[12:59] <mpt> Gooooooooooooooood morning Launchpadders!
[01:58] <fasdfasdf> how to get karma with support?
[02:00] <fasdfasdf> i made comments but did not get karma,what does to have happen so that i get karma?
[02:18] <oohlaf> "This Source has been certified and is now unmodifiable."
[02:18] <oohlaf> hehe, ehm, how can I change the CVS repo url for https://launchpad.net/products/yate/devel/+source
[02:19] <oohlaf> upstream changed hosts
[02:53] <jamesh> oohlaf: email launchpad-users
[03:03] <oohlaf> jamesh: oki, do I need to be subscribed, or is it open?
[03:03] <oohlaf> wiki:MailingLists, which wiki? wiki.ubuntu.com?
[03:03] <jamesh> oohlaf: you need to be subscribed.  It is low volume
[03:03] <jamesh> https://lists.ubuntu.com/mailman/listinfo/launchpad-users
[03:32] <oohlaf> hmm, that's weird, both yate.null.ro and voip.null.ro point to the same ip
[03:33] <oohlaf> yet, if I access the cvs repo at voip the last update is a few months ago
[03:33] <oohlaf> if I use yate.null.ro it is recent
[03:34] <jamesh> I don't think CVS had support for name based vhosts
[03:41] <jamesh> lifeless: ping?
[04:01] <Ubugtu> New bug: #59280 in soyuz "queue tool does not display target pocket" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59280
[04:01] <Ubugtu> New bug: #59281 in soyuz "queue tool can't filter on source package" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59281
[04:31] <lifeless> jamesh: pong
[04:33] <jamesh> lifeless: I've been working on the product series branch stuff, and there was a question that came up with ddaa's mini-review.
[04:34] <lifeless> shoot
[04:34] <jamesh> lifeless: in my branch, I have series.import_branch being the attribute that importd uses for all its stuff and series.user_branch being setable by the user.  The series branch is considered to be user_branch if set, falling back to import_branch
[04:36] <jamesh> ddaa mentioned wanting to make sure the series branch is unique, and it isn't trivial to do multi-field unique constraints like this (i.e. saying that user_branch is different to all import_branch values)
[04:36] <lifeless> why should it be unique ?
[04:36] <jamesh> and I was wondering if things might be simpler if series.user_branch _is_ the product series branch
[04:37] <jamesh> and have importd's branch creation look like this:
[04:37] <jamesh> (1) create branch and set series.import_branch to it
[04:37] <jamesh> (2) if series.user_branch is NULL, assign this branch to that attribute too
[04:38] <jamesh> I'm not sure about the uniqueness constraint.  He brought it up because the existing field is set to be unique
[04:38] <lifeless> ok
[04:38] <lifeless> its important that importd only have one reference to each branch
[04:38] <lifeless> and its separately important than users not be able to commit to importd branches
[04:39] <jamesh> I mentioned that in my reply -- I can see the problems importd would face if that constraint broke
[04:39] <lifeless> but I dont think its important that a users branch not be mentioned by two series
[04:39] <jamesh> Okay.
[04:39] <lifeless> its not very useful, but it may be frustrating to have an arbitrary 'you must fix this somewhere else in the system' error
[04:40] <jamesh> what do you think about getting importd to set user_branch if it is NULL?
[04:40] <lifeless> particularly if its not /easy/ to get to the other series that happens to be referencing the branch you are trying to set
[04:40] <lifeless> coming to that
[04:41] <lifeless> so I think its much nicer for users to let them do what they want here - but if we choose to require uniqueness on the series branch, we should make it easy - as a priority IMO - to get to the other branch to change it. A direct link in the error for instance.
[04:41] <lifeless> as for implementing uniqueness by having importd assign to user_branch - sure. But rather than checking for NULL, make it not NULL.
[04:41] <jamesh> (I've done this in my branch, btw -- the error includes a link to the other series)
[04:41] <lifeless> jamesh: nice
[04:42] <lifeless> so make it not NULL, that way we'll never have a bug where its not set and we dont fall back for some reason
[04:42] <jamesh> make what not null?
[04:43] <jamesh> there will be series without Bazaar branches, so user_branch can't be a NOT NULL column
[04:43] <lifeless> mmm
[04:43] <lifeless> I see a potential for bugs:
[04:44] <lifeless> importd_branch is set
[04:44] <jamesh> I was also thinking that getting importd to set user_branch would simplify the UI a bit
[04:44] <lifeless> series_branch is not set
[04:44] <lifeless> but we dont fallback 
[04:44] <lifeless> so it thinks there is no series branch, but there should be
[04:45] <lifeless> thats my only concern
[04:45] <jamesh> so that if the user sets up CVS import details and a branch gets created, they'll see that branch in the product series edit form
[04:45] <jamesh> at the moment they'd see a blank field
[04:46] <lifeless> let me put it another way
[04:46] <lifeless> we're saying that there should be an invariant : if there is an importd branch, there must always be a series branch
[04:46] <lifeless> the series branch -may- be NULL IFF the importd branch is NULL
[04:47] <lifeless> perhaps a constraint on the record, to preserve that invariant is all thats needed to trap bugs
[04:47] <lifeless> grabbing food
[05:31] <mpt> fasdfasdf, you should be getting karma from helping with support, but that might not start happening for another week or two
[05:35] <Ubugtu> New bug: #59291 in soyuz "queue tool seems confused about NEWness" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59291
[05:40] <Ubugtu> New bug: #59292 in soyuz "Confusion between +packages and +distributions pages" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59292
[05:54] <jamesh> lifeless: so, I guess removing the uniqueness constraint for series.user_branch would solve most of these issues
[05:55] <jamesh> lifeless: also, have you had a chance to look at my product-release-finder branch?
[05:55] <lifeless> jamesh: oh, no. I'll do so tonight I think
[05:56] <jamesh> lifeless: I put up a second product-release-finder branch (jamesh/launchpad/bug-58847) that includes a merge of the first.  It might be easier to just review that.
[05:57] <jamesh> (it includes the fix for the UnicodeError we ran into most recently)
[05:57] <lifeless> sweet
[05:57] <lifeless> shall do
[05:58] <jamesh> once that's in, hopefully we'll get a full successful run
[06:11] <lifeless> whats the review url?
[06:11] <lifeless> (i'm in deep hack mode)
[06:11] <jamesh> https://devpad.canonical.com/~jamesh/pending-reviews/jamesh/launchpad/bug-58847/full-diff
[06:14] <lifeless> whats the vocab for ?
[06:18] <jamesh> I added the dbschema vocab in the previous branch, but forgot to include the registration for it.  I guess it'd be worth adding a verifyObject() test to make sure the file info object is being created properly
[06:34] <lifeless> ;)
[06:34] <lifeless> +1 with that
[06:38] <jamesh> thanks.
[08:04] <SteveA> morning
[08:08] <Fujitsu> Afternoon, SteveA.
[08:10] <Ubugtu> New bug: #59301 in blueprint "Don't give me a vague "URL is registered by another specification" error" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59301
[08:12] <SteveA> hi Fujitsu.  morning!
[08:12] <Fujitsu> Not in Austuralia!
[08:13] <Burgundavia> SteveA: is there a spec for "every false url is a search"?
[08:15] <lifeless> Burgundavia: I dont think so
[08:15] <lifeless> Burgundavia: if you write one, please consider http correctness as part of it
[08:16] <Burgundavia> http correctness?
[08:16] <SteveA> it should be a 404
[08:16] <SteveA> unlike what moin does
[08:16] <Burgundavia> right, I love what moin does
[08:16] <Burgundavia> 404 are very very human unfriendly
[08:16] <SteveA> um
[08:16] <SteveA> no
[08:16] <SteveA> it *should* be a 404
[08:16] <SteveA> a 404 page can show search results
[08:17] <SteveA> moin returns a 200 for a page that doesn't exist
[08:17] <Burgundavia> ah
[08:17] <SteveA> meaning that your browser remembers it in the history
[08:17] <Burgundavia> ah
[08:17] <SteveA> and caches cache it
[08:17] <SteveA> etc.
[08:17] <jamesh> while internet explorer often hides 404 pages, it will display them if they are big enough
[08:17] <jamesh> bit of a weird behaviour
[08:17] <SteveA> jamesh: good point.
[08:17] <Burgundavia> ok, we are talking at different levels. I am talking about what the user sees
[08:17] <SteveA> jamesh: there are similar heuristics in parts of outlook, I've found.
[08:18] <SteveA> jamesh: MS seem to have a habit of using flakey heuristics.
[08:18] <SteveA> Burgundavia: so am I
[08:18] <jamesh> Burgundavia: they see whatever the web server sends, usually
[08:18] <SteveA> Burgundavia: if the user sees lots of pages that don't exist in their history, or out of date searches because they're cached
[08:18] <SteveA> then that's a problem
[08:19] <Burgundavia> ok
[08:19] <jamesh> I like the sound of microsummaries in Firefox 2
[08:19] <SteveA> there maybe a a bug on this, but not a spec yet
[08:19] <lifeless> jamesh: theres a documented trigger for IO
[08:19] <lifeless> IE I mean
[08:19] <lifeless> its in the techbase
[08:19] <Burgundavia> anyway, it would be nice if LP was smart when users are dumb
[08:20] <jamesh> if we put the right file on our server, you'd be able to bookmark a bug on launchpad, and the bookmark title gets updated when people edit the bugs title
[08:20] <SteveA> Burgundavia: If you write a spec for this, I will review it, and get it prioritized on the path to implementation
[08:20] <Burgundavia> SteveA: will do
[08:20] <SteveA> jamesh: nice
[08:20] <jamesh> SteveA: http://wiki.mozilla.org/Microsummaries
[08:20] <SteveA> jamesh: do you know if FF2 will be in edgy?
[08:21] <Fujitsu> SteveA, the beta is.
[08:21] <jamesh> SteveA: if Edgy really is 6.10, I doubt it
[08:21] <Burgundavia> b1 is already in edgy
[08:21] <SteveA> great.  I'll install edgy soon, and try it out
[08:21] <Fujitsu> Burgundavia, isn't it b2?
[08:21] <Fujitsu> Hm.
[08:21] <Fujitsu> You're right, it is only b1.
[08:21] <jamesh> Burgundavia: as default, or as an extra package?
[08:21] <Burgundavia> b2 is out but not yet packaged
[08:21] <Burgundavia> default
[08:22] <Burgundavia> it breaks ephy is horrible horrible ways
[08:22] <jamesh> lifeless: seems the condition is that the 404 page be larger than 512 bytes
[08:22] <SteveA> jamesh: why do they need xslt for microsummaries?
[08:23] <SteveA> jamesh: seems overdesigned to me
[08:23] <jamesh> SteveA: it is designed to be able to extract a title from the page
[08:23] <jamesh> SteveA: and to allow writing microsummary generators for other people's websites
[08:23] <SteveA> jamesh: can you ignore the xslt part, and just give them the summary?
[08:24] <jamesh> SteveA: no idea.  The XSLT could be a simple <xslt:value-of select="...."/> though
[08:24] <lifeless> jamesh: I dont recall precisely, but it is well documented by MS.
[08:26] <SteveA> jamesh: for launchpad, it would obviously be much much cheaper to render a u-summary than to render the whole page
[08:26] <mpt> hi SteveA 
[08:26] <jamesh> SteveA: yeah.  It isn't clear the spec allows you to direct them to a different URL to retrieve the summary
[08:28] <SteveA> jamesh: I can see why they've done it this way -- to make it easy for people who just have file-upload ability to sites to write such summaries, and to allow third parties to write summaries.  but...
[08:29] <SteveA> I think that should be a fall-back to having a <link> in the document, to a page that has the microsummary
[08:29] <SteveA> or having FF request the microsummary with a different mime type
[08:30] <SteveA> their spec doesn't give any indication about where to send comments
[08:30] <jamesh> I was just looking for such a link
[08:30] <SteveA> the only one I see is [edit]  ;-)
[08:30] <jamesh> myk@mozilla.org might be appropriate
[08:31] <jamesh> the links to the examples are on his people.mozilla.com page, and has his email at the bottom
[08:31] <jamesh> http://people.mozilla.com/~myk/microsummaries/generators/
[08:32] <carlos> morning
[08:33] <SteveA> jamesh: do you know "myk" ?
[08:33] <mpt> I know his/her name is Myk Melez
[08:34] <SteveA> I'm concerned that this feature is like an RSS reader...
[08:34] <SteveA> except that it's not reading RSS
[08:34] <SteveA> the web server is given no opportunity to just return the minimum necessary
[08:34] <SteveA> so it may encourage webservers to add "number of requests per minute" limitations to all pages
[08:35] <SteveA> where such things were previously common just for dynamically generated RSS
[08:36] <SteveA> http://www.melez.com/mykzilla/
[08:37] <jamesh> It'd also be nice to be able to include multiple microsummary generators in a single file
[08:37] <jamesh> but that doesn't seem possible
[08:40] <SteveA> jamesh: would you be interested in blogging about microsummaries?
[08:40] <jamesh> sure.
[08:40] <SteveA> maybe that's a good way to give public feedback
[08:42] <SteveA> personally, I think it's a cool idea, but I'm concerned about the "encouraging a lot of full page renders on a web app" aspect.  For example, on launchpad, serving up a simple xml view of the data would make the xsl-t more reliable.
[08:42] <SteveA> and we could make it not a resource concern
[08:45] <mpt> SteveA, voice call?
[08:46] <SteveA> mpt: sure, but in 15 mins?
[08:46] <mpt> ok
[08:47] <mpt> I'll see if I can fix another bug in 15 minutes ...
[08:48] <mpt> In the meantime you might want to pull mpt/launchpad/2006-08-ui to see progress
[09:00] <SteveA> mpt: i'll be ready in a few minutes.  ekiga or skype?
[09:02] <mpt> I haven't had a successful call with Ekiga yet, so it depends how much troubleshooting time you have :-)
[09:02] <SteveA> I had a good ekiga call with mpool
[09:03] <SteveA> so, let's give it a go, and if not, use skype
[09:03] <mpt> ok
[09:03] <SteveA> ill wait until I've merged your branch
[09:06] <SteveA> it is half way through merge phase zero
[09:08] <SteveA> and now it is done
[09:08] <SteveA> that phase zero... it's the big one
[09:08] <mpt> Big fat zeroes
[09:10] <SteveA> mpt: I'm running ekiga now
[09:11] <SteveA> let me know when you are, and I will call you
[09:11] <mpt> I am
[09:12] <SteveA> funny... my headset connects and reconnects for each ring
[09:12] <mpt> I clicked "Accept" and immediately got "Remote user has stopped calling"
[09:12] <SteveA> so it goes beep-ring-beep (pause) beep-ring-beep (pause)
[09:13] <SteveA> i'll call again
[09:13] <SteveA> ok, let's try skype
[09:13] <mpt> ok
[09:34] <carlos> jamesh: hi, could you do  a fast review for me?
[09:34] <carlos> jamesh: I forgot to add support for koffice in the branch you reviewed yesterday
[09:34] <carlos> jamesh: https://devpad.canonical.com/~andrew/paste/fileK6LECC.html
[09:37] <jamesh> carlos: looks okay
[09:37] <carlos> ok, thanks
[09:51] <jordi> carlos: the Punjabi stuff is interesting
[09:52] <carlos> jordi: and confusing...
[09:52] <jordi> because we did unify the effort a while back :(
[09:52] <jordi> I'm not happy about splitting it again.
[09:52] <carlos> well, if there is a valid language code (seems to it)
[09:52] <carlos> and both teams are not able to work together...
[09:52] <carlos> we cannot force them to do it
[09:53] <carlos> It's not exactly like es_ES vs es
[09:53] <jordi> I know
[09:53] <carlos> it's quite similar, but not the same
[09:53] <carlos> so...
[09:53] <jordi> I need more info regarding their inability to work together
[09:53] <jordi> there were argentinians claiming they couldn't contribute to es at some point
[09:54] <jordi> I want to make sure they really can't "understand each other"
[09:54] <carlos> sure
[09:54] <carlos> we should try to convince them, but if we reach a point when they say 'no way' we cannot force them to do it
[09:55] <jordi> woa, and kopete is also sweet
[09:55] <jordi> yeah, I agree
[09:55] <jordi> I want more opinions from the area though
[09:56] <carlos> sure
[09:56] <jordi> Debian or GNOME Punjabi teams
[09:56] <carlos> that's also something to talk with that guy, if he really sees that we should split, he should start doing it in GNOME and KDE too
[09:57] <carlos> so we  don't start having problems with upstream
[09:59] <carlos> stub: hi, if you already saw my cherrypick request, I would like to ask you some extra minutes to get another merger request accepted by PQM, it's related to that cherry pick, but I forgot to add that code there, it adds support for KOffice layout
[10:20] <SteveA> danilos, carlos: short meeting in 10 mins?
[10:21] <carlos> SteveA: sure
[10:21] <danilos> SteveA: sure
[10:21] <SteveA> ok, thanks.
[10:25] <jordi> SteveA: should I be there?
[10:25] <danilos> "jordi, the social guy" :P
[10:26] <SteveA> jordi: this is a specific meeting with mpt about some 1.0 ui stuff
[10:26] <SteveA> so you don't need to be involved
[10:27] <SteveA> it's more about the technicalities of searching and when some specs are scheduled
[10:27] <jordi> SteveA: okay
[10:27] <jordi> danilos: dude shuddup
[10:27] <jordi> :)
[10:32] <sivang> morning
[10:32] <SteveA> bug 44
[10:32] <Ubugtu> Malone bug 44 in rosetta "Messages should be searchable." [Wishlist,Confirmed]  http://launchpad.net/bugs/44
[11:07] <stub> carlos: ok
[11:07] <carlos> stub: the merge is done, and the new revision number in your inbox
[11:08] <carlos> stub: thank you
[11:50] <Ubugtu> New bug: #59318 in launchpad ".changes file should be displayed too" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59318
[11:59] <SteveA> malcc: ping
[12:06] <malcc> SteveA: pong
[12:18] <SteveA> malcc: I had an idea about the __eq__ sets thing.  see list
[12:19] <malcc> SteveA: I saw. I don't think it can be that, as that would have fooled my debugging too - I was only able to confirm this was caused by them being different objects because they repr()ed with different ids
[12:22] <SteveA> in that case, my recommendation is to work around it explicitly using obj.id for now, and know that this will be fixed when we upgrade to a better ORM
[12:23] <SteveA> the correct way to implement __eq__ involves taking into account object class, object id and database connection id.
[12:23] <SteveA> but, given the hairiness of sqlobject connection handling, i think implementing it properly will be error-prone
[12:23] <SteveA> so, it's better to use the minimum needed in the code where it is needed
[12:24] <SteveA> maybe implement __eq__ based on object id, and make it raise a warning, or even raise an exception
[12:24] <SteveA> what do you think?
[12:24] <malcc> Hmm, yes
[12:25] <malcc> Making it raise would be bad
[12:25] <malcc> No more SQLObjects in sets at all, even if you can be sure of the comparison issues in your particular case...?
[12:26] <stub> Is there any reason why we can't implement __eq__ according to those criteria?
[12:26] <malcc> I think your ideal comparison is the right one
[12:26] <stub> type comparison, object.id comparison, check the connection cache to ensure both objects are in the current cache
[12:26] <malcc> Yes, stub's question is the right one - if we can do the right thing, we should. If we can't, I'm not sure which of the workarounds is best, possibly just leaving well alone and noting the issues
[12:28] <SteveA> malcc: we shouldn't use sqlobjects in sets if they're not implementing __eq__ properly
[12:29] <SteveA> stub: I don't know about checking theyre in the current cache.  we'd just check they have the same connection
[12:29] <SteveA> but, it is a bug today, right now, to use __eq__ on a database object
[12:29] <SteveA> unless you specifically mean "this very object Ihave here"
[12:29] <SteveA> rather than "the object from the current connection with this class and this id"
[12:30] <malcc> SteveA: If that's how SQLObject is supposed to work, then yes, using __eq__ is a bug.
[12:31] <malcc> SteveA: I thought it was supposed to unique these things, in which case we've exposed a SQLObject uniquing bug here instead
[12:31] <SteveA> what does "to unique these things" mean?
[12:32] <malcc> To ensure that there can only be one object in any one transaction representing any one database row
[12:32] <SteveA> ok, to ensure uniqueness
[12:32] <SteveA> sqlobject doesn't do that
[12:32] <SteveA> it does it *a bit*
[12:32] <SteveA> enough to lull you into a false sense of security
[12:32] <malcc> Cool!
[12:32] <SteveA> but it isn't designed to do it thoroughly
[12:32] <stub> SteveA: If there is a back reference to the connection in the object, sure. But I thought we would have to ask the connection if the object is valid (which means it is in the connections cache) (?)
[12:32] <SteveA> another reason to use a better-designed ORM
[12:33] <SteveA> stub: there is a kinda back-reference, but now we're getting into horrible hairy territory.
[12:33] <SteveA> stub: so maybe your plan of making it only work for the current cache is a good one
[12:33] <SteveA> stub: I still think that this issue is too hairy for sqlobject, and we should make people use obj.id
[12:34] <SteveA> but I think I may be too extreme to be practical there
[12:34] <malcc> Well, given it doesn't work properly, we might not actually have that many places comparing SQLObjects with ==
[12:35] <stub> I don't think we will be worse off if we just implement __eq__ checking for just id and class equality either. If the developer uses objects from previous transactions, then the obj.id == ob2.id comparison won't stop that happening either.
[12:35] <stub> malcc: Wouldn't that fall back to id(obj) == id(obj2), which would often work by accident?
[12:39] <malcc> stub: Yes, I guess it would often work
[12:39] <BjornT> i think comparing SQLObject with == is quite frequent in our code base, so most of the time it seems to work.
[12:41] <malcc> I think ignoring connections and making == mean "represents the same row" is winning at the moment, for me
[12:42] <malcc> Hmm, or forcing everyone to use id. I'm so indecisive today!
[12:42] <malcc> I'm having one of those days when I shouldn't be allowed to develop
[01:07] <SteveA> ok, let's define that two database objects are equal if they are of the same class and have the same obj.id.
[01:07] <SteveA> at least then, we know what we want to happen, rather than have a kinda random situation.
[01:07] <SteveA> who wants to implement this, with tests of __eq__, __ne__ and __hash__ ?
[01:10] <malcc> SteveA: I suggest a further email to the launchpad list before implementation; several people on that thread aren't awake yet
[01:11] <SteveA> let's cover it in the meeting
[01:47] <sabdfl> evening all
[01:47] <jordi> hey mark
[01:50] <stub> Argh... its Thursday!
[01:50] <stub> Meeting in 10 mins I guess
[01:50] <stub> Or do we get to cancel due to a lack of brazilians?
[01:51] <stub> yo
[01:52] <SteveA> stub: there's still the meeting
[01:52] <SteveA> and there will be some brazilians here
[01:54] <jordi> oh it is thursday
[01:55] <Ubug2> New bug: #59330 in rosetta "Request: Mark string as "translation is not necessary"" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59330
[01:57] <bradb> stub: You saw the query timings I sent you? Will removing the distinct speed up the query that much?
[01:58] <bradb> i.e. old one: https://devpad.canonical.com/~andrew/paste/fileQJI0RU.html
[01:58] <stub> bradb: Nothing significant.
[01:58] <bradb> vs. new one: https://devpad.canonical.com/~andrew/paste/fileYuSRa7.html (less the distinct)
[01:58] <bradb> the new one is quite a bit slower
[01:59] <SteveA> ooh
[01:59] <SteveA> it's that time of week again
[01:59] <SteveA> LAUNCHPAD DEVELOPMENT MEETING
[01:59] <SteveA> welcome to today's meeting
[01:59] <SteveA> who's present today?
[01:59] <bradb> M.E.
[01:59] <malcc> me
[01:59] <BjornT> me
[02:00] <SteveA>  * Andrew: on leave (2006-09-07)
[02:00] <SteveA>  * Matsubara: on leave (2006-09-07)
[02:00] <SteveA>  * Kiko: on leave
[02:00] <SteveA>  * James
[02:00] <cprov> me
[02:00] <malcc> bradb: Myalgic Encephalomyelitis?
[02:00] <danilos> me
[02:00] <mpt> me
[02:00] <bradb> sounds painful
[02:01] <SteveA> stub: ?
[02:01] <danilos> carlos: ping
[02:01] <stub> here
[02:01] <SteveA> jordi: ?
[02:01] <ddaa> here
[02:01] <SteveA> hi ddaa 
[02:01] <SteveA> let's go
[02:01] <bradb> no francis?
[02:01] <carlos> me
[02:01] <jordi> hey!
[02:01] <SteveA> == Agenda ==
[02:01] <SteveA> * Roll call * Agenda * Next meeting * Activity reports * Actions from last meeting * Oops report (Matsubara) * Bug report report (mpt) * Production and staging (Stuart) * Launchpad 1.0 status reports * Sysadmin requests
[02:01] <jordi> I0m here
[02:01] <SteveA> ---- * SQL object equality (Steve) * (other items)
[02:01] <SteveA> hmm
[02:01] <SteveA> irssi...
[02:01] <SteveA> ---- * Keep, Bag, Change * Three sentences
[02:02] <SteveA> next meeting -- next thursday as usual?
[02:02] <SteveA> any public holidays etc?
[02:02] <SteveA> mpt: would you set the channel title accordingly please?
[02:02] <SteveA> (I don't know how to drive that in irssi)
[02:02] <SteveA> hey kiko
[02:02] <SteveA> wasn't expecting you today
[02:02] <kiko> I missed you all
[02:02] <jordi> (stevea: /topic <tab>)
[02:02] <danilos> kiko: :)
[02:02] <kiko> and I came to pimp my picture up at www.bigbiker.com.br
[02:03] <kiko> anyway sorry for interrupting
[02:03] <kiko> me
[02:03] <SteveA>  * Activity reports
[02:03] <flacoste> me
[02:03] <SteveA> who's hot an dwho's not
[02:03] <mpt> up to date (finally!)
[02:03] <flacoste> up to date
[02:03] <kiko> not hot but on vacation
[02:03] <stub> up to date
[02:03] <danilos> behind again
[02:03] <carlos> I'm hot!
[02:03] <malcc> I'm up to date
[02:03] <bradb> up to date
[02:03] <BjornT> up to date
[02:03] <danilos> (just that I caught up :()
[02:03] <ddaa> up to date
[02:04] <cprov> not up, will send summary
[02:04] <SteveA>  * Actions from last meeting
[02:05] <SteveA>  * SteveA to update infrastructure specs if /$name is needed for 1.0
[02:05] <SteveA> still to do.  it is needed.
[02:05] <SteveA>  * stub to check that bug 57474 isn't an SQL injection attack vector
[02:05] <Ubug2> Malone bug 57474 in launchpad "Passing a list as the query string in the product search field crashes ftq" [High,Confirmed]  http://launchpad.net/bugs/57474
[02:05] <SteveA>  * SteveA to put up a wiki page for the launchpad project to note disaster scenarios on, and mail the list about it
[02:05] <SteveA> still to do
[02:05] <stub> Didn't do that :-(
[02:05] <SteveA> actions for steve and stub
[02:05] <SteveA>  * Oops report (Matsubara)
[02:06] <kiko> SteveA, he sent us the report through mail, will you proxy?
[02:06] <SteveA> yes
[02:07] <SteveA> the following is from matsubara:
[02:07] <SteveA> here's the OOPS bugs that I think it would be worth mentioning in the meeting:
[02:07] <SteveA> This is a really weird bug that causes repeated calendar subscribers to appear
[02:07] <SteveA> in the portlet because of that it issues insane amounts of SQL queries:
[02:07] <SteveA> https://launchpad.net/products/launchpad-cal/+bug/57762
[02:07] <Ubug2> Malone bug 57762 in launchpad-cal "Calendar subscription portlet shows lots of repeated subscriptions." [Medium,Confirmed]  
[02:07] <SteveA> 
[02:07] <SteveA> we need to remove the calendar from view, as part of 1.0
[02:07] <SteveA> so I propose simply removing that portlet
[02:08] <SteveA> Broken links to sourcepackages in the +packages page:
[02:08] <SteveA> https://launchpad.net/products/soyuz/+bug/50399
[02:08] <Ubug2> Malone bug 50399 in soyuz "Broken links at /people/$person/+packages" [Medium,Confirmed]  
[02:08] <carlos> SteveA: will we lose the calendar?
[02:08] <SteveA> carlos: yes
[02:08] <mpt> hurrah
[02:08] <carlos> ok
[02:08] <SteveA> carlos: we'll bring it back when it works well enough
[02:08] <SteveA> and we have time to do it
[02:08] <carlos> I know we are not using it too much...
[02:08] <carlos> oh I see
[02:08] <carlos> that's fine, then
[02:08] <mpt> Who should be assigned to the removal?
[02:08] <kiko> mpt, wanna do it?
[02:08] <SteveA> mpt: for the portlet alone, you
[02:09] <mpt> ehehe
[02:09] <mpt> ok
[02:09] <SteveA> for the whole lot... how about you again?
[02:09] <mpt> all right
[02:09] <SteveA> thanks!
[02:09] <mpt> Removing code, I can do
[02:09] <SteveA> you are a star
[02:09] <SteveA> we'll leave the guts in
[02:09] <SteveA> but remove it from the UI
[02:09] <SteveA> malcc, cprov: comment on but 50399?
[02:10] <SteveA> 
[02:10] <SteveA> more from matsubara:
[02:10] <SteveA> I'd suggest Bjorn to take this one:
[02:10] <SteveA> https://launchpad.net/products/launchpad/+bug/44919
[02:10] <Ubug2> Malone bug 44919 in launchpad "UnicodeDecodeError while registering a new account." [Medium,Confirmed]  
[02:10] <SteveA> I can take this one if it can wait until monday:
[02:10] <SteveA> http://launchpad.net/bugs/59249
[02:10] <Ubug2> Malone bug 59249 in launchpad-bazaar "Edit branch details form need input validation for non-existent product" [High,Confirmed]  
[02:10] <SteveA> ddaa: comment?
[02:10] <kiko> malcc, cprov: is that the bug where we have depends, etc on unpublished packages?
[02:11] <ddaa> SteveA: to me, it looks like a bug in the form machinery
[02:11] <SteveA> ddaa: can it wait until monday?
[02:11] <malcc> kiko: Don't think so
[02:11] <kiko> oh
[02:11] <kiko> it's in that page I fixed up
[02:11] <malcc> SteveA: On 50399, just haven't proiritized it, haven't looked in any detail
[02:11] <ddaa> SteveA: it's ugly, but I do not see it as critical.
[02:12] <kiko> malcc, ah, I know what it is
[02:12] <kiko> that package name has an epoch
[02:12] <BjornT> i've assigned myself to bug 44919. i'll see if there's a quick fix for it. otherwise it'll probably take a while, since it requires changes (with a proposal) in upstream zope3.
[02:12] <Ubug2> Malone bug 44919 in launchpad "UnicodeDecodeError while registering a new account." [Medium,Confirmed]  http://launchpad.net/bugs/44919
[02:12] <kiko> so canonical_url is busted
[02:12] <kiko> duh
[02:12] <SteveA> bug 59249 can wait until monday, and be done by matsubara, then
[02:12] <Ubug2> Malone bug 59249 in launchpad-bazaar "Edit branch details form need input validation for non-existent product" [High,Confirmed]  http://launchpad.net/bugs/59249
[02:13] <kiko> malcc, I'll do it, no worries
[02:13] <malcc> kiko: Cool, thanks
[02:13] <SteveA> The top time out (+translate page):
[02:13] <SteveA> https://launchpad.net/products/rosetta/+bug/30602
[02:13] <Ubug2> Malone bug 30602 in rosetta "Timeout errors in +translate" [Critical,Confirmed]  
[02:13] <jordi> I'm up to date (again) on activity reports
[02:13] <SteveA> Stuart need to create a cache table so Carlos or Kiko can continue on the code
[02:14] <SteveA> modification to optimize the +translations page:
[02:14] <SteveA> https://launchpad.net/products/rosetta/+bug/2497
[02:14] <Ubug2> Malone bug 2497 in rosetta "/people/*/+translations times out for prolific translators" [Critical,Confirmed]  
[02:14] <danilos> 30602 state still same as last week
[02:14] <kiko> yes!
[02:14] <kiko> yes!
[02:14] <stub> Its on top of my todo
[02:14] <danilos> (i.e. not much work done on it; it's currently conflicting some clean-up work carlos is doing, so we're coordinating that between ourselves)
[02:15] <SteveA> ok
[02:15] <SteveA> that's all from matsubara's oops report
[02:16] <SteveA> BjornT: on 44919, you can make a band-aid fix locally in our zope, seeing as this oops happens from time to time.  it is an option.
[02:16] <SteveA>  * Bug report report (mpt)
[02:16] <mpt> Launchpad has 18 Critical bugs without released fixes. Well done to cprov, malcc, and kiko for committing fixes for several of them recently. The oldest remaining eight are:
[02:16] <mpt>  * Bug #1558 (Export request form should check for uniqueness of entry), Critical, Confirmed, danilos
[02:16] <Ubug2> Malone bug 1558 in rosetta "Export request form should check for uniqueness of entry" [Critical,Confirmed]  http://launchpad.net/bugs/1558
[02:16] <mpt> danilos, this is newly Critical, and it's not clear why from the bug report. Is it really?
[02:16] <mpt>  * Bug #2497 (/people/*/+translations times out for prolific translators), Critical, Confirmed, stub, which SteveA has already mentioned
[02:16] <Ubug2> Malone bug 2497 in rosetta "/people/*/+translations times out for prolific translators" [Critical,Confirmed]  http://launchpad.net/bugs/2497
[02:16] <mpt>  * Bug #30602 (Timeout errors in +translate), Confirmed, Critical, danilos, which SteveA has already mentioned
[02:16] <Ubug2> Malone bug 30602 in rosetta "Timeout errors in +translate" [Critical,Confirmed]  http://launchpad.net/bugs/30602
[02:17] <danilos> mpt: matsubara said that those type of errors should be deemed critical since a little while ago
[02:17] <mpt>  * Bug #31308 (Cannot set branch associated to a product series), Critical, In Progress, jamesh
[02:17] <Ubug2> Malone bug 31308 in launchpad-bazaar "Cannot set branch associated to a product series" [Critical,In progress]  http://launchpad.net/bugs/31308
[02:17] <mpt> jamesh, tell us the good news
[02:17] <kiko> mpt, it's critical because it's an oopser.
[02:17] <mpt> ok
[02:18] <mpt>  * Bug #42760 (Exception NameNotAvailable raised while trying to create a new msgset from submitted translation), Confirmed, Critical, carlos
[02:18] <mpt>  * Bug #44214 (We need to add code to prevent POFiles being in the same path), Confirmed, Critical, carlos
[02:18] <mpt>  * Bug #46982 (Rosetta does not accept correct KDE plural forms when there are more than 2), Confirmed, Critical, carlos
[02:18] <mpt> carlos, will you have time for those three this week? Should danilos get one of them? :-)
[02:18] <ddaa> mpt: jamesh working on 31308, I reviewed his code once. Will review again then it's up to the review team and pqm.
[02:18] <Ubug2> Malone bug 42760 in rosetta "Exception NameNotAvailable raised while trying to create a new msgset from submitted translation." [Critical,Confirmed]  http://launchpad.net/bugs/42760
[02:18] <Ubug2> Malone bug 44214 in rosetta "We need to add code to prevent POFiles being in the same path" [Critical,Confirmed]  http://launchpad.net/bugs/44214
[02:18] <Ubug2> Malone bug 46982 in rosetta "Rosetta does not accept correct KDE plural forms when there are more than 2" [Critical,Confirmed]  http://launchpad.net/bugs/46982
[02:18] <mpt> ddaa, great
[02:18] <mpt> and finally
[02:18] <mpt>  * Bug #48860 ("Also notified" makes difficult to unsubscribe), Critical, In Progress, bradb
[02:18] <Ubug2> Malone bug 48860 in malone ""Also notified" makes difficult to unsubscribe" [Critical,In progress]  http://launchpad.net/bugs/48860
[02:18] <mpt> bradb, do you have a solution sorted out yet?
[02:18] <carlos> mpt: I'm fixing 42760 right now
[02:18] <danilos> kiko: specifically, it's because it's raising IntegrityError, according to matsubara
[02:18] <kiko> right
[02:18] <bradb> mpt: No. Some discussion on list, but no conclusion yet.
[02:18] <carlos> mpt: I will look into 44214 later
[02:19] <kiko> which causes the Retry exceptions to kick in
[02:19] <carlos> mpt: and I need to meet with danilo before starting with 46982
[02:19] <danilos> kiko: right
[02:19] <bradb> mpt: the other big issue related to it is fine-tuning who gets notifications about when a bug is marked a dupe.
[02:19] <kiko> bradb, mpt: we'll sort parts of this out next week, I think.
[02:19] <mpt> bradb, do you think it's urgent enough to just do a temporary workaround (like, don't do subscriptions from dups at all) until there's time to implement a more elegant solution?
[02:19] <kiko> BjornT had a good suggestion
[02:19] <kiko> and we should do it
[02:20] <bradb> mpt: hm, almost
[02:20] <kiko> so let's reconvene next week and get a mini-spec for the mitigation strategy
[02:20] <kiko> sans ignore subs for now
[02:20] <mpt> carlos, ok, I've marked 42760 as In Progress then :-)
[02:20] <carlos> mpt: oh, right, I forgot it...
[02:20] <carlos> O:-)
[02:20] <mpt> And that's all from me, SteveA 
[02:21] <SteveA> thanks mpt
[02:21] <SteveA>  * Production and staging (Stuart)
[02:21] <SteveA> (stub)
[02:22] <stub> Nothing thrilling is happening on staging except that a new database has been setup for language pack exports on Carbon so asuka should become much more reliable.
[02:22] <stub> I'll grant access via sodium after the meeting to carlos and danilos so it isn't blocked on shell accounts on carbon.
[02:22] <stub> Production systems just had some cherry picks requested by carlos rolled out. I would like to push for fortnightly rollouts again, and skip next weeks rollout unless people cry.
[02:22] <SteveA> any urgent things to roll out?
[02:22] <SteveA> good
[02:22] <stub> So no rollout next tuesday, with next scheduled 19th Sep
[02:23] <SteveA> stub: would you add that date to the status page?
[02:23] <stub> will do
[02:23] <SteveA> ta
[02:23] <SteveA> I'm going to jigger around the agenda a bit... don't be alarmed
[02:23] <SteveA>  * Sysadmin requests
[02:23] <flacoste> stub: did you cherrypicked 4016?
[02:23] <SteveA> say "I have one!" if you have one, before the end of the count
[02:23] <SteveA> 6
[02:24] <SteveA> 5
[02:24] <SteveA> 4
[02:24] <SteveA> 3
[02:24] <SteveA> 2
[02:24] <ddaa> have one
[02:24] <stub> flacoste: no?
[02:24] <SteveA> 1
[02:24] <SteveA> ddaa: yes?
[02:24] <carlos_> sorry, my wireless network went down
[02:24] <ddaa> SteveA: same old thing, sudo -u supermirror on vostok
[02:24] <SteveA> carlos_: any sysadmin requests?
[02:24] <flacoste> stub: i sent an email yesterday about that: Cherry pick request: revision 4016
[02:24] <SteveA> ddaa: RT number?
[02:24] <carlos> no
[02:24] <SteveA>  * SQL object equality (Steve)
[02:24] <kiko> SteveA, there's that one for carlos to get shell access on carbon
[02:25] <ddaa> rt 16533
[02:25] <danilos> kiko: that's what stub mentioned in "production and staging" above, no?
[02:25] <kiko> ah right. just that it's an RT.
[02:25] <carlos> kiko: hmm, right, but I don't have the number... I forgot it...
[02:25] <SteveA> after some discussion by email and irc, I think we should implement __eq__, __ne__ and __hash__ for our databvase objects, so that the type and obj.id is taken into account.
[02:25] <SteveA> to be totally correct, we should take the connection id into account too
[02:25] <SteveA> but that is way way flaky in sqlobject
[02:25] <stub> flacoste: Argh.... I didn't highlight the email. 
[02:26] <kiko> SteveA, right, right.
[02:26] <SteveA> so I'd rather have a simple contract we can keep to than have a contract we can't rely on
[02:26] <SteveA> so, that is my proposal.
[02:26] <malcc> SteveA: +1
[02:26] <SteveA> any serious worries?
[02:26] <SteveA> (I realize we don't have spiv or jamesh here, but jamesh was part of the earlier discussion)
[02:27] <SteveA> anyone interested in implementing it, with tests?
[02:27] <kiko> SteveA, +1, but unlikely to be able to implement in the short term.
[02:27] <danilos> if it's not short term, I'd be interested
[02:27] <stub> I'm happy to do it but also can't do it short term
[02:28] <SteveA> ok.  I don't think this is harming anything except one recent soyuz issue by its absence
[02:28] <kiko> SteveA, which I've worked around/
[02:28] <SteveA> so, if it fits into soyuz bugfixing, then great
[02:28] <SteveA> otherwise, Action on me to document what we'll do
[02:28] <SteveA> and we'll do it when convenient
[02:29] <SteveA> thanks for the discussion, particularly stub, kiko, malcc and jamesh
[02:29] <SteveA> (did I miss anyone?)
[02:29] <SteveA>  * Keep, Bag, Change
[02:29] <SteveA> 8
[02:29] <SteveA> 7
[02:29] <SteveA> 6
[02:29] <SteveA> 5
[02:29] <SteveA> 4
[02:29] <SteveA> 3
[02:29] <SteveA> 2
[02:29] <SteveA> 1
[02:29] <SteveA> 0
[02:29] <SteveA> thnku
[02:29] <SteveA>  * Launchpad 1.0 status reports
[02:29] <bradb> Malone 1.0
[02:29] <bradb> [02:29] <bradb> Release management: Quite a few more test failures to fix before landing.
[02:29] <bradb> Keeping bugs concise: Nothing new.
[02:29] <bradb> Guided filebug: Implemented "most common bugs" API with Francis last week.
[02:29] <bradb> Malone docs: Nothing new.
[02:29] <danilos> Rosetta 1.0:
[02:29] <danilos> - opening edgy for translation: DONE!
[02:29] <danilos> - firefox import/export: slow progress (getting up to speed this week again)
[02:29] <danilos> - oo import/export: blocked on firefox
[02:29] <danilos> - translation review: slow progress (due to langpack improvements, automatic approval of less common import layouts)
[02:29] <bradb> Bug tags: Nothing new.
[02:29] <danilos> - essential docs: assigned to danilos, need to discuss with jordi
[02:29] <danilos> - checks not to upload wrong language PO file using "too many changes" check: not started
[02:29] <danilos> - ui fixes: discussed
[02:29] <danilos> - outstanding issues: none
[02:29] <ddaa> supermirror-smart-server: spiv on vacation this week, reported on target. The goal is to get the smart-server up on launchpad for October 8th althought it's not expected to perform very well at that time.
[02:29] <ddaa> importd-bzr-native: ddaa started on the cleanups, on track to be complete for lp-1.0.
[02:29] <ddaa> bzr-roundtrip-svn (postponed): discussion to make bzr-svn acceptable to vcs-imports made good progress. Should yield improved bzr consistency checks for bzr-2.0.
[02:30] <cprov>  = Soyuz-1.0 Report =
[02:30] <cprov>  * PPA: blocked on ArchiveRework.
[02:30] <cprov>  * Archive Rework: blocked on SoyuzTestSystem.
[02:30] <cprov>  * SoyuzTestSystem: Good progress, currently held up by all the bugs we've already
[02:30] <cprov>    found in the new code from the sprint.
[02:30] <cprov>  * Code quality: nothing big since sprint-cleanup.
[02:30] <cprov>  * General Fixing: good progress,
[02:30] <cprov>    #30264 (P-a-s fix, needs test in dogfood),
[02:30] <cprov>    #31609 (Build-notification, need-review),
[02:30] <cprov>    #35965 (Exceptions in process-upload not communicated to uploader, commited),
[02:30] <cprov>    #58144 (Backport is rejected if an older backport is already there, needs review)
[02:30] <cprov>    #58187 (Uploads to frozen should land in unapproved, not be rejected, needs review)
[02:30] <flacoste> Question Tracker 1.0
[02:30] <flacoste> ---------------------------------
[02:30] <flacoste> - SupportTrackerWorkflow: guided request submission in production as of last Thursday. Spec is reviewed and implementation was just started.
[02:30] <flacoste> - SupportTrackerViews: waiting on completion of SupportTrackerWorkflow.
[02:30] <flacoste> - SupportTrackerKarma is implemented and in production as of last Tuesday.
[02:30] <flacoste> - Localization has been dropped as a 1.0 target.  Salgado finished rearranging it into other specs so that we can decide what will be a 1.0 goal and what's not.  Needs input from SteveA.
[02:30] <flacoste> Random Things 1.0
[02:30] <flacoste> -------------------------------
[02:30] <flacoste> - KarmaContext is implemented and in production.
[02:30] <flacoste> - PersonCreationRationale is almost finished. there are some small issues to sort and salgado needs to write a script to guess the creation rationale for existing profiles.
[02:30] <flacoste> - DirectPersonRegistration has a tricky issue blocking its implementation, so it needs discussion.
[02:32] <SteveA> anything else?
[02:32] <SteveA> thanks
[02:32] <flacoste> SteveA: we need to decide what part of infrastructure localization will be part of 1.0
[02:32] <SteveA> catch up with me next week on that please, if that's okay
[02:32] <flacoste> SteveA: no problem
[02:32] <SteveA> ta
[02:32] <SteveA>  * Three sentences
[02:32] <malcc> DONE: Fixed couple of live bugs in rf, lots of work on mawson testing (inc. fixing some undeployed bugs).
[02:32] <malcc> TODO: Finish system test on sprint code, deploy, start on ArchiveRework.
[02:32] <malcc> BLOCKED: No
[02:32] <danilos> DONE: firefox support, bug management and testing, imports, discussions
[02:32] <danilos> TODO: bug 30602, ff integration (in progress), other bug fixing
[02:32] <danilos> BLOCKED: no
[02:32] <Ubug2> Malone bug 30602 in rosetta "Timeout errors in +translate" [Critical,Confirmed]  http://launchpad.net/bugs/30602
[02:33] <flacoste> DONE: Test and fix bugs in tt-search on staging; SupportTrackerWorkflow discussion; got fix to bug 52671 reviewed
[02:33] <flacoste> TODO: SupportTrackerWorkflow implementation
[02:33] <flacoste> BLOCKED: no, (yeah!)
[02:33] <Ubug2> Malone bug 52671 in launchpad-support-tracker "Support contact implementation shortcomings" [High,In progress]  http://launchpad.net/bugs/52671
[02:33] <mpt> DONE: bugfixing, new templates, administrivia
[02:33] <mpt> TODO: RosettaSearch spec, lots of template work
[02:33] <mpt> BLOCKED: no CSS or images from Usman yet
[02:33] <carlos> DONE: TranslationReview, bug #58168, OO.org language packs, POMsgSetViewRestructuration, bug #58556, Language Pack performance improvements, Debian installer updates, Detected and debugged some missing files in Dapper lang packs.
[02:33] <carlos> TODO: bug #42760, POMsgSetViewRestructuration, TranslationReview
[02:33] <carlos> BLOCKED: no
[02:33] <Ubug2> Malone bug 58168 in rosetta "packages with .po files in different directories are not imported automatically" [High,Fix committed]  http://launchpad.net/bugs/58168
[02:33] <Ubug2> Malone bug 58556 in rosetta "New upstream translations of k3b not imported to Rosetta" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/58556
[02:33] <ddaa> DONE: much email discussion, spec and code review. Email catchup. Importd rollout. Started on python-import and remove-gnuarch
[02:33] <ddaa> TODO: More of the same
[02:33] <ddaa> BLOCKED: no
[02:33] <Ubug2> Malone bug 42760 in rosetta "Exception NameNotAvailable raised while trying to create a new msgset from submitted translation." [Critical,In progress]  http://launchpad.net/bugs/42760
[02:33] <BjornT> DONE: code reviews. various discussions. finished product-bugtracker branch.
[02:33] <BjornT> TODO: code reviews. bug fixes. continue work on upstream forwarding workflow.
[02:33] <BjornT> BLOCKED: no
[02:33] <bradb> DONE: Various bits of release management and guided filebug. Upstream status filter fixes. Long weekend.
[02:33] <bradb> TODO: Fixed RM test failures so I can land it. Get guided filebug up for review.
[02:33] <bradb> BLOCKED: No.
[02:33] <jordi> DONE: email, IRC
[02:33] <jordi> TODO: product queue
[02:33] <jordi> BLOCKED: no
[02:33] <stub> TODO: Cache for +translations 
[02:33] <stub> DONE: Caught up from leave, firefighting
[02:33] <stub> BLOCKED: No
[02:33] <bradb> s/TODO: Fixed/TODO: Fix/
[02:33] <SteveA> matsubara: DONE: bug triage, oops report analysis, support tickets gardening, fixed leftovers from # 52880
[02:33] <cprov> DONE: bug fixing, assist SoyuzTestSystem, stub.test_emails discussion, demise a-f discussion
[02:33] <cprov> TODO: help malcc to finish SoyuzTestSystem, get my branches reviewed, check performance of new version of a-f in mawson
[02:33] <cprov> BLOCKED: solution for stub.test_emails (stub),  demise a-f solution (kiko)
[02:33] <SteveA> matsubara: TODO: more of the same plus more code fix
[02:33] <kiko> DONE: launchpad report, soyuz fixes, malone python comment rendering, other small bugfixes
[02:33] <SteveA> jamesh: DONE: code review, product-release-finder fixes (bugs 58332 and 58847), productseries branch support (bug 31308), spec out AutomaticBugBranchLinks
[02:33] <Ubug2> Malone bug 58332 in launchpad "Remove the URL cache from product-release-finder" [Untriaged,Fix committed]  http://launchpad.net/bugs/58332
[02:33] <Ubug2> Malone bug 58847 in launchpad "product-release-finder HTTPWalker " [Untriaged,Fix committed]  http://launchpad.net/bugs/58847
[02:33] <SteveA> matsubara: BLOCKED: no
[02:33] <Ubug2> Malone bug 31308 in launchpad-bazaar "Cannot set branch associated to a product series" [Critical,In progress]  http://launchpad.net/bugs/31308
[02:33] <SteveA> jamesh: TODO: code reviews, AutomaticBugBranchLinks, URL class stuff
[02:33] <SteveA> jamesh: BLOCKED: no
[02:33] <kiko> TODO: help Soyuz and Malone out some more
[02:33] <kiko> BLOCKED: no
[02:34] <SteveA> DONE: management, ui stuff
[02:34] <SteveA> TODO: management, ui stuff
[02:34] <SteveA> BLOCKED: no
[02:35] <SteveA> cprov, stub: can you unblock?
[02:35] <SteveA> kiko: demise a-f solution for cprov ?
[02:35] <stub> SteveA: I'll try and have a look tomorrow
[02:35] <stub> cprov: ^^
[02:35] <kiko> SteveA, read your email <wink>
[02:35] <cprov> stub: thanks 
[02:35] <SteveA> mpt: about "blocked" lines, would it help if there were a standard "BLOCKED: no" for not being blocked?
[02:35] <SteveA> then you could grep it out?
[02:35] <mpt> that would help
[02:36] <SteveA> mpt: note it on the agenda for next week, and we'll ask people to do that
[02:36] <danilos> we've had them today with grep -i ;)
[02:36] <mpt> yeah, case-insensitive no(.*) should work
[02:36] <malcc> Blocked: [Nn] o\.?
[02:36] <SteveA> One more item:
[02:36] <malcc> Hmm, yes, my regex sucks, I'll go back to sleep
[02:37] <danilos> except mpt's, whose has been "BLOCKED: no images..." ;)
[02:37] <bradb>  /[Nn] on?/
[02:37] <mpt> aha, good catch danilos :-)
[02:37] <stub> (?i)no.*
[02:37] <SteveA> from jamesh on the python bugtracker competition
[02:37] <SteveA> Python bug tracker competition status:
[02:37] <SteveA> The Python infrastructure committee plans to make a decision on 1st
[02:37] <SteveA> October, and send a recommendation to python-dev.
[02:37] <SteveA> This week Brett has been doing mini-reviews of the various aspects of
[02:37] <SteveA> the different trackers.  I sent out an email with his thoughts on
[02:37] <SteveA> Launchpad.  If there are specific info people want me to forward to
[02:37] <SteveA> him, please email me.
[02:37] <SteveA> 
[02:37] <kiko> SteveA, I thought of replying to his python-dev email. should I?
[02:38] <SteveA> kiko: sure, just make sure jamesh knows
[02:38] <kiko> SteveA, did you paste in matsubara's 3 sentences?
[02:38] <SteveA> I did
[02:38] <kiko> sonderful
[02:38] <SteveA> okay that's all
[02:38] <SteveA> thanks for coming.
[02:38] <SteveA> MEETING ENDS
[02:39] <kiko> YAWN
[02:39] <stub> flacoste: I'm running tests with 4016 cherry picked now
[02:39] <flacoste> stub: thanks!
[02:39] <kiko> thanks stub, another topcrasher bites the dust
[02:39] <carlos> thanks dudes
[02:39] <carlos> stub: thanks for the cherry picks!
[02:40] <flacoste> stub: btw, I sent you an email asking for a review of my similarity search algorithm when you were away, do you still have it?
[02:40] <jordi> thanks stub. the hordes of KDE translators are now happy :)
[02:40] <kiko> malcc, did you manage to test my patch for bug 59186?
[02:40] <Ubug2> Malone bug 59186 in soyuz "buildd-queue-builder broken with odd SQLObject problems" [Critical,In progress]  http://launchpad.net/bugs/59186
[02:41] <stub> flacoste: I replied earlier today I think.
[02:41] <stub> flacoste: I'm afraid this may now make you the resident text search expert btw ;)
[02:41] <flacoste> stub: lol
[02:42] <flacoste> stub: hold on, i have some more requests coming along the way regarding the fti implementation
[02:42] <kiko> flacoste, ooooh! I have a bug for you to fix then!
[02:42] <malcc> kiko: Yes, it works
[02:43] <kiko> malcc, ah! grand. and kudos to SteveA for pointing out that security proxies were likely to be the issue here.
[02:43] <malcc> kiko: Still getting some crackful results in which builds are queued, although getting better
[02:43] <kiko> malcc, really? architecture issues?
[02:44] <malcc> kiko: Mostly just regressions. First version re-queued every arch-all build done on previous releases, my first attempt to fix that introduced a different issue I've now also fixed
[02:45] <malcc> kiko: The latest version seems to have re-queued about 40 builds which were already done on dapper i386, I'm still looking at why and whether or not they're right
[02:45] <malcc> Well, still looking after lunch
[02:45] <kiko> malcc, okay, cool. send me a patch once you're happy so I can see what was up
[03:02] <carlos> SteveA, kiko: I have a question for you (or anyone that think is able to answer it)
[03:02] <kiko> the answer is 42
[03:02] <carlos> kiko: ;-)
[03:04] <carlos> would be possible that we create a new row/SQLObject as part of the traversal code and then, try to use it from the view got after that traversal but due cache problems we don't see it yet?
[03:04] <carlos> it's all in the same transaction
[03:05] <SteveA> you'd need to flush updates after adding it
[03:05] <SteveA> until we use a better ORM
[03:05] <carlos> so that's possible...
[03:05] <carlos> ok
[03:05] <carlos> kiko: that's the problem with https://launchpad.net/products/rosetta/+bug/57312
[03:05] <Ubug2> Malone bug 57312 in rosetta "Translation form fails with NameNotAvailable exception" [Untriaged,Unconfirmed]  
[03:05] <carlos> SteveA: thanks
[03:06] <kiko> carlos, hmmm. can you explain?
[03:06] <carlos> kiko: we don't have yet a POMsgSet for one of the messages we have in the submitted form
[03:06] <carlos> we create that object
[03:06] <carlos> and in other part of the code, we try to fetch it
[03:07] <carlos> we don't get it, so we create it again
[03:07] <carlos> and then, we get a duplicate
[03:07] <carlos> (the exception name is completely clueless)
[03:08] <carlos> kiko: at least' that's the only explanation I'm able to find
[03:08] <kiko> carlos, object creation isn't cached in SQLObject.
[03:08] <kiko> so that doesn't seem to be the case
[03:08] <kiko> only updates are cached
[03:08] <carlos> hmm
[03:08] <kiko> and flush_database_updates solves that
[03:09] <carlos> kiko, SteveA: Thanks
[03:09] <kiko> carlos, anytime
[03:25] <Ubug2> New bug: #59339 in launchpad "Right clicking on menu bar item behaves unexpectedly and loads a page" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59339
[03:29] <stub> flacoste: 4016 is in production
[03:29] <flacoste> stub: great, thanks a lot!
[03:35] <kiko> thanks stub 
[03:35] <kiko> one more oopser bites the dust
[05:12] <tomveens> hey
[06:18] <carlos> kiko-zzz: sleeping? so early?
[06:57] <carlos> Is there any reviewer with some time to do a fast review ?
[06:57] <carlos> BjornT, SteveA?
[06:58] <BjornT> carlos: how large is it?
[06:59] <carlos> 98 lines included tests
[06:59] <carlos> it's quite short
[06:59] <carlos> https://devpad.canonical.com/~andrew/paste/file4uLlzv.html
[06:59] <ddaa> BjornT: if you are in reviewing mood, it would be nice if you could do https://devpad.canonical.com/~jamesh/pending-reviews/david/launchpad/remove-gnuarch/full-diff
[07:00] <ddaa> it's the last functional change to importd before starting actual removal of arch support. And since I'm already started on the removal in another branch, it would help to get that branch landed soon.
[07:01] <carlos> It's a tentative to fix https://launchpad.net/products/rosetta/+bug/57312 because I'm not able to find a way to reproduce it in our local launchpad, but it fails on staging or production...
[07:01] <Ubug2> Malone bug 57312 in rosetta "Translation form fails with NameNotAvailable exception" [Untriaged,Unconfirmed]  
[07:03] <BjornT> ddaa: we'll see. i am in reviewing mood, but my review queue is quite long at the moment. it's a small diff, so i should get around to it, though.
[07:03] <BjornT> carlos: ok, i'll review it after i've finished the review i'm doing now.
[07:03] <carlos> BjornT: thank you very much
[07:03] <ddaa> The season of huge diffs with more - than + is opening, but it's not yet time to reap the fruits :)
[07:10] <danilos> later
[07:11] <carlos> danilo[out] : see you!
[07:15] <Ubug2> New bug: #59368 in soyuz "Retrying failed builds in later releases" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59368
[08:00] <Ubug2> New bug: #59377 in rosetta "Orignal author names are not keeped in PO files" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59377
[08:00] <BjornT> carlos: i sent the review via email.
[08:00] <carlos> BjornT: ok, thanks
[08:05] <Ubug2> New bug: #59379 in malone "Need to allow a third party to set bug contact for a package" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59379
[08:27] <bradb> BjornT: just wondering, why did you reject and add a task on bug 52330 instead of retargeting?
[08:27] <Ubug2> Malone bug 52330 in malone "Reassign bug to binary package should just work" [Critical,Confirmed]  http://launchpad.net/bugs/52330
[08:29] <bradb> that problem happens often enough that I think we might have to make it more obvious that retargeting is a possibility
[08:29] <BjornT> bradb: i didn't. if you look at the activity log, you can see that both tasks existed before i rejected one of them.
[08:30] <bradb> ah, maybe it was matsubara
[08:33] <BjornT> seems like it was some random user who did it
[08:36] <ddaa> bradb++
[08:37] <ddaa> When I moved bugs from launchpad to launchpad-bazaar, I did not realize I could retarget bugs at first.
[08:37] <ddaa> However, after going to the trouble of creating+deleting bug reports, I looked a bit further and found it.
[08:37] <bradb> BjornT: oh, indeed, johnlynn
[08:38] <ddaa> So, when you look for it, it's easy to find in my experience.
[08:38] <bradb> ddaa: yeah, i'll file a bug on that
[08:39] <ddaa> According to the theory that LoC should be considered a liability and not an asset
[08:40] <ddaa> I am going to noticeably increase the net worth of Launchpad in the next few days :)
[08:53] <BjornT> ddaa: i reviewed your branch. r=me
[08:55] <ddaa> BjornT: got another not-quite-trivial-but-almost patch (for cscvs) going down the pipe. Think you could look at it tonight?
[08:57] <BjornT> ddaa: sorry, no. i need to take a break now, and go to sleep soon.
[08:57] <ddaa> okay. Thanks for the review, the patch is now in pqm
[09:01] <Ubug2> New bug: #59389 in malone "The possibility of retargeting a bug should be made more obvious" [Untriaged,Unconfirmed]  http://launchpad.net/bugs/59389
[10:14] <bradb> mpt: ping
[10:24] <Seveas> @nick Ubugtu
[10:52] <ddaa> @nick Mordor
[10:52] <ddaa> nice, privileges
[10:54] <lifeless> orning
[10:55] <ddaa> hey lifeless
[10:55] <ddaa> putting up a branch I'm sure you'll love to review ;)
[10:57] <ddaa> sftp://devpad.canonical.com/home/warthogs/archives/david/launchpad/remove-gnuarch
[10:57] <ddaa> removes all arch stuff from importd
[10:58] <LarstiQ> oeh!
[11:00] <ddaa> hey LarstiQ, wassup?
[11:01] <LarstiQ> ddaa: I'd vote +1 on that branch ;)
[11:01] <ddaa> LarstiQ: well, it does not affect users in any way
[11:01] <ddaa> it's not introducing any functional change, it's just garbage collection