas 
[03:00] <jordi> 15.00. The magic time of the day.
[03:00] <spiv> Oh, that's unfortunate.  I can see daf's password in the error report :/
[03:00] <jordi> kiko: let's talk about this later today
[03:00] <kiko> cool
[03:00] <carlos> jordi, I don't think Mark will agree with that "Spanish" concept :-P
[03:00] <jordi> heh
[03:00] <daf> oh, nice :)
[03:01] <kiko> spiv, I can see jordi naked in ANOTHER error report
[03:01] <spiv> daf: Because it's part of the form, I guess.
[03:01] <salgado> that's right, because your password is in the request.form
[03:01] <daf> spiv: yep
[03:01] <salgado> OMG, everybody is beating me today
[03:01] <carlos> kiko, dude, you didn't show me the pornlet!!!
[03:01] <spiv> kiko: I think I'll pretend I never heard that.
[03:01] <jordi> "mark, so... we have this thing called Siesta. We will be on stand-by mode from 14:00 to 17:00 every day. And then, there's the "bridges".
[03:01] <spiv> salgado: It's your slow internet ;)
[03:02] <carlos> jordi, WTF? three hours????
[03:02] <carlos> jordi, dude!!!
[03:02] <salgado> but it's fast now. two times faster than it was
[03:02] <salgado> and even so, it's still slow. :-(
[03:02] <jordi> carlos: a good siesta, :)
[03:02] <daf> bandwidth vs. latency?
[03:02] <jordi> ok, off
[03:03] <kiko> salgado, IRC goes over the bad-latency link
[03:03] <salgado> even the latency is good (okay, okay, acceptable) now
[03:03] <salgado> ah, that's true
[03:15] <jamesh> cprov: if you can reliably reproduce the crash BjornT brought up, could you try the steps I mentioned in the followup email?
[03:15] <cprov> jamesh: sure, I'm on it 
[03:16] <stub> Kinnison: BATCH_SIZE = 1 can be found in HEAD's lib/canonical/librarian/librariangc.py, which indicates you are not merged up as much as you think you are.
[03:18] <kiko> "What I do know, is that these fucking idiots claim there is free wireless in this spot of Terminal 2F, but it's not true at all."
[03:20] <stub> Free wireless in an airport? I'll believe that when the stop asking for my credit card :-/
[03:20] <jamesh> stub: Changi airport in Singapore has free wireless
[03:21] <jamesh> (well, it did when I was there last time)
[03:21] <stub> web browsing only, or actual internet?
[03:21] <jamesh> full internet access
[03:22] <stub> cool
[03:22] <BjornT> vilnius airport has free wireless as well :)
[03:22] <jamesh> you need to pay if you don't have your own laptop/wireless card though
[03:35] <SteveA> https://wiki.launchpad.canonical.com/DevelopmentMeeting20051124  <-- summary of thursday's meeting
[03:37] <SteveA> carlos: how's that diff going?
[03:38] <bradb> BjornT: Has that change landed to make control commands automatically operate on the correct task if there only is one task?
[03:39] <BjornT> bradb: no, i'm working on it now. should be able to submit it for review today or tomorrow.
[03:39] <bradb> ok, cool
[03:45] <salgado> daf, will you try that merge again, for us to check if it fails in the same query?
[03:47] <daf> timed out again
[03:47] <daf> by the way: s/bellow/below/ on the merge form
[03:49] <stub> bugger
[03:50] <daf> the other merge I did at almost the same time worked fine first time
[03:50] <salgado> now it failed when merging the pofiles
[03:50] <daf> hmm, maybe it's just that that account has a huge amount of stuff to merge?
[03:51] <salgado> I guess so
[03:51] <kiko> you should have been less active
[03:51] <daf> sorry kiko 
[03:51] <jamesh> BjornT: if it's not too much trouble, could you try running gdb again but slightly differently?
[03:51] <salgado> kiko, do you have access to the staging db?
[03:52] <jamesh> BjornT: instead of waiting for a core file, run "gdb python2.4", then at the gdb prompt type "run -vvf --test=..."
[03:52] <salgado> kiko, I'll try merging your account into mine, to see if it fails while reassigning all the bugs you reported to me
[03:52] <jamesh> BjornT: it should drop back to the gdb prompt at the crash, where you can do a back trace
[03:53] <BjornT> jamesh: sure, i'll do that.
[03:55] <jamesh> BjornT: if the trace is identical then don't bother reposting it
[03:56] <jamesh> I'm just wondering if it is a case of unreliable symbol names due to stripping or a corrupt stack
[03:57] <kiko> salgado, I do have access to the staging db, and don't you go stealing my karma
[03:58] <salgado> kiko, it's on staging. stub will give your karma back tomorrow. :)
[03:58] <BjornT> jamesh: the backtrace is identical (both bt and pystack)
[03:58] <jamesh> BjornT: okay, thanks.
[03:59] <sivang> kiko: you're hosting an Xmas party with naked aliens? :-)
[04:00] <kiko> salgado, I think staging isn't being updated
[04:00] <kiko> sivang, maybe.
[04:00] <kiko> it depends if they accept my invitation
[04:00] <sivang> kiko: rotfl, girl aliens?
[04:01] <kiko> bradb, my comment to bug 3620 refers to the guided-bug-entry form that we specified in, uhm, JANUARY 
[04:01] <Ubugtu> Malone bug #3620: Default bug list should include Pending Upload bugs In: malone (upstream), Severity: Normal, Assigned to: Nobody, Status: NeedInfo http://launchpad.net/bugs/3620
[04:02] <carlos> SteveA, I think I will have it ready in next hour or so, I think I fixed most productseries issues I got after move it to the new notification code
[04:02] <salgado> kiko, I just realized it's impossible to even request a merge in staging --it times out
[04:04] <kiko> salgado, argh
[04:04] <bradb> kiko: I think much of the UI design from UbuntuWayBelowZero would still stand up to today's use cases
[04:05] <kiko> bradb, for the record, I don't really like the idea of a footer for bugmail
[04:05] <salgado> stub, https://chinstrap.ubuntu.com/~dsilvers/paste/filepJRPMZ.html <= this query is timing out in staging. is it possible to optimize it?
[04:06] <bradb> kiko: Why not?
[04:06] <kiko> bradb, it generated a lot of negative feedback in bugzilla, basically because it looks bad and doesn't answer the imporant question. I'd suggest doing it as another header
[04:06] <kiko> s/the imporant/any important/
[04:07] <kiko> bradb, if you added a header it would be easily filterable and still provide what you want
[04:07] <bradb> kiko: A Malone email is pretty hard to identify unless you know what to look for but if, for someone like elmo, you (reasonably) don't even have any idea why you're receiving the email to begin with, it's just downright ARGH.
[04:08] <bradb> kiko: A header would not have solved the problem that was reported.
[04:08] <kiko> bradb, bug 3797? it would. read his last comment.
[04:08] <Ubugtu> Malone bug #3797: Bug mails should be more descriptive In: malone (upstream), Severity: Normal, Assigned to: Nobody, Status: NeedInfo http://launchpad.net/bugs/3797
[04:09] <bradb> kiko: elmo didn't comment on that bug.
[04:11] <kiko> bradb, I am really not thrilled by the footer. It is bad UI to say "or", to start off with. but let mpt comment on it
[04:11] <kiko> "You are receiving this message because you are on the Cc list of this bug, or are a member of a team that is, or you are assigned to fix it."
[04:11] <kiko> you /could/ say "This email was send to you by Launchpad's bug tracker"
[04:12] <bradb> kiko: The problem that people were complaining about (not referring to 3797, which is a different issue, probably solved by X-Launchpad-Bug, already) is not knowing why they were receiving this mail.
[04:12] <bradb> kiko: "Why?"
[04:13] <kiko> bradb, I understand the problem you're trying to solve. But saying "A or B or C" isn't a good answer to Why?
[04:13] <kiko> and people have flamed bugzilla for doing it
[04:13] <bradb> kiko: That doesn't mean it's bad.
[04:13] <kiko> it's pretty bad
[04:13] <kiko> if you want a /good/ answer then do the extra work necessary
[04:13] <kiko> find out /why/
[04:14] <bradb> kiko: I could have made it give a more precise answer, and if mpt and others think it's worth the extra work, then I'd more than happily do it.
[04:14] <kiko> You are assigned to this bug.
[04:14] <kiko> You are a member of the MOTU team, assigned to this bug.
[04:15] <bradb> Yes, I fully agree with you that that would be excellent. It just takes about 2 days longer to implement :)
[04:15] <kiko> I just think it's not worth doing anything but that
[04:15] <kiko> (and you could have brought it up on launchpad-list..)
[04:16] <kiko> anyway
[04:16] <kiko> Jorge's reply is reasonable -- the footer doesn't help.
[04:17] <kiko> (doesn't help with bug 3797)
[04:17] <Ubugtu> Malone bug #3797: Bug mails should be more descriptive In: malone (upstream), Severity: Normal, Assigned to: Nobody, Status: NeedInfo http://launchpad.net/bugs/3797
[04:24] <bradb> kiko: Do you want me to fix the footer today? It might be doable, and I think it would be a nice step forward in despamifying bugmail.
[04:26] <kiko> bradb, mpt is in the house, and after lunch I am pretty sure I can get him to give you an opinion on it :-)
[04:26] <bradb> ok
[04:32] <mhz> kiko: bradb: hi. I case you dont't remember me, I had trouble yesterday with 'confirming email' after creating 2 teams in LP
[04:32] <bradb> mhz: How did it turn out?
[04:33] <mhz> no success :)
[04:33] <kiko> bradb, I'd suggest the 5c fix to bug 4595 is just not linkifying it
[04:33] <Ubugtu> Malone bug #4595: Don't auto-linkify non-existent bug reports In: malone (upstream), Severity: Normal, Assigned to: Brad Bollenbach, Status: Accepted http://launchpad.net/bugs/4595
[04:33] <mhz> however, the teams info is there
[04:33] <kiko> mhz, interesting. I need to have lunch now, though. growl. salgado, mhz's problem is in your area -- can you give him some user support?
[04:34] <mhz> kiko: thx again
[04:34] <mhz> and i'm sorry to be bugging so much
[04:35] <bradb> We don't mind being bugged. There's not much point doing this if not for the users.
[04:35] <kiko> nah, I want to fix this, I'm just having a @#!@! busy week
[04:35] <mhz> my major concern is for others who will join. I don't want them to feel this is more complex that they are used  to (mainly teachers) and it does not work properly :)
[04:35] <mhz> bradb: thx
[04:36] <bradb> mpt!
[04:36] <mhz> bradb: BTW, just fyi, so far many of the people who will join are learning moin wiki and they felt it was difficult
[04:37] <bradb> mhz: I agree with them.
[04:38] <bradb> The day I first saw a version of the Moin wiki where you logged in by going to the "UserPreferences" link, I knew there was trouble ahead.
[04:38] <mhz> bradb: hehehehe
[04:39] <mhz> bradb: hence I am starting to work on a KISS how to work with Moin
[04:39] <mhz> bradb: however, that UserPreferences can be 'named' log in if admin wanted to.
[04:40] <mhz> it's a small tweak
[04:40] <bradb> yeah
[04:40] <mhz> actually, there's a theme that uses only login instead
[04:41] <mhz> once I finish that KISS doc, i'd love to help writing a KISS howto LP
[04:41] <mhz> IMHO, LP is a whole new concept for team-work
[04:42] <mhz> .oO(if LP was just 'connected' with Moin...)
[04:46] <salgado> mhz, you had problems creating teams or validating the team email address?
[04:49] <mhz> salgado: the latter
[04:50] <salgado> mhz, was it you who reported https://launchpad.net/products/launchpad/+bug/5122 ?
[04:50] <Ubugtu> Malone bug #5122: Validate email in team In: launchpad (upstream), Severity: Normal, Assigned to: Diogo Matsubara, Status: Accepted http://launchpad.net/bugs/5122
[04:51] <mhz> salgado: nope, but it looks similar
[04:51] <mhz> similar prob, I mean
[04:56] <stub> salgado: It is much faster if you do it as two seperate queries and join them, which can be done by rewriting it to use UNION
[05:07] <bradb> BjornT: I've forwarded you an email about the Malone UI seemingly swallowing some of my bugmail. Any ideas?
[05:22] <BjornT> bradb: hmm. so, sending mail in tests and in production works differently. the thing is that an email notifications was sent, then the transaction was aborted due to an error. (you should have gotten an error message, and no comment should have been added)
[05:22] <BjornT> bradb: i guess i have to ensure that the events are fired off after the email has been fully processed.
[05:23] <BjornT> (the 'no error message' is another issue, which i will be working on in the next few days)
[05:25] <daf> salgado: I tried the merge again and it worked
[05:25] <BjornT> bradb: ah, missed the part that the bug report was swallowed as well, i'll have to look into that
[05:25] <bradb> BjornT: the no error messages thing is particularly deadly. :/ I could be losing a lot of work that I'm doing this morning, for all I know. it's good to know that you're working on it though.
[05:28] <stub> daf: Could you please try that person merge again?
[05:29] <stub> Ahh...
[05:29] <salgado> stub, did you change anything that could make the merge to succeed?
[05:30] <daf> my karma just went way up
[05:30] <stub> daf: do you remember roughly when the merge worked? I made some changes over the last 10 mins.
[05:30] <daf> 16:25ish UTC
[05:31] <stub> salgado: I've created a stack of indexes that should fix it
[05:31] <stub> daf: Ok. I think you managed to do it just as I'd finished adding a load of indexes ;)
[05:32] <daf> heh
[05:33] <daf> I'm probably not exceptional in having a lot of extant translation work under different email addresses, so I imagine we'll hear about it again if it's not fixed
[05:33] <daf> in the meantime, I reckon we can close the bug
[05:34] <salgado> stub, cool. should this be enough or do you think we should do the merges asynchronously?
[05:36] <bradb> BjornT: Also, what do you think of removing the requirement to GPG-sign mail messages?
[05:36] <stub> salgado: The current code will last a while longer
[05:37] <stub> salgado: But we will eventually need to redo it asynchronously.
[05:38] <salgado> fair enough
[05:38] <salgado> stub, did you see my message where I ask you if it's possible to optimize another query?
[05:39] <salgado> this is the query: https://chinstrap.warthogs.hbd.com/~dsilvers/paste/filepJRPMZ.html
[05:40] <BjornT_> bradb: for bug submission, i think we should definitely do it. for editing bugs i'm not sure, i don't think i would mind removing the requirement, but as i remember there was a great deal of discussion last time this was suggested.
[05:40] <BjornT_> bradb: for example, we'd have to special case private bugs
[05:40] <stub> salgado: Yes - I looked at that.
[05:41] <bradb> BjornT: right
[05:41] <salgado> stub, that query fails in staging all the time. that's why I asked
[05:41] <stub> salgado: It is much faster if you split it into two queries, joined with a UNION (select Person.whatever from person,emailaddress where person.id = emailaddress.person and person.fti @@ ftq('kiko') union select person.whatever from person,emailaddress where person.id = emailaddress.person and lower(email) like 'kiko%%')
[05:42] <stub> Oh... and the 'merged is null and teamowner is null' clauses in there too
[05:42] <stub> That query looked like it was generated with SQLObject - can you make SQLObject to the UNION in that fashion?
[05:43] <salgado> stub, right, I'll try to fix this today. would you review it?
[05:43] <stub> salgado: If I'm still up ;)
[05:43] <salgado> stub, btw, I have a small db patch, can you review it now?
[05:44] <salgado> https://chinstrap.ubuntu.com/~dsilvers/paste/filehSe6wg.html
[05:45] <stub> salgado: continents have codes?
[05:45] <stub> Or are you just making them up?
[05:45] <salgado> stub, this db patch comes with sampledata to fill the continent table and the links from the country to the continent table
[05:45] <salgado> stub, I got the codes and everything from http://en.wikipedia.org/wiki/List_of_countries_by_continent_%28data_file%29
[05:46] <stub> Cool - I never realized there was some sort of a standard
[05:46] <stub> I'll need the sampledata as a .sql in database/schema/pending as normal
[05:46] <salgado> I'm not sure that's a standard, actually. but it's something...
[05:47] <salgado> stub, hmmmm. you need a .sql that updates the existing data and inserts new stuff?
[05:47] <carlos> SteveA, ok, I have my branch free of test failures
[05:48] <stub> Approved pending table comments (including a link to that URL). patch-40-07-0.sql
[05:48] <carlos> SteveA, I will prepare the diffs and send you them by email
[05:48] <stub> salgado: ^^^
[05:48] <salgado> stub, or can it have inserts that are supposed to replace existing data
[05:48] <SteveA> carlos: cool.  i'll start the review
[05:49] <stub> salgado: inserts the continent records and updates the country records. What I need to run on production (and probably what you use to update the sample data)
[05:49] <salgado> (before he forgets)
[05:49] <carlos> SteveA, I'm sure we will need another review because I had to change a lot of code and I didn't fix all issues the old code had.
[05:49] <salgado> stub, right, I'll write that too
[05:49] <carlos> SteveA, but If you don't mind... I prefer to fix the issues introduced by me and then, after merge it into rocketfuel, address the others
[05:50] <carlos> so we are sure this update lands into production next week
[05:50] <carlos> SteveA, if you don't mind, of course
[05:52] <SteveA> carlos: i understand what you're saying.  let's see what's in there, and then we'll talk about it.
[05:52] <BjornT> bradb: as for your bug submission problem, i haven't found the exact problem yet, but i'm quite sure it has to do with the attached zip file. if you don't attach a file it should work.
[05:52] <carlos> SteveA, ok
[05:53] <jblack_> Gah. Big windstorm here. Lost power.
[05:55] <carlos> SteveA, hmm, I had to merge from rocketfuel so I'm not sure I can provide you a diff of the changes since the initial review
[05:55] <carlos> any hint?
[05:56] <SteveA> carlos: just give me the diff against RF then, but also mail rob collins to ask the question for future reference
[05:57] <carlos> ok
[05:57] <salgado> stub, the .sql file to update existing data should be left in schema/pending/ right?
[06:10] <carlos> SteveA, mail sent
[06:10] <stub> salgado: yes
[06:12] <salgado> stub, but then the tests will break, no? I have a NOT NULL constraint on Country.continent
[06:12] <salgado> actually, it'll fail to load the sampledata
[06:15] <carlos> see you later
[06:38] <Kinnison> stub: On dogfood, to speed up librarian GC, I added an index on libraryfilealias over the content column
[06:38] <Kinnison> stub: is this a bad thing to have done?
[06:39] <stub> libraryfilealias already has an index on content
[06:40] <Kinnison> It didn't on dogfood
[06:40] <Kinnison> most odd
[06:41] <stub> Kinnison: schema creep.
[06:41] <Kinnison> then it wasn't a bad thing to have done
[06:41] <Kinnison> hurrah
[06:42] <stub> It looks like an initial run on production will take over 5 days. Although disk on the Librarian will be going up to 4TB soon, so we could just not bother for a year or two.
[06:43] <Kinnison> I think it's better to keep running this sort of cleanup
[06:43] <Kinnison> because regardless of how much space we have, keeping the ratio of live data to garbage as sane as possible should be a priority
[06:43] <elmo> FWIW, the morgue on jackass is still only 270GB or so, that's since 2004/06
[06:43] <elmo> all binaries, all source
[06:43] <stub> I suspect you are right. If the time scales linearly the run would take over 20 days if we do let it get up to 4TB ;)
[06:44] <Kinnison> elmo: aye, but the librarian has huge amounts of stuff in it which aren't package-related
[06:44] <Kinnison> elmo: In particular, as rosetta usage ramps up, we'll be creating more and more
[06:44] <elmo> stub: also I've no idea what it is you're doing, but you could probably half that runtime if you wait for the new machine
[06:44] <elmo> the machine itself and the array are like, an order of magnitude faster
[06:44] <Kinnison> elmo: s'mostly database
[06:44] <elmo> oh, meh, well never mind then
[06:45] <Kinnison> It's very cute database stuff
[06:45] <stub> elmo: Removing 10 rows from one of the tables takes 7 seconds. And there are 500,000 dangling waiting to be tidied ;)
[06:46] <elmo> stub: take the constraints off, that'll speed up removing the rows ;-P
[06:46] <elmo> \o/ <-- winn0r
[06:46] <stub> Look ma, no hands!
[06:48] <Kinnison>  Look ma, ain't life grand?
[06:49] <Kinnison>  Didn't I turn out, oh didn't I turn out to be.. everythin' you wanted ma? Ain't you proud of meeeee?!
[06:49] <Kinnison> NP: "Look ma, no hands"
[06:53] <ddaa> There's this not-funny jokes that ends with "Look ma, no theeth!"
[06:55] <ddaa> stub: can you please add bzrtool at revision 261 (abentley branch) into ./sourcecode/bzr/bzrlib/plugins?
[06:55] <ddaa> mh...
[06:55] <ddaa> nevermind that, I'll need to add it locally, but lifeless is supposed to put a custom baz2bzr in launchpad for me.
[07:42] <Kinnison> ciao all
[09:14] <ddaa> The deity you have summoned has not been implemented yet.
[09:16] <lifeless> _keturn: what do you need ?
[09:20] <_keturn> I'm interested in your application of OpenID.  Then I found some braindumps on the spec page, and that helped.
[09:21] <ddaa> Hey lifeless.
[09:22] <ddaa> Hozit goin?
[09:23] <ddaa> lifeless: I'm zeroing in on importd2bzr, I'm going to be blocking on your baz2bzr customisation very soon.
[09:23] <ddaa> lifeless: also, I'd like your opnion on an implementation point.
[09:24] <ddaa> Do you think we should commit directly to the http-served bzr branch, or should we commit to a scratch branch and push to the http-served one?
[09:29] <lifeless> for importd bzr style, I think it should use a temporary branch
[09:29] <lifeless> for the same reasons that we used a wipeable master in the arch world
[09:29] <lifeless> its going well, busy as, but well.
[09:29] <lifeless> re baz2bzr I understand, I'm juggling priorities :[
[09:30] <lifeless> for the baz2bzr run, we don't need to have a temp branch because baz2bzr has that built in.
[09:30] <ddaa> right, that was what I was asking
[09:31] <ddaa> Okay, then I just need to sprinkle a few logging statements, adjust the test suite, and and it's done, I think.
[09:31] <lifeless> cool.
[09:31] <ddaa> But cannot be merged before launchpad has baz2bzr
[09:31] <lifeless> yup
[09:31] <lifeless> You are using the bzrtools-launchpad branch ?
[09:32] <ddaa> hu... no, I stuffed a bzrtool 0.6 in my launchpad checkout.
[09:32] <ddaa> I do not know about bzrtools-launchpad
[09:32] <lifeless> ok, bzrtools 0.6 has teh wrong interface
[09:32] <lifeless> http://people.ubuntu.com/~robertc/baz2.0/bzrtools-launchpad
[09:32] <lifeless> that has the same api to drive that you will want
[09:32] <ddaa> great
[09:33] <lifeless> I'll be branching off that to make the copy in rf
[09:33] <ddaa> you'll be putting it in sourcecode/bzr/bzrlib/plugins, right?
[09:33] <lifeless> no
[09:33] <ddaa> :(
[09:33] <lifeless> sourcecode/bzrtools
[09:34] <lifeless> we try not to nest branches as it makes it unclear
[09:34] <lifeless> there will be a symlink in sourcecode/bzr/bzrlib/plugins to it
[09:34] <ddaa> hu... what will be the symlink then, lib/bzrtools -> ../sourcecode/bzrtools?
[09:34] <lifeless> you will be able to import it as bzrlib.plugins.bzrtools
[09:34] <ddaa> hu?
[09:34] <lifeless> which is its official name
[09:35] <lifeless> hu?
[09:35] <ddaa> able to import it as bzrlib.plugins.bzrtools
[09:35] <lifeless> right
[09:35] <ddaa> how comes?
[09:35] <lifeless> you want to know how to get at the api right ?
[09:36] <ddaa> oh yes, I have enough information for that now. But I'm just wondering how you get bzrtools to appears within a package where it does not live.
[09:36] <lifeless> bzrtools is a plugin, it logically lives at bzrli.plugins.bzrtools
[09:37] <ddaa> sure
[09:37] <lifeless> but to make it easy to see that its in our sourcecode tree we put it at sourcecode/bzrtools
[09:37] <ddaa> sure
[09:37] <lifeless> to make it appear we just symlink it from the bzrlib/plugins dir
[09:37] <lifeless> ln -s ../../bzrtools
[09:37] <ddaa> ha, right
[09:37] <ddaa> so no link to bzrtools in ./lib
[09:38] <lifeless> right
[09:38] <lifeless> 'import bzrtools' is not guaranteed to work.
[09:38] <ddaa> ?
[09:38] <lifeless> if there was a link in lib you would be able to do 'import bzrtools'
[09:38] <ddaa> ha, yes
[09:39] <ddaa> sorry, I read "import bzrlib is not guaranteed to work"
[09:39] <lifeless> hah!
[09:39] <ddaa> Thanks, I'll fix for the right API very soon.
[09:39] <lifeless> np
[09:40] <ddaa> BTW, I have to say I find doctests quite fun to write.
[09:40] <ddaa> less painful than unittest
[09:40] <lifeless> interesting
[09:40] <ddaa> at least when one does not need multiple fixtures
[09:40] <lifeless> I find them great for documenting how an api should be used
[09:40] <lifeless> terrible for doing TDD beyond the trivial 'this works' point
[09:42] <ddaa> But then I'm a bit sick you know. I once wrote a non-trivial gtk+ app using funnelweb (a literate programming tool).
[09:42] <lifeless> ;)
[09:42] <ddaa> I like the literate programming aspect of doctests.
[09:42] <lifeless> so do I 
[09:42] <ddaa> I just wish they had better support for fixtures.
[09:42] <lifeless> well
[09:43] <ddaa> anyway, I'll probably change my mind in a few weeks :)
[09:43] <lifeless> I think that the things that make doctest great - that it makes it easier to write comments than code - cause a diametric tension with the things that make xUnit great - making it easier to write tests than code
[09:43] <lifeless> so what I want to do is to make it easier to switch between them
[09:44] <lifeless> to say 'now I want to document an API'
[09:44] <lifeless> and then 'bugger, I have a bug in the behaviour, I want a unittest with precise semantics
[09:44] <lifeless> *easily*
[09:44] <ddaa> Mh... I see your point.
[09:45] <ddaa> sort of
[09:45] <ddaa> I do not yet have good feel of doctest shortcomings.
[09:45] <ddaa> When I mean doctest, I mean "doctest in a text file".
[09:45] <ddaa> Not in a docstring.
[09:45] <lifeless> yes
[09:45] <lifeless> I figured that ;)
[09:46] <lifeless> I see a pattern in doctest tests - they tend to be inadequate ;)
[09:46] <lifeless> I don't mean that as a slight on the testers, but the tool encourages a particular set of tests
[09:46] <ddaa> Want to review my branch?
[09:47] <ddaa> I think i'm stubborn enough to test what I want to test, regardless.
[09:47] <lifeless> Thanks, but I'd rather get the other stuff I have on my todo for you done ;0
[09:47] <lifeless> I'm sure you are testing that.
[09:47] <lifeless> And I'll definately look at it to see how you are meeting the challenges I'm thinking off
[09:47] <ddaa> Well, that's pretty trivial code all in all.
[09:48] <ddaa> Probably not a good benchmark.
[09:48] <lifeless> so, the launchpad list has had a number of threads about 'how do I test foobar effectively'
[09:49] <lifeless> where the answer in unittest is trivial, but in doctest you have a harder time - for instance you can choose between testing equality, or having the mismatched object printed on error
[09:49] <lifeless> you can't get both
[09:49] <lifeless> if there was a 'self' in the doctest that had the value of a unittest TestCase, then you could. I think that that would be interesting to play with
[09:50] <ddaa> Ho right. I had that once already. I code a helper funciton with a conditional.
[09:50] <ddaa> if object == expected:
[09:50] <ddaa>   print "object ok"
[09:50] <ddaa> else:
[09:50] <lifeless> so you are implementing machinery that unittest provides.
[09:50] <ddaa>   print "got bad object %r" % (object,)
[09:50] <lifeless> except because its not 'stock', its not a language that everyone reading your test knows
[09:50] <jblack> ddaa! Been a while. How are you? 
[09:51] <ddaa> lifeless: but that's only really a problem when the object does not have a useful enough repr.
[09:52] <ddaa> But, you are right. Most of my tests would not do good diagnostic output if they start failing.
[09:52] <ddaa> (most of non-trivial ones)
[09:52] <lifeless> ddaa: hah! Its quite possible that two objects with different repr compare ==
[09:52] <lifeless> ddaa: anyway, I need to get nose down. ciao
[09:52] <jblack> You're thinking the sm __eq__ ? 
[09:53] <ddaa> lifeless: probably bad repr then...
[09:53] <ddaa> hello jblack
[09:53] <ddaa> I'm feeling a bit tired. Could use some holiday.
[09:54] <ddaa> And I'm having my first experience in living with a woman. So far we did not have to buy new dishes.
[09:54] <ddaa> That's a good thing, I guess :)
[09:54] <jblack> congrats!
[09:56] <bradb> BjornT: The Malone email UI doesn't support "duplicate" yet, right? If not, I'll open a bug.
[09:56] <ddaa> jblack: and how are you doing?
[09:56] <jblack> Pretty good. The supermirror is close to done
[09:57] <jblack> I've got a handler for format4, and bzrlib support for 5 and 6
[09:57] <ddaa> do you _really_ want to support format4?
[09:57] <jblack> No. Its disabled.
[09:57] <ddaa> anybody in his right mind would have upgraded, or would soon.
[09:58] <jblack> but when there's a format7, I can move format5_6 to format 7, and format4 to 5 and 6
[09:58] <jblack> aye. 
[09:58] <ddaa> hu?
[09:58] <ddaa> -ENOSENSE
[09:59] <jblack> Never mind. 
[10:00] <jblack> Just think "supermirror with bzrlib for format 5 and 6
[10:01] <ddaa> Anyway, there should not any watershed format change in the future...
[10:01] <jblack> Ohhh. interesting
[10:01] <jblack> bzr: WARNING: bzrlib version doesn't match the bzr program.
[10:01] <jblack> This may indicate an installation problem.
[10:02] <ddaa> Your PYTHONPATH
[10:02] <jblack> I guess I have to set path and PYTHONPATH at the same time
[10:02] <ddaa> launchpad contains bzrlib
[10:02] <ddaa> if you do "PYTHONPATH=$PWD/lib ; bzr" is launchpad, you'll get that.
[10:08] <ddaa> lifeless: *cries*
[10:08] <ddaa> lifeless: bzrtool-launchpad is incompatible with the bzrlib in launchpad!
[10:09] <lifeless> ddaa: theres a bzr-launchpad too 
[10:09] <lifeless> if you could update the bzr in launchpad that would be great
[10:09] <lifeless> just send in a merge request ;)
[10:09] <ddaa> rs=lifeless
[10:09] <ddaa> right?
[10:09] <lifeless> [Trivial]  Update to new upstream
[10:10] <ddaa> ack will do
[10:15] <BjornT> bradb: right, the email ui doesn't support 'duplicate' yet
[10:16] <bradb> BjornT: ok...maybe I should just implement it now, or would it conflict with something you're doing?
[10:21] <BjornT> bradb: hmm, i think it would be best if you waited until my changes are done.
[10:22] <bradb> ok, no prob
[11:11] <SteveA> Seveas: hi, are you around?
[11:11] <Seveas> SteveA, yes
[11:12] <SteveA> the guys in brazil are having an irc issue that maybe you can help with
[11:12] <Seveas> #ubuntu-br?
[11:12] <SteveA> there's an office with 5 or 6 people who want to be on this channel, and about 8 in total 
[11:12] <SteveA> they have a single dynamic IP address for the office
[11:13] <SteveA> and freenode servers keep k-lining them
[11:13] <Seveas> ah
[11:13] <Seveas> too bad it's dynamic :/
[11:13] <SteveA> beacuse they all appear to join from the same place on the same channel at the same time
[11:13] <SteveA> yeah.
[11:13] <SteveA> fixed ones cost R$$$$
[11:13] <SteveA> they all get identified
[11:13] <SteveA> with nickserv
[11:13] <SteveA> is there anything that can be done?
[11:14] <Seveas> maybe, are they all in #ubuntu-br? 
[11:14] <SteveA> no
[11:14] <SteveA> this channel
[11:14] <Seveas> ah ok
[11:14] <SteveA> but there are others
[11:14] <SteveA> like the canonical-related channels
[11:14] <Seveas> If the IP range is narrow enough, I might get lilo to whitelist that range
[11:15] <SteveA> i'll see if i can find out
[11:15] <Seveas> can you give me a (few) nick(s) of affected people
[11:16] <SteveA> apparently the class A varies.
[11:16] <SteveA>  kiko, salgado, cprov
[11:17] <Seveas> none of them online now
[11:18] <SteveA> kiko is going to log the ip addresses for a while
[11:18] <SteveA> and see how they vary
[11:18] <Seveas> Maybe their ISP can give a more precise answer?
[11:19] <SteveA> maybe.  their ISP is not very communicative 
[11:19] <SteveA> it is the former goverment monopoly in brazil
[11:19] <Seveas> well, RIPE may help :)
[11:19] <SteveA> so has something of a monopoly position in most of the country
[11:20] <Seveas> what is the name of the ISP?
[11:20] <SteveA> telefonica, or something that sounds like that
[11:20] <SteveA> * mpt (n=mpt@201-27-71-186.dsl.telesp.net.br) has joined #launchpad
[11:21] <SteveA> mpt was either at that office, or at home
[11:21] <SteveA> i'm not sure which
[11:33] <Seveas> hmm, that is 201.27/16
[11:34] <Seveas> I'm not sure whether lilo will whitelist such a large range
[11:35] <mdke> can they use some kind of proxy?
[11:35] <Seveas> if they are irssi users, I might be able to give them a shell account on the same server as Ubugtu is on
[11:36] <mdke> good plan
[11:36] <Seveas> or a dircproxy if they set it up themselves ;)
[11:36] <Seveas> and a whitelist for 3 IP addresses (the machine has 3) is doable
[11:38] <Seveas> SteveA, could you suggest that to them
[11:42] <OneFix_Work> I've got a question that I'm sure is asked here often, but I just placed an order for Ubuntu CDs from ShipIt...and I know it says 4-6 weeks, but does it usually take that long or are they generally faster?
[11:46] <Seveas> OneFix_Work, launchpad and shipit are not related (except that shipit uses launchpad logins)
[11:49] <SteveA> Seveas: thanks, i'll pass that along.
[11:50] <Seveas> SteveA, the dircproxy can be used with any irc client btw ;)
[11:54] <OneFix_Work> Seveas: oh, ok thanks
[11:55] <lifeless> OneFix_Work: they are generally that long