[02:20] <stub> lifeless: In case you are keeping score, I had to kill -9 that same buildbot test again for ddaa last night
[04:09] <lifeless> stub: thanks
[09:26] <Burgundavia> help, I am trying to report a bug on a package LP claims doesn't exist
[10:02] <carlos> hi
[10:15] <Kinnison> stub: has gina been running against asuka?
[10:17] <stub> Kinnison: Yes
[10:18] <stub> Kinnison: It has been dying with a Librarian upload failure though, so may not have completed for a while
[10:20] <Kinnison> ARGH
[10:21] <stub> I remember you asking about the database synchronization, but not Gina :-(
[10:21] <Kinnison> Urgh
[10:21] <Kinnison> Not the end of the world
[10:21] <Kinnison> but not ideal
[10:21] <Kinnison> Can you make sure it's definitely not trying now?
[10:21] <stub> Yup. 
[10:21] <stub> Done
[10:21] <Kinnison> ''cos it keeps killing the publisher
[10:21] <stub> Deadlock?
[10:23] <Kinnison> bizarre deadlock or else librarian errors
[10:41] <SteveA> morning
[10:41] <Kinnison> hi steve
[10:57] <kiko> good morning
[10:57] <kiko> hello SteveA 
[10:57] <kiko> how are you?
[10:58] <kiko> stub, is staging.ubuntu.com working?
[10:58] <kiko> hmm, just slow
[10:58] <Kinnison> We are doing a full publish
[10:58] <stub> Poor thing only has one CPU - be nice!
[10:58] <kiko> on staging?
[11:00] <Kinnison> indeed. This is uberslow
[11:29] <Kinnison> stub: I just killed gina on drescher in order to have more ram for the publishing runs
[11:29] <stub> Yup
[11:29] <stub> Feel free to disable it in lp_import's crontab
[11:30] <stub> since it will kick off again in a few hours otherwise
[11:31] <Kinnison> okay
[11:32] <kiko> stub, I was talking to SteveA about all the queries for teamparticipation
[11:32] <kiko> and he was telling me about CrowdControl
[11:32] <kiko> what do you think of that angle?
[11:34] <carlos> kiko, hi, is CrowdControl already implemented?
[11:34] <kiko> nope.
[11:34] <carlos> ok
[11:34] <kiko> we're considering it now, but I am unsure of how this will affect caching of participation
[11:35] <stub> kiko: Depends on when crowd control lands
[11:35] <stub> We will still need caching - just need to implement it in crowds or in teamparticipation
[11:35] <kiko> ah.
[11:35] <kiko> that makes more sense.
[11:35] <kiko> so I was thinking that we could have inTeam() be caching
[11:36] <kiko> using a similar strategy as we used for launchpad celebs
[11:37] <stub> Sounds good. Might be some edge cases when removing yourself from a team, but the foaf tests should point them out (?)
[11:37] <kiko> I spoke to salgado about that
[11:37] <kiko> and he and I agree that it's silly
[11:38] <stub> Would this shave hundreds of queries off some pages, or just a few?
[11:38] <kiko> that you can definitely cache inTeam for a request
[11:38] <kiko> about 20 for a few pages, ten for most
[11:38] <kiko> it's not a massive win like celebs were
[11:38] <stub> There are more urgent areas then
[11:38] <kiko> well
[11:38] <stub> eg. the bug report I filed earlier today :-(
[11:39] <kiko> it can probably be done cheaply, but tell me about your bug?
[11:39] <kiko> also, did you see my oopses email?
[11:39] <stub> Bug 29725
[11:40] <kiko> Ubugtu, wake up
[11:40] <kiko> Ubugtu, bug 29725
[11:40] <Ubugtu> (bug <abbreviation> <number>) -- Look up bug <number> in the bugtracker associated with <abbreviation>.
[11:40] <Ubugtu> Malone bug 29725: "Timeout trying  to register a Launchpad branch" Fix req. for: launchpad (upstream), Severity: Major, Assigned to: Nobody, Status: Confirmed http://launchpad.net/bugs/29725
[11:40] <stub> kiko: Nope - not that I recall
[11:40] <kiko> stub, can you check it out?
[11:41] <kiko> there's a query in there i'd like you to look at
[11:41] <stub> Where is the email?
[11:41] <kiko> launchpad list?
[11:41] <SteveA> stub: what i want to do is this:
[11:41] <kiko> aren't you reading that any more? :)
[11:41] <SteveA>  - land crowdcontrol, so there's just one query per security lookup
[11:41] <SteveA>  - make security queries cached on GETs
[11:42] <SteveA>  - see whether security queries still need cacheing on POSTs
[11:42] <stub> What was the title of the email 'oopses email' is a bit vague.
[11:42] <kiko> it as "A set of oopses"
[11:43] <kiko> and it was the latest email to the list
[11:43] <stub> hasn't arrived yet. Last I have is one from Bjorn
[11:43] <kiko> something's wrong with your mail
[11:44] <kiko> I sent it last night, about 12h ago
[11:44] <kiko> To: The Soyuz Team at Launchpad <launchpad@lists.canonical.com>,
[11:44] <kiko>         Stuart Bishop <stuart@stuartbishop.net>
[11:45] <kiko> SteveA, can you make sure the __len__ patch progresses (and lands?) today?
[11:45] <kiko> I fear it is stuck
[11:46] <kiko> where is jamesh? ah, I know
[11:47] <stub> kiko: Last thing from you on the launchpad mailing list was Launchpad Report for 2006-01-25
[11:47] <stub> https://lists.ubuntu.com/mailman/private/launchpad/2006-January/author.html
[11:47] <stub> kiko: Check your mail queue
[11:47] <kiko> stub, uhm, no, you're mistaken
[11:48] <kiko> https://lists.ubuntu.com/mailman/private/launchpad/2006-January/007401.html
[11:48] <kiko> however
[11:48] <kiko> I did send it with a different From: header than usual
[11:48] <SteveA> kiko: jamesh is at the linux conf.  i have the __len__ patch on my list for this week.
[11:48] <kiko> great.
[11:49] <SteveA> kiko: so, maybe not today
[11:49] <kiko> you should aim to get it in early so we can check it on staging before rollout..
[11:49] <Kinnison> hat's up with asuka being so slow all of a sudden?
[11:49] <kiko> SteveA, what else is on your plate this week?
[11:49] <Kinnison> s/^/w/
[11:52] <Kinnison> carlos: can you not run your hideous selects on asuka?
[11:53] <Kinnison> carlos: It's killing my stuff and I'm already 36 hours behind on a deadline
[12:07] <Kinnison> carlos: It would be really nice if asuka wasn't almost constantly io-blocked because of whatever query you're running.
[12:07] <Kinnison> hey sabdfl
[12:07] <kiko> sab d f l, take a bow
[12:08] <carlos> Kinnison, my scripts are executed at 4:00AM we should not have anything running now...
[12:08] <carlos> let me check
[12:09] <sabdfl> kiko: ?
[12:09] <carlos> oh, I see, it's still running. But should be finish soon, anyway, I can kill it. I forgot to chante it to run once per week 
[12:09] <kiko> how are you, sabdfl?
[12:09] <matsubara> good morning!
[12:10] <Kinnison> carlos: if it could be knocked dead I'd really appreciate it
[12:10] <carlos> Kinnison, it's already killed
[12:10] <kiko> hello matsubara 
[12:10] <jordi> kiko: https://launchpad.net/malone/bugs/5952 <--- I don't quite understand what you ask in your comment
[12:10] <Ubugtu> Malone bug 5952: "Group "owners" should be able to add translators" Fix req. for: rosetta (upstream), Severity: Major, Assigned to: Nobody, Status: Confirmed
[12:10] <kiko> jordi, I have a request for you today
[12:10] <jordi> Sr. Kiko, qu tal?
[12:10] <kiko> muy bien
[12:10] <matsubara> kiko: hi kiko
[12:11] <Kinnison> carlos: hopefully the postgresql child will notice soon and die too
[12:12] <kiko> jordi, you know translation teams?
[12:12] <carlos> Kinnison, I just updated the crontab to be executed on Sundays
[12:12] <kiko> jordi, well, they can be shared between multiple products and projects and distros
[12:12] <Kinnison> Also, anyone know what the debbugs mirror is all about?
[12:12] <Kinnison> that's chewwing CPU/disk
[12:12] <jordi> kiko: yes
[12:12] <carlos> kiko, I don't think it's a big issue
[12:13] <kiko> carlos?
[12:13] <carlos> if you trust the list of translators... you should trust the 'owner' of that list...
[12:13] <kiko> carlos, that's not the point
[12:13] <Kinnison> stub: ping?
[12:13] <kiko> who is the owner of the list, carlos?
[12:13] <carlos> kiko, atm, the Rosetta experts
[12:14] <kiko> carlos, the bug calls for context owners being able to appoint new translators
[12:14] <carlos> oh, context owners?
[12:14] <carlos> no, that's broken
[12:14] <kiko> read the bug, carlos.
[12:14] <carlos> kiko, I read it some days ago and forgot the 'context' word
[12:14] <carlos> sorry
[12:14] <kiko> sure.
[12:15] <carlos> kiko, the title is broken then
[12:15] <jordi> kiko: well, it may only apply to the very specific groups like KDE or Plone.
[12:15] <kiko> jordi?
[12:15] <carlos> "Group owners" are not context owners
[12:15] <kiko> jordi, even so, any product owner can go and assign a pre-existing translation team to themselves, right?
[12:15] <kiko> so let's say I create a product
[12:15] <kiko> say "make the plone translation team my translation team"
[12:15] <Kinnison> woohoo, pgsql noticed and now my publishing is going way faster
[12:16] <Kinnison> carlos: thanks dude
[12:16] <kiko> then I would be able to add people at will
[12:16] <kiko> and potentially displease the plone people
[12:16] <kiko> salgado!
[12:16] <jordi> kiko: aha. We somehow shouldn't assume this random dude creating the "plone-fake-project-to-take-over-real-plone-muhaha" can do it.
[12:17] <kiko> jordi, can you say that again?
[12:17] <jordi> "muhaha", I know you'd like that part.
[12:17] <jordi> I mean:
[12:17] <carlos> jordi, that has an easy fix, just add an owner to the translation group instead of using the product's owner
[12:18] <jordi> If Plone translators has an "owner", then nobody else can be the owner unless it's changed by Launchpad management
[12:18] <jordi> carlos: correct
[12:18] <carlos> so Only the maintainers of the translation group can change it, for Plone, only people that the plone project tell us, will be able to appoint new translators
[12:18] <jordi> yes
[12:18] <kiko> carlos, what about shared teams? but okay..
[12:18] <kiko> carlos, jordi: can you update the bug with the suggested plan?
[12:18] <jordi> I can
[12:19] <carlos> kiko, it's not a big problem
[12:19] <carlos> translators will be appointed only by the ones that created that team
[12:19] <kiko> sounds okay.
[12:19] <carlos> so if the project Foo uses the Plone team
[12:19] <sabdfl> kiko: well thanks. sick today but enjoying LCA nonetheless, timezones a bit bollocksed but hey this is the asia tour :-)
[12:19] <carlos> will get any translator that the Plone admins want
[12:20] <kiko> sabdfl, how sick is "sick"? mdz and I unscrewed the showerhead at the hotel today, for improved personal hygiene
[12:20] <kiko> carlos, somehow, I think we should have used groups for TTs but..
[12:20] <kiko> I mean, groups have admins, owners, etc -- everything we would have needed :)
[12:21] <jordi> TT, expand?
[12:21] <kiko> translation teams?
[12:21] <jordi> err :P
[12:21] <kiko> doh
[12:21] <jordi> thanks
[12:21] <kiko> jordi, so, I have a special request for you
[12:21] <kiko> can you come up with the top 10 FAQs for launchpad?
[12:21] <jordi> yes?
[12:21] <kiko> just questions
[12:21] <jordi> for Launchpad?
[12:22] <kiko> I can produce answers
[12:22] <kiko> yes
[12:22] <kiko> see staging.ubuntu.com/faq ?
[12:22] <lbm> i get timeouts on https://launchpad.net/products/gnomebaker/+series/main/+pots/gnomebaker/da/+translate
[12:22] <kiko> that's daf's 15-minute fix
[12:22] <carlos> kiko, the problem is that they lack the language information
[12:22] <lbm> OOPS-26D138
[12:22] <kiko> carlos, we could have annotated them
[12:22] <jordi> Not sure. I have little input for some parts of launchpad
[12:22] <carlos> kiko, for the language teams we are using groups
[12:22] <carlos> so it's only to associate a language with a team
[12:22] <jordi> I mostly do rosetta and do some malone and doap at user level
[12:23] <carlos> kiko, that what's the initial plan, but it was decided to go this way....
[12:23] <kiko> jordi, that's okay
[12:24] <carlos> kiko, btw, current FAQ page points to the legal page as a Rosetta specific page, but we have information there for Malone too
[12:24] <jordi> kiko: I'll think about it.
[12:24] <carlos> and soyuz 
[12:24] <kiko> carlos, I'll fix that faq as soon as I have more inputs, friday the latest
[12:24] <carlos> ok
[12:25] <jordi> kiko: is there ongoing navigation improvement work for launchpad?
[12:25] <jordi> I still get many people saying they get lost
[12:26] <kiko> jordi, well, ongoing is a difficult word. we definitely know it needs to be improved
[12:26] <salgado> BjornT_, around?
[12:27] <jordi> ok, the log of this conversation is in malone now.
[12:28] <jordi> carlos: I submitted some import queue bugs
[12:28] <carlos> jordi, cool, thanks
[12:28] <carlos> I will try to handle them as soon as possible
[12:28] <carlos> jordi, did you set priorities on them?
[12:29] <carlos> about which ones are the "must have fixed" and which ones are the wishes?
[12:30] <BjornT_> hi salgado 
[12:31] <salgado> hi BjornT_, any news on that review I replied to you yesterday?
[12:34] <BjornT_> salgado: yes, i'll reply to your mail after the meeting, i have to get something to eat now
[12:34] <salgado> BjornT, great. thanks. :)
[12:35] <lbm> argh, i need a complete export of gnomebaker (main) translations
[12:35] <lbm> i get timeouts
[12:36] <kiko> lbm, oops id? please don't stop giving them to me, I am collecting evidence :)
[12:36] <lbm> kiko: OOPS-26D138, OOPS-26A181
[12:37] <lbm> kiko: i get timeout when i try accessing https://launchpad.net/products/gnomebaker/+series/main/+pots/gnomebaker/XX/+translate
[12:37] <lbm> which is the only place i can export afair
[12:38] <kiko> carlos, is lbm right?
[12:39] <carlos> lbm, https://launchpad.net/products/gnomebaker/+series/main/+pots/gnomebaker/XX should allow you to export too
[12:40] <lbm> carlos: of all languages?
[12:40] <carlos> lbm, a full export?
[12:40] <carlos> lbm, https://launchpad.net/products/gnomebaker/+series/main/+pots/gnomebaker/
[12:40] <carlos> there you have a link for a full export
[12:40] <lbm> carlos: yes, where can i find the link in the future?
[12:41] <carlos> lbm, selecting that template from gnomebaker
[12:41] <carlos> you get that url
[12:41] <lbm> your right :)
[12:41] <lbm> how could i not see that?
[12:41] <carlos> SteveA, seems like the suggestions are not fast enough, the problem lbm has is related with the suggestions :-(
[12:41] <lbm> well, thanks :)
[12:41] <carlos> lbm, you are welcome
[12:42] <kiko> lbm, I'll look into that timeout today, hopefully something we can improve.
[12:42] <jordi> carlos: no I didn't
[12:42] <lbm> kiko: great
[12:42] <carlos> btw... how is that we are showing back traces with the timeouts?
[12:42] <SteveA> suggestions don't need to be right up to date.  maybe some kind of cacheing will help
[12:42] <kiko> carlos?
[12:42] <jordi> carlos: well, both ar high for me, as they will improve my performance
[12:42] <carlos> kiko, we did an optimization already
[12:42] <kiko> and it is still timing out
[12:42] <lbm> another thing, gnomebaker is coded in en_GB, so such translation shouldn't be possible
[12:42] <carlos> kiko, I got a back trace with the timeout of https://launchpad.net/products/gnomebaker/+series/main/+pots/gnomebaker/es/+translate
[12:43] <SteveA> another option would be... ajax
[12:43] <lbm> we need a place to specify upstream language
[12:43] <SteveA> have the suggestions come in with separate queries after the page has loaded
[12:43] <carlos> kiko, right, and I think we need to implement what SteveA said, caching
[12:43] <carlos> hmmm
[12:43] <kiko> carlos, you get backtraes always
[12:43] <carlos> SteveA, I like that more than caching
[12:43] <carlos> kiko, because I'm an admin?
[12:43] <kiko> carlos, yes, a launchpad devel
[12:44] <carlos> oh, it's not the first time I see it but I was not sure if it was on production or staging... 
[12:44] <carlos> kiko, nice feature ;-)
[12:45] <lbm> did you guys understand my feature request?
[12:46] <kiko> carlos, have you seen the oops reports?
[12:46] <kiko> it has EVERY SINGLE QUERY issued
[12:47] <kiko> that is rock and roll suicide
[12:47] <kiko> lbm, that's an interesting topic. can you make sure that carlos understands it and we have a bug report filed?
[12:48] <lbm> carlos: did you understand my request?
[12:50] <carlos> kiko, ?
[12:51] <carlos> lbm, not really, could you explain it a bit more?
[12:51] <kiko> it is almost time
[12:51] <carlos> kiko, what do you mean by 'Every single query'? from gnomebaker?
[12:52] <kiko> carlos, have you seen the oops reports?
[12:52] <carlos> kiko, I planned to do it this afternoon so no, I didn't see them yet
[12:53] <lbm> carlos: atm it's possible to translate en_GB in gnomebaker (main), which is the language used in the upstream code
[12:53] <kiko> carlos, I mean, the format of them
[12:53] <lbm> carlos: it shouldn't be possible
[12:55] <carlos> kiko, dude, sorry but I don't understand you. I can only think on the big query that raises the timeout....
[12:56] <carlos> lbm, Why not? it's a good way to fix typos ...
[12:56] <kiko> carlos, if you look at an oops report, you will see at the end of it a log of all queries issued during that request.
[12:56] <carlos> oh
[12:56] <ddaa> Is there something like a meeting in 5 mins?
[12:56] <carlos> kiko, I'm looking at the website bt
[12:56] <carlos> let me look at the OPPS directly...
[12:56] <kiko> right
[12:56] <kiko> there is, ddaa 
[12:57] <SteveA> meeting in 2 mins
[12:57] <carlos> kiko, that's normal
[12:58] <carlos> kiko, not the time it takes, we should improve it, but the amount of queries...
[12:58] <carlos> kiko, We have 3 kinds of suggestions and 10 entries per page...
[12:58] <AlinuxOS> coff and some chokolade in meating room
[12:58] <lbm> carlos: you have a point, but i think these should go in upstream code instead
[12:58] <AlinuxOS> thank you.
[12:58] <carlos> kiko, at least 30 queries for suggestions
[12:58] <kiko> yeah
[12:59] <carlos> lbm, right, but in the mean time... and also, the source code is not updated after an Ubuntu release
[12:59] <kiko> we should do 3 queries, carlos, I think
[12:59] <carlos> kiko, 3 queries?
[12:59] <SteveA> MEETING TIME
[01:00] <SteveA> who is here today for the launchpad development meeting?
[01:00] <jblack> Here
[01:00] <daf> me
[01:00] <salgado> here
[01:00] <ddaa> here
[01:00] <mpt_> me
[01:00] <gneuman> here
[01:00] <carlos> me
[01:00] <bradb> here
[01:00] <matsubara> here
[01:00] <BjornT> here
[01:00] <cprov> here
[01:01] <SteveA> apologies from australian residents (public holiday)
[01:01] <kiko> here
[01:01] <kiko> carlos, yes, fewer queries, not more
[01:01] <lbm> carlos: i think about main translation, not ubuntu package translations
[01:01] <SteveA> mpt_: you are on vacation, no?
[01:01] <carlos> lbm, let's talk about it after the meeting, ok?
[01:02] <lbm> carlos: okay, good meeting
[01:02] <mpt_> SteveA, yes, but I said I'd be at the meeting anyway
[01:02] <SteveA> ok
[01:02] <SteveA> == Agenda ==
[01:02] <SteveA>  * Roll call
[01:02] <SteveA>  * Agenda
[01:02] <SteveA>  * Next meeting
[01:02] <SteveA>  * Activity reports
[01:02] <SteveA>  * Items from last meeting
[01:02] <SteveA>  * Production / staging (stub)
[01:02] <SteveA>  * General policy about how should we deal with broken user input: Raise an exception, ignore it? (CarlosPerelloMarin)
[01:02] <carlos> kiko, I will appreciate any comment about how to reduce it to 3 single queries... (after the meeting)
[01:02] <SteveA>  * Keep, Bag, Change
[01:02] <SteveA>  * Three sentences
[01:02] <SteveA> 
[01:02] <SteveA> any further agenda items
[01:02] <SteveA> ?
[01:02] <kiko> not from me.
[01:02] <SteveA> ok.  next meeting, same time next week?
[01:03] <SteveA> it is done
[01:03] <SteveA>  * Activity reports
[01:03] <jblack> up to date
[01:03] <ddaa> up to date
[01:03] <SteveA> not up to date
[01:03] <carlos> up to date
[01:03] <bradb> up to date
[01:03] <BjornT> i'm up to date
[01:03] <matsubara> up to date
[01:03] <kiko> somewhat up to date, unfortunately accumulated and flaky during sprint.
[01:03] <cprov> not up to date 
[01:03] <salgado> up to date
[01:03] <kiko> last week was perfect
[01:03] <Kinnison> not up to date due to sprint
[01:04] <gneuman> up to date
[01:04] <mpt_> n/a
[01:04] <jordi> I'm here
[01:04] <jordi> sorry
[01:04] <niemeyer> And is almost up to date..
[01:04] <kiko> niemeyer, don't use /me :-P
[01:04] <jordi> oops
[01:05] <jordi> I'm not up to date on activity reports
[01:05] <jordi> :P
[01:05] <niemeyer> promises even
[01:05] <daf> up to date
[01:05] <SteveA>  * Items from last meeting
[01:05] <SteveA>     *
[01:05] <SteveA>       MeetingAction: Andrew and James to try Kiko's suggested time logging workflow.
[01:05] <SteveA> they're not here, so for next week's meeting
[01:05] <SteveA> MeetingAction: Kiko to check reliability of OOPS syncs.
[01:05] <SteveA> kiko: ?
[01:06] <kiko> it's working reliably.
[01:06] <kiko> every 10 minutes.
[01:06] <SteveA> MeetingAction: Kiko and James to discuss implementing a validator.
[01:06] <SteveA> (for page templates / pages)
[01:06] <kiko> did not happen.
[01:06] <SteveA> kiko: is 10 minutes okay?
[01:06] <kiko> it's okay. 5 would be better.
[01:06] <SteveA> Znarl can be asked to make it 5.  i don't think it would be a problem.
[01:06] <kiko> I load the page, get a not found, wait for the :X0 minute and reload.
[01:07] <SteveA> i'll submit an RT request for it to be 5
[01:07] <SteveA> MeetingAction: James B to review DatabaseSetup. 
[01:07] <SteveA> jblack: ?
[01:07] <SteveA> MeetingAction: Matthew to document the decision in the FixingProjects spec.
[01:07] <SteveA> mpt: ?
[01:07] <kiko> jblack, SteveA: perhaps of note is that we fixed launchpad-database-setup to work in dapper too.
[01:08] <SteveA> that's cool
[01:08] <SteveA> do we now have two launchpad dependencies packages?
[01:08] <jblack> Yes, I need to do that
[01:08] <mpt_> SteveA, I recorded it briefly, and need to finish writing it up
[01:08] <jblack> I forgot. 
[01:08] <SteveA> one in breezy-updates, one in dapper-updates?
[01:08] <SteveA> MeetingAction: Kiko to delegate optional-branch-title.
[01:08] <SteveA> kiko: to whom was it delegated?
[01:08] <ddaa> daf
[01:08] <kiko> daf
[01:08] <kiko> it was done.
[01:09] <SteveA> cool
[01:09] <SteveA> MeetingAction: David to isolate the bzr problem he encountered and file a bug report.
[01:09] <SteveA> ddaa: ?
[01:09] <jblack> I'll do it immediately after the meeting.
[01:09] <SteveA> thanks jblack 
[01:09] <ddaa> SteveA: nm
[01:09] <kiko> SteveA, mdz is taking care of launchpad-dependencies
[01:09] <SteveA> okay, great
[01:09] <kiko> well, different versions in different distros. yes
[01:09] <SteveA> ok
[01:09] <mdz> well, more accurately
[01:09] <mdz> doko is fixing twisted
[01:09] <mdz> and launchpad-dependencies is going to remain unchanged
[01:10] <SteveA> that is all the meeting actions from the last meeting
[01:10] <kiko> oh.
[01:10] <daf> ddaa: "nm"?
[01:10] <ddaa> nm = nevermind
[01:10] <SteveA> stub: are we using the launchpad dependencies package in production?
[01:11] <SteveA> there have been problems in the past with dependencies missed from production machines.  using a package means we can manage these things in just one place.
[01:11] <kiko> yes, indeed.
[01:11] <daf> Stuart is not here
[01:11] <daf> (yet)
[01:12] <carlos> SteveA, if the package is from dapper, we will not be able to use it with breezy soon
[01:12] <SteveA> let's move on, and catch up with stuart when he turns up
[01:12] <SteveA> carlos: launchpad developers are generally using breezy, and launchpad is deployed on breezy, so it must work with breezy
[01:13] <kiko> sabdf1, connection bothersome
[01:13] <SteveA> it will be good for people to start using dapper soon, to get the system well tested
[01:13] <SteveA> so, the package should work with dapper too
[01:13] <carlos> SteveA, but it makes no sense that the package is with dapper if it's for breezy ....
[01:13] <sabdf1> kiko: yup
[01:13] <mdz> in fact, twisted is already fixed
[01:14] <kiko> SteveA, move on?
[01:14] <SteveA> mdz: is this straightforward to arrange, if different packages are required for breezy and for dapper?
[01:14] <mdz> https://lists.ubuntu.com/archives/dapper-changes/2006-January/005353.html
[01:14] <mdz> SteveA: yes, it's not a big deal if and when we need to change it
[01:14] <SteveA> ok, thanks
[01:14] <mdz> but we don't need to branch it yet
[01:14] <SteveA>  * General policy about how should we deal with broken user input: Raise an exception, ignore it? (CarlosPerelloMarin)
[01:14] <SteveA> carlos: can you talk about this?
[01:15] <carlos> Ok
[01:15] <carlos> with the review that spiv did for one of my branches
[01:15] <doko> the new python-twisted packages are in the archive
[01:15] <SteveA> thanks doko
[01:16] <carlos> he saw that I'm raising an AssertionError if the user gives us broken data from the form (manually generated URLs)
[01:16] <carlos> and he said that we should not do that
[01:16] <carlos> because the problem could be with a proxy changing/breaking the URLs or the POST information
[01:16] <carlos> instead we should either ignore or raise a concrete exception for that kind of error
[01:17] <carlos> I like the idea of the concrete exception, but I don't want to do my own stuff here
[01:17] <carlos> as I'm sure it's a problem in other parts of launchpad
[01:17] <SteveA> we have had this situation in a lot of code in launchpad
[01:17] <mpt> I reported bug 28655 on what I think is the same thing, so it can be fixed gradually
[01:17] <carlos> so, could we agree on a standard procedure to deal with this?
[01:18] <mpt> (where's Ubugtu when you need her)
[01:18] <kiko> she has problems
[01:18] <SteveA> in general, we should make form handlers accept data that has been submitted by launchpad forms
[01:18] <SteveA> but not data that has been submitted by other means.
[01:19] <carlos> SteveA, we still have the problem of the arguments on the URL
[01:19] <SteveA> as in, we are not under an obligation to behave very nicely with data submitted by other means
[01:19] <carlos> SteveA, "+translate?show=foo"
[01:19] <SteveA> what's the problem, carlos?
[01:20] <carlos> SteveA, if the user puts there invalid values
[01:20] <SteveA> if the user does, then getting an error page is okay
[01:20] <kiko> a 500 error or an invalid input error?
[01:20] <kiko> perhaps the latter would ne nice
[01:20] <SteveA> because we may want to give specific error pages for this in the future, or handle such errors in a different way
[01:20] <carlos> kiko, yeah, I prefer an invalid input error
[01:20] <SteveA> we should use a specific UnexpectedFormData error
[01:20] <SteveA> rather than AssertionError
[01:21] <kiko> that's great
[01:21] <kiko> UnexpectedFormData
[01:21] <SteveA> a nice validation error would be nice, but not something we should spend effort on
[01:21] <kiko> I'm happy with UFD
[01:21] <SteveA> so, carlos, have you landed the code in question yet?
[01:21] <carlos> SteveA, not yet
[01:21] <SteveA> okay.  so, we'll talk after the meeting about adding an UnexpectedFormData exception
[01:21] <carlos> but I'm planning to do it today or tomorrow
[01:21] <SteveA> and you can use that, as the first case of that
[01:21] <carlos> ok
[01:21] <carlos> thanks
[01:21] <mpt> How is that different from "Constraint not satisfied"?
[01:22] <SteveA> we'll make it derive from AssertionError at first
[01:22] <SteveA> mpt: for this kind of situation, it says to a programmer or code maintainer exactly what has happened
[01:22] <SteveA> mpt: we might make it derive from ConstraintNotSatisfied later on
[01:22] <SteveA> when we have examined the situation more
[01:23] <SteveA> we've used AssertionError for this purpose in the past
[01:23] <SteveA> so, making it a more specific AssertionError means that we can convert existing cases as we find them
[01:23] <SteveA> and then see what we want to do next
[01:23] <ddaa> I think ConstraintNotSatisfied means "programmer was too lazy to give a proper error" and UnexpectedFormDatat means "programmer does not have to support you doing weird things like that".
[01:23] <SteveA> for example, we could have a special section in OOPS reports
[01:23] <BjornT> more use of zope widgets should minimize this problem, since we would get the validation of data for free, and wouldn't have to validate it manually
[01:23] <bradb> A UFD will make it harder for users to write screen scrapers, and to understand what went wrong if they edit a URL incorrectly.
[01:23] <SteveA> to see where we've had UnexpectedFormData errors
[01:24] <SteveA> and decide whether to support the cases we've seen
[01:24] <SteveA> BjornT: agreed.
[01:24] <mpt> a special section in the OOPS reports is what bug 28655 is driving at
[01:24] <mpt> so that's cool
[01:24] <SteveA> bradb: i think it will make it easier for them to see what is wrong, compared to an AssertionError
[01:25] <bradb> It would be easier than an AssertionError, I agree.
[01:25] <SteveA> we should be nice, but only when it is no effort to do so
[01:25] <SteveA> or when there is obvious demand for it
[01:25] <kiko> ddaa++
[01:25] <SteveA> and, an OOPS report provides evidence of demand
[01:25] <SteveA> ddaa: thanks for the clarification.
[01:25] <mpt> exactly
[01:26] <SteveA> carlos: has your agenda item been fully dealt with?
[01:26] <carlos> yeah, I'm happy with this solution
[01:26] <SteveA> ok
[01:26] <SteveA>  * Keep, Bag, Change
[01:26] <carlos> thank you
[01:26] <ddaa> BAG: test suite run by pqm
[01:26] <ddaa> CHANGE: importd->bzr announced date
[01:27] <SteveA> i'm going to do a countdown, and then we can talk a little about the KBC comments
[01:27] <SteveA> 6
[01:27] <SteveA> 5
[01:27] <kiko> oh my god
[01:27] <SteveA> 4
[01:27] <kiko> really?
[01:27] <stub> Here
[01:27] <SteveA> 3
[01:27] <stub> Up to date
[01:27] <SteveA> 2
[01:27] <SteveA> 1
[01:27] <stub> (Man the lag is bad tonight!)
[01:27] <mpt> haha
[01:27] <SteveA> ddaa: what do you mean "test suite run by pqm" ?
[01:28] <SteveA> stub: even the clocks in thailand are laggy tonight
[01:28] <ddaa> pqm running launchpad test suite, occasionnally hanging, occasionally failing because of unrelated problems preventing good code from getting in.
[01:28] <SteveA> bug 28655
[01:28] <ddaa> I'm really really sick of my merges getting rejected or blocking pqm.
[01:28] <Ubugtu> An error has occurred.
[01:28] <SteveA> ddaa: okay.  noted.
[01:28] <ddaa> and I think in the end it's currently causing more harm than good to our development process.
[01:29] <SteveA> that needs more of an agenda item, so item for next week, unless we deal with it before then
[01:29] <stub> Is it only David's merges getting blocked?
[01:29] <kiko> no, stub 
[01:29] <bradb> It'd be nice if pqm could do proper setup/teardown, using a chroot jail.
[01:29] <kiko> it happens with me too.
[01:29] <bradb> and me
[01:30] <bradb> I wrote to lp@ about the 70 hour wait on a merge request the other day.
[01:30] <SteveA> ddaa: we should discuss the importd->bzr project after this meeting
[01:30] <daf> I think the specific problem is the buildbot tests
[01:30] <stub> bradb: It does. Just that it currently won't kill the tests if they hang for, say, 70 hours.
[01:30] <ddaa> bradb: that was because of one my merge caused it to hang, because of buildbot bitrot, see launchpad mailing list.
[01:30] <jblack> stevea: I'd like to be present for that meeting.
[01:30] <kiko> ddaa, and we can't remove buildbot completely? :)
[01:30] <SteveA> jblack: i'd like lifeless to be present too
[01:30] <daf> fixing the buildbot tests and adding a timeout are both good short-term steps, I think
[01:30] <ddaa> kiko: working on it, but needs time
[01:31] <SteveA> jblack, ddaa: so "after this meeting" means "tomorrow, effectively"
[01:31] <bradb> ddaa: pqm can (and should) be made smart enough to never hang for 70 hours, for any reason
[01:31] <jblack> stevea: I'll email the four of us for a meeting
[01:31] <SteveA> thanks jblack 
[01:31] <mpt> We seem to have a braindump from stub on the issue carlos raised: https://wiki.launchpad.canonical.com/PresentingDatabaseErrors
[01:31] <mpt> that's from November
[01:32] <kiko> ddaa, about delaying the bzr migration.
[01:32] <kiko> tell me more.
[01:32] <cprov> bradb: don't know, pqm don't need to be clever, but the tests do ;)
[01:32] <SteveA> i don't think we can make progress on pqm and testing issues without lifeless here
[01:32] <SteveA> kiko: please, after the meeting
[01:32] <ddaa> kiko: ask jblack to be invited to tomorrow's meeting
[01:33] <SteveA> ddaa: can you give kiko and me and jblack a quick summary right after this meeting?  and then we'll have a proper meeting to make decisions as jblack organises is
[01:33] <carlos> mpt, not really, that's specific for DB errors, I was talking about submitted/input data
[01:33] <SteveA>  * Production / staging (stub)
[01:33] <stub> SteveA: I believe we are using launchpad-dependancies on at least one box (gandwana). I don't know about the others.
[01:33] <mpt> carlos, so is it
[01:34] <carlos> mpt, dude, input data is not always DB data...
[01:34] <kiko> ddaa, I can't, I'm too busy this week
[01:34] <kiko> I'll have to talk to you about it next week.
[01:34] <ddaa> kiko: meeting discipline!
[01:35] <stub> Staging is currently tied up for publishing tests. Hopefully Daniel and Carlos have agreed on a timeshare of Asuka's CPU
[01:35] <Kinnison> I own 100% of it
[01:35] <Kinnison> Otherwise soyuz deployment is gonna be delayed even more
[01:35] <stub> Gina was not running on production until recently (which we didn't notice as error reports were not getting out). But it is now, and running every three hours.
[01:36] <carlos> Kinnison, I only need it for 8 hours on Sunday... O:-)
[01:36] <Kinnison> (currently disabled, but will re-enable after soyuz deployment is going nicely)
[01:36] <Kinnison> carlos: *nod* should be done by then
[01:36] <stub> Production was updated happily yesterday. Downtime was around 5 minutes. We should be able to keep the window small for most rollouts now.
[01:37] <stub> Session timeout has been increased to 60 days, so nobody should be complaining about being logged out any more. If they are, it may be a bug but it appears to be working fine.
[01:37] <stub> That is pretty much all I can think of on staging/production
[01:37] <kiko> drescher is being used as the soyuz production box
[01:38] <kiko> the uploader, publisher and build master will run on it
[01:38] <kiko> more details on the rollout will go out via EMAIL to the list this week
[01:38] <SteveA> stub: next production rollout -- is that tied to the soyuz rollout?
[01:39] <stub> Probably, yes, as the soyuz rollout will involve a database patch
[01:39] <kiko> I would rather we split out that patch.
[01:39] <kiko> the patch will need to be applied to production RSN or we will be blocked on that.
[01:40] <stub> If you can land the patch sooner, that would be good.
[01:40] <kiko> all right.
[01:40] <SteveA> all done?
[01:40] <kiko> stub, I'll have cprov pastebin it
[01:40] <SteveA>  * Three sentences
[01:40] <stub> kiko: I've seen the patch. It just needs to be landed. (unless it has been modified since I approved it)
[01:41] <kiko> stub, I'll do it.
[01:41] <jblack> DONE: 70% of wiki rewrite
[01:41] <SteveA> three sentences... fire when ready
[01:41] <mpt> DONE: MaloneFrontPages design work; LCA
[01:41] <mpt> TODO: LCA; MaloneFrontPages, DuplicateBugHandling etc specs
[01:41] <mpt> BLOCKED: no
[01:41] <bradb> DONE: Landed +filebug package guesser. Bug contact bug report designfest.
[01:41] <bradb> TODO: Implement on of the proposed bug contact bug report prototypes.
[01:41] <bradb> BLOCKED: Nope.
[01:41] <Kinnison> DONE: Soyuz deployment sprinting
[01:41] <stub> DONE: Zope3.2 migration work
[01:41] <stub> TODO: Zope3.2 migration work
[01:41] <stub> BLOCKED: Nope
[01:41] <kiko> DONE: soyuz rollout, perf analysis, management
[01:41] <matsubara> DONE: implemented an admin interface to merge accounts.
[01:41] <matsubara> TODO: fix validator problem on request fix pages, re-think the solution to canned search for commented bugs.
[01:41] <matsubara> BLOCKED: nope
[01:41] <Kinnison> TODO: Soyuz deployment sprinting
[01:41] <kiko> TODO: more of the same
[01:41] <jblack> TODO: 30% of wiki rewrite, presentation design for drupal
[01:41] <ddaa> DONE: handed optional-branch-title to daf. cscvs fix from Kamion + svn-1.2 compat fix. Partial design spec for buildd-ng.
[01:41] <ddaa> TODO: remove gnarly from cscvs, add bzr to cscvs
[01:41] <ddaa> BLOCKED: buildbot test suite, importd2bzr merge, bzr signing, baz2bzr failures
[01:41] <niemeyer> DONE: Arrived from Sprint, organizing, researching, documenting, Smart maintenance, ...
[01:41] <niemeyer> TODO: Move forward
[01:41] <niemeyer> BLOCKED: Nope
[01:41] <jblack> BLOCKERS: None
[01:41] <BjornT> DONE: some work on SupportTrackerViews. reviews. looked at email
[01:41] <BjornT> interface issues, found some bugs.
[01:41] <BjornT> TODO: finish SupportTrackerViews implemenation. fix a few bugs in the email interface. hopefully get started on bug watches improvements.
[01:41] <Kinnison> BLOCKED: time in the day, speed of asuka
[01:41] <kiko> BLOCKED: by various things, but I'm working on all of them
[01:41] <BjornT> BLOCKED: no
[01:41] <salgado> DONE: Finished fixing bugs #5394 and #5324, reviewed cprov's uploader-tests branch, some random fixes and more MirrorManagement work
[01:41] <salgado> TODO: Review cprov's branch again, get fixes for #5394 and #5324 merged, fix people vocabs as kiko suggested and discuss some MirrorManagement issues with cprov
[01:41] <salgado> BLOCKED: No
[01:41] <daf> DONE: summary-o-matic, optional-branch-title, bug triage, meeting summary
[01:41] <daf> TODO: Malone bug filtering, meeting summary
[01:41] <daf> BLOCKED: no
[01:41] <jblack> BLOCKED: None
[01:42] <Kinnison> jblack: use 'BLOCKED' not 'BLOCKERS' please
[01:42] <Kinnison> jblack: ta
[01:42] <gneuman> DONE: few fixes
[01:42] <carlos> DONE: language packs, PoMsgSetPage review, user support
[01:42] <carlos> TODO: more language packs, Finish PoMsgSetPage, performance issues with suggestions
[01:42] <carlos> BLOCKED: No
[01:42] <SteveA> DONE: vacation days in london, catching up after london meetings
[01:42] <SteveA> TODO: land jamesh's __len__ work, crowd security optimisation
[01:42] <SteveA> BLOCKED: no
[01:42] <gneuman> BLOCKED: no
[01:42] <jblack> kinnison: aye. Sometimes I mislabel on accident. I rewrite when that happens.
[01:42] <jordi> DONE: cleaning import queue of series files
[01:42] <SteveA> spiv: DONE: SFTP tests passing, lots of reviews, some librarian issues diagnosis.
[01:42] <SteveA> spiv: TODO: Get all the little niggling things done to get SFTP ready for rocketfuel (needs SVN Twisted, branches need joining, etc).  Sort out the librarian.
[01:42] <SteveA> spiv: BLOCKED: no (co-ordinating with Robert on the SFTP issues)
[01:43] <gneuman> TODO: finish 2 bug fixes
[01:43] <jordi> TODO: cleanup import queue for distro files, update wiki
[01:43] <jordi> BLOCKED: none
[01:43] <SteveA> spiv: My activity reports are up to date
[01:43] <jblack> whoah
[01:44] <SteveA> ddaa:  BLOCKED: buildbot test suite, importd2bzr merge, bzr signing, baz2bzr failures
[01:44] <SteveA> i guess we should talk about that a little later
[01:44] <ddaa> sure
[01:44] <SteveA> anyone else blocked?
[01:44] <cprov> DONE: Soyuz deploymet sprint
[01:44] <cprov> TODO: more Soyuz
[01:44] <cprov> BLOCKED: none
[01:45] <SteveA> okay
[01:45] <SteveA> that's it
[01:45] <SteveA> MEETING ENDS
[01:45] <kiko> thanks steve
[01:45] <SteveA> thanks everyone -- we made it start and end on time.
[01:45] <jblack> SteveA: DatabaseSetup is consistant with RocketfuelSetup and PQMSetup.
[01:45] <SteveA> jblack: great, thanks
[01:46] <ddaa> jblack: SteveA: kiko: if you want to talk about importd->bzr now, I suggest we move to #canonical-meeting
[01:46] <jblack> neimeyer: ping
[01:46] <SteveA> ddaa: agreed
[01:46] <jblack> ddaa: lifeless isn't here
[01:46] <carlos> SteveA, when do you want to talk about the UnexpectedFormData exception?
[01:46] <jblack> I sent an email and all.. :) 
[01:46] <niemeyer> jblack: pnog :)
[01:46] <SteveA> jblack: i want us to see what the issues are, and have a proper meeting with lifeless later
[01:47] <jblack> niemeyer: You did fedora packages for bzr. The wiki says that bzr is still at 0.0.7 or 0.1.1 or somesuch. Can you make a new set? 
[01:47] <niemeyer> jblack: Of course
[01:47] <niemeyer> jblack: They're rpm packages, not specifically to Fedora
[01:47] <jblack> Thank you sir. I appreciate it a lot
[01:48] <SteveA> carlos: put it in canonical/launchpad/interfaces/launchpad.py, make it derive from AssertionError, give it a docstring saying that it is for when form data is not what is expected by a form handler, so most likely a hand-crafted URL.
[01:48] <carlos> SteveA, ok, thanks
[01:48] <niemeyer> jblack: Glad to help
[01:49] <ddaa> kiko: -> #canonical-meeting
[01:49] <kiko> ddaa, hmmm ok
[01:49] <SteveA> daf: would you do a write up of this meeting?
[01:57] <daf> SteveA: yes
[01:57] <SteveA> thanks daf
[01:57] <kiko> hello mantiena
[01:57] <kiko> and his alter ego mantiena-baltix 
[01:58] <mantiena-baltix> hi kiko ;)
[01:58] <kiko> hey there
[01:58] <kiko> have you been able to +addmilestone, mantiena?
[01:59] <mantiena-baltix> kiko, yes, but I'm still not able to +addrelease :(
[01:59] <kiko> mantiena-baltix, +addrelease is more complicated, I would need to think about that somewhat.
[02:00] <daf> SteveA: first draft: https://wiki.launchpad.canonical.com/DevelopmentMeeting20060126
[02:02] <mantiena-baltix> kiko, strange :-/
[02:02] <kiko> mantiena-baltix, it's because the way launchpad manages the repository. it may be silly. I need to talk to Kinnison and cprov a bit.
[02:04] <Seveas> what's the status on the XML interfaces to malone? I am getting pretty tired of having to fix Ubugtu after every update
[02:05] <mantiena-baltix> kiko, maybe there is some bug reported about this ? I wanna to subscribe :)
[02:06] <kiko> mantiena-baltix, maybe there is, care to try and find it?
[02:06] <SteveA> Seveas: i proposed a text/plain page on a bug that returns the info ubugtu needs
[02:06] <SteveA> this can be done in 30 mins
[02:07] <SteveA> rather than the big task of getting an xmlrpc specced out and written
[02:07] <Seveas> that would rock, as long as the format of that page doesn't change :)
[02:08] <SteveA> Seveas: can you think about what kind of text/plain output you'd want for a bug?  stick it on a wiki page, BugPageForUbugtu
[02:08] <daf> I could do with stuff like that for some of the screen scraping I've done
[02:08] <SteveA> and that can form a test for someone implementing it soon
[02:08] <cprov> salgado: ping
[02:08] <SteveA> daf: can you work with Seveas on this, and make something quick and easy?
[02:08] <daf> SteveA: yes
[02:09] <salgado> cprov, pong
[02:09] <daf> (using BeautifulSoup works, but it's still a little brittle)
[02:10] <SteveA> this is basically fingering a bug
[02:10] <SteveA> if anyone remembers the finger protocol...
[02:12] <cprov> salgado: there is a branch of mine that has the concept proff of the mirror-management daemon.
[02:12] <cprov> salgado: it's listed in the spec (cprov/launchpad/mirror-management)
[02:12] <cprov> salgado: hope it helps you ;)
[02:14] <SteveA> niemeyer: difflib patch?
[02:16] <niemeyer> SteveA: What about it?
[02:17] <siretart> I'm getting an error OOPS-26B201 when trying to create a poll. Is this a known bug or shall I file a new one?
[02:17] <SteveA> can you get it to ddaa?
[02:17] <ddaa> niemeyer: actually, it wouldbe best if you could send me a wholesale patched difflib
[02:17] <ddaa> so I can just drop it in place
[02:18] <niemeyer> SteveA: I've sent it monday
[02:18] <niemeyer> ddaa: Didn't you get it?
[02:19] <ddaa> niemeyer: yeah, I saw a patch.
[02:19] <ddaa> nevermind, I'll do whatever it takes
[02:19] <ddaa> just pick a random python install and check the patch applies
[02:20] <niemeyer> ddaa: I'll send you the patched difflib again
[02:20] <niemeyer> ddaa: With the version I'll probably apply upstream
[02:20] <mantiena-baltix> SteveA, look at baltix bug 29194 :)
[02:20] <Ubugtu> moo
[02:20] <Ubugtu> Error: Could not parse data returned by Malone: 'NoneType' object has no attribute 'group'
[02:20] <niemeyer> ddaa: Sent!
[02:21] <siretart> ok. my date field was wrong
[02:21] <ddaa> niemeyer: thank you
[02:21] <SteveA> mantiena: is it related to bug 1 perhaps?
[02:21] <mantiena-baltix> kiko-fud, maybe you know aproximatly when bug with +addrelease will be fixed ?
[02:22] <mantiena-baltix> SteveA, yes, this is expanded translation to Lithuanian of bug 1 :)
[02:22] <salgado> stub, around?
[02:22] <SteveA> you could alternatively add baltix as a target for bug 1
[02:22] <SteveA> or additionally even
[02:22] <stub> salgado: yes
[02:23] <mantiena-baltix> SteveA, in bug 29194 there is sections about open standards, because it Lithuania main problem with Linux is widely spread closed document formats in public sector and government institutions :(
[02:24] <SteveA> isn't there a law now saying that government institutions must use open standards?
[02:24] <salgado> stub, I'm trying to find out what's making the queries in the people vocabs so slow... how much do we think we could gain by not doing that substring match on email addresses?
[02:24] <mantiena-baltix> SteveA, main difference between bug 1 and 29194 is, that bug 29194 is mainly about Lithuania and Latvia
[02:24] <SteveA> i see.  different place, different flavours of problem
[02:26] <mantiena-baltix> SteveA, hehe, this law is about 3 years, but government institutions doesn't follow this law :( More, than 90% government instututions publishes documents only in Microsoft Office formats and requires documens in Microsoft Office formats from citizens and enterprises :((((
[02:26] <mantiena-baltix> I don't know a way to force government institutions to follow the law :(((
[02:27] <SteveA> idea: compile a list of the situations, and see if a journalist would be interested in running a short article on it
[02:28] <stub> salgado: The substrig match on email address only has a wild card on the right hand side (lower(email) like 'foo%' ). So it is using an index, and we won't gain much.
[02:29] <sabdf1> bug 29722
[02:29] <mantiena-baltix> SteveA, I know this idea, but I don't know any journalists :(
[02:29] <sabdf1> no moo
[02:29] <SteveA> someone at AKL must know journalists
[02:30] <stub> salgado: Have you seen https://launchpad.net/products/launchpad/+bug/29725 ?
[02:32] <mantiena-baltix> SteveA, it seems noone at AKL can't help here, because I told this idea several times to akl-wg (akl-workgroup) mailing list
[02:32] <salgado> stub, no, I'm looking at it now
[02:33] <mantiena-baltix> SteveA, one person from AKL sometimes helps with such things, but he is very busy now :(
[02:33] <salgado> stub, but that time out is because of the query #31, isn't it? (31  	10619ms  	SELECT DISTINCT * FROM ( SELECT DISTINCT Person.id, Person.displayname FROM...)
[02:34] <SteveA> mantiena-baltix: it's tough.  i guess that if one group of people collect the information, then it can wait on some weblog, for someone else who knows journalists to point it out
[02:34] <stub> There are also a few hundred other queries in there... that one is taking 10 seconds. The timeout is 25 seconds.
[02:34] <BjornT> mantiena-baltix: if you send me a short summary (in lithuanian) about what the story would be about, i could pass it on to a few people i know.
[02:34] <stub> salgado: But yes, that query should be optimized if possible.
[02:35] <Seveas> SteveA, daf, https://wiki.ubuntu.com/BugPageForUbugtu
[02:37] <lbm> can i get a direct link to upload single po-file?
[02:38] <daf> Seveas: that looks good
[02:38] <salgado> stub, I was expecting that you'd tell me if that's possible ;)
[02:38] <SteveA> Seveas: okay.  actually, it should be on the launchpad wiki.  looks very easy to implement.  we should think about what the page name will be, and also put whether the bug is private or not in there.
[02:38] <Seveas> SteveA, I have no access to the lp wiki :)
[02:38] <daf> Seveas: you should do
[02:38] <stub> salgado: I'm looking at it now. I recall we already optimized this, but that would have been for 7.4. 
[02:38] <daf> https://wiki.launchpad.canonical.com/
[02:39] <Seveas> SteveA, ubugtu does not login, so it can't see private bugs anyway
[02:39] <daf> Seveas: may I suggest that the format make it clearer where one task ends and another begins?
[02:39] <salgado> stub, indeed, the optimization we did was to make it use an UNION instead of an OR
[02:39] <daf> I could use that format with a little modification for lists of bugs
[02:40] <Seveas> daf the idea is that a line that starts with affects: indicates the start of a new block, but i'm open for suggestions 
[02:40] <daf> ah, that's simple enough
[02:40] <daf> Seveas: can you check whether you can access the wiki?
[02:41] <Seveas> "You are not allowed to edit this page."
[02:41] <daf> ah, you need to log in first
[02:41] <daf> (same as Ubuntu wiki, I think)
[02:41] <Seveas> right
[02:42] <Seveas> ok, that works, I'll move it
[02:42] <daf> thanks
[02:43] <stub> salgado: Why are we only matching on preferred email address?
[02:43] <salgado> stub, these double selects on the email address table (https://chinstrap.ubuntu.com/~dsilvers/paste/file3sy2mp.html) are issued by this code: https://chinstrap.ubuntu.com/~dsilvers/paste/fileRfojnz.html
[02:46] <stub> salgado: So that would be to populate the select box on the form with all possible matches. We should not do that if there are too many matches I guess.
[02:46] <stub> (for some value ot 'too many')
[02:46] <salgado> stub, I don't think we have a reason for only matching on preferred emails
[02:48] <salgado> stub, we could even do that if the batching were working as expected, but it seems to me that we're retrieving all results from the database every time
[02:49] <stub> salgado: So can you see why we can't remove this entire clause: 'teamowner IS NOT NULL OR (teamowner IS NULL AND password IS NOT NULL AND merged IS NULL AND EmailAddress.status = 4 )', and just replace it with 'merged is NULL' ?
[02:49] <salgado> stub, no, we can't do that because that would return people without a preferred email address
[02:49] <stub> ok
[02:50] <salgado> AIUI, to do the search on all email addresses of a person would make things even more complex.
[02:50] <salgado> am I wrong?
[02:50] <SteveA> Seveas: sure, but someone can make a tool that does log in.  Launchpad understands basic auth over HTTPS
[02:51] <janimo> hey
[02:51] <janimo> can an existing product be deleted?
[02:51] <daf> currently, no
[02:51] <janimo> it should be a project instead
[02:52] <Seveas> SteveA, ah, nice
[02:52] <salgado> stub, if at least we had a newer version of sqlobject with support to JOINS, I'd be able to rewrite that to make the batch works as expected
[02:52] <janimo> another q: I go to a product page
[02:52] <janimo> is there a way I can see which project(s) it belongs to?
[02:53] <janimo> I know actually but would like a link to it instead of me typing it in
[02:55] <daf> SteveA: I'm about to go out for lunch
[02:55] <daf> SteveA: shall we arrange a time for bug wrangling?
[02:55] <SteveA> i'm going for lunch too
[02:56] <daf> shall we provisionally say 15:30 UTC?
[02:56] <SteveA> shall we say at 1600 UTC?
[02:56] <SteveA> sure, 1530 works too
[02:56] <daf> yes
[02:56] <daf> ok, either is fine for both of us
[02:56] <SteveA> 1530 then
[02:56] <daf> ok
[02:58] <SteveA> Seveas: for the names of people, it would be good to include their launchpad name.  maybe like this:    reported-by: seveas (Dennis Kaarsemaker)
[02:58] <SteveA> what do you think?
[02:58] <Seveas> I see no reason why it would be good (neither do I see a reason not to include it)
[03:00] <SteveA> having the name there allows you to look think up on that person in launchpad, if you want to
[03:00] <SteveA> or, if it is a team, on that team
[03:00] <Seveas> hmm, yes, that would be useful
[03:00] <Seveas> bon apetit
[03:00] <SteveA> ai
[03:13] <salgado> was that cachedproperty (I think I heard about it some time ago) removed?
[03:13] <kiko> not that I know of
[03:13] <kiko> canonical.cachedproperty?
[03:13] <salgado> oh, right. it doesn't seem to be used anywhere
[03:14] <kiko> it is actually in cprov's branch
[03:14] <kiko> I'm not sure if elsewhere
[03:14] <salgado> right, I remember seeing it there
[03:15] <Kinnison> cachedproperty is used in the uploader
[03:15] <Kinnison> lots
[03:17] <mantiena-baltix> BjornT, you know lithuanian language ?
[03:17] <mantiena-baltix> kiko, maybe you know aproximatly when bug with +addrelease will be fixed ?
[03:18] <kiko> mantiena-baltix, did you find the bug?
[03:18] <ddaa> jblack: sent you mail with unixodbc failure details
[03:18] <mantiena-baltix> kiko, no, I still don't tried to search ;)
[03:19] <BjornT> mantiena-baltix: yeah, i know some lithuanian
[03:19] <jblack> ddaa: Thank you
[03:20] <bradb> BjornT: Is there a test bug report or something else you have available for doing manual functional tests of email UI?
[03:20] <ddaa> jblack: BTW I do not understand the mail you sent about "dogfooding supermirror"
[03:20] <jblack> ddaa: ok. can you reply to the email saying you don't understand, so that I can respond when I wake?
[03:21] <ddaa> okay
[03:23] <kiko> mantiena-baltix, I think the best solution for that, right now, is to request to launchpad-users@lists.canonical.com that your releases be added.
[03:23] <BjornT> bradb: no. there was talk about enabling the email system on staging, but it hasn't been done yet. i should look into that again, since it's quite useful.
[03:24] <bradb> BjornT: That might be tricky, because we wouldn't want to spam people with test email.
[03:25] <BjornT> bradb: of course, staging wouldn't send mail to people :) all mail would go to a special mailbox, which you could inspect.
[03:25] <bradb> ah, ok
[03:27] <jbailey> I'm going to raise the severity of 29444 - it seems to be infinitely reproducable on any package now.  Do we have any guidelines on the difference between 'major' and 'critical'?
[03:28] <kiko> bug 29444
[03:28] <Ubugtu> Malone bug 29444 in launchpad: "Timeout querying DistributionSourcePackageCache when searching for the evolution source package" [Normal,Unconfirmed]  http://launchpad.net/bugs/29444
[03:28] <Seveas> major: affects lots of people and is serious. critical: prevents release of Ubuntu
[03:28] <kiko> jbailey, critical is stuff which doesn't let you do your job at all
[03:29] <jbailey> Seveas, kiko: Thanks.  Major should do.  It only prevents me from filing bugs.
[03:29] <Seveas> or in this case: prevents the use of launchpad
[03:30] <elmo> is there a function in LP to SQL-regex escape a string?
[03:31] <kiko> lemme see
[03:32] <ddaa> there's a function to escape like-patterns, but it does not really do something useful
[03:33] <elmo> ddaa: oh?
[03:33] <ddaa> it does not give something that can be interpolated in a like pattern
[03:33] <ddaa> which is the use case for such escaping...
[03:41] <bradb> BjornT: To report a bug in more than one package, do I simply add another "affects" stanza to my email? I couldn't find documentation on that use case.
[03:47] <BjornT> bradb: no, that's not possible atm, but i think it should. currently it will change the source package of the existing task, but that's is better handled by some other command in the future.
[03:47] <bradb> BjornT: So one must send two emails currently?
[03:47] <kiko> BjornT, how does one change the text ValidPersonOrTeam in the vocabulary popup?
[03:48] <BjornT> bradb: no, that will have the same effect as one email. it's quite a small fix to make it work, though, and i'm currently fixing a bug in the email system, so i could probably fix it at the same time.
[03:49] <BjornT> kiko: let me take a quick look
[03:49] <bradb> BjornT: Is it possible to open a bug on more than one package with the email UI?
[03:49] <kiko> or SteveA 
[03:51] <BjornT> bradb: no, that use case isn't covered. but as i said, i think it should, and i could fix that. (that means that it won't be possible to change the package until we add another command for it, though)
[03:51] <bradb> ok, thanks, just updating the docs
[03:55] <stub> salgado: I'm having difficulty optimizing that query with one that is reliably fast (btw. the existing one is sometimes fast too! Depends on what is in cache). I'll look into setting up a materialized view using triggers tomorrow to sort this out.
[03:56] <salgado> stub, that'd be great. thank you
[03:56] <BjornT> kiko: the only way to change the text ValidPersonOrTeam is to rename the vocabulary. i haven't thought much about it yet, but one option would be to add a title attribute to our vocabularies, should be a quite small fix.
[03:57] <kiko> BjornT, that would make us look less like a hack job, yes
[03:57] <salgado> stub, I'm checking what I can do to get rid of the unnecessary queries that the people vocabs are generating
[03:57] <kiko> cool.
[03:58] <stub> BjornT: Just add the attribute to IHugeVocabulary would be simplest
[03:59] <BjornT> stub: yeah
[04:00] <kiko> stub, salgado: make sure you coordinate on the fixes.. email is good
[04:00] <BjornT> kiko: do you want me to send an email to someone who has time and explain what to do?
[04:01] <kiko> sounds good
[04:01] <salgado> I can do that, as I'm already changing some things in the vocabs
[04:04] <kiko> carlos, want to talk about the +translate timeout we saw there?
[04:04] <carlos> Sure
[04:04] <carlos> kiko, here?
[04:04] <kiko> why not?
[04:05] <carlos> ok
[04:06] <kiko> ok
[04:06] <kiko> for one
[04:07] <carlos> kiko, https://chinstrap.ubuntu.com/~jamesh/oops.cgi/2006-01-26/A184
[04:07] <kiko> why don't we issue queries with in (XX,XX,XX...) clauses?
[04:07] <carlos> for example....
[04:08] <kiko> https://chinstrap.ubuntu.com/~dsilvers/paste/file12sKfr.html
[04:08] <kiko> look at that
[04:08] <kiko> there are many queries there that are identical, only varying per id
[04:08] <mantiena-baltix> BjornT, what is your email ? I will inform you when the story about Lithuanian government institutions and open standards law will be finished
[04:09] <kiko> I believe issueing one query with in (X,Y,Z) is much better than issueing 3 queries with X, Y and Z.
[04:09] <kiko> and stub can confirm
[04:09] <carlos> kiko, hmmm, and then split them again using python code?
[04:09] <kiko> well
[04:09] <carlos> every query is for a single message
[04:09] <BjornT> salgado: ok, bug 29782
[04:09] <kiko> don't you iterate over a set of messages anyway?
[04:09] <Ubugtu> Malone bug 29782 in launchpad: "SinglePopupWidget shouldn't use the vocabulary's name as title" [Normal,Confirmed]  http://launchpad.net/bugs/29782
[04:09] <kiko> thanks BjornT 
[04:10] <kiko> thanks salgado
[04:10] <carlos> kiko, it's not so simple
[04:10] <carlos> kiko, it's a IPOTMsgSet.foo call
[04:10] <BjornT> mantiena-baltix: bjorn.tillenius at gmail.com
[04:10] <kiko> carlos, what does that mean?
[04:10] <carlos> kiko, with your suggestion we cannot do that anymore as the context disappear and instead of one context, we get 10
[04:11] <kiko> I don't quite understand what you mean
[04:11] <carlos> kiko, if you want to do the IN (XX, XX, XX)
[04:11] <kiko> oh.
[04:11] <carlos> we need to move the context from IPOTMsgSet to a IPOFile or IPOTemplate
[04:11] <kiko> or to IPOTMsgSetCollection
[04:12] <carlos> kiko, that's what IPOTemplate does
[04:12] <carlos> a collection of IPOTMsgSets
[04:12] <carlos> it makes no sense to add a new class for this
[04:12] <kiko> whatever
[04:12] <carlos> I don't have any problem to move to that kind of optimization
[04:12] <kiko> that would avoid a gazillion queries there
[04:12] <kiko> however
[04:13] <carlos> but I will do some performance checks first because I'm not sure if the timeout will disappear...
[04:13] <kiko> the distinct problem
[04:13] <kiko> that's different, I believe.
[04:13] <kiko> is there a bug filed on +translate timeouts?
[04:14] <carlos> we have one about suggestions
[04:14] <carlos> but I think I closed it
[04:14] <carlos> we need to reopen it again
[04:15] <kiko> carlos, you are issuing the query 3 times
[04:15] <kiko> which makes the page 3 times as likely to time out 
[04:15] <carlos> kiko, https://launchpad.net/products/rosetta/+bug/5751
[04:15] <Ubugtu> Error: I cannot access this bug.
[04:15] <kiko> or perhaps it's more than 3 times
[04:15] <kiko> and we just managed to render 3 stanzas
[04:15] <carlos> kiko, it's 10 times
[04:16] <kiko> right.
[04:16] <carlos> three different queries
[04:16] <carlos> 10 times per page
[04:16] <mantiena-baltix> kiko, btw, I still can't find a way how to register an official baltix mirror (primary download location))
[04:16] <kiko> mantiena-baltix, talk to salgado
[04:17] <kiko> carlos, I see.
[04:18] <kiko> you could a) build a view to join the information you need into a single sqlobject, b) do a in (X, Y, Z..) query to avoid doing that query N times 
[04:20] <kiko> carlos, do you really need the LEFT OUTER JOIN there?
[04:20] <carlos> kiko, yes
[04:21] <carlos> it's the only way to get the row even when that field does not exists
[04:21] <kiko> I see
[04:21] <carlos> I will do some testing first using raw SQL queries on staging
[04:22] <kiko> sure.
[04:22] <kiko> less queries may reduce the risk of contention however
[04:23] <carlos> kiko, but if the query is still slow... we will see...
[04:23] <cprov> stub: ping
[04:23] <salgado> mantiena-baltix, for now, I can mark it as official for you, but you should be able to do that once bug 29785 is fixed
[04:23] <Ubugtu> Malone bug 29785 in launchpad: "The mirror administrators should not be the same for all distributions" [Normal,Unconfirmed]  http://launchpad.net/bugs/29785
[04:25] <cprov> kiko: can you review the soyuz_prodution diff in https://chinstrap.ubuntu.com/~dsilvers/paste/filefs1Rlo.html ?
[04:26] <cprov> kiko: the DB patch was approved and has a official number, just in case, review it too,  https://chinstrap.ubuntu.com/~dsilvers/paste/fileLJU8xN.html
[04:32] <cprov> kiko: so, do I have r=kiko for it ?
[04:32] <kiko> yes
[04:33] <daf> SteveA: yo
[04:33] <SteveA> daf: hi
[04:47] <stub> cprov: pong
[04:50] <cprov> stub: could you have a look on that DB patch  again ? preform ...
[04:51] <cprov> stub: -> https://chinstrap.ubuntu.com/~dsilvers/paste/fileLJU8xN.html
[04:51] <kiko> pro-forma? 
[04:52] <stub> Looks like it did last time I looked at it.
[04:52] <kiko> cool
[04:52] <kiko> cprov, land it and give stub a revision number
[04:54] <cprov> kiko: ok
[04:58] <cprov> yes, 'pro forma' as in http://www.thefreedictionary.com/pro+forma, local dict sucks
[05:05] <ddaa> kicking conversion of difflib failures using local difflib hack
[05:07] <ddaa> now, let's proceed with a lobotomy on buildbot's test suite...
[05:13] <kiko> is PQM not accepting requests?
[05:18] <cprov> stub: PQM is not listen my mail, you probably better rollout directly from my branch, is it possible ?
[05:18] <stub> Just land that DB patch you mean? Sure.
[05:18] <kiko> stub, there are code changes if you want tests to pass
[05:19] <stub> Ahh.
[05:19] <stub> How about I ask PQM to merge it on your behalf?
[05:19] <stub> What is the branch?
[05:20] <kiko> that would be okay, if PQM actually works for you
[05:20] <kiko> it currently hates me
[05:20] <cprov> stub: chinstrap.warthogs.hbd.com:/home/warthogs/archives/cprov/launchpad/small-fixes/
[05:22] <stub> You need to use chinstrap.ubuntu.com for a start...
[05:23] <daf> SteveA: hello?
[05:24] <kiko> heh
[05:24] <stub> With pqm now
[05:24] <kiko> that's better
[05:25] <stub> echo star-merge sftp://chinstrap.ubuntu.com/home/warthogs/archives/cprov/launchpad/small-fixes sftp://chinstrap.ubuntu.com/home/warthogs/archives/rocketfuel/launchpad/devel | gnome-gpg --clearsign | mail -s "[r=kiko]  DistroReleaseQueue table updates" pqm@pqm.ubuntu.com
[05:26] <kiko> stub, yeah, warthogs.hbd.com is dead
[05:28] <cprov> stub: oh man .. did it too, sorry ... once isn't enough, we are going to have it twice :( 
[05:28] <ddaa> that's weird, I have a bunch of ssh connections to within the DC
[05:28] <ddaa> and they work
[05:31] <iwj> Hi again.  Does anyone here know why Malone pages sometimes don't set the browser window title ?
[05:32] <bradb> iwj: Probably a usability bug. URL?
[05:32] <kiko> I've seen this happen before -- I get a ???
[05:32] <iwj> Any Malone page seems to do it for me.  Eg,  https://launchpad.net/distros/ubuntu/+source/firefox/+bug/29412
[05:32] <Ubugtu> Malone bug 29412 in firefox: "Firefox don't look for plugins in $MOZ_PLUGIN_PATH" [Normal,Unconfirmed] 
[05:32] <kiko> iwj, that page has a title for me
[05:32] <bradb> me too
[05:32] <iwj> It seems to depend how you get to it.
[05:33] <kiko>     <title>Bug #29412 in firefox (Ubuntu): Firefox don't look for plugins in $MOZ_PLUGIN_PATH</title>
[05:33] <ddaa> me too (with firefox)
[05:33] <kiko> I've seen that happen on dapper
[05:34] <bradb> iwj: Presumably other non-bug-specific pages do this for you too?
[05:34] <ddaa> thought that dapper was the rigid and boring release... how comes I seem to hear so much about mysterious mystical breakage...
[05:34] <iwj> bradb: I haven't completely tracked it down but it seems most common with bugs pages.  I don't use many other LP pages very much.
[05:34] <kiko> iwj, I think this is a browser bug, in firefox.
[05:34] <iwj> kiko: I think so too.
[05:34] <kiko> if you load the page in lynx, do you get a broken title?
[05:34] <iwj> Viuew
[05:34] <iwj> View Source seems to show what ought to be a title.
[05:34] <kiko> my test doesn't at least
[05:35] <kiko> it's not all pages, oddly enough
[05:35] <salgado> daf, why did you assign bug 1953 to me?
[05:35] <Ubugtu> Malone bug 1953 in launchpad: "E-mail addresses page should always have one radiobutton preset" [Minor,Confirmed]  http://launchpad.net/bugs/1953
[05:35] <daf> salgado: Steve told me to
[05:36] <kiko> daf, perhaps it would make sense to assign it to mpt first since he needs to detail the design 
[05:36] <salgado> hmmm. it was mpt who assigned it to himself, because first of all he needs to design the new UI
[05:36] <daf> kiko: it was assigned to MPT
[05:36] <kiko> right
[05:36] <kiko> and mpt needs to specify what he wants
[05:36] <daf> I thought it was just a matter of making the current email address selected
[05:36] <iwj> Umm, I think it might be related to the UTF-8 quotes.
[05:36] <kiko> might be
[05:37] <kiko> daf, but we need more information
[05:37] <iwj> What wm do you use ?
[05:37] <daf> Steve's interenet is down, by the way
[05:37] <kiko> iwj, I use ion2.
[05:37] <daf> ISP problems
[05:38] <iwj> kiko: Hmm, it WFM with dapper's metacity.
[05:38] <iwj> But not with my ancient vtwm.
[05:39] <iwj> Is LP supposed to be useable, in general, with non-UTF-capable software ?
[05:40] <kiko> I don't think so.
[05:40] <daf> that's not a goal currently
[05:41] <iwj> OK then.
[05:41] <stub> I doubt it ever will be - we already have people with Kanji names.
[06:14] <dilys> Merge to devel/launchpad/: [r=kiko]  DB patch for Production Soyuz and test fixing. (r3033: Celso Providelo)
[06:16] <salgado> BjornT, around?
[06:19] <kiko> stub, merged!
[06:20] <cprov> stub: you can rollout that patch.
[06:21] <stub> Onto staging or production? And if production, what it the reason?
[06:25] <kiko> stub, production. the reason is that when we move to soyuz pointing to the production db, we need the database to be at the correct schema version.
[06:25] <Kinnison> and on staging please
[06:26] <stub> So we don't need production straight away, just before we switch soyuz
[06:26] <kiko> right -- it could be tomorrow
[06:26] <stub> Kinnison: Will I screw you up if I drop the staging DB now?
[06:26] <kiko> staging needs to be now because we're blocked on this
[06:26] <Kinnison> stub: drop?
[06:26] <kiko> stub, by drop do you mean "shut down"?
[06:27] <stub> Yes
[06:27] <Kinnison> why shut down?
[06:27] <stub> To ensure it all applies happily. I can try applying the db patch manually, which will work if nothing is accessing that table
[06:28] <kiko> why not try it?
[06:28] <stub> db patch has been applied
[06:28] <Kinnison> nothing will be touching that table
[06:28] <Kinnison> thanks
[06:29] <stub> kiko: because things work smoother if I follow procedures rather than improvise ;)
[06:29] <cprov> stub: ehe
[06:56] <cprov> $python scripts/process-upload.py  -MNvv -C buildd -b 9 /home/cprov/www/bad_binaries/
[07:18] <dilys> Merge to devel/launchpad/: [trivial]  add package bug contact sample data (r3034: Brad Bollenbach)
[07:21] <bradb> pqm is like a rocket the last few days.
[07:21] <jbailey> bradb: All launches colide with something, and a fair chance of exploding?
[07:22] <bradb> I actually *prefer* to just use pqm to run my tests, instead of locally, because I get the results much quicker.
[07:22] <bradb> jbailey: heh
[07:29] <salgado> BjornT, ping?
[07:32] <kiko> bradb, I do that too
[07:32] <kiko> it's not a crime to use balleny
[07:40] <BjornT> hi salgado 
[07:42] <salgado> BjornT, there's an issue related to having a getInitialValuesFromSearchParams(), because in the case of  AdvancedBugTaskSearchView, getExtraSearchParms() will call getWidgetsData()
[07:42] <salgado> and in turn, the getInitialValuesFromSearchParams() is called before setUpWidgets()
[07:48] <BjornT> salgado: my thought was that there should be a getInitialValues(), or maybe even an attribute, initial_values. then you would set the initial values when you create AdvancedBugTaskSearchView using the new converter function.
[07:48] <BjornT> salgado: to clarify, if a view doesn't use widgets, it shouldn't set any initial values in its initialize()
[07:54] <salgado> BjornT, hmmm, now I think I got it
[08:05] <stub> SteveA: LaunchpadRootNavigation is being registered as a view with no name on RootObject, correct?
[08:08] <SteveA> stub: the navigation is registered as an IBrowserPublisher view
[08:08] <SteveA> with no name
[08:08] <SteveA> so it isn't a "page to publish" view with no name
[08:08] <SteveA> but rather, a "I do URL traversal" view
[08:09] <stub> Which is exactly the same as the Resources object. So I understand why things are failing. I don't understand how they used to be working under 3.0 :-/
[08:09] <SteveA> the Resources object?
[08:09] <SteveA> that would be odd
[08:10] <SteveA> resources in 3.0 are done as a magic namespace thing
[08:10] <SteveA> and don't interfere with regular URL traversal
[08:10] <stub> Have a look at lib/canonical/launchpad/webapp/configure.zcml -- search for @@
[08:10] <stub> Ahh... I probably changed that to the new way...
[08:11] <SteveA> i see
[08:11] <stub> Nope... that was how it was registered under 3.0 too...
[08:11] <SteveA> um
[08:11] <SteveA> that is odd
[08:12] <SteveA> hang on
[08:12] <SteveA> it is a brwoser:page
[08:12] <SteveA> so, it is not being registered for IBrowserPublisher at all
[08:12] <SteveA> it is registered for whatever views are registered for
[08:12] <SteveA> so, /@@/ is a page, a view in its own right
[08:13] <SteveA> whereas the RootNavigation is the traversal component used to traverse the root
[08:13] <SteveA> i'd be tempted though to remove the Resource view
[08:13] <SteveA> and add that to the RootNavigation perhaps
[08:13] <stub> Under 3.2, /@@/launchpad.css is attempting to traverse to launchpad.css via the ToorNavigation
[08:14] <SteveA> i see
[08:15] <SteveA> is there a tree i can grab overnight, and look at tomorrow?
[08:15] <stub> Sure. I'll push it up.
[08:15] <SteveA> ok
[08:16] <SteveA> there could be many things going on here
[08:16] <SteveA> like, issues with exactly how we're overriding certain zcml directives
[08:16] <SteveA> or a change in the view lookup interface a navigation component needs to work with
[08:17] <kiko> cprov, chinstrap.ubuntu.com:/home/warthogs/archives/kiko/launchpad/cprov-hacking/
[08:17] <hannosch> hi. could somebody do a quick approval of two new products for rosetta or should I write to the mailing list?
[08:18] <kiko> jordi, carlos?
[08:19] <carlos> hannosch: jordi will do it as soon as possible. But we can do now an easy check
[08:19] <carlos> hannosch: are you one of the maintainers of that product?
[08:19] <hannosch> carlos: yep, it's too little plone add-om products
[08:20] <carlos> oh, ok
[08:20] <hannosch> I can do the import stuff, but need somebody to allow me to do it ;)
[08:20] <cyberix> Do you think someone would mind, if we'd start translating Ubuntu into Lojban language. I don't think there are lot of people who would want to put serious effort in this.
[08:20] <cyberix> But there are few people who might translate something every now and then, just for fun.
[08:21] <carlos> hannosch: the new system allows you to do the initial upload directly into Rosetta and will wait for an admin review
[08:21] <carlos> hannosch: just select the productseries where you want to attach it
[08:21] <cyberix> Probably would not hurt anyone to have such project?
[08:21] <LarstiQ> cyberix: good way to learn it too
[08:21] <cyberix> Even, if it was not very productive
[08:22] <carlos> cyberix: dude, if you know Lojban, just translate ;-)
[08:22] <cyberix> I'm just beginning to learn it.
[08:22] <carlos> cyberix: there isn't a Lojban translation team atm
[08:23] <cyberix> I should set up a translator team in Launchpad then?
[08:23] <carlos> so you can go ahead and translate without doing anything else
[08:23] <carlos> cyberix: you only need to request a translation team
[08:23] <cyberix> How is that done?
[08:23] <carlos> if you want to coordinate with other translators 
[08:23] <carlos> but as soon as one of the translators request it, you will not be able to translate more
[08:23] <LarstiQ> #lojban might have some interested people
[08:24] <carlos> until you join it so it's up to you
[08:24] <carlos> cyberix: if you want to create it, send an email to rosetta@launchpad.net requesting it
[08:24] <cyberix> LarstiQ: I was going to ask them, as soon as I have everything ready
[08:25] <LarstiQ> cyberix: I think it is a nice idea, it would provide a focus for me at least to seriously start learning Lojban, instead of just attending lectures
[08:25] <hannosch> carlos: thx, I added the first template on both
[08:25] <cyberix> How do I add a new language for a package?
[08:25] <carlos> cyberix: just select to translate into that language and add a translation ;-)
[08:28] <cyberix> LarstiQ: Maybe it is added by translating one string
[08:28] <LarstiQ> cyberix: try it! :)
[08:30] <cyberix> Where can I find the places string that appears in menubar
[08:30] <cyberix> "places"
[08:31] <salgado> BjornT, I just replied to your review, and it'd be great if you could have a look at the diff I attached and tell me if I got the idea right this time
[08:36] <carlos> LarstiQ: it should...
[08:37] <carlos> LarstiQ: I see it there....
[08:38] <LarstiQ> carlos: at https://launchpad.net/rosetta/prefs ?
[08:38] <carlos> LarstiQ: yes
[08:38] <seb128> hi
[08:38] <carlos> LarstiQ: just after Portuguese (Brazil)
[08:38] <LarstiQ> carlos: I'm presuming the list is in alphabetical order, is that a faulty assumption?
[08:39] <seb128> lauchpad gives some oops while trying to add an upstream task to some bug, somebody interested? :)
[08:39] <LarstiQ> gah
[08:39] <carlos> seb128: file bugs? :-P
[08:39] <bradb> seb128: We have a bug open on that already.
[08:39] <LarstiQ> carlos: I have Punjabi there
[08:39] <seb128> bradb: it happens with all the tasks you mean?
[08:40] <bradb> seb128: this one: https://launchpad.net/products/malone/+bug/5757
[08:40] <Ubugtu> Malone bug 5757 in malone: "Oops from making a second fix request for the same product" [Normal,In Progress] 
[08:40] <seb128> hum
[08:40] <seb128> I don't do a second fix request
[08:41] <carlos> LarstiQ: so there is no problems, right?
[08:41] <LarstiQ> carlos: Punjabi != Lojban
[08:41] <bradb> seb128: What's the OOPS id?
[08:42] <carlos> LarstiQ: ok, I misread you....
[08:42] <carlos> sorry O:-)
[08:42] <seb128> bradb: don't bother you are right
[08:42] <seb128> there is one bug but it's on an upstream task
[08:42] <seb128> not a distro bug
[08:42] <BjornT> salgado: ok, you got the idea right. there's one small thing to do, i'll reply to the mail
[08:42] <seb128> you guys should have different colors for upstream/distro tasks :p
[08:42] <salgado> BjornT, good, thanks
[08:42] <seb128> bradb: thank you
[08:42] <bradb> seb128: no prob
[08:43] <carlos> hmmm
[08:43] <carlos> LarstiQ: it's in our database...
[08:43] <LarstiQ> carlos: likewise, I don't see Latin Serbian 
[08:44] <bradb> seb128: I prototyped a UI to improve that part of the page. I think you might have seen it already: http://flickr.com/photos/84096161@N00/88275619/.
[08:44] <bradb> But I'm not too serious on thinking it'll go much further than a prototype. :) It might stimulate thinking in the right direction.
[08:44] <carlos> Ok, I see the problem
[08:45] <carlos> LarstiQ: it should appear tomorrow
[08:45] <carlos> LarstiQ: about Latin Serbian...
[08:45] <carlos> is a limitation on our side
[08:45] <seb128> bradb: yeah, current page is way better than this mockup
[08:45] <carlos> LarstiQ: we don't fully support the locales like sr@Latn
[08:46] <LarstiQ> carlos: yeah, you said something about that a few days back
[08:46] <seb128> bradb: as I said by mail, you waste half of the page for what takes 1 line atm, you drop informations from the pages, and you push comments out of the page on that mockup
[08:46] <LarstiQ> carlos: what makes support difficult for that?
[08:46] <carlos> LarstiQ: the @Latn thing (a.k.a. variants)
[08:47] <carlos> We have half support for them
[08:47] <carlos> we store the .po imported
[08:47] <carlos> and we can export them
[08:47] <carlos> but we don't have a UI to reach them
[08:47] <LarstiQ> carlos: but when one requests a tarball, you do get them?
[08:48] <lifeless> moin moin
[08:49] <carlos> LarstiQ: Hmmm, I think so, yes
[08:49] <carlos> LarstiQ: but only if it existed upstream
[08:50] <seb128> bradb: how do you require a fix for dapper by example?
[08:50] <LarstiQ> carlos: good enough for me 
[08:50] <LarstiQ> carlos: any idea on when you will support it?
[08:50] <seb128> bradb: "Select distribution" knows about Ubuntu but not dapper
[08:51] <cyberix> LarstiQ: https://launchpad.net/distros/ubuntu/dapper/+source/menu/+translations
[08:51] <carlos> not sure... It depends on having a team on Rosetta that needs it working to update the translations
[08:51] <bradb> seb128: On the current bug page, you mean?
[08:51] <cyberix> LarstiQ: Translated a string
[08:51] <carlos> LarstiQ: let me check for the bug report ...
[08:52] <LarstiQ> carlos: oh, I can provide you with one
[08:52] <LarstiQ> cyberix: sweet :)
[08:52] <seb128> bradb: no, I click on "Request fix: in distribution"
[08:52] <bradb> seb128: Interesting you ask that, because on the new bug page, I bet it's much clearer to get a hint. :)
[08:53] <bradb> seb128: Currently, you do that from +editstatus
[08:53] <bradb> er, sorry, s/new bug page/new bug page prototype/
[08:54] <seb128> oh, it's an another package for the system?
[08:54] <carlos> LarstiQ: I thought we had it... but I don't see it so please, file a bug...
[08:54] <kathi_> hi
[08:54] <LarstiQ> carlos: a general report on variants?
[08:55] <carlos> LarstiQ: yes, noting that you need that functionality
[08:55] <bradb> seb128: You need only set the "Milestone" value on +editstatus. Sorry, I'm not sure what you mean about another package.
[08:56] <seb128> bradb: I was just looking on the new mockup and there is "report this bug in another Ubuntu package...", but it was not that :p
[08:57] <bradb> oh, no. perhaps the mental model put forth by Malone is even more broken than I thought.
[08:58] <seb128> I still don't get how doing that
[08:58] <seb128> I've a bug on 
[08:58] <seb128> package (upstream)
[08:58] <seb128> package (Ubuntu)
[08:58] <bradb> seb128: What bug are you looking at?
[08:58] <seb128> and now I want to request a fix on ... let's say hoary
[08:58] <seb128> because it's bugged on hoary and dapper and I want 2 taks
[08:58] <seb128> tasks
[08:58] <seb128> how do I do that?
[08:59] <seb128> bradb: no particular bug, just a wondering on how to that if the case happens one day :)
[08:59] <bradb> Ah, backporting, that's different than the dapper use case. Use the "Target Fix to Releases" link in the actions portlet.
[09:01] <LarstiQ> carlos: I hope that is readable
[09:01] <bradb> seb128: Does the help?
[09:01] <bradb> s/the/that/
[09:02] <seb128> bradb: it does, thanks ... it's not really obvjous, I would expect the "request fix in ..." doing that
[09:02] <seb128> like I request a fix in hoary
[09:02] <bradb> seb128: I can understand why that'd be confusing.
[09:02] <SteveA> stub: please mail me where the tree is pushed to.  i'm going offline for a while to upgrade packages.
[09:03] <seb128> bradb: anyway, now I know it's so it's fine enough for me :)
[09:03] <bradb> seb128: I'll file a bug on this problem.
[09:04] <seb128> ok, thank you
[09:04] <bradb> no prob
[09:04] <bradb> thanks for the feedback, as always
[09:05] <LarstiQ> wooh, big bug numbers!
[09:06] <salgado> BjornT, is it possible to run the docstring doctests from browser/bugtask.txt?
[09:07] <salgado> s/.txt/.py
[09:10] <BjornT> salgado: yes. add browser/tests/test_bugtask.py and do the same as is done in tests/test_datetimeutils.py
[09:10] <LarstiQ> cyberix: mind to check my contribution?
[09:11] <bradb> seb128: bug 29801
[09:11] <Ubugtu> Malone bug 29801 in malone: "seb128 reported fix targeting to be confusing" [Normal,Unconfirmed]  http://launchpad.net/bugs/29801
[09:16] <cyberix> LarstiQ: Where is it
[09:16] <seb128> bradb: I've subscribed to it :)
[09:16] <cyberix> LarstiQ: I'm no guru, but ofcourse I can look at it
[09:16] <bradb> seb128: cool
[09:17] <LarstiQ> cyberix: I'm trying to figure out how to easily display my change
[09:17] <LarstiQ> cyberix: need review filter is the trick
[09:17] <LarstiQ> cyberix: https://launchpad.net/distros/ubuntu/dapper/+source/menu/+pots/menu-sections/jbo/+translate
[09:17] <LarstiQ> cyberix: and then show: need review
[09:22] <cyberix> LarstiQ: I guess it is correct
[09:23] <salgado> BjornT, the test and the modified function: https://chinstrap.ubuntu.com/~dsilvers/paste/fileDlpYgI.html
[09:24] <LarstiQ> cyberix: hah, that sounds waay more confident than I did ;P
[09:25] <salgado> BjornT, actually, I think this one is better: https://chinstrap.ubuntu.com/~dsilvers/paste/filey0N98b.html
[09:26] <cyberix> Why doesn't this update? https://launchpad.net/distros/ubuntu/dapper/+lang/jbo
[09:29] <BjornT> salgado: yes, the else clause with a comment makes it clearer. i'd like to see some more tests, though, i said the test should include, not consists of, what i suggested ;) every code path should be tested.
[09:30] <LarstiQ> carlos: is bug 29800 what you need?
[09:30] <Ubugtu> Malone bug 29800 in rosetta: "variant (sr@Latn) support" [Normal,Unconfirmed]  http://launchpad.net/bugs/29800
[09:30] <salgado> BjornT, okay, but then I'd prefer to move these tests to bugtask-pages.txt. is that okay with you?
[09:33] <BjornT> salgado: if you think it becomes too big, sure
[09:34] <carlos> LarstiQ: yeah, thanks
[10:03] <LarstiQ> lamont: https://lists.ubuntu.com/archives/bazaar-ng/2006q1/007999.html might be of interest to you
[10:28] <salgado> BjornT, how about this one: https://chinstrap.ubuntu.com/~dsilvers/paste/fileHt25Up.html
[10:28] <salgado> (I decided not to moving the tests to a doctest, as they're not so big)
[10:28] <salgado> s/moving/move
[10:34] <BjornT> salgado: the first test should use any(), shouldn't it?
[10:35] <salgado> yes, it definitely should
[10:37] <salgado> BjornT, fixed. wanna look at the patch one last time?
[10:38] <BjornT> salgado: no, it should be ok
[10:41] <salgado> great. thanks
[10:45] <lifeless> salgado: generally, use the test type you find easiest to express the test, IMO.
[10:46] <carlos> LarstiQ: Lojban should appear now when we list languages
[10:47] <salgado> lifeless, usually I tend to write tests in the docstring if they're really small, but in this specific case I couldn't find any place where that test would "fit" better than in the docstring
[10:47] <LarstiQ> carlos: it does, thanks
[10:48] <carlos> you are welcome
[11:04] <dilys> Merge to devel/launchpad/: Fix https://launchpad.net/products/launchpad/+bug/1281 (Implement an admin interface to merge accounts). r=salgado (r3035: Diogo Matsubara, Guilherme Salgado)