[01:45] <kiko> still lookin
[01:47] <kiko> Kinnison, cprov: it seems there were updates to breezy-updates that we missed
[01:47] <kiko> did they get uploaded? stull like flashplayer and libast2
[01:47] <kiko> stuff
[01:48] <kiko> yeah, just those two.
[01:49] <Kinnison> kiko: not sure
[01:49] <Kinnison> kiko: I'm currently looking at some rejected stuff
[01:49] <kiko> okay.
[01:50] <kiko> but note those two -- they were breezy-updates.
[01:50] <kiko> and breezy-security
[01:50] <kiko> Kinnison cdebconf is a major culprit of problems
[01:50] <kiko> -Version: 0.97ubuntu1
[01:50] <kiko> +Version: 0.95ubuntu1
[01:50] <kiko> wireless-toops is another
[01:50] <kiko> usbutils is another yet
[01:51] <kiko> util-linux another
[01:51] <Kinnison> I'm currently investigating util-linux
[01:51] <kiko> I also think some removals are needed
[01:51] <kiko> nic-restricted-modules-2.6.12-9-amd64-generic-di
[01:52] <kiko> nic-restricted-firmware-2.6.12-9-amd64-generic-di
[01:52] <kiko> these two are extras
[01:52] <Kinnison> listing things here for me right now isn't going to help because I'm concentrating on one at a time
[01:52] <Kinnison> working from the buildd reject list
[01:52] <kiko> well, store them somewhere.
[01:53] <kiko> I have listed a set of things worth investigating based on comparator output
[01:53] <kiko> pango1.0
[01:53] <kiko> dhcp3
[01:53] <daf> kiko: dude, send him an email
[01:53] <Kinnison> ciau
[01:53] <kiko> very funny
[01:54] <daf> it is, but seriously
[01:55] <kiko> well, I'm half-asleep myself, so he can pick it up tomorrow
[01:55] <kiko> I can relook at the output if we lose it anyway
[01:55] <kiko> fucking massive diffs, bummer
[01:56] <kiko> I suspect few packages however because they keep reappearing
[01:56] <kiko> like 10 max
[01:57] <dilys> Merge to devel/launchpad/: [trivial]  Present pending builds according the processing order. (r3071: Celso Providelo)
[02:37] <mpt> Goooooood afternoon Launchpadders!
[02:41] <ajmitch> hi mpt 
[06:19] <dilys> Merge to devel/launchpad/: [trivial]  Fix massively inneficient specifications query (r3072: Stuart Bishop)
[06:41] <dilys> Merge to devel/launchpad/: [trivial]  Add Person.teams_participated_in for recursive team emblem display (Bug 30306) (r3073: Stuart Bishop)
[07:40] <dilys> Merge to devel/launchpad/: [trivial]  Convert peopleCount and teamsCount to use cached statistics instead of expensive table scans (r3074: Stuart Bishop)
[08:45] <carlos> morning
[09:05] <mpt_> hi carlos
[09:06] <carlos> mpt_: How is going?
[09:42] <mpt_> carlos, annoyingly slowly :-)
[09:44] <stub> Kinnison: Can I update stagings code and drop the librarian for a few minutes?
[09:46] <Keybuk> Kinnison: would it be possible for Launchpad's Packages page to also include packages which you've uploaded/Changed-By?
[09:46] <Keybuk> Maintainer doesn't mean much for distro people, but "things I've touched last" could be
[09:47] <Keybuk> (also production doesn't appear up-to-date wrt version numbers and stuff -- I guess that's "IT WILL TOMORROW!!!!muahahahah"? :p)
[09:51] <fabbione> morning
[09:51] <fabbione> stub: hey dude
[09:51] <stub> yo
[09:52] <fabbione> stub: I need some of your db foo :)
[09:52] <fabbione> stub: http://people.ubuntu.com/~fabbione/x-src-main.list
[09:52] <mdz> stub: fabio needs a bulk setting of bug contact on a few hundred packages
[09:53] <mdz> stub: and also a bulk subscription to existing bugs against those packages
[09:53] <fabbione> stub: and the list of src is at that url
[09:53] <fabbione> the team that will take care of these pkgs is ubuntu-x-swat
[09:55] <SteveA> hi
[09:55] <stub> Ok. So set the packagebugcontact on all of them to ubuntu-x-swat, and add ubuntu-x-swat to all bugs already targetted to any of those source packages.
[09:55] <mdz> stub: that's it exactly
[09:55] <SteveA> jamesh: ping
[09:56] <fabbione> stub: it would be neat if we can avoid malone to spit a gazzillion emails for this process
[09:57] <stub> fabbione: I'm going directly to the DB. Malone won't even notice.
[09:57] <fabbione> ok great
[09:57] <stub> mdz: Do you think this will happen often?
[09:58] <fabbione> stub: might
[09:59] <stub> Ok - reusable scripts then :)
[09:59] <fabbione> or make a UI for it 
[09:59] <fabbione> where admins can ask the UI to load a list from a file and do some processing on it
[10:02] <stub> We make a UI for it when we *know* it will be wanted often. Might or occasionally can be done without adding more features to Launchpad
[10:03] <mdz> stub: it is unlikely to happen often, but likely to happen at least once more
[10:03] <sivang> morning
[10:04] <mdz> stub: in fact, I bet Kamion has a similar list for installer-related packages which could be processed at the same time
[10:05] <stub> Sure. Just get a list of packages assembled and email me all the details if I'm not around.
[10:12] <mdz> stub: I will, as soon as I locate Colin
[10:13] <kiko> morning
[10:13] <SteveA> hi kiko
[10:15] <kiko> hey SteveA 
[10:15] <kiko> how's it going
[10:15] <SteveA> i had a good day yesterday.  overslept today, and woke up to the world covered in snow!
[10:16] <kiko> I overslept today too, need to get to the office
[10:16] <SteveA> see you later then
[10:16] <kiko> SteveA, mdz: bugzilla and logincookies should be closed for good now
[10:17] <kiko> SteveA, if you'd like to give jamesh the go-ahead on importing the bugs we missed, I'd appreciate it
[10:25] <mdz> kiko_afk: you are my hero
[10:25] <jamesh> kiko_afk: I'm guessing we could probably identify the comments in question by date
[10:25] <jamesh> are there new bugs too?
[10:26] <jamesh> SteveA: pong
[10:26] <mdz> jamesh: I am not sure whether there are new comments, but there are new bugs
[10:28] <jamesh> kiko_afk: do you know how people were able to file new bugs?
[10:32] <lalo> for the record - tabindexes in the add pro{je,du}ct form are funny
[10:32] <jordi> mdz: how frozen is dapper now?
[10:32] <jordi> mdz: no new versions, but fixes allowed?
[10:32] <mdz> jordi: wiki.ubuntu.com/DapperReleaseSchedule
[10:33] <mdz> jamesh: they had login cookies
[10:33] <mdz> jamesh: weeks old, but still valid
[10:35] <jordi> mdz: ok, I'll be uploading a nano revision to Debian that fixes a very longstanding crasher
[10:37] <Kinnison> Morning
[10:37] <Kinnison> stub: can it wait? have you already done it?
[10:37] <stub> Kinnison: Already did it
[10:37] <Kinnison> Keybuk: Possibly. File a bug on the soyuz ui?
[10:37] <Kinnison> Keybuk: the data model would allow the search to be done I believe
[10:38] <Kinnison> stub: fortunately the majority of the test was last night so it didn't affect us
[10:38] <stub> Kinnison: Hmm... need to do it again. My connection died most way through :-(
[10:38] <stub> All the builds were idle like you said
[10:38] <Kinnison> Right
[10:38] <Keybuk> Kinnison: what product is that?
[10:38] <Kinnison> Keybuk: if it's something under distros/ it's the soyuz product
[10:39] <Kinnison> Keybuk: if it's something under people/ then it's the launchpad product
[10:40] <Keybuk> bug 30383
[10:40] <Ubugtu> malone bug 30383 in soyuz "Provide "last touched" packages" [Normal,Unconfirmed]  http://launchpad.net/bugs/30383
[10:41] <Kinnison> your request sounds possible, however I don't know when there'll be time to do it. It'll be triaged as the reports come in I imagine
[10:45] <SteveA> jamesh: hello
[10:45] <jamesh> hi
[10:45] <kiko_afk> jamesh I'm not sure. 
[10:46] <kiko_afk> it may be that disabledtext wasn't set for everybody, or that the code is broken
[10:46] <SteveA> jamesh: kiko told me about people being able to file bugs after it was supposed to be not possible
[10:46] <kiko_afk> I just added a hack to disable it completely
[10:46] <kiko_afk> in the beginning of Auth/Login/WWW.pm
[10:46] <SteveA> it may be that disabledtext stops logins, but not access while already logged in
[10:46] <kiko_afk> there were logincookies for 2006-01-27
[10:46] <jamesh> kiko_afk: mdz was just saying that what we did prevented logins but didn't disable use
[10:47] <kiko_afk> jamesh but we cleared outlogincookies the day after the migration when we noticed this
[10:47] <SteveA> maybe having a cookie in your browser automatically means it is valid?
[10:48] <kiko_afk> no, we check that part
[10:48] <Kinnison> kiko_afk: the comparator output isn't desperately useful because there's a bunch of stuff in NEW we forgot to process
[10:48] <kiko_afk> the funny part is that people /created/ new logincookies over that period
[10:48] <kiko_afk> Kinnison ah, that is a good thing. can I re-run?
[10:48] <Kinnison> kiko_afk: I'm just merging codelines with celso and then I'll be processing the queue
[10:49] <kiko_afk> I'm about to leave. okay, see you in 20m
[10:49] <Kinnison> I need someone with madison powers to tell me for sure where things ended up after new
[10:49] <kiko_afk> mdz has them
[10:49] <kiko_afk> as does elmo -- I imagine he shouldn't take too long to come in today
[10:49] <SteveA> whoever added the lightbulb emblem for the launchpad-infrastructure team, thanks so much for not making it a spanner!
[10:50] <jamesh> so there are 40 rows which have disabledtext=''
[10:50] <Kinnison> SteveA: You do realise that having said that, someone will change it to a spanner, right?
[10:50] <SteveA> i thought they were all meant to have that?
[10:50] <jamesh> I think we disabled all accounts without preventing people from creating accounts
[10:51] <jamesh> the user IDs are sequential and have the highest IDs
[10:51] <mdz> Kinnison: we're moving a lot of things from main to universe during the sprint; is that likely to confuse gina?
[10:52] <Kinnison> mdz: Won't confuse gina when we run her tonight, no
[10:53] <Kinnison> mdz: but may confuse me if I ask "so, where should <blah> be"
[10:53] <Kinnison> I'
[10:53] <Kinnison> I'll use the Packages files locally and just prod
[10:53] <Kinnison> rather than asking for madison powahs
[10:53] <SteveA> jamesh: aha... so new accounts could still log in
[10:53] <mdz> Kinnison: madison-lite?
[10:53] <Kinnison> mdz: can I run that against an arbitrary archive?
[10:54] <mdz> Kinnison: yes, that's the idea
[10:54] <Kinnison> right
[10:54] <Kinnison> thanks
[10:57] <jamesh> SteveA: yeah
[10:59] <SteveA> jamesh: is that fixed now, or do we need further hacking of bugzilla?
[11:03] <jamesh> SteveA: kiko was saying he changed some things, so it might be.  If he joins again I'll ask him
[11:04] <SteveA> well, we could do an empirical test
[11:07] <jamesh> it looks to be fixed
[11:08] <jamesh> I created an account, but when I go to the URL in the confirmation email, I get a "Logins are disabled. Use Launchpad." message
[11:08] <SteveA> okay
[11:09] <SteveA> do you know if the new users have logins disabled?
[11:11] <jamesh> no
[11:11] <jamesh> but I am guessing that they'd get caught by the same message I just saw
[11:14] <SteveA> do the 40 rows with disabledtext='' have contiguous ids ?
[11:15] <SteveA> if so, they are all new signups, and can have disabledtext set to something like the rest
[11:17] <jamesh> yeah
[11:17] <jamesh> I don't have the permissions to do that though (I've only got read-only access to the DB)
[11:17] <SteveA> i see
[11:17] <SteveA> kiko should be around in a few minutes
[11:20] <SteveA> when we're changed those remaining fields, can you easily import the remaining bugs?
[11:21] <SteveA> if any preexisting bugs were commented on or had statuses changed, i think we just need to lose those changes
[11:21] <SteveA> because they may conflict with the changes made in launchpad since
[11:24] <jamesh> Importing the remaining bugs should be easy, since they'll have consecutive IDs
[11:25] <jamesh> and I can tell from the Laucnhpad bugwatches what the highest bug number I imported previously was
[11:26] <SteveA> do you have logs from the previous import?
[11:26] <kiko> snow snow snow
[11:26] <kiko> good morning
[11:27] <fabbione> hey kiko
[11:27] <fabbione> kiko: so we rollout today.. what about binaries upload from the external buildds?
[11:28] <kiko> well
[11:28] <kiko> funny you should ask
[11:29] <SteveA> kiko: can you write to the bugzilla db?
[11:29] <jamesh> SteveA: yes.
[11:29] <kiko> SteveA, no, but elmo can. what's up?
[11:29] <SteveA> jamesh worked out what's going on
[11:29] <SteveA> and what we need to do next
[11:29] <jamesh> kiko: I think I know what happened: we disabled all existing accounts, but did not disable creation of new accounts
[11:29] <kiko> I see.
[11:30] <fabbione> kiko: i am serious now.. are we going to have an exception for the external buildds or do we need to stop them?
[11:30] <kiko> jamesh, and can that still work?
[11:30] <jamesh> kiko: I just tried creating a new account, but got your "go away, use Launchpad" message when trying to go to userprefs.cgi to complete the account creation process
[11:31] <kiko> fabbione, a) external buildds will not work yet b) we will have sparcs coming in soon
[11:31] <fabbione> kiko: a) ok. b) i need to tell lamont to stop hppa. it's not just sparc
[11:32] <fabbione> c) thanks!
[11:32] <kiko> b) depends on when we get them delivered -- rumor has it should be soon
[11:32] <fabbione> b) yes, but we can stop building from this evening till they will hit DC
[11:32] <kiko> right.
[11:32] <fabbione> it's pointless to increase entropy in the universe
[11:33] <Kinnison> fabbione: No, it's pointless to try and decrease it
[11:33] <Kinnison> fabbione: it's inevitable that it'll increase
[11:33] <kiko> external buildds are not difficult to support but we need to figure out how to authenticate them and then tighten some of the locking
[11:33] <Kinnison> and, perhaps more importantly, make the slave processor more parallel
[11:33] <Kinnison> otherwise things will block on slower links
[11:34] <fabbione> kiko: via gpg?
[11:34] <fabbione> kiko: since the uploads are still signed
[11:34] <Kinnison> fabbione: In *theory* we could do gpg signed binary uploads on a different port or something
[11:34] <Kinnison> fabbione: in practice this is untested currently
[11:34] <fabbione> Kinnison: it's still useful to slow down the speed of increase in entropy :=
[11:34] <fabbione> Kinnison: we could test it :) 
[11:35] <Kinnison> fabbione: perhaps, but not immediately
[11:35] <lucasvo> is it possible to change the primary email address in launchpad?
[11:35] <kiko> lucasvo, why, certainly.
[11:35] <fabbione> Kinnison: of course.
[11:35] <kiko> just visit your personal page and change them.
[11:35] <Kinnison> fabbione: it *should* just work
[11:35] <lucasvo> kiko: wasn't sure
[11:36] <fabbione> Kinnison: i do believe it will. i am quite confident in your superskills
[11:36] <Kinnison> fabbione: Essentially it'll involve writing a special binary port and I'll need to work out how to restrict it to given keys
[11:36] <Kinnison> fabbione: because currently all we can do is a per-component restriction on sources
[11:39] <kiko> stub, one thing we need to talk about is production bouncing (on rollout day) versus soyuz
[11:41] <kiko> SteveA, I need your help here
[11:41] <ddaa> mpt> It should not be Launchpad's goal to completely model the structure of the open source community (which is fuzzier than any data model could represent). Rather, Launchpad should aim to maximize productivity of Free Software and open source development through collaboration.
[11:41] <SteveA> kiko: what's up?
[11:42] <ddaa> Damn, I'm going to put that in the canonical wiki fortunes!!
[11:42] <SteveA> kiko: and also, can we get someone with bugzilla DB access to do a query before jamesh finishes for the day, so jamesh can run the script today?
[11:42] <kiko> SteveA, basically, production rollouts that bounce emperor and the librarian will upset soyuz.
[11:42] <kiko> SteveA, yes. mail me the query.
[11:42] <SteveA> jamesh: please mail kiko the query
[11:43] <SteveA> kiko: you mean, once we've rolled soyuz out, we can't do any more work on emperor or the librarian ever?
[11:43] <kiko> very funny
[11:43] <SteveA> soyuz needs to be designed to expect the librarian to disappear occassionally, and for the database to be bounced occassionally
[11:43] <kiko> no, that's not what I am saying
[11:44] <jamesh> kiko: would your bz changes prevent people with disabledtext='' from making edits/new bugs?
[11:44] <SteveA> what specifically in soyuz will get upset?
[11:44] <kiko> jamesh, I believe so -- any authorized access is denied.
[11:44] <kiko> SteveA, let's say that for now, it's best to disable soyuz while bouncing production.
[11:44] <jamesh> if that's the case, we don't need to wait for someone with mysql db write access
[11:45] <SteveA> kiko: um.  gah.  it ain't going to happen all the time.  look at the librarian problems in the past week or so.
[11:46] <SteveA> jamesh: can we test this?  i guess not easily unless we temporarily undo kiko's "no new accounts" hack and create a new account.
[11:47] <jamesh> SteveA: That should be possible.  I've got a half created account to test with
[11:47] <jamesh> kiko: is that easy for you to do?
[11:49] <kiko> jamesh, not until elmo arrives
[11:50] <SteveA> i think znarl has write access to that database too
[11:50] <jamesh> kiko: okay.  In that case we may as well just get disabledtext sety
[11:51] <jamesh> do we delay another import run til then?
[11:51] <kiko> jamesh, i wouldn't delay if I were you, I think the risk is low and elmo should be in soon.
[11:51] <kiko> as soon as he's in I'll set disabledtext
[11:51] <kiko> but I'm pretty sure the change I made will forbid changes
[11:53] <SteveA> hmm, okay.  i agree with kiko.  the worst is we lose some changes from one of the people who got a new account in the last few weeks
[11:55] <kiko> SteveA, could you place an RT request for Znarl?
[11:56] <kiko> so he can set disabledtext again
[11:58] <SteveA> yes.
[11:58] <kiko> thanks
[11:58] <SteveA> jamesh: please tell me what the query karl should run is
[11:58] <kiko> update profiles set disabledtext = 'Bugzilla disabled';
[11:58] <SteveA> ok
[11:58] <SteveA> cool
[12:01] <jamesh> from my checks, we are talking about 11 new bugs
[12:02] <kiko> wonderful
[12:02] <kiko> there are probably new comments on old bugs..
[12:03] <Kinnison> Why don't we remove the bugzilla instance and simply use mod_rewrite to rewrite the relevant URLs into malone
[12:04] <jamesh> Kinnison: as time goes on, that's probably the right solution.
[12:04] <jamesh> we'd need to make sure the bugzilla view doesn't provide any needed information not found in LP though
[12:06] <Kinnison> so put the bugzilla instance on closed-bugzilla.ubuntu.com
[12:06] <Kinnison> :-)
[12:07] <kiko> I think jamesh is right. Let it bake for a while.
[12:08] <jamesh> Kinnison: that was the idea. (and one of the reasons we imported all bugs rather than just the open ones)
[12:08] <Znarl> kiko, SteveA : Done.
[12:08] <kiko> thank you Znarl 
[12:08] <kiko> Znarl, by the way, we leave tomorrow and you STILL have not come by to say hello
[12:08] <kiko> this is highly unsatisfactory
[12:08] <Kinnison> jamesh: heh
[12:09] <Kinnison> Znarl is just a figment of elmo's imagination
[12:09] <Kinnison> clearly
[12:11] <SteveA> thanks Znarl !
[12:11] <Znarl> kiko : I know.  :(
[12:11] <kiko> Znarl, should I file an RT request?
[12:13] <SteveA> kiko: put it directly in the launchpad queue, priority 99
[12:20] <daf> stub, SteveA: re bug 30370: I know that putting the Librarian behind HTTPS is undesirable (performance, certificate tax), but I don't see any other way of addressing this problem
[12:21] <daf> Ubugtu: bug #30370
[12:21] <Ubugtu> malone bug 30370 in launchpad "Graphics from Librarian over HTTP cause browser warnings on Launchpad over HTTPS" [Normal,Unconfirmed]  http://launchpad.net/bugs/30370
[12:21] <kiko> daf, maybe allow the librarian to answer over HTTPS and HTTP equally?
[12:21] <SteveA> daf: what we should do is offer the librarian behind both http and https
[12:21] <SteveA> and then link to it as https when the page embedding the image is served over https
[12:21] <kiko> and then only use https:// for images (and is there other inline content)?
[12:22] <kiko> such a second-guesser
[12:22] <daf> cool, I'll confirm the bug
[12:22] <SteveA> kiko: we are SO on form today
[12:22] <SteveA> daf: and make it an infrastructure "not spanner" team task too
[12:23] <daf> "not spanner"?
[12:23] <SteveA> SteveA> whoever added the lightbulb emblem for the launchpad-infrastructure team, thanks so much for not making it a spanner!
[12:23] <daf> ha
[12:24] <daf> you want me to assign to infrastructure?
[12:24] <SteveA> yes
[12:24] <SteveA> and cc too
[12:24] <SteveA> or whatever we need to do to simulate keywords
[12:24] <daf> hmm, then it will get filtered out by my script, even though it impacts the user experience
[12:24] <SteveA> it should be a 1.1 thing too
[12:25] <daf> done
[12:28] <SteveA> thanks daf
[12:28] <SteveA> do you have a list of bugs for us to talk through?
[12:29] <daf> yes -- the same list as before
[12:29] <daf> I'll just sync it
[12:30] <daf> I'm looking forward to not having to do that any more
[12:30] <daf> maybe I should just cron it
[12:30] <SteveA> we'll have the new view rolled out soon
[12:30] <SteveA> did you get that reviewed?
[12:30] <daf> but it's only until Tuesday
[12:31] <daf> yes, Steve, you reviewed it
[12:31] <SteveA> cool
[12:31] <SteveA> i don't even know what i'm doing...
[12:32] <daf> you're SO on form
[12:33] <daf> https://chinstrap.warthogs.hbd.com/~daf/bugs/scrape.py
[12:33] <SteveA> daf: okay, voice call in 5?
[12:33] <daf> 108 confirmed bugs sans milestone
[12:33] <daf> sure
[12:34] <daf> a three-way with matsubara would be nice -- I'm looking forward to asterisk
[12:34] <SteveA> we'll try doing it differently this time...
[12:34] <SteveA> you make the call about what to do about the bugs, and then we'll discuss
[12:35] <SteveA> kiko: do you know why matsubara can't get skype at async?  can we do anything about that?
[12:35] <kiko> I know, yes.
[12:35] <kiko> I can do something about it, but not today.
[12:35] <SteveA> ok
[12:36] <SteveA> daf: do you remember what lifeless said in answer to my question about asterisk in the meeting yesterday?
[12:39] <daf> I don't, I'll check
[12:39] <daf> he didn't
[12:45] <SteveA> daf: i'm running skype now
[12:45] <daf> waiting for it to load...
[12:54] <Kinnison> kiko: https://chinstrap.ubuntu.com/~dsilvers/paste/file261YmK.html
[01:05] <kiko> In [10] :guess("fo").encode("ascii", "replace")
[01:05] <kiko> Out[10] :'f?o'
[01:08] <kiko> In [24] :ascii_smash(guess("fo"))
[01:08] <kiko> Out[24] :'foeo'
[01:15] <Kinnison> u'Rapha\xebl Pinson <raphink@ubuntu.com>'
[01:16] <kiko> In [37] :ascii_smash("aou")
[01:16] <kiko> TypeError: normalize() argument 2 must be unicode, not str
[01:17] <Kinnison> kiko: https://chinstrap.ubuntu.com/~dsilvers/paste/fileKtvtdJ.html
[01:18] <SteveA> kiko: of course.
[01:19] <SteveA> the in [37]  one is in whatever encoding your console / defultencoding is in
[01:19] <SteveA> which is why the TypeError
[01:19] <kiko> yeah
[01:19] <SteveA> because the TypeError is stopping a programmer accidentally being ambiguous
[01:19] <kiko> ascii_smash() needs unicode
[01:19] <SteveA> and rightly so.  
[01:19] <SteveA> if python did things like java
[01:19] <kiko> and guess() needs non-unicode
[01:19] <SteveA> and had no "ascii string" concept
[01:19] <SteveA> or "unencoded string" concept
[01:19] <Kinnison> it'd be nice if instead of guess complaining, it just acted as identity when given a unicode
[01:20] <SteveA> we'd all be better off
[01:20] <SteveA> Kinnison: that's a good point, and i think that is reasonable, but only just
[01:20] <SteveA> as it can hide errors in the program at large
[01:20] <Kinnison> here, guess what this is... *sniff sniff* it's a unicode!
[01:20] <kiko> or /already/ a unicode as it is.
[01:21] <SteveA> the issue is this: in python there is a historical thing of treating "strings" as a lump
[01:21] <daf> that's a bit like mixing ints and floats and not expecting rounding errors
[01:21] <SteveA> and considering str and unicode being roughtly the same
[01:21] <SteveA> but i think this causes more errors in the long run
[01:21] <SteveA> and i think we should treat str and unicode as totally different
[01:21] <SteveA> and not fit for substitution
[01:21] <daf> ++
[01:21] <SteveA> so, based on that, guess() should acccept a unicode only if some arg is given explicitly allowing it
[01:22] <SteveA>   guess(u'foo', accept_unicode=True)
[01:22] <SteveA> so that we know where the larger application is being flaky about the difference between str and unicode
[01:22] <SteveA> this difference has done, and continues to, cause problems for rosetta
[01:22] <SteveA> because parts of rosetta were written tollerantly of this
[01:23] <SteveA> another approach would be to allow unicode into guess, but issue a warning
[01:23] <SteveA> i now think i am against having guess() take a unicode implicitly, without complaint
[01:23] <daf> I wonder if we can use decorators to make enforcing this easier
[01:23] <SteveA> you mean like painters and plasterers?
[01:23] <daf> yes
[01:24] <SteveA> i think they'd agree with me for the most part
[01:24] <daf> big ornate signs saying "NO STRINGS HERE"
[01:24] <SteveA> let's get back to the bugs, daf ;-)
[01:26] <stub> guess used to accept Unicode, but I changed that because it appeared that people where not thinking about their string encodings and just shoving 'whatever' into guess to cast to Unicode. Which is sweeping the bugs under the carpet.
[01:28] <stub> guess should only be used when your source data is broken (unknown encoding) and we can't repair it properly.
[01:30] <SteveA> stub: +1
[01:30] <SteveA> no carpet-sweeping-under in launchpad please
[01:30] <kiko> yeah, I thought that too
[02:24] <kiko> SteveA, what of system-error@launchpad.net
[02:26] <SteveA> kiko: what of it?
[02:26] <SteveA> kiko: i think it is a good mail address
[02:26] <kiko> SteveA, did you not get my email?
[02:26] <SteveA> and we should use it on the main template and shipit template
[02:26] <kiko> yes...
[02:27] <SteveA> so, i just mailed a test message
[02:27] <kiko> it needs Bug in the Subject you know?
[02:28] <SteveA> yes
[02:28] <SteveA> i sent exactly what you asked for in your email
[02:29] <kiko> great
[02:29] <kiko> and it worked
[02:29] <kiko> Old-From: steve@z3u.com (steve)
[02:29] <kiko> From: www@launchpad.ubuntu.com
[02:30] <kiko> so let's use it.
[02:30] <SteveA> okay, cool
[02:30] <kiko> did jamesh import the remaining bugs?
[02:33] <SteveA> i have not heard about success or failure from him yet
[02:34] <mdke> carlos, thanks for your reply on the ubuntu-docs mess. i personally think that they should go into the ubuntu distro-specific section, but I think mpt disagrees. In any event, if we decide on the distro-specific way, will they be imported automatically into dapper, if they are in the source tarball for ubuntu-docs?
[02:45] <kiko> SteveA, do you have any suggestion on how to fix my template problem?
[02:50] <carlos> mdke: yes
[02:51] <mdke> carlos, cool thanks
[03:02] <Kinnison> kiko: wnck, vte and eel2 building now
[03:02] <kiko> great
[03:16] <Seveas> Will Launchpad be present on the SANE "freedom of cooperation" conference in Delft, Holland in May this year? If not - would you allow/appreciate a talk about it from me (I'll be there representing Ubuntu NL)?
[03:32] <SteveA> kiko: the macros stuff?
[03:32] <SteveA> kiko: a workaround is to register your macros place as some standard view or resource, and refer to it by that name from within itself 
[03:33] <SteveA> Seveas: I'm not familiar with that conference.  Got a URL for me to look at?
[03:36] <Seveas> SteveA, www.sane.nl/sane2006
[03:39] <SteveA> Seveas: that's an interesting programme.
[03:39] <SteveA> i don't think the launchpad project will have any official representation there, but please do talk about launchpad in your talk about Ubuntu NL
[03:40] <Seveas> SteveA, the plan is to do a 10-minute promotional "Here's why you should love Ubuntu" talk and a BOF about launchpad
[03:40] <SteveA> and ask here or on launchpad-users if you need any other information about what there is, or what's planned
[03:41] <SteveA> daf: reconvene in 15 mins?
[03:41] <daf> yes
[03:47] <kiko> https://staging.ubuntu.com/distros/ubuntu/+source/opal/2.1.2-0ubuntu2
[03:54] <ddaa> Sorry SteveA, I was a bad boy.
[03:54] <ddaa> I scraped the python conversion stuff I worked before, and wrote a xslt sylesheet instead.
[03:54] <ddaa> https://chinstrap.warthogs.hbd.com/~david/bzr-launchpad/bzr-launchpad.html
[03:55] <daf> s/scraped/scrapped/?
[03:55] <ddaa> whatever, deleted it...
[03:55] <ddaa> I must  have gone crazy trying to do xml transforms in python...
[03:56] <SteveA> ddaa: it's fine, so long as you get the other work done first
[03:57] <daf> SteveA: I'm ready
[03:57] <ddaa> Gotta finish this documentation now.
[03:58] <SteveA> ddaa: i'd like to talk through the docs with you.  how about in 1. 5 hrs we do that?
[03:59] <ddaa> Fine with me. Do you want me to prepare something in the meantime?
[03:59] <SteveA> you were planning to do more on the docs
[03:59] <SteveA> so, whatever you were planning to do
[03:59] <ddaa> okay
[04:01] <matsubara> hi SteveA , not being able to delete a launchpad account is by design or a bug?
[04:02] <SteveA> matsubara: hi
[04:02] <SteveA> basically, it is a a design thing.  we cannot delete a launchpad account *entirely* when it is linked to other things
[04:03] <SteveA> so, we could delete an account if nothing had been done to it
[04:03] <SteveA> and we could delete one that has been merged into another account
[04:04] <SteveA> matsubara: it can be a tricky issue, though
[04:04] <SteveA> because some things we do at the moment *are* bugs
[04:04] <SteveA> and also, we must consider that launchpad is about cataloguing information about the people and teams involved in open source
[04:05] <SteveA> and people with accounts are those people who choose to log in to launchpad
[04:05] <daf> what sort of things do we do that are bugs?
[04:05] <matsubara> SteveA: will there be an option to the users to close a launchpad account in the future?
[04:05] <SteveA> but, we still want to catalogue the publicly available information about a person, even if they don't want to log in
[04:06] <SteveA> matsubara: i expect so, but we need to carefully define what we mean by "close an account"
[04:06] <kiko> indeed.
[04:06] <SteveA> daf: there was a bug that appeared to show that people we catalogue information about, who have never used launchpad, were launchpad users
[04:06] <SteveA> this is an important distinction between a User and a Person in launchpad
[04:07] <daf> and perhaps one which we don't make clear enough
[04:07] <SteveA> all Users are also Persons
[04:07] <SteveA> only some Persons are Users
[04:07] <daf> we know whether a Person is a User by whether they have a password or not, I think
[04:07] <daf> but we don't make that information clear
[04:07] <SteveA> basically, yes, daf
[04:07] <SteveA> actually, if they have a prefered email address
[04:08] <SteveA> but it is a related concept
[04:08] <SteveA> maybe we should say clearly on a person page
[04:08] <SteveA>  "Foo is a launchpad user" or "Foo is not a launchpad user"
[04:08] <daf> or even have a different emblem for Users vs. Persons
[04:08] <SteveA> then, it makes no sense to say "ubuntero: whatever" for Persons who are not Users
[04:08] <matsubara> SteveA: so should I confirm bug 2773 and target it to the future milestone?
[04:09] <Ubugtu> malone bug 2773 in launchpad "Deleting or removing or closing an account" [Normal,Unconfirmed]  http://launchpad.net/bugs/2773
[04:09] <SteveA> becuase "ubuntero" is something that can apply only to Users, not to Persons
[04:09] <daf> perhaps we can start a spec about turning a User back into a Person
[04:10] <SteveA> matsubara: this is an emotive issue, and i think that in this case, you should raise the issue with me and kiko (as you have done)
[04:10] <SteveA> i'll reply to the bug
[04:10] <kiko> DELETE MY ACCOUNT KTHXBYE
[04:11] <matsubara> SteveA: ok then.
[04:12] <daf> bug #1590 makes that distinction, without using the same terms
[04:12] <Ubugtu> malone bug 1590 in launchpad "Differentiate between validated and unvalidated accounts when people try to register emails that are already registered" [Normal,Confirmed]  http://launchpad.net/bugs/1590
[04:13] <SteveA> what i've done is left a basic comment saying that we should offer the option of closing a launchpad account.
[04:13] <SteveA> this may or may not mean removing the launchpad Person associated with that account.
[04:13] <SteveA> but it does mean making removing that user.  if you get the distinction.
[04:13] <SteveA> i've confirmed the bug, and targetted it to 1.1
[04:13] <SteveA> we'll see if it stays as 1.1, after the 1.1<-->future cut
[04:14] <SteveA> daf: i don't like the wording in 1590
[04:14] <SteveA> saying "validated account" and "unvalidated account" speaks to them both being launchpad accounts
[04:14] <SteveA> it is offensive to create a launchpad account for people who have not asked for one
[04:14] <SteveA> it is okay to create a launchpad Person record for anyone who has done open source things
[04:15] <daf> https://launchpad.net/products/launchpad/+bug/3958 was rejected
[04:15] <Ubugtu> malone bug 3958 in launchpad "Delete my own account?" [Normal,Rejected]  
[04:15] <SteveA> we must not think of a launchpad Person record as an account or as a user, unless that Person record has a prefered email address / password
[04:15] <daf> Person.is_user()? Person.has_account()?
[04:16] <SteveA> is_user is confusing
[04:16] <SteveA> because we talk about "a person", "a user", and "the user"
[04:16] <SteveA> the latter is short for "the currently logged-in user" 
[04:16] <daf> outrageous suggestion: we should have a separate Account table to make the distinction perfectly clear to developers
[04:16] <SteveA> "a user" is short for "a person who has a launchpad account"
[04:16] <SteveA> let's not
[04:17] <SteveA> but, let's write a spec on "launchpad Persons, launchpad Users"
[04:17] <SteveA> to make this clear
[04:17] <SteveA> to point people at
[04:17] <SteveA> and to motivate development to make this clearer in the launchpad UI
[04:17] <daf> sounds good
[04:17] <daf> shall I kick that off?
[04:17] <SteveA> please do
[04:19] <daf> matsubara-lunch: do you have time to talk this afternoon?
[04:29] <SteveA> ddaa: maybe we can talk in 10 mins time?
[04:29] <ddaa> fine
[04:30] <SteveA> ok
[04:31] <daf> bradb, BjornT: it occurs to me it would be useful to have a +bugs-new page
[04:32] <daf> this would be like +bugs-untriaged
[04:32] <daf> but would only list bugs where last-modified-time == creation-time
[04:32] <daf> i.e. nobody has changed anything about the bug or even added a comment since it was filed
[04:32] <daf> my use case:
[04:33] <daf> I'm triaging bugs and I want to look at bugs which nobody else has looked at yet
[04:33] <daf> as opposed to bugs which haven't yet been confirmed, but which have had some attention
[04:33] <daf> the alternative to doing this by date would be to have an extra status New, in addition to Unconfirmed, but that strikes me as more cumbersome
[04:34] <daf> sorting the +bugs-untriaged page by Newest First sort of gets me what I want, but not quite
[04:36] <bradb> daf: Something like bug 6041?
[04:36] <Ubugtu> malone bug 6041 in malone "Search to support answering every bug reported" [Normal,Confirmed]  http://launchpad.net/bugs/6041
[04:37] <daf> yes, something like that
[04:40] <SteveA> ddaa: 
[04:40] <ddaa> pong
[05:01] <matsubara> daf: sure.
[05:02] <bradb> BjornT: I sent a filebug mail and two bugmails got delivered. Known bug?
[05:03] <bradb> (For bug 30419)
[05:04] <BjornT> bradb: the other mail was a 'also affects' notification, right? if so, yes, it's a known bug which has been fixed.
[05:04] <daf> matsubara: great, when do you have time?
[05:04] <bradb> BjornT: Yep, that's the one, thanks.
[05:06] <dilys> Merge to devel/launchpad/: [r=BjornT]  bug contact package bugs overview report and searchable bug (r3075: Brad Bollenbach)
[05:06] <matsubara> daf: i'll fix a conflicting branch and ping you, is that ok?
[05:07] <daf> matsubara: that's fine
[05:10] <kiko> ROCK ON BRADB!
[05:11] <bradb> I'm glad that one's in. :)
[05:12] <bradb> Hopefully we can infect other bug listings with this new design. Two-column, linkified search filter display.
[05:13] <bradb> There are a couple things I still need to add, like column-click sorting, which I was planning on this morning, and an advanced search (there's a message on the page saying "Advanced search coming soon" :)
[05:14] <daf> what URL's this page at?
[05:15] <bradb> $person/+packagebugs.
[05:15] <daf> cool
[05:46] <bradb> BjornT: here's why I wasn't getting a UnicodeEncodeError:
[05:46] <bradb> >>> urllib.urlencode((('foo', u'\u2122'),), doseq=True)
[05:46] <bradb> 'foo=%3F'
[05:46] <bradb> That seems retarded to me.
[05:47] <bradb> vs.
[05:47] <bradb> >>> urllib.urlencode({'foo': u'\u2122'})
[05:47] <bradb> Traceback (most recent call last):
[05:47] <bradb>   File "<stdin>", line 1, in ?
[05:47] <bradb>   File "/usr/lib/python2.4/urllib.py", line 1152, in urlencode
[05:47] <bradb>     v = quote_plus(str(v))
[05:47] <bradb> UnicodeEncodeError: 'ascii' codec can't encode character u'\u2122' in position 0: ordinal not in range(128)
[05:49] <SteveA> file a python bug
[05:51] <BjornT> bradb: hmm, strange. seems like they replace non-ascii characters with ? in the first example.
[05:53] <bradb> Bah, my SF account uses an email I no longer have access to. That's how much I enjoy using SF.
[05:55] <bradb> I can get access to it by re-adding a brad@ alias to that domain. I'll file this bug later.
[06:09] <BjornT> bradb: http://sourceforge.net/tracker/index.php?func=detail&aid=1349732&group_id=5470&atid=105470
[06:11] <daf> even Malone is better than the SF.net tracker
[06:21] <carlos> so
[06:21] <carlos> is there anyone else getting the test errors I got?
[06:21] <SteveA> matsubara: your email to mpt.  are you looking for feedback on the layout of the scrape.py page, or on his input into particular bugs?
[06:21] <SteveA> matsubara: be specific in what you're asking for, and you're more likely to get it!
[06:22] <kiko> indeed.
[06:22] <dilys> Merge to devel/launchpad/: [trivial]  fix a small character encoding issue with the bug contact (r3076: Brad Bollenbach)
[06:23] <daf> kiko: matsubara just pointed out that the bug text page is O(n)
[06:24] <daf> bradb: maybe you have some ideas about how to make this use less queries
[06:24] <matsubara> SteveA: i would like him to give some general hints on bug triage process and contribute to the LaunchpadBugTriage wiki 
[06:25] <SteveA> matsubara: so, point him specifically at resources you would like him to review 
[06:25] <SteveA> but, maybe we should arrange a meeting for this?
[06:25] <SteveA> it would need to be first thing in the morning, brazil time
[06:25] <SteveA> so that mpt and i and daf and you can make it
[06:26] <SteveA> we can all talk about the bug triage process
[06:26] <SteveA> and issues we've found
[06:26] <matsubara> SteveA: ok
[06:26] <daf> bradb: https://chinstrap.warthogs.hbd.com/~dsilvers/paste/fileMcJYqq.html <-- this is the code in question; specifically, the second render() method is the problem
[06:26] <SteveA> use the launchpad list, rather than ccing people in particular
[06:26] <SteveA> is O(n) bad?
[06:27] <daf> sorry, let me clarify
[06:27] <SteveA> i mean, O(n^2) is bad
[06:27] <SteveA> order of O(nlogn) is pretty bad
[06:27] <SteveA> O(n) is okay usually
[06:27] <daf> the number of queries executed is linear in the number of bugs
[06:27] <daf> (bug tasks)
[06:27] <daf> this is not scalable
[06:27] <SteveA> does it time out?
[06:27] <daf> sometimes
[06:28] <SteveA> how many bugs are we trying to display at once?
[06:28] <daf> this is on staging
[06:28] <daf> ~980
[06:28] <SteveA> if we have a need to display summary information for a lot of bugs at once, then doing it in a single query is a good idea
[06:28] <SteveA> that's a lot of bugs to be displaying on a single apge
[06:28] <SteveA> page
[06:29] <SteveA> that sounds like the page needs a redesign
[06:29] <SteveA> rather than the queries made better
[06:29] <daf> ideas:
[06:30] <daf>  - filter out fixed bugs
[06:30] <daf>  - allow filtering by query parameters
[06:30] <daf>  - return a list of bugs
[06:31] <daf> (i.e. bug IDs; the scrape script would have to fetch details of each bug individually)
[06:32] <daf> but I think that whatever we do, it should be in conjunction with reducing the number of queries
[06:32] <SteveA> what is the page
[06:32] <daf> https://staging.ubuntu.com/products/launchpad/+bugs-text
[06:33] <SteveA> bugs-text ?
[06:33] <SteveA> what an odd name for a page
[06:33] <SteveA> oh, that page ;-)
[06:34] <daf> yes, that page :)
[06:35] <SteveA> i think it should be just a list of bug ids and urls and statuses
[06:35] <SteveA> a single query
[06:35] <SteveA> maybe add milestone too
[06:35] <SteveA> the page is too big otherwise, really
[06:35] <daf> I wonder if we can include timestamp information
[06:36] <daf> to avoid unneccessary queries
[06:36] <SteveA> keep it simple
[06:36] <SteveA> a single query that gives minimal information for each bug to allow a script to decide which ones are interesting
[06:36] <SteveA> should be enough
[06:38] <daf> bradb, BjornT: what does "Community Timestamp" mean?
[06:38] <daf> and, for that matter, "Hits Timestamp" and "Activity Timestamp"?
[06:38] <BjornT> daf: where did you find that?
[06:38] <daf> IBug
[06:39] <daf> """The core bug entry.""" ???
[06:40] <BjornT> daf: hmm, my guess is that those things are unused and should be removed, but let's hear if bradb knows more about it.
[06:40] <salgado> SteveA, is SelectResults.__len__ going to be removed soon?
[06:40] <daf> perhaps you can improve that docstring while you're at it :)
[06:40] <SteveA> salgado: yes, jamesh should be doing it, if it isn't done already in RF
[06:41] <daf> jamesh/sqlos/select-results-len-fix has not yet been merged
[06:41] <daf> (merge-conditional)
[06:42] <salgado> SteveA, right, because that method uses FailedLenWarning, which is not imported nor defined in that file. I was going to fix it, but I think I don't need to bother, then
[06:42] <SteveA> "that method" ?
[06:42] <SteveA> oh, you mean __len__
[06:42] <salgado> SelectResults.__len__
[06:42] <SteveA> um
[06:42] <SteveA> so, that case isn't tested :-(
[06:42] <SteveA> anyway, it is okay
[06:42] <SteveA> it will go away soon
[06:44] <salgado> right, I just wanted to make sure there's no need to fix it
[06:50] <dilys> Merge to devel/launchpad/: [r=kiko]  Remove the indirection to buildlog and changes and put them directly on the build-index page. (r3077: James Troup)
[06:50] <kiko> rock and fucking roll
[07:04] <bradb> BjornT: ah, thanks for the sf link
[07:05] <bradb> daf: I don't know what the bug text page is. I'll have to look at it.
[07:07] <bradb> timeout, doh ;)
[07:09] <bradb> daf: re: *timestamp. What BjornT said. And if they're not used anywhere, we should remove them.
[07:16] <dilys> Merge to devel/launchpad/: [trivial]  Small UI fixes and more DB tweaks for queue announcement process. (r3078: Celso Providelo)
[07:53] <dilys> Merge to devel/launchpad/: [r=kiko]  Cleanup build-index page some more. (r3079: James Troup)
[08:05] <carlos> hmm
[08:05] <carlos> lifeless: around?
[08:06] <carlos> nothing
[08:06] <carlos> grr I was downloading the rockefuel mirror to the wrong place...
[08:11] <zyga> carlos: hello, how is development? :-)
[08:13] <carlos> zyga: fine, thanks. Closing most of the open branches I had pending to merge
[08:13] <zyga> carlos: what are you using to merge to a central branch?
[08:14] <carlos> zyga: pqm
[08:14] <zyga> hmm :-)
[08:16] <carlos> zyga: http://people.ubuntu.com/~robertc/pqm/trunk/
[08:16] <zyga> carlos: I was about to ask...
[08:16] <carlos> I think it supports bzr, bazaar 1.x and tla
[08:17] <zyga> fetching
[08:17] <zyga> I'm only interested in bzr
[08:24] <zyga> carlos: thanks
[08:24] <carlos> zyga: you are welcome
[08:25] <zyga> carlos: looks nice alas slighly complex 
[08:26] <carlos> zyga: I haven't setup it, I'm just an user
[08:26] <carlos> so I don't know
[08:26] <zyga> carlos: I want to set it up for my personal project
[08:27] <carlos> me too
[08:27] <carlos> ;-)
[08:45] <carlos> por?
[08:45] <carlos> upps...
[08:45] <carlos> sorry, wrong window
[08:50] <ddaa> zyga: pretty soon you should be able to set up multi-commiter branches easily with launchpad
[08:50] <zyga> ddaa: ooooh
[08:50] <zyga> ddaa: most interesting! :-)
[08:50] <ddaa> two bits involved
[08:51] <ddaa> first bit will be the sftp server on bazaar.launcphad.net
[08:51] <ddaa> which will allow you push branches to launchpad
[08:51] <ddaa> second bit will be setting the branch owner to a team (you'll need an admin to do that for you ATM)
[08:51] <ddaa> that will allow all team members to push
[08:52] <zyga> ddaa: and conflits?
[08:52] <ddaa> then that should automagically support all the future bzr goodness, like bound branches.
[08:52] <zyga> ddaa: only clean merges apply?
[08:52] <ddaa> zyga: conflicts do not happen on push
[08:52] <zyga> ddaa: bound branches?
[08:52] <zyga> ah, right
[08:53] <ddaa> ask on #bzr about bound branches, or look into the wiki, I'm pretty sure it's documented
[08:53] <zyga> k
[08:53] <ddaa> not there yet, but should be in a few weeks
[08:53] <ddaa> unless spiv is hit by a bus...
[08:56] <zyga> ddaa: do you want to tell me that bound branches are implemented today?
[08:57] <ddaa> zyga: I meant to say that launchpad's sftp server will be rolling out in a few weeks
[08:57] <ddaa> zyga: j-a-meinel has a branch that implements bound branches
[08:57] <zyga> awsome feature :>
[08:57] <ddaa> it's been waiting for a number of important infrastructure changes before getting merged into the mainline
[08:58] <ddaa> stuff like versioned format support and internal reorganisations
[09:30] <dilys> Merge to devel/launchpad/: Fix https://launchpad.net/products/malone/+bug/4201 (Bugs with Needs Info status should be displayed on open bugs query). r=salgado (r3080: Diogo Matsubara, Guilherme Salgado)
[09:52] <carlos> pmjdebruijn: dude, sorry, I had network problems at the time you asked me for it
[09:52] <pmjdebruijn> carlos, thankyou :)
[09:52] <carlos> and forgot it completely....
[09:53] <pmjdebruijn> carlos, hehe, no big deal...
[09:53] <pmjdebruijn> carlos, but i'll slap you silly next time you forget :p
[09:53] <carlos> :-P
[09:59] <pmjdebruijn> carlos, btw the {de|fr|nl} files are borked...
[09:59] <carlos> broken, how?
[10:00] <pmjdebruijn> carlos, huh, well you said it had something to do with the encoding? between ASCII and UTF-8
[10:00] <carlos> oh
[10:00] <carlos> sorry, I thought you were talking about other files
[10:00] <carlos> no, the content is not broken
[10:01] <carlos> or it should not be broken...
[10:13] <dilys> Merge to devel/launchpad/: [trivial]  comment cleanups of builddmaster.py (r3081: James Troup)
[10:15] <carlos> pmjdebruijn: please, try the download again, if the workaround worked... you should be able to get them
[10:16] <pmjdebruijn> carlos, I requested the download...
[10:20] <pmjdebruijn> carlos, sweet...
[10:20] <carlos> pmjdebruijn: did it work?
[10:20] <pmjdebruijn> carlos, yep
[10:21] <carlos> It's the first time I tried the workaround ;-)
[10:21] <carlos> cool
[10:21] <pmjdebruijn> carlos, thankyou!
[10:22] <carlos> you are welcome
[10:36] <dilys> Merge to devel/launchpad/: [trivial]  small CSS fixes to the bug contacts reports (r3082: Brad Bollenbach)
[11:22] <mpt> bradb or BjornT, do either of you know why there are some bug numbers without bugs?
[11:23] <kiko> huh?
[11:23] <kiko> mpt, can you tell me more about this?
[11:24] <mpt> kiko, see stub's message "Re: More Malone statistics needed"
[11:24] <kiko> I am on an email holiday
[11:25] <mpt> kiko, ok, he returned the table I asked for showing batched results for bugs #1~#100, #101~#200, #201~#300, and so on
[11:25] <kiko> yeah
[11:25] <mpt> and many of those batches don't have 100 bugs in them
[11:25] <kiko> I saw that
[11:25] <kiko> ah hmmm.
[11:26] <mpt> Coincidentally I found an example a couple of days ago: https://launchpad.net/bugs/2 is a 404
[11:28] <BjornT> mpt: yes, that happens if a bug gets created in the database, and then later the database transaction gets aborted.
[11:28] <mpt> BjornT, explain to me the relationship between "later" and "the" :-)
[11:29] <mpt> which database translation is it that gets aborted?
[11:29] <bradb> mpt: It might have had something to do with a bug in the email interface, where email was getting processed twice.
[11:29] <mpt> bradb, the e-mail interface didn't exist when bug 2 was reported
[11:30] <BjornT> mpt: 1. start database transaction. 2. create bug. 3. abort database transaction
[11:30] <mpt> iirc it didn't exist until at least bug 2000, and there are dozens of apparently-unused bug numbers from before then
[11:30] <Ubugtu> malone bug 2000 in launchpad "System error on team hierarchy pages" [Normal,Fix released]  http://launchpad.net/bugs/2000
[11:30] <Kinnison> Ubugtu: everyone loves you
[11:31] <mpt> BjornT, is that a bug that should be fixed?
[11:32] <BjornT> mpt: no, it's not a bug. if it didn't work like that, two parallel transaction could create two different bugs with the same id. it's the way the database works, nothing we can do about it.
[11:33] <mpt> So how does Bugzilla avoid this? Does it just not have parallel transactions?
[11:35] <BjornT> does bugzilla avoid this? well, you can avoid it by making sure that you check that nothing can go wrong with the transaction after you create the bug (not avoid it entirely, but almost)
[11:41] <kiko> before you create the bug you mean
[11:45] <BjornT> yeah, sort of. i see now that i didn't communicate it well :) you check, before you create the bug, that nothing can go wrong with the transaction after that the bug has been created.
[11:47] <BjornT> the biggest cause for us 'loosing' ids is the email interface, which aborts the transaction if anything goes wrong. if needed, it could be re-written to check that everything is ok before the bug is created, but it adds complexity.
[12:00] <dilys> Merge to devel/launchpad/: [trivial]  Add a simple test for SourcePackage.releasehistory (r3083: James Troup)
[12:02] <FaBMak> sorry for the newbie question, but can i remove a launchpad account ?