[12:50] <lifeless> jamesh: had a chance to look at the patch yet ?
[12:58] <BjornT> kiko: pong
[12:58] <kiko> BjornT!
[12:58] <kiko> it's too late to IRC
[12:58] <kiko> go to bed
[12:58] <kiko> we can chat tomorrow
[01:00] <BjornT> well, i fell asleep earlier today and woke up again just now
[01:00] <kiko> BjornT, ok, I can give you a midnight chore. 
[01:01] <kiko> run "time bzr branch https://launchpad.net/products/bugzilla"
[01:01] <kiko> and tell me tomorrow how long it took
[01:01] <kiko> it takes 27 minutes here :-(
[01:01] <kiko> and on that note
[04:32] <mpt> Goooooooooooooooooood afternoon Launchpadders!
[04:34] <jamesh> stub: ping?
[04:34] <stub> jamesh: pong
[04:35] <jamesh> I was looking at the logs for the production/1.76 branch, and don't see a cherry pick of r4275
[04:35] <jamesh> did it actually get cherry picked?
[04:36] <Ubugtu> New bug: #73689 in launchpad "Malone doesn't take Debian bug tracker URL's" [Undecided,Unconfirmed]  http://launchpad.net/bugs/73689
[04:37] <stub> jamesh: Your right - looks like I missed it
[04:40] <jamesh> should I move it back to the "requested cherry picks" section, or can you handle it?
[04:44] <stub> I'm handling i
[04:44] <stub> t
[04:44] <stub> Just rerunning the tests now.
[04:55] <Ubugtu> New bug: #73695 in malone "Attachment description should be optional (taken from filename)" [Undecided,Unconfirmed]  http://launchpad.net/bugs/73695
[06:25] <stub> jamesh: The missing patch should now be live
[06:28] <jamesh> cool.  We seem to be able to render things like https://launchpad.net/distros/ubuntu/+source/xawtv/+bug/73077/comments/0 now :)
[06:28] <Ubugtu> Malone bug 73077 in xawtv "When running scantv I get a segmentation fault (core dumped) when ut gets to channel 77" [Undecided,Unconfirmed]  
[06:30] <jamesh> stub: btw, I fixed up the last few issues with my tzfile branch of pytz
[06:31] <jamesh> I can unpickle old pickles for time zones like W_minus_SU
[06:31] <stub> Cool. I was looking at it yesterday.
[06:32] <stub> I'll add some tests for that too - it is particularly important for ZODB applications that have pytz instances stored as persistent pickles.
[06:32] <jamesh> the test_tzinfo and test_docs tests pass
[06:32] <jamesh> I haven't checked test_zdump one because it doesn't work on x86-64
[06:33] <stub> I'll generate it here and give it a spin
[07:24] <lifeless> jamesh: thanks!
[08:53] <carlos> morning
[08:58] <mdke> morning carlos
[08:58] <carlos> morning
[08:58] <carlos> mdke: how's going?
[08:58] <mdke> carlos: good thanks. you?
[08:59] <carlos> fine, thanks :-D
[09:37] <jordi> mdke: had your session already, right? How did it go?
[09:38] <mdke> jordi: it's this evening, after yours I think
[10:35] <jordi> hmm, carlos, how could a template be imported as German?
[10:35] <jordi> there's even no Language field when importing templates
[10:36] <carlos> jordi: because for some reason, we imported de.po as the first file for that template
[10:36] <carlos> so the potemplate got de.po as its name
[10:36] <jordi> weird. I'm always careful about doing the template first
[10:36] <jordi> anyway, thanks for fixing
[10:36] <carlos> np
[10:38] <jordi> and thanks for the silva request. As soon as stub executes it, I'll mail thisfred.
[10:38] <jordi> it'd be good if you could remind me of the hidden translation-group links so I can teach thisfred on how he needs to manage it
[10:39] <jordi> https://launchpad.net/rosetta/groups/silva-translators
[10:43] <ddaa> jordi: haha!
[10:43] <ddaa> gtetrinet!
[10:43] <ddaa> this is so last year, you know? ;)
[10:50] <jordi> ddaa: I'm adding your email to the MAINTAINERS file so you share the pain mmmmkay :)
[10:51] <jordi> ddaa: not last year. This is 1990's :)
[11:00] <ddaa> mkay... the most intriguing pysvn exception ever
[11:01] <ddaa> pysvn._pysvn.ClientError: Code: 00020f58
[11:01] <ddaa> talk of XXth century technology...
[11:01] <jordi> heh
[11:02] <lifeless> double triple-wrappers
[11:03] <ddaa> you have no idea
[11:03] <ddaa> since it's actually for the cups import, which is itself a wrapper of wrappers of sorts :)
[11:12] <ddaa> lifeless: can you look for cscvs--insane now?
[11:13] <lifeless> ddaa: what do you need it for again?
[11:14] <ddaa> just want to get the complete history of cscvs back to the last public code
[11:14] <lifeless> /home/david/.arch-cache/archives/robertc@robertcollins.net--warty/cscvs--insane--1.0--patch-172
[11:14] <lifeless> on chinstrap
[11:14] <ddaa> that will also help me set the copyright notices, though it's obviously not indispensable for this
[11:14] <ddaa> ?
[11:15] <ddaa> dude, that's an arch-cache entry
[11:15] <ddaa> I'm looking for the actual archive
[11:15] <lifeless> yes
[11:15] <lifeless> I know
[11:15] <lifeless> I figure that means theres a location registered on chinstrap
[11:16] <ddaa> i doubt so, but I'll check
[11:17] <ddaa> so, it's not in my registered archives on chinstrap, I checked
[11:20] <ddaa> and as far as I can tell, this arch-cache entry is the baz equivalent of a ghost revision
[11:20] <lifeless> ddaa: its in /home/warthogs/archives now I hope
[11:22] <ddaa> looks like it's there
[11:22] <ddaa> thanks
[11:22] <lifeless> np
[11:35] <ddaa> in case anybody cares, we now have an import of iptables
[11:35] <ddaa> Enough importd babysitting for today
[11:46] <jordi> ddaa: hm, gtetrinet failed
[11:57] <ddaa> yeah
[11:57] <ddaa> no time to diagnose it today
[11:57] <ddaa> apparently, it tried to retrieve a file that's not present at all in the cvs repo
[11:59] <ddaa> ATM, I'm spending one hour a day herding imports, diagnosing and cataloguing failures
[11:59] <mpt> hi salgado, got some time to talk about pillargotchis?
[11:59] <salgado> hey mpt
[11:59] <salgado> sure
[11:59] <mpt> salgado, are you doing it on the 1-0 branch?
[11:59] <ddaa> jordi: is this for a one-off conversion or for tracking an uncaring upstream?
[11:59] <salgado> mpt, no. should I?
[12:00] <mpt> salgado, some of the fallback icons for things that haven't specified their own icons don't exist on mainline
[12:00] <mpt> For example, /@@/team
[12:00] <salgado> ah, right
[12:00] <mpt> or /@@/team-heading, rather
[12:00] <salgado> I'm using the nyet-something for now
[12:00] <mpt> ok
[12:01] <mpt> well if that's fairly easy to change, that's fine
[12:01] <mpt> but mainline doesn't show pillargotchis anyway
[12:01] <mpt> so it might be difficult to test :-)
[12:01] <mpt> At least, it shows them for people and teams, but not products and projects
[12:02] <salgado> we show the pillargotchis on the +editicon page of products, projects, etc
[12:02] <salgado> and yes, it's fairly easy to change
[12:03] <jordi> ddaa: I just registered the cvs info when I created the product
[12:03] <salgado> mpt, now I have a question about SearchingProjects
[12:03] <mpt> ok
[12:03] <jordi> I don't need it personally
[12:03] <ddaa> okay
[12:03] <ddaa> scratch one for separating vcs imports from registry :(
[12:04] <salgado> mpt, it may not be possible for us to get a batched list before we start using storm, so I was thinking of somehow limit the number of results and show a info message in case we had more results than the limit
[12:04] <salgado> mpt, would that be okay with you?
[12:04] <ddaa> this conflation is just annoying to me and confusing to users
[12:05] <mpt> salgado, I think that would be ok
[12:06] <mpt> though we have problems sometimes with the obvious thing not appearing top in the search results
[12:07] <ddaa> for very large values of sometimes, yes
[12:07] <salgado> mpt, we'll be sorting the results by their fti rank, which means the most relevant ones should show up on top of the list
[12:08] <salgado> doesn't matter if they're products, projects or distros
[12:08] <mpt> salgado, well, for example, https://launchpad.net/products?text=firefox returns Firefox as #3 :-)
[12:09] <salgado> mpt, I'm pretty sure that page sorts the results by displayname or something like that
[12:09] <salgado> maybe not
[12:12] <BugMaN> hi
[12:12] <salgado> mpt, the query used on that page doesn't seem to order the results at all
[12:12] <BugMaN> in a translation page is possible to have  more than 10 string?
[12:20] <salgado> mpt, that problem should not exist on the new page
[12:20] <salgado> mpt, btw, we need to sort out the sizes of icons/emblems. specially for teams
[12:20] <mpt> salgado, that seems to be bug 1022, bug 1501, and bug 42311
[12:20] <Ubugtu> Malone bug 1022 in launchpad "Full-text searches require ranking." [Medium,Confirmed]  http://launchpad.net/bugs/1022
[12:20] <Ubugtu> Malone bug 1501 in launchpad "The full-text search criteria and the search results are often confusingly mismatched." [Medium,Confirmed]  http://launchpad.net/bugs/1501
[12:20] <Ubugtu> Malone bug 42311 in launchpad "Exact matches should come first in searches" [Medium,Confirmed]  http://launchpad.net/bugs/42311
[12:21] <carlos> BjornT: ping
[12:21] <salgado> mpt, the query I have for the new page takes both 1022 and 42311 into account
[12:22] <mpt> great!
[12:23] <mpt> salgado, the sizes are all explained on /+graphics
[12:23] <salgado> ahhh
[12:23] <mpt> pillargotchi is the "heading" size, maximum 96px * 96px
[12:23] <mpt> but advertised as being 64px * 64px
[12:23] <carlos> I wonder which timezone is WST
[12:23] <carlos> and why are we using it for pagetests
[12:24] <mpt> salgado, and mugshots 170 * 170
[12:24] <salgado> great
[12:25] <mpt> carlos, might be Western Standard Time, if it was code written by jamesh :-)
[12:25] <salgado> that /is not/ mentioned anywhere in the spec
[12:26] <mpt> oh dear, the spec has no Design section
[12:26] <mpt> Shall I fill it in?
[12:26] <mpt> including the sizes?
[12:27] <carlos> mpt: the problem is that zope.app.datetimeutils doesn't accept it as a valid timezone
[12:28] <salgado> mpt, that'd be nice. from there I can update the schema changes to match what is needed
[12:29] <BjornT> carlos: pong
[12:29] <carlos> BjornT: have you read the problem I'm having with zope.app.datetimeutils and the default timezone used for pagetests?
[12:30] <SteveA> mpt: conferences have logos
[12:30] <mpt> true enough
[12:31] <BjornT> carlos: can you paste the test and the traceback?
[12:32] <carlos> well, problem is in new browser code I wrote, I don't think the test itself will help you a lot. Let me show the value I get and the traceback...
[12:35] <carlos> BjornT: https://devpad.canonical.com/~andrew/paste/fileTF4hAu.html
[12:36] <jamesh> carlos: if this date isn't for display to the user, then don't use fmt:datetime
[12:37] <carlos> jamesh: no, it's not for display to the user
[12:37] <carlos> jamesh: will that leave it as UTC?
[12:37] <jamesh> carlos: datetime.isoformat() should provide a form that can be parsed reliably
[12:37] <carlos> ok
[12:38] <jamesh> or "value/isoformat" in TALES syntax
[12:38] <carlos> ok
[12:38] <carlos> thanks for the tip
[12:39] <jamesh> carlos: in fact, you could get rid of the need to parse dates by changing your code to format the current timestamp with isoformat() and compare it against lock_timestamp
[12:40] <mpt> salgado, https://launchpad.canonical.com/PillarGotchis?action=diff
[12:42] <salgado> mpt, no small emblems anymore? the schema changes on the spec has them
[12:42] <mpt> and thanks for that discussion :-)
[12:43] <mpt> hmmm.
[12:43] <mpt> Did sabdfl write the schema changes?
[12:43] <salgado> I think so
[12:43] <mpt> Maybe we are using them after all, then
[12:43] <mpt> though I don't know in what
[12:44] <jamesh> carlos: as you found out, the short time zone abbreviations like "WST" are not unique, so shouldn't really be used on their own when parsing dates
[12:44] <carlos> jamesh: not really, I need to know whether a date is newer or older 
[12:44] <jamesh> for Australia, WST now means something different to WST next week
[12:44] <salgado> mpt, okay, then I have one last question
[12:45] <mpt> salgado, we could keep them around in the database until you find out from sabdfl what they should be used for
[12:45] <jamesh> now, WST == Western Standard Time.  Next week, WST == Western Summer Time
[12:45] <mpt> heh
[12:45] <sabdfl> hi
[12:46] <sabdfl> the emblems are for listings
[12:46] <carlos> jamesh: how's that? we use CEST and CET depending on whether is summer time or not
[12:46] <sabdfl> so, if you do a search, and there are a bunch of products and projects, then those that have emblems would use that rather than product_mini
[12:46] <carlos> I thought most people do that...
[12:46] <jamesh> carlos: with our system, you don't need to remember which abbreviation to use
[12:46] <mpt> sabdfl, the emblems are 16*16, and iirc the reason we have -mini icons at all was because 16*16 was too large for listings
[12:46] <sabdfl> right
[12:46] <carlos> jamesh: anyway, a UTC date/time is enough to do what I'm implementing right now
[12:47] <mpt> or is it ok for team listings to be a little taller?
[12:47] <sabdfl> mpt: feel free to tweak, or drop the pillar emblem idea, as you see fit
[12:48] <salgado> mpt, what I have implemented here allows me to easily provide a page for products/projects/distros/etc in which they can add/change/remove an image of any size, with the size being specified in each different page, together with some labels. this makes it very easy to extend the existing images we have associated with each pillar, but it has the disadvantage of having a separate page for each of these images
[12:48] <mpt> yeah
[12:49] <salgado> I thought about combining two images in a single page, but that wouldn't scale nicely for more than 2
[12:49] <mpt> and if you combine them you run into the problem I had when trying to fix bug 2421 and bug 52765
[12:49] <Ubugtu> Malone bug 2421 in launchpad "Hackergotchi field should be on Edit Details page" [Medium,Confirmed]  http://launchpad.net/bugs/2421
[12:49] <Ubugtu> Malone bug 52765 in launchpad "No way to remove hackergotchi" [Low,In progress]  http://launchpad.net/bugs/52765
[12:50] <salgado> I actually had a way that would fix both
[12:50] <mpt> because if you edit anything else, it assumes you want to delete the image you didn't specify this time
[12:50] <mpt> unless you use a custom form
[12:50] <salgado> yeah, using a custom form, with radio buttons
[12:50] <mpt> cool
[12:51] <salgado> but as I said, that doesn't scale well if we have more than two images
[12:52] <mpt> It could be a <select> instead
[12:52] <mpt> if the issue is compactness
[12:52] <mpt> Emblem: [Leave unchanged :^] 
[12:52] <mpt> Icon: [Remove :^] 
[12:52] <mpt> Mugshot: [Add     :^]           ( Choose File... )
[12:53] <mpt> hmmm
[12:53] <salgado> that looks good to me
[12:54] <salgado> any objections?
[12:54] <mpt> sabdfl, I think inviting people to upload an image in two sizes is ok, but three sizes might be overdoing it
[12:54] <salgado> can't we resize the mugshot for them?
[12:55] <mpt> Last year when we were brainstorming PersonNamePresentation etc I was also imagining that, for example, in a list of subscribers "important" people would have customized icons
[12:55] <mpt> or tweaked generic icons, rather
[12:55] <salgado> they could provide only the mugshot. we save it in its original size and also save a resized copy of it as the icon
[12:55] <mpt> e.g. a helmet for a driver of relevant software
[12:56] <mpt> A beret for RosettaExperts and MaloneExperts (remember those?)
[12:56] <mpt> salgado, that would work
[12:57] <mpt> if the resizing code is decent :-)
[12:59] <salgado> gtk gives us different options when resizing. I'm checking if the Image module has something similar
[12:59] <danilos> salgado: PIL has decent support for resizing, afair
[12:59] <mpt> SteveA, pushed 2006-08-ui with latest icons and other page updates
[12:59] <salgado> yeah, it has a filter argument
[01:01] <salgado> so, what's the decision, mpt? allow users to upload a large (200x200 px) one and a small one (16x16 or whatever) and also save an intermediate size one, obtained by resizing the large one?
[01:02] <mpt> That seems like a good plan for now, salgado 
[01:03] <mpt> Once all the templates are revised for 1.0 we can reconsider whether we're using emblem for anything useful
[01:03] <mpt> and thanks very much for fixing bug 2421 :-D
[01:03] <Ubugtu> Malone bug 2421 in launchpad "Hackergotchi field should be on Edit Details page" [Medium,Confirmed]  http://launchpad.net/bugs/2421
[01:03] <salgado> I haven't fixed that yet
[01:13] <salgado> stub, around?
[01:13] <stub> salgado: yes
[01:17] <kiko-zzz> hello
[01:19] <salgado> stub, I wanted a constraint to ensure that people can't have emblems, but I couldn't come up with one that works
[01:19] <SteveA> if people had emblems, we could do fanclubs
[01:19] <SteveA> so, if you're in the RMS fanclub, you get his emblem, etc.
[01:19] <kiko> like orkut?
[01:22] <stub> CHECK (emblem IS NULL OR teamowner IS NOT NULL)
[01:22] <stub> salgado: ^^^
[01:23] <niemeyer> Heya
[01:24] <salgado> thanks stub
[01:24] <niemeyer> stub: Have you had a chance to roll your eyes on the packaging-related issue I mailed you?
[01:25] <Keybuk> SteveA: aren't emblems going away with the new UI?
[01:25] <kiko> niemeyer, I looked at it and may have an opinion
[01:25] <kiko> Keybuk, SteveA: I am -1 on emblems going away fwiw
[01:25] <stub> niemeyer: I see no email
[01:25] <salgado> stub, btw, I have a db patch on my pillar-gotchis branch which I added to your queue
[01:26] <salgado> kiko, Keybuk, no, they're not going away
[01:26] <salgado> the plan is to use them on listings, I think
[01:26] <niemeyer> stub: Hmm.. weird
[01:26] <niemeyer> stub: You were CC'd
[01:26] <stub> Subject?
[01:26] <niemeyer> stub: I've just resent it
[01:27] <kiko> I have a great resent too
[01:27] <stub> Got it
[01:27] <niemeyer> kiko: Woot!
[01:27] <niemeyer> kiko: Tell me
[01:27] <kiko> niemeyer, not yet. :)
[01:28] <niemeyer> Pleeeease :)
[01:31] <stub> niemeyer: The timing differences may be just because solution 1 has a foreign key constraint and solution 2 doesn't. You should drop the foreign key constraint for a more realistic comparison.
[01:32] <niemeyer> stub: Both of them have no foreign key constraints in the test
[01:32] <sabdfl> mpt: then you would also need to give an indication of the context in which they are important
[01:32] <sabdfl> so, for example, the driver of ubuntu may not be a driver in x.org
[01:32] <sabdfl> i don't think we should do that
[01:32] <sabdfl> the pillar-gotchis were really for listings of specs, say, where the specs might be from multiple products
[01:33] <sabdfl> specs in a sprint
[01:33] <sabdfl> or on listings of bugs across multiple products/distros
[01:33] <sabdfl> to indicate which product/distro the report is in
[01:33] <niemeyer> stub: It is true, however, that option 1 has indexes while 2 doesn't, but that's intrinsic to how they store and search for information.
[01:33] <stub> niemeyer: GIST or the new GIN indexes in 8.2 might allow you to query solution 2 efficiently.
[01:34] <niemeyer> stub: Never heard of these.. will look for info
[01:34] <niemeyer> stub: Hmm, actually I've herad of GIST
[01:34] <niemeyer> stub: But not of GIN
[01:34] <stub> GIST is used as the basis of tsearch2. GIN is a replacement for GIST.
[01:34] <stub> Well... different. Slower inserts, faster queries.
[01:35] <niemeyer> stub: Would a 3rd solution exists?
[01:35] <stub> I'm not 100% sure they support the operations you need though for arrays.
[01:35] <niemeyer> exist
[01:36] <niemeyer> stub: I was a bit concerned with iteration over the list of packages, since it'd mean sending a blob of 180k over the wire
[01:36] <niemeyer> stub: But I found a way to iterate over them individually in the server side
[01:37] <niemeyer> So this shouldn't be an issue
[01:37] <stub> Solution 1 is the only sane structure that lets you say 'which computers have package foo'. Do you need to do that ever?
[01:38] <niemeyer> "
[01:38] <niemeyer> Gin comes with built-in support for one-dimensional arrays (eg. integer[] , 
[01:38] <niemeyer> text[] ), but no support for NULL elements.  The following operations are
[01:38] <niemeyer> available:
[01:38] <niemeyer> "
[01:38] <stub> I can't really think of alternative representations.
[01:39] <stub> That means you are targetting 8.2 unless GIST lets you do it too
[01:39] <niemeyer> stub: We may have to do it from batch processing
[01:39] <stub> (Which should be fine for your work)
[01:40] <niemeyer> stub: But the timing there doesn't look awful
[01:40] <niemeyer> stub: Right
[01:40] <stub> So the GIN or maybe GIST indexes look like they will make the array approach just as usable as solution 1
[01:40] <niemeyer> stub: Well, we can still target 8.1, since it's gonna be a while before this becomes a major issue.. and when it does, "CREATE INDEX" gives instantaneous bonus points.
[01:41] <stub> Just a little more complex as you might need stored procedures to filter the arrays rather than sending the entire blobs to the app
[01:41] <niemeyer> stub: Right
[01:41] <niemeyer> stub: As far as the logic is correctly hidden behind an API, shouldn't be an issue as well
[01:43] <niemeyer> stub: Thanks for helping
[02:22] <mni> hi
[02:39] <mni> hi all
[02:57] <kiko> hey mni 
[02:57] <kiko> what's cooking?
[03:01] <mni> men nothing
[03:02] <mni> just want to disscuss on ubuntu
[03:23] <kiko> flacoste!
[03:23] <flacoste> hi kiko!
[03:23] <kiko> how's it going
[03:24] <kiko> ddaa, what is the frequency in which vcs imports are updated?
[03:25] <ddaa> between 6 and 24 hours, depends... but the actual published branches are only published daily.
[03:25] <flacoste> good even I I recevied too much spam for junk stock this morning
[03:25] <ddaa> So, if you are really unlucky, the import may be up to 48h out of date.
[03:26] <kiko> ddaa, depends on what?
[03:26] <ddaa> on the value of productseries.syncinterval
[03:29] <kiko> mpt, ping?
[03:29] <kiko> ddaa, how is that set?
[03:29] <ddaa> kiko: https://launchpad.net/products/launchpad-bazaar/+bug/60418
[03:29] <Ubugtu> Malone bug 60418 in launchpad-bazaar "configurable default syncinterval for vcs imports" [High,Confirmed]  
[03:29] <kiko> okay but.. what's the default?
[03:30] <ddaa> kiko: at some point I set the syncinterval for svn import to 6 hours, and 12h for cvs imports. But all new imports still get 1day syncinterval.
[03:30] <kiko> okay! that sounds fine -- I was just curious.
[03:30] <ddaa> But anyway, that's all pretty moot since anyway the actual publishing happens daily.
[03:30] <kiko> ddaa, if the interval was much lower would these boxes die?
[03:31] <ddaa> The svn interval can be lowered pretty much arbitrarily. The limiting factor ATM would be the cost to "bzr branch" the import branch every time (that can be optimised out), and how much we find it acceptable to spam community servers.
[03:32] <ddaa> The cvs interval should not be lowered too much as it increases the probability of capturing an in-progress cvs commit.
[03:32] <ddaa> note, in-progress cvs commits _are_ a practical problem for imports, remember? That's why the cscvs catalog is precious.
[03:33] <ddaa> So, 6h and 12h are just about the best reasonable values in the current setup.
[03:34] <ddaa> given some optimisations, we could lower svn sync interval to just about anything (say, every 10 mins).
[03:35] <kiko> ddaa, the interval doesn't really mean too much for race conditions, though I guess doing it more frequently increases the risk of a race proportionally.
[03:35] <ddaa> that's correct
[03:36] <ddaa> the rate of races is proportional to the frequency as long as the interval between sync is >> than the commit time.
[03:36] <ddaa> Note that in some repos, the commit time can be pretty large.
[03:36] <kiko> yeah, for mozilla for instance.
[03:37] <ddaa> Then the rate of races goes through the roof.
[03:38] <ddaa> But again, that's pretty much moot until we have Branch.pull_now, and corresponding support in importd and the branch puller, so importd can request immediate mirroring of a branch.
[03:42] <kiko> right
[04:05] <Ubugtu> New bug: #73756 in rosetta "Keychain requries translations" [Undecided,Unconfirmed]  http://launchpad.net/bugs/73756
[04:39] <ddaa> kiko-fud: how do you like this? https://launchpad.canonical.com/DeleteBranchesBasic
[04:45] <Ubugtu> New bug: #73761 in soyuz "Missing support for any-i386 and linux-i386" [Undecided,Unconfirmed]  http://launchpad.net/bugs/73761
[04:48] <Keybuk> ^ please mark that bug as Critical
[05:02] <carlos> kiko-fud, SteveA: I have bug #68014 ready to review
[05:02] <Ubugtu> Malone bug 68014 in rosetta "Rosetta reverts translation fixes to old faulty values" [Critical,In progress]  http://launchpad.net/bugs/68014
[05:02] <carlos> is there any reviewer with time to do an urgent review?
[05:04] <salgado> carlos, how big is it?
[05:05] <carlos> salgado: the diff has 1629 lines but code changes are around 200-300 lines the rest are test modifications/additions
[05:06] <carlos> https://devpad.canonical.com/~andrew/paste/fileqYPMN8.html
[05:06] <salgado> carlos, do you have a bug for it? are you going to request a cherry pick?
[05:06] <carlos> salgado: the usual pending review page is not yet updated
[05:06] <carlos> salgado: I do
[05:06] <carlos> it's third step to fix that bug
[05:06] <carlos> #68014
[05:08] <salgado> third step or third try?
[05:08] <carlos> hmmm, in fact, I forgot to enable again the .po upload form. I disabled it as first step to fix this bug, but it's just a matter of remove some code so it shouldn't affect your review
[05:08] <carlos> salgado: third step
[05:09] <carlos> salgado: we did it in three steps: 1. disable upload form, 2. tag .po files and store more metadata in our database
[05:10] <carlos> and now, we use .po files tagging and that new metadata and allow our users to see conflicts while editing the same translation in parallel
[05:13] <salgado> carlos, so, the conflict notification will be done when somebody uploads a new po file?
[05:13] <carlos> salgado: so, will you be able to review it? I have an email ready to send to launchpad mailing list asking for a reviewer
[05:14] <salgado> I can review, but first I need to know more about this fix
[05:15] <salgado> carlos, does your branch also include a fix for bug 73509?
[05:15] <Ubugtu> Malone bug 73509 in rosetta ".po file export doesn't update cached files" [Critical,Confirmed]  http://launchpad.net/bugs/73509
[05:15] <carlos> no
[05:15] <carlos> salgado: that's another thing
[05:15] <salgado> and the conflict notification?
[05:15] <carlos> but this branch will prevent problems caused by that other bug
 carlos, so, the conflict notification will be done when somebody uploads a new po file?
[05:16] <carlos> salgado: look at it as two persons editing the same wiki page
[05:16] <carlos> salgado: either uploading a new po file
[05:16] <carlos> salgado: or translating using the web interface
[05:17] <salgado> when they have a stale page, I guess?
[05:17] <salgado> something like they start translating and then somebody else uploads a new pofile before they submit their translations?
[05:17] <carlos> salgado: btw, with your question, I just realise that I forgot to add a test for the notification email
[05:17] <carlos> salgado: yeah, kind of
[05:17] <salgado> kind of?
[05:18] <carlos> the po file would be already uploaded, but still waiting in the import queue
[05:18] <carlos> or someone else loading the same translation form
[05:18] <carlos> and submits it faster than the other user
[05:19] <salgado> wouldn't it be possible for the pofile be approved after I start translating and before I submit?
[05:19] <salgado> s/be/to be/
[05:19] <carlos> yes, that's possible, it's not usual, but it's possible
[05:20] <carlos> our main focus was to fix when someone translates something in the UI and then, we get a .po upload with old information
[05:20] <carlos> with our current code, that .po file upload reverts the changes done using the web UI
[05:20] <salgado> I'm trying to think of all possible scenarios that would cause translations to be reverted, to make sure we're really safe this time
[05:21] <carlos> I think we covered all those, I had a meeting that would be taken as a preimplementation call  with SteveA and kiko about this issue
[05:22] <carlos> but thanks for check it again, do you need more info about the workflow?
[05:22] <salgado> let's say I start translating tomboy and in the meantime somebody approves the upload of a tomboy pofile which touches the same strings that I translated...
[05:22] <salgado> am I going to get a warning when I submit my translations?
[05:23] <carlos> yes
[05:23] <carlos> your translation will be stored as suggestions
[05:23] <carlos> and you will get an error message asking you to check it again
[05:23] <salgado> ah, cool
[05:25] <carlos> danilos: ping
[05:25] <danilos> carlos: pong
[05:26] <carlos> danilos: the only thing I'm missing for the branch that salgado is going to review is the help text
[05:26] <carlos> danilos: but I really don't know what to put there. any suggestion?
[05:26] <carlos> danilos: the form is the translation form
[05:28] <danilos> carlos: hum, we'd probably want a couple of paragraphs for that page... something like "You can translate by filling the empty fields with appropriate translations. For your convenience, you can also use 'copy' buttons next to each suggestion, which are ideas others have submitted."
[05:28] <carlos> well, at this point, we don't have (yet) copy buttons ;-)
[05:28] <danilos> carlos: and then simply "When you're satisfied with translations on this page, click on 'Save & Continue' to go to the next page."
[05:28] <carlos> but I get the idea
[05:29] <danilos> carlos: right, but you can do something like "You can copy them from other suggestions, which are displayed below each field."
[05:29] <danilos> carlos: I am not sure on the terminology we should use here (but I am sure we should not use terms like 'input box', 'textarea' etc. :)
[05:30] <carlos> yeah, anyway, TranslationReview should land this week too
[05:30] <carlos> danilos: right
[05:30] <carlos> danilos: ok, thanks for the input!
[05:31] <danilos> carlos: great, ff stuff is almost cleaned up as well, hopefully up for review later
[05:31] <carlos> danilos: cool
[05:32] <carlos> danilos: I'm leaving for holidays next week and I want to land and cherrypick bug #68014 and TranslationReview before leaving
[05:32] <Ubugtu> Malone bug 68014 in rosetta "Rosetta reverts translation fixes to old faulty values" [Critical,In progress]  http://launchpad.net/bugs/68014
[05:32] <carlos> danilos: please, tell me anything you need from me before Friday night
[05:32] <danilos> carlos: just an invitation letter to come make use of your flat in Alicante if you'll be away during your holidays :P
[05:34] <carlos> :-P
[05:59] <Anatoly> Hi, I have a problem zope3 on Ubuntu 6.1. TypeError: readline() takes exactly 1 argument (2 given). Python 2.4 installed from packages. Zope 3.3 compiled
[05:59] <salgado> carlos, why did you add lots of "potemplate = POTemplate.get(1)" lines to rosetta-karma.txt?
[05:59] <carlos> salgado: because jamesh told me that I need to fetch again objects after a commit
[06:00] <carlos> salgado: I didn't change all tests, just the ones that I was having problems with
[06:00] <carlos> but, at some point, I guess we should change all tests 
[06:01] <salgado> Anatoly, I think this is a known bug in zope and is fixed in their svn (r70587 and r70593 according to our logs)
[06:01] <salgado> Anatoly, btw, this is not the right channel for these questions. #ubuntu is
[06:02] <Anatoly> salgado, Thank's a lot
[06:03] <salgado> carlos, I see... maybe we should not allow transaction.commit() to be used in tests?
[06:04] <salgado> carlos, please add comments on each of those lines you added, since it's not clear why you need them
[06:09] <carlos> salgado: well, without the commits, there are some cache issues and librarian problems
[06:09] <salgado> carlos, the utilities don't have any persistent attributes, so they can be used across multiple transactions. can you remove the extra getUtility() calls you added?
[06:09] <carlos> salgado: I try to use as less as possible the commits, specially now that I know I should fetch again the SQLObjects...
[06:09] <carlos> salgado: ok, I was not sure
[06:10] <carlos> salgado: will do (add comment + that)
[06:10] <salgado> carlos, right, I was asking you to add comments on the POTemplate.get(1) lines
[06:10] <salgado> I mean, no need to add comments for the commits -- just for the extra POTemplate.get() lines
[06:11] <carlos> don't worry, I got it
[06:12] <ssam> i have just spotted two bugs in launchpad that both link to the same upstream bug. could some magic be put in place to automatically find and mark these as duplicates?
[06:12] <ssam> bug 67970 and bug 58373
[06:12] <Ubugtu> Malone bug 67970 in xorg "Endianness problem with AIGLX(BSOD on ppc)" [Unknown,Fix released]  http://launchpad.net/bugs/67970
[06:12] <Ubugtu> Malone bug 58373 in xorg-server "Blue compiz for PowerPC" [Unknown,In progress]  http://launchpad.net/bugs/58373
[06:13] <ssam> i have manually marked them as dupes
[06:14] <salgado> ssam, I'm not sure we should mark them as dupes when we find they're linked to the same upstream bug, but we definitely could tell the user adding the link that there's another bug already linked to that upstream bug
[06:15] <salgado> and maybe even show that in the bug's main page
[06:15] <ssam> salgado, yeah that could work
[06:15] <salgado> ssam, can you file a bug for that?
[06:15] <ssam> yep, against launchpad right?
[06:15] <salgado> malone, actually. launchpad.net/products/malone
[06:16] <ssam> thanks
[06:17] <salgado> you're welcome
[06:20] <carlos> salgado: I need to go out, but will be back in around 30 minutes. Btw, thanks for handling this review!
[06:20] <salgado> np
[06:20] <matsubara> ssam: bug 29206
[06:20] <Ubugtu> Malone bug 29206 in malone "indicate when another bug already uses the same watch" [Medium,Confirmed]  http://launchpad.net/bugs/29206
[06:21] <ssam> matsubara, thanks
[06:22] <ssam> would it be possible for someone to make a list of all the existing bugs with the same watch, so that someone could look through them all?
[06:25] <salgado> ssam, you mean as a one-off or to have a page listing that?
[06:27] <ssam> salgado, maybe just a one-off to see if its worth the effort
[06:30] <salgado> ssam, that's trivial. let me see if I can get it
[06:31] <ssam> salgado, thanks
[06:43] <cprov> BjornT: ping
[07:22] <matsubara> shouldn't https://launchpad.net/products/launchpad-integration/ use malone as its official bug tracker?
[07:24] <salgado> ssam, it was not as easy as I thought, but here it is: http://rafb.net/paste/results/7jP84A64.html
[07:30] <jordi> danilo[out] : ping?
[07:31] <ssam> salgado, thanks
[07:31] <ssam> salgado, thats quite a few
[07:31] <jordi> carlos: rosetta talk in 30 mins. If you're around, great :)
[07:32] <salgado> ssam, indeed, but there might be some that are already marked as dupes. it's not easy for me to exclude them. :/
[07:34] <xhaker> I have a question about team memberships: when a team is a member of another team, it works just like subclasses right?
[07:35] <ssam> salgado, i'll have a look through and see if any are not already duped
[07:47] <ssam> salgado, there are a few in there that are not marked as dupes yet
[07:47] <salgado> ssam, hmmm. and are they really dupes?
[07:51] <ssam> salgado, i think so like, bug 54602 and bug 3512
[07:51] <Ubugtu> Malone bug 54602 in gnome-terminal "quote filename on dnd" [Unknown,Fix released]  http://launchpad.net/bugs/54602
[07:51] <Ubugtu> Malone bug 3512 in gnome-terminal "escape spaces in paths of files dragged from Nautilus" [Low,Rejected]  http://launchpad.net/bugs/3512
[07:52] <ssam> salgado, though one was rejected and the other fixed
[07:52] <salgado> one is rejected and the other is fix released
[07:52] <salgado> heh
[07:53] <ssam> salgado, bug 63796 and bug 68291, i am not completely sure so i commented on them
[07:53] <Ubugtu> Malone bug 63796 in xorg-server "X doesn't launch with voodoo 5 5500" [Unknown,In progress]  http://launchpad.net/bugs/63796
[07:53] <Ubugtu> Malone bug 68291 in xorg-server "GDM Restarts after Desktop loads while using tdfx driver" [Unknown,In progress]  http://launchpad.net/bugs/68291
[07:54] <ssam> i guess its pointless to mark dupes for fixed bugs
[07:54] <salgado> ssam, I'm not sure... if there's a regression and they change one of the bug's status to confirmed, for instance
[07:55] <ssam> yeah that makes sense
[07:56] <salgado> ssam, also, you may want to talk to sfllaw about what you're doing. he may have some good advice about what to do in these cases
[07:56] <ssam> ok, thanks
[08:25] <sabdfl> matsubara: yes, launchpad-integration should do
[08:29] <matsubara> sabdfl: already sorted. I asked mvo and he flagged it as an official malone product. thanks anyway. 
[08:44] <xhaker> sabdfl or kiko, care to explain this? Team memberships works just like subclasses right?
[08:44] <ssam> salgado, sfllaw likes the idea of automatically marking these as potential dupes, maybe by adding a comment
[08:44] <xhaker> i'm taking analogies here.
[08:45] <kiko> xhaker, that's a pretty far analogy you're stretching there.
[08:45] <kiko> xhaker, but team membership is recursive, yes
[08:45] <flacoste> xhaker: they are more akin to set
[08:45] <kiko> so you can be a member of team C indirectly by being a member of team B which is itself a member of team C.
[08:45] <flacoste> xhaker: set A contain set B when team B is a member of team A
[08:47] <xhaker> kiko, not the other way around for sure? i just don't want to have people member sof the most general team being members of the other aut
[08:47] <kiko> not the other way around, for sure.
[08:48] <xhaker> thanks kiko, needed to clear that thing up for my LoCo team
[08:48] <kiko> sure thing
[08:59] <sabdfl> xhaker: we should use the term "member" and "participant" to distinguish between direct membership, and membership in a team which is in turn a member
[09:00] <sabdfl> so, the "participants" in a team are ALL the members, and members-of-members, recursively
[09:05] <xhaker> sabdfl, neat. I'm trying to shake things up a bit in the PT LoCo team. We're mostly dedicated to Rosetta translations. I'm trying to gather more "participants", those that would rather like to do other things. 
[09:18] <ddaa> yay, kdepim import
[09:18] <ddaa> only one year late
[09:27] <kiko> matsubara, that bug reported by ssam, bug 29206, is kind of a dupe.
[09:27] <Ubugtu> Malone bug 29206 in malone "indicate when another bug already uses the same watch" [Medium,Confirmed]  http://launchpad.net/bugs/29206
[09:28] <kiko> oh sorry, I am on crack. of course.
[09:30] <kiko> salgado, can you email the query you used to obtain that to launchpad? that's a neat feature.
[09:31] <lumpki> i have a launchpad account, does that give me an ubuntu forums account also?
[09:31] <Ubugtu> New bug: #73813 in malone "fdisk is not a package" [Undecided,Unconfirmed]  http://launchpad.net/bugs/73813
[09:31] <mdke> lumpki: no, not at the moment
[09:32] <salgado> kiko, that's not just the result of a query. the query is quite simple, but it includes two rows for each two bugs that have the same watch, and I couldn't find a way to get rid of the extra ones using SQL
[09:34] <salgado> kiko, https://devpad.canonical.com/~andrew/paste/fileeqmte7.html is the query I used, anyway
[10:54] <mdz> cprov-out: https://launchpad.net/distros/ubuntu/+search?text=cdrkit shows no results, though https://launchpad.net/distros/ubuntu/+source/cdrkit exists
[11:00] <kiko> mdz, hmmm. matsubara is there a bug filed for that? I thought I fixed the last of those
[11:00] <cprov-out> kiko: looks like it's not in the indexes
[11:04] <matsubara> kiko: I found only the ones you fixed. I file a new one with that specific issue and assign to you, ok?
[11:05] <cprov-out> matsubara: 'SELECT * from distributionsourcepackagecache where name = 'cdrkit';' returns no result
[11:05] <kiko> cprov-out, matsubara: could it be that the damned cronscript isn't running?
[11:05] <kiko> I HATE JUPITER
[11:06] <cprov-out> kiko: yes, could be, I don't know where update-pkgcache is running
[11:11] <cprov-out> kiko: for sure it hasn't been running since 24th this mounth, when 'cdrkit' was published the first time in ubuntu, it's definitely not in DistributionSourcePackageCache. 
[11:11] <kiko> cprov-out, can you email stub CC: launchpad? 
[11:11] <kiko> matsubara, I'm not sure this is worth a bug, but perhaps
[11:11] <cprov-out> kiko: sure, will do
[11:12] <matsubara> cprov-out, kiko: I glanced over lp-errors but there's nothing unusual there. 
[11:12] <matsubara> 06:57:25 INFO    No need to rebuild full text index on
[11:12] <matsubara> +distributionsourcepackagecache
[11:12] <matsubara> but that's for the staging db
[11:16] <kiko> yeah
[11:16] <kiko> probably not running on production
[11:16] <matsubara> bug 73836
[11:16] <Ubugtu> Malone bug 73836 in soyuz "Sourcepackage search for 'cdrkit' returns no results but sourcepackage exists." [Undecided,Unconfirmed]  http://launchpad.net/bugs/73836
[11:20] <cprov-out> kiko: matsubara: email sent
[11:20] <kiko> thanks cprov-afk 
[11:20] <cprov-afk> kiko: np, see you later
[11:24] <ddaa> kiko: http://www.fantasticfiction.co.uk/a/isaac-asimov/buy-jupiter.htm
[11:25] <Ubugtu> New bug: #73836 in soyuz "Sourcepackage search for 'cdrkit' returns no results but sourcepackage exists." [Undecided,Confirmed]  http://launchpad.net/bugs/73836
[11:32] <sigurdga> will new translations done to dapper be merged over to edgy and feisty?
[11:32] <sigurdga> i think the answer is no, myself