doesn't have a height [10:27] a height declared [10:27] and it's a float [10:27] stub: could you run that script on staging now, and tell me how long it takes to run? [10:27] jblack: Do you have a time frame for supporting baz archives on the supermirror with branches containing that CONTINUATION marker? (I don't know what version of baz these appeared in) [10:27] mpt: the CSS uses max-height: 20em, IIRC [10:27] daf: Yes, that's a max-height, it's not a height. [10:27] hum [10:28] it seems to work intermittently [10:28] one would think it would default to a minimum height of the natural height of the content [10:28] iz gecko bug [10:28] yes [10:28] if you could come up with a test case and report a bug, that'd be lovely [10:28] daf: you could set an explcit height in em's [10:28] kthxbye [10:29] jamesh: the problem with that is that it looks ugly if you only have 3 items [10:29] jamesh: but it would work in this particular case, yes [10:29] daf: you don't have three items though ... [10:29] since there's always going to be many items [10:29] we do use the multi-selector in the PO export pages [10:29] and they sometimes do have only 3 options [10:29] daf: the div also needs a solid border [10:30] stub, hi, could you execute again the SQL sentences I gave you yesterday to know the po/pot import status? [10:30] jamesh: I'll let mpt be the judge of that ;) [10:30] stub, btw, thanks for the production fix [10:31] daf: the div also needs a solid border [10:32] launchpad_prod=# SELECT COUNT(*) FROM POTemplate WHERE rawimportstatus=2; [10:32] count [10:32] ------- [10:32] 102 [10:32] (1 row) [10:32] launchpad_prod=# SELECT COUNT(*) FROM POFile WHERE rawimportstatus=2; [10:32] mpt: feel free to [trivial] that in then :) [10:32] count [10:32] ------- [10:32] 16590 [10:32] (1 row) [10:32] carlos: If you need that regularly, I can cron it. [10:33] ok, so we are fucked, the poimport script is not importing anything... [10:33] stub, that would be a good idea until breezy is imported completely [10:45] SteveA: Did you see http://lists.ubuntu.com/archives/ubuntu-devel/2005-July/008688.html ? [10:45] (I'd ask him to report a bug, but the problem is that he can't report bugs) [10:45] oh [10:46] he's calling malone "bugzilla" [10:46] i don't understand the report [10:46] hmm [10:46] okay\ [10:46] maybe i do [10:46] registration mail isn't reaching the guy [10:47] that's on production, so that requires looking at production logs to work out if the mail was sent [10:47] we'd need to know if the mail he registered with is the one he is mailing with [10:47] and then stub would need to look at the logs [10:48] stub: have there been any problems with launchpad production sending out mail? [10:50] stub: i've added a few more bits to BrazilTopics for you [10:54] Nothing I can see in the logs re: problems sending email. At least for the last couple of hours. [10:55] stub: are we passing things off to the MTA directly on the appserver, or are we trying to smarthost it out straight away? [10:56] sabdfl: The script took about 25 minutes [10:56] Kinnison: it goes SMTP to localhost [10:57] sabdfl: in the add bounty browser code, it raises Unauthorized if there is no logged in user. is that view protected by a launchpad.AnyPerson permission? if so, the test and raise of Unauthorized can that be replaced with an assert. [10:57] stub: thanks, could we cron that every 12 or 24 hours please? [10:57] sabdfl: Yes. [10:57] stub: Thu Jul 7 05:10:59 CDT 2005 was when the mail http://lists.ubuntu.com/archives/ubuntu-devel/2005-July/008688.html was sent [10:57] SteveA: ok, will clean bount stuff out in due course [10:58] i'm reviewing a small imports clean up from daf, and noticed this near by. it does no particular harm as it is, just not squeaky clean. [10:59] SteveA: quick questions about the zodb [10:59] okay [10:59] - would that be a good way for a cronscript to leave behind cached values for the appservers? [10:59] - do scripts have access to it? [10:59] - is the api to store/retrieve items in the zodb documented anywhere? [11:00] stub: so it's getting rid of the mails ASAP to a proper MTA, that's good [11:00] it is not a good way for a cronscript to leave behind cached values for the appservers [11:00] because we're using the zodb just for browser session information [11:00] and there is one zodb storage per app server process [11:01] SteveA: https://launchpad.ubuntu.com/malone/bugs/1303 [11:01] seems bug mail is lagged [11:01] affinity is used to keep a browser always hitting the same app server [11:01] When are we moving to www.launchpad.net ? [11:01] SteveA: Can't see anything in the logs with that email address [11:02] stub: thanks. any other ideas based on the report http://lists.ubuntu.com/archives/ubuntu-devel/2005-July/008688.html ? [11:02] SteveA: would the best way to cache system-wide value be to create a LaunchpadCache table with a single row? [11:02] sounds like the cleanest way to me [11:02] or name/value paired rows? [11:02] what kind of value? [11:02] a blob? an int? [11:03] SteveA: The usual - user error, browser cache problems. We could try doing the request using that email address and see what happens. [11:03] true, i guess it will just have to have a lot of columns, though i only care about int's at the moment [11:04] maybe cleaner to use a specific table for each type of caching [11:07] Blech. We can do it with multiple columns, but I've seen what that approach looks like after 6 months of extensions... [11:08] Stick with just ints for the time being, and if we need to refactor it later we can. [11:09] sabdfl: I suggested a LaunchpadCache (id serial, name integer, updated timestamp, value integer) table to Daf/Carlos originally [11:09] 'name' from a DBSchema? [11:09] Yup [11:10] stub: I was going to go with that, but sabdfl had the idea of calculating them at startup time in a utility, which means you don't need to have to worry about persistent storage [11:10] the first page load is slow, but it's fine after that [11:11] Which makes detecting slow pages on staging problematic ;) [11:11] true [11:12] I guess we do restart production a fair bit [11:12] which means that users will still get a very slow page fairly often [11:12] cron is better. Any query with response times measured in seconds rather than milliseconds doesn't belong in the web app. [11:14] fair enough [11:20] lifeless: PQM is hung [11:23] mpt: can you invite the universe "bugzilla" reporter to come onto #launchpad and be helped to log in here? [11:23] ok [11:28] done === Keybuk [~scott@syndicate.netsplit.com] has joined #launchpad [11:29] Keybuk: you should know better [11:30] than? [11:30] oh, wait [11:31] mpt: it's not Keybuk, is it? [11:31] huh? [11:31] daf: No, someone called Frans Cool [11:31] daf: what are you waffling on about? [11:32] Email seems pretty screwed on production. I'm not seeing log messages and stuff isn't appearing in the queues. [11:33] stub: do you need an elmo? [11:33] I don't think messages are getting that far. [11:34] oh, so they're getting lost in the launchpad [11:34] Kool, rather [11:35] stub: do you need assistance looking into this, or can i go and get lunch? [11:37] I'll have a look. Messages are going out SMTP just fine, so things are being lost internally. [11:37] ah, that's why dilys is not announcing bugs reliably === lamer1 [~lamer1@lamer1.albalug] has joined #launchpad === lamer1 [~lamer1@lamer1.albalug] has left #launchpad ["Leaving"] === mpt prods PQM [12:08] jamesh: thanks for that review, it looks great. I'll ask celso to do as much of what you asked as he can before we merge today === sabdfl heads to helsinki [12:12] cheers all [12:12] great work, launchpad team, i think it's really coming together nicely [12:12] mpt: please could you play with the thoughts i laid out for the top bar? [12:12] have fun in HEL sab [12:12] see you in SC [12:12] the tabsl should go in nicely where the facets are currently if the facets shunt to the right [12:13] stub: looking fwd to seeing the next production update, it should be looking very slick === sabdfl [~mark@sabdfl.silver.supporter.pdpc] has left #launchpad [] === elmo [~james@82.211.81.249] has joined #launchpad [12:14] hey elmo. [12:15] hi Kinnison [12:15] I need to go out for a bit [12:15] back later [12:35] stub, I'm improving Rosetta's poimport script and I'm adding support to use config files [12:35] stub, which one should I update? default/launchpad.conf ? [12:35] stub, anyone else? [12:36] carlos: staging/launchpad.conf would be good too for when I get those things running on staging. I like to configure the production config myself though. [12:36] ok [12:36] which one is used with tests? [12:37] default/launchpad.conf in the section [12:37] stub, there are also staging[1-4] directories ... [12:37] what about them? [12:38] carlos: I should remove them - they were being used for load balancing testing [12:44] ok === terrex [~terrex@84-122-69-8.onocable.ono.com] has joined #launchpad === palmTree [~chatzilla@213.210.241.111] has joined #launchpad === palmTree [~chatzilla@213.210.241.111] has left #launchpad [] [01:19] SteveA, I have a friend that wants to blog about launchpad's wiki now that it's open [01:19] SteveA, is it ok or should wait for an official announcement? === SteveA_ [~steve@213.226.190.85] has joined #launchpad [01:54] SteveA_, did you see my question about launchpad's wiki? [02:00] anyone know how I can select distinct BugTasks (using BugTask.select()), while ordering on, for example, Bug.id? === cprov [~cprov@200-171-140-32.dsl.telesp.net.br] has joined #launchpad [02:01] if I set distinct=True, it complains that Bug.id is not in the select fields [02:01] carlos: no [02:02] carlos SteveA, I have a friend that wants to blog about launchpad's wiki now that it's open [02:02] carlos SteveA, is it ok or should wait for an official announcement? [02:02] BjornT: that's a good question for andrew, once he gets back next week [02:04] SteveA_: yeah. i was hoping that someone else would know. but it can wait until next week. [02:04] carlos: i would prefer waiting for an official announcement, but if your friend is keen she (or he) should go ahead. the wiki has been opened in a bit of an ad-hoc manner. i'd like to check out with kiko what policies we need on there. [02:04] ok [02:07] BjornT: i recall that spiv did some work in that area before, and i'm kind of surprised you're having a problem with it [02:09] SteveA_, distinct=True fails under some conditions, perhaps this is one of them [02:10] at least I remember that spiv said that === salgado [~salgado@200-171-140-32.dsl.telesp.net.br] has joined #launchpad [02:34] stub, around? [02:34] salgado: Yup [02:35] do you remember the name change widget we discussed? === debonzi [~debonzi@200-171-140-32.dsl.telesp.net.br] has joined #launchpad [02:36] salgado: Yup [02:37] so, I'm stuck in a problem: [02:37] I can see at least two things that I want to validate [02:37] 1) If the name follow the convention (no spaces, starting with a letter, etc) [02:37] 2) if the name is not already used by someone else === Kinnison lunches. ciao [02:38] but as we can specify only one constraint in our schema I can't see how to have validators for these two things [02:39] And there is a reason you can't combine them into a single constraint 1) The name is not already in use and follows conventions ? [02:40] morning [02:41] mpt: not sure if somebody later responded to your comment, but there was mail to the list about the intermittent mail failures *goes to read morning mail* [02:42] stub, then the validator function would have to receive a third argument, which is the current name [02:43] stub, and unless I make this function raise some exceptions, I will not be able to tell, for sure, what was the problem (if the names was already in use or invalid) [02:44] I don't see what the difference between needing two validators and using a single validator. If you validate 'name not in use by someone else' then you either need to pass the current username in or pick it up from context somehow. [02:46] what does a validator get in order to do its job? [02:46] it's validating new input into a field [02:46] so, it should have access to the new input, and to the field [02:46] the field should be a bound field, of course, and so give access to the current value [02:46] stub, sorry, I'm not following you [02:47] SteveA, AFAICT a validator function will get only the input value [02:47] salgado: the issue is not that you need two separate validators. it is that you need to get certain information to a validator. [02:48] salgado: can you point me at a validator, and i'll take a look [02:48] SteveA, launchpad.interfaces.person.valid_person_name() [02:48] The problem you said you had was "we can specify only one constraint in our schema I can't see how to have validators for these two things". The solution to that is you just have a single validator that performs both functions. [02:49] Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial] Fix bugs 1284, 1285 - no links to product releases from product or productseries (patch-2050: morgan.collett@canonical.com) [02:49] but as it is now, I can't perform the second validation in a validator function, because it doesn't get the old value [02:49] salgado: i will look at the validation infrastructure and tell you what you can do [02:50] the validator has access to the context I think, which is the field. You might be able to pull the initial value from the field, or traverse back to the object the field was generated from. [02:50] (Which beats cheating and pulling it from the Launchbag [02:50] which would be wrong [02:51] oh,and i need to make the fascist forbid use of the launchbag in database code [02:51] yep. /me requested that to Andrew. :) [02:52] This isn't database code - it is browser code [02:52] stub, the validators I'm aware of have access only to the input value. nothing else [02:52] urgh.... actually it *is* database code :-( [02:53] i think it's a valid database constraint [02:53] it doesn't depend on the current person [02:53] just the current value of the field [02:54] agreed === SteveA_ does an evil python experiment with constraints [02:56] salgado: Vocabularies get context. If you model your control on a Choice you can use the Choice validation, which simply asks the Vocabulary. [02:57] yeah [02:57] to do this with a normal constraint, you need to jump up a stack frame [02:58] inspect the bound field [02:58] (Note that a vocabulary does not have to represent a finite set. All our existing Vocabulary instances implement IIterableVocabulary. The base vocabulary instances don't require it) [02:58] Yech! [02:58] Bad Steve! [02:58] that, or implement a TextField subclass to support this [02:59] overriding _validate [02:59] that is probably the best way to do it, if the vocabulary stuff doesn't work out [02:59] just implement a PersonNameField [03:00] class PersonNameField(TextField): [03:00] def _validate(self, value): [03:00] That would be cleanest [03:00] TextField._validate(self, value) [03:00] # salgado's validation goes here [03:01] this way it would first call the user-specified validator (the constraint argument) and later my validation? [03:01] stick it in interfaces/person.py, but don't include it in the __all__ [03:01] yes [03:02] and, you get the current state from 'self' [03:02] and the new state from 'value' [03:03] but I'd like to make this widget generic enough so it can be used for things that have a unique name other than a person [03:03] do we have any uses for this yet? [03:04] I can see uses for products/projects/polls at least [03:04] i think you should do this for person now [03:04] and then make it generic once you have it working well for person [03:04] let's actually get it into production for Person first [03:05] or at least, on staging [03:05] looks like a plan. I'll do this [03:05] i am not aware of any specs or open bugs that require it for other things right now [03:05] but i'm sure we can generalize it easily enough [03:07] salgado: You could make it generic by using a Vocabulary to say 'is this name in use', or by adapting self to something that provides a function to say 'is this name in use' [03:07] salgado: So it could be refactored easily once you get the basics working [03:08] and once we get a standard IFooSet base === salgado haven't thought about that [03:08] it may be that "is this name in use" or "get thing with this name" will be a standard api [03:09] in which case, you'd just give it an interface that extends IContentObjectSet === ddaa [~ddaa@marvin.xlii.org] has joined #launchpad [03:10] hi david [03:10] thanks SteveA and stub. :) === ddaa grumbles indistinctly [03:11] various people i know who speak some french were impressed with the example of shining wit you explicated yesterday. [03:11] stub, btw, I need to bother you some more: is there anything missing for the karmacache updater cronscript to be run? [03:11] SteveA_: hu? [03:11] salgado: Someone with a better memory to do the rollouts? === stub goes and switches it on [03:11] the french version of spoonerism taken to the Nth degree [03:12] stub, heh. thanks [03:12] SteveA_: there must be a slight desync... [03:13] "choix dans les dates" and all that [03:13] SteveA_: the contrepteries I remember mentioning yesterday were quite ordinary... [03:13] they caused a snigger or two in lithuania [03:13] but really, i just wanted an example to say "shining wit" in that context. [03:13] s/example/opportunity/ [03:13] :) [03:14] ah! [03:14] maybe you can help me, Steve === Mez [~mez@cpc2-lich4-3-0-cust115.brhm.cable.ntl.com] has joined #LAUNCHPAD [03:14] I'm debugging a test failure [03:14] daf: what's up === Mez [~mez@cpc2-lich4-3-0-cust115.brhm.cable.ntl.com] has left #LAUNCHPAD ["Leaving"] [03:14] you evil bastard === Mez [~mez@cpc2-lich4-3-0-cust115.brhm.cable.ntl.com] has joined #launchpad [03:14] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=salgado] some cleanup of calendar templates (patch-2051: mpt@canonical.com) [03:14] so, there's a page on PublishedPackageSet [03:15] and the view class for this page calls .query() on self.context [03:15] ddaa: it's the standard british response to a mention of spoonerism ;-) [03:15] PublishedPackageSet implements IPublishedPackageSet [03:15] and IPublishedPackageSet defines a query method [03:15] we just call that "l'art de dcaler les sons" [03:15] but the page test fails with a ForbiddenAttribute error === SteveA_ "l'art"s ddaa [03:16] I'm not totally clear on why the salinity of female organs is that important... but nevertheless [03:17] aha [03:17] I think I know what it is [03:17] your colloquialisms go over my head [03:17] daf: it sounds like it should work. so, i guess something is amiss in the configuration. [03:17] "l'art de dcaler les sons", that's self-referential [03:18] SteveA_: the was missing an [03:18] "secaler" is to make dry? [03:18] daf: aha [03:18] salgado: It is very noise. I'm seeing lots of: [03:18] Exception exceptions.AttributeError: "'NoneType' object has no attribute 'cache'" in > ignored [03:18] Exception exceptions.AttributeError: "'NoneType' object has no attribute 'cache'" in > ignored [03:18] SteveA_: thanks for rubber ducking :) [03:18] stub: i have a fix to that coming [03:19] stub: it's part of my "rip parts out of sqlos" [03:19] "dssaler" is to purge from its salt (as in sailing equipment) [03:19] ddaa: you can stop there [03:20] I know you are familiar with the rest of it :) [03:20] I can go with only explaining the non-dirty bit === ddaa go looking for a way to increase shareholder value [03:22] stub: basically, rather than returning None under these circumstances, the connection descriptor will return a dummy transaction that has a dummy cache [03:23] stub, please don't go to sleep. the karmacache updater just triggered a bug [03:24] random thought... [03:25] did you guys considered compressing page (as in eliminating all non-significant whitespace, comments and other cruft) before sending them on the wire? [03:25] salgado: I can purge the KarmaCache table. Probably best. [03:25] stub, please do so then [03:26] Done [03:26] https://launchpad.ubuntu.com/errors/showEntry.html?id=1120828917.60.850024609079 [03:26] I'm sure I have a test for that [03:26] ddaa: why bother? [03:26] ddaa: gzip content transfer encoding will do that [03:27] SteveA_: either CE or TE, there is no CTE. [03:28] i obviously meant "we can GZIP the CONTENT for TRANSFER ENCODING" ;-) === SteveA_ goes all revisionist [03:29] SteveA_: TE is very sparsely supported, for a number of reasons, and CE is seriously problematic for others - browser support has been heinous - such as send a CE gif to a netscape 4.x and crash it === SteveA_ invents CTE as a new internet technology and goes to seek VC money [03:30] i'm pretty sure that firefox <--> apache will gzip stuff [03:31] slashdot does it. i just looked there with firefox [03:31] not opportunisitically it won't [03:31] SteveA_: apache 2 has a replace mod_gzip of its own, the original mod_gzip guys who I did some work with are not very positive about it. The 1.3 mod_gzip will work fine with firefox yes - but you need to consider the impact of non compliant proxies (like the majority of soho ones) :[ [03:31] but you can configure apache to do it (e.g. www.ubuntu.com does) [03:31] and it's doing gzip content encoding of the page [03:32] one of the nastier things about CE is that it fucks with the content - is a gzip CE on a tar.gz meant to give you a tar on disk ? Its semantically unclear. === lifeless stops before he starts on a http-1.1 bugs rant. [03:33] my point is, i think we should turn on gzip encoding for our pages, before we look at stripping whitespace from our pages [03:33] pages are identified by their mime type [03:33] SteveA_: sure. [03:33] lifeless: when are you writing http 2.0 ? [03:33] SteveA_: heh. [03:34] SteveA_: there likely won't be a 2.0. There may be a 1.2 if the current errata gets big enough. [03:34] the extension mechanism in 1.1 does work well - see dav et al. [03:35] there's an "et al"? [03:37] daf: you need a t-shirt saying "Daf, pronounced according to RFC 2518" [03:38] that's a good idea [03:38] it would save a lot of bother === ddaa does not find the string "daf" in rfc2518 [03:39] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=BjornT] DisplayingParagraphsOfText (patch-2052: steve.alexander@canonical.com, mpt@canonical.com) === ddaa does not find the string "pron" for "pronounciation" or some such === ddaa will pronounce "daf" "PROPFIND" [03:41] wow, that took its time === daf wonders what happened to the merges he submitted a few hours ago [03:42] perhaps PQM is just backlogged [03:42] it would be nice to be able to see the PQM queue somehow [03:42] lifeless: where is that code you wanted me to look at about archive locations? [03:43] debonzi, Kinnison 13:11:38 ERROR FMO courtesy of TROUP & TROUT inc. on gadfly (1:1.0.0-8ubuntu3) [03:43] ? [03:43] lifeless: I guess you meant "code I want you to write" [03:43] carlos: that's a gina error [03:43] I know [03:43] but I don't understand the error ;-) [03:44] FMO being the technical term [03:44] for when something or other happens to a source package [03:44] carlos: it means that when tring to insert the binarypackage for gadfly the sourcepackage was not found [03:45] oh, ok [03:49] debonzi: did you see daf's message about broken searches [03:49] ? [03:51] SteveA_: not yet.. [03:51] SteveA_: will take a look now [03:56] 13:13:08 ERROR FMO courtesy of TROUP & TROUT inc. on python-kde3 (3.11.3-1) 13:13:08 [03:56] WARNING Sourcepackage python-kde3 (3.11.3-1) not found for python2.3-kde3 (3.11.3-1) [03:56] debonzi: Do either of these actually deserve to be WARNING or ERROR? We have rather a lot of them... [03:58] stub: does "direct delivery" mean non-transactional ? [03:59] SteveA: It is transactional. It means email is delivered as part of commit rather than queued, which slows down pages that trigger emails. [03:59] stub: Well.. personaly I think it could be WARNING... since it is just telling that a binarypackage could not be imported because gina could not find a binarypackage to it. [03:59] sorry [03:59] stub: a sourcepackage to it [03:59] stub: do we run queued mail delivery on staging in any way at all? [04:00] New Malone bug 1306 filed on product FOAF by Guilherme Salgado: PersonSet.createPerson should produce nicer exceptions than IntegrityError [04:00] https://launchpad.ubuntu.com/malone/bugs/1306 [04:00] debonzi: So can we fix it so it doesn't happen? We need to shrink the output Gina sends - if they are actual problems that we need to fix, they should stay high level. If we just have to put up with them, they should become INFO or DEBUG. [04:02] Or maybe they should stay warnings and we run gina with -qq and only output ERROR messages. [04:04] this sounds like the kind of information that the owner of the buildds will want to know [04:05] stub: I believe it happens in two cases: 1) The sourcepackage is not available on archive; 2) It is on archive, was imported into launchpad, but the binarypackage info has some problem on the SOURCE field and gina can't define the sourcepackage. The first one I think we can do nothing. The second one, we could try to improve the sourcepackage lookup but I don't know precisily how to do that. [04:06] SteveA_: agree.. I think it will be a problem for all the packages that build-depends on this one that could not be imported. [04:06] ok. So it sounds lilke the FMO one becomes an INFO or DEBUG and the WARNING stays the same [04:06] or both waning... [04:06] probably both warnings... [04:10] debonzi: Let me know if you do any work on gina today and I'll give her a run tomorrow against staging. [04:10] stub: trying to customize SinglePopupWidget like: [04:10] field="sourcepackagename" [04:10] class="canonical.widgets.popup.SinglePopupWidget" [04:10] extra="tabindex=1" /> [04:10] i get this exception when viewing the page: [04:10] ... [04:11] * Module zope.app.form, line 91, in __call__ [04:11] instance = self._widget_factory(*args) [04:11] TypeError: __init__() takes exactly 4 arguments (3 given) [04:11] stub: sure.. does she still stoped on the createPerson problem? [04:11] stub: i can customize a TextWidget in the same way just fine though (no errors, autofocus Just Works, etc.) any ideas? [04:11] bradb: i guess the SinglePopupWidget doesn't except extra args [04:11] debonzi: That was the last run I have tried. I got tied up with other stuff today. [04:11] bradb: probably a simple signature change [04:11] SteveA_: its base class has an extra attribute [04:12] SteveA_: the extra arg is actually vocabulary === SteveA_ goes afk for a short while [04:12] bradb: I don't know if the extra attribute is wired up or was ever working. Feel free to fix it ;) [04:12] (i.e. the arg that the __init__ expected but didn't get is vocabulary) [04:12] stub: right.. no problem.. I saw it happens once and was a problem with Upper case letters on emails.. I fixed that.. will check why is happening again. [04:13] debonzi: When did you fix it? Am I running an old version? [04:13] ItemsWidgetBase.__init__'s sig is: [04:13] def __init__(self, field, vocabulary, request) [04:13] i don't understand why it expects both field /and/ vocabulary [04:14] stub: I don't believe so.. IIRC was just before I ask you to merge my branch into your [04:14] stub: would hacking SPW's __init__ to be self, field, request be horrid and wrong? [04:14] ok. [04:15] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=jamesh] testrunner improvements (?) (patch-2053: stuart.bishop@canonical.com) === bradb tries it just for kicks [04:15] you guys better bring warm cloathes when comming to Brazil. this place is getting quite cold. [04:15] bradb: I can't remember what it is supposed to be - I've forgotten how all that stuff works ;) [04:16] s/cloathes/clothes [04:16] And nobody wanted the conference in Bangkok! === bradb wants a conf in Star City [04:20] salgado: how cold? [04:21] bradb, 10 degrees sometimes [04:21] the problem is that we get this inside the office [04:22] stub: I did! [04:31] salgado: buddy, stop crying ! buy some descent closes or work outside, under sun ;) in the common world sense *BR HASN'T COLD WEATHER* [04:33] heh [04:37] so, i'm leaving sunny warm lithuania for 3 weeks to go to sub zero brazil? [04:38] Ahh stuff it. You guys can go to Brazil. I'm going to Bangkok and attending via IRC :) [04:39] we have a swedish that's here for 7 months and he's complaining about the cold [04:40] salgado: hey.. stop telling lies :) [04:40] SteveA: your DisplayingParagraphsOfText landing is great, but it puts <p> and not

in my product page... [04:40] the problem is that we don't have heating stuff in houses and offices (we don't need them in 95% of the days in a year) [04:40] salgado: don't trust those half-swedish people ;) [04:40] morgs: did you use the 'structure' keyword? [04:40] SteveA: no, I'll try that === debonzi -> BRB [04:41] okay, okay. I gave my advice [04:41] page templates automatically escape stuff [04:41] now I'm going to make some tea, so I can hold the cup to get my hands warm [04:41] SteveA: Aah... that works. [04:41] you need to say tal:replace="structure foo/fmt:..." [04:41] cool [04:41] Nice, thanks very much === morgs goes to sprinkle some fmt:text-to-html all over Registry... === debonzi [~debonzi@200-171-140-32.dsl.telesp.net.br] has joined #launchpad [04:44] morgs: one thing to note [04:45] soon it will be linkifying too [04:45] SteveA: hmm? [04:45] SteveA: OK [04:45] so note if there are any places you don't want that [04:45] OK [04:48] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=SteveA] add addLanguage/removeLanguage to IPerson, pagetest for the Rosetta prefs page (patch-2054: daf@canonical.com) [04:48] ooh [04:51] SteveA, do you think I can review daf's menus branch? (I'd like to do it because I was reading about the menus yesterday and this seems like a good oportunity to see some code of it) [04:51] of course, there's the problem that I don't know too much about them [04:51] salgado: Steve already did it this morning :/ [04:52] oh. :-( [04:52] I haven't went through my launchpad-reviews folder yet === daf updates the wiki [04:52] I should have done so === salgado does [04:52] salgado: nothing stopping you from looking at it [04:52] yes, Steve's right [04:53] indeed. I was definitely going to use that as example when doing the menus for foaf [04:54] stub: oi, you forgot to put a status/[@DATE@] on your cronscripts branch [04:54] salgado: the same for soyuz, I guess [04:54] (I've added them now) [04:55] SteveA: Are we supposed to use fmt:url for all urls? [04:55] ta [04:55] I see some portlets have it and most don't... [04:55] morgs: no [04:55] Steve is going to add a feature which is going to remove the necessity for the _parent_url hacks I added [04:55] so that part is not a great example of menuing [04:55] morgs: use it for urls to objects within launchpad [04:56] OK [04:56] Well, that's still *most* URLs in portlets etc... [04:57] stub, what patchlevel are we running in staging? [04:58] salgado: 2047. Should I push the upgrade button? [04:59] stub: is it big and red? [04:59] Of course. [05:00] with a plastic flip-back cover and two keys? [05:00] stub, that (supposed) fix is already there, then. :-( [05:02] stub, that problem you saw with gina (the IntegrityError) was raised recently? (I mean after the upgrade to patch-2047) [05:02] salgado: Oh.... gina is running her own branch because the recent changes arn't in rocketfuel yet. [05:03] Looks like she is only up to 2011 [05:04] hmmm. then if you cherrypick patch-2042 into it that problem should be fixed [05:04] No need to cherry pick - I'll just sync it. [05:04] debonzi, ^^ [05:05] in the last 20 lines is the answer to your email. :) [05:05] salgado, cool.. thanks [05:15] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=salgado] clean up templates with names starting with 'c' (patch-2055: mpt@canonical.com) [05:16] everytime I see dilys name in yellow I think it's my branch that get merged... but no, it's someone else's branch which I reviewed. :-( [05:19] salgado, ;-) [05:20] and again... it's not my turn on pqm. [05:20] come on pqm, I have some things that I need to merge here === bradb heads out to meet up with jbailey. bbl. [05:41] Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial] Switch production to directDelivery temporarily (patch-2056: stuart.bishop@canonical.com) [05:42] well, I have two merges queued and soon a third... [05:42] so you guys do not be too impatient :) === debonzi -> Lunch === morgs thinks we need a beowolf cluster of PQMs... [05:51] morgs: it's entirely unclear whether that is parallelisable... [05:51] Clearly, running the test suite, should be parallelisable. [05:51] But we still need to have only one merge checked at a time. [05:52] morgs: btw, while you are here :) [05:52] Well at least a faster machine... [05:53] not sure that would help that much, in test suites I'm familiar with, a lot of time is spent on i/o and explicit waits (to work around bugs in cvs). [05:54] clearly a faster maching would help for running Keybuks test of the form "test that 2+2 = 4" [05:54] dunno what is the bottleneck for launchpad tests... [05:56] Oh well, let's see how well the LP team scales, especially in deadline season! [05:56] morgs: in series +edit pages where the rcs is svn, the page is initially loaded showing (empty) cvs details even though the dropdown has SVN selected. One has to reslect SVN so the JS will show the svn details. [05:57] Are you aware of that bug, or should I file it? [05:57] That is new to me, please file it. === morgs will have to dust off his JS skills [05:57] Okay. Will do. === ddaa goes for a little break [06:03] Merge to rocketfuel@canonical.com/launchpad--production--1.24: Update production mail config (patch-5: stuart.bishop@canonical.com) [06:17] New Malone bug 1309 filed on product Registry by David Allouche: +source does not show svn details on load, mouse action required [06:17] https://launchpad.ubuntu.com/malone/bugs/1309 [06:17] morgs: here it is === ddaa is very annoyed by people who disregard keyboard actions [06:30] Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial] Fix bugs 1284, 1285 - no links to product releases from product or productseries (patch-2057: morgan.collett@canonical.com) [06:38] ddaa: thx === Kinnison decides he's incredibly tired [06:59] Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial] Add KarmaActionName to dbschema.__all__ and add a distinct to IKarmaActionSet.selectByCategoryAndPerson() (fix https://launchpad.ubuntu.com/malone/bugs/1260). (patch-2058: guilherme.salgado@canonical.com) [07:10] Kinnison: then go get some rest :-) [07:11] sivang: tempting [07:11] sivang: fairly soon I'll be able to stop for the weekend === Luciph3r [~Luciph3r@217.221.80.130] has joined #launchpad [07:21] re-salve [07:29] Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial] RDF formatting changes per bug 1248 (patch-2059: morgan.collett@canonical.com) === debonzi [~debonzi@200-171-140-32.dsl.telesp.net.br] has joined #launchpad === bradb [~bradb@modemcable134.23-201-24.mc.videotron.ca] has joined #launchpad === bradb waves from jbailey's [07:35] hey brad [07:36] hey Kinnison === Kinnison heads off [07:37] time for rest and then fridge moving [07:37] ciao === debonzi [~debonzi@200-171-140-32.dsl.telesp.net.br] has joined #launchpad === jbailey [~jbailey@modemcable134.23-201-24.mc.videotron.ca] has joined #launchpad [07:43] bradb: *poke* [07:43] ouch [07:57] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=SteveA] another round of import fascism and other cleanups (patch-2060: daf@canonical.com, morgan.collett@canonical.com) === SteveA merges from RF to see where the fascism is up to [07:59] last I checked, traversers.py importing from DB and about 15 imports of things not in __all__ [08:00] many of the cases of importing from outside __all__ that i have seen have been when module A imports N from module B [08:00] module B has N in its __all__ [08:01] then module C imports name from module A [08:01] rather than from module V [08:01] rather than from module B === sabdfl [~mark@pc-n253.wlan.inet.fi] has joined #launchpad [08:03] hey lunchpadders [08:03] salgado: ping [08:04] sabdfl, pong [08:04] salgado: hi [08:04] i'm finally getting to work through foaf [08:04] cleanup an ui [08:04] some initial feedback: [08:04] generally great work, thank you! [08:04] you've made too much use of interfaces [08:05] IPerson and ITeam really don't need to be distinct [08:05] in zcml/person.zcml things were a mess because of this [08:05] i'm cleaning it up [08:05] all of the portlets really belong on IPerson [08:05] and you can use if you need to [08:06] there was a bit of a mess in portlet-team-actions.pt for example [08:06] all of that is going going gone [08:06] ack? [08:07] I have been blocked in some other stuff. I don't know what the status of the portlets and this stuff [08:07] are you going to get rid of ITeam? [08:07] salgado: i haven't gotten rid of it entirely [08:08] I think we need it to have distinct pages for persons and teams [08:08] just slimmed it down [08:08] do you have something in mind to manage this? [08:08] salgado: there are some pages we need for teams that we don't need for people [08:08] but we don't actually need a different team-index.pt [08:08] that's goooone [08:08] most of the stuf is either: [08:09] - a completely separate page [08:09] - a tal:condition [08:09] it's crackful to have a completely differnet set of portlets, duplicated, as you had [08:09] now, on the memberships [08:09] it's a little confusing ;-) [08:09] the interface that you have has [08:09] ahh, I see. store both templates in a single file and choose them with a tal:condition? [08:09] - allmembers [08:10] - active/admin/etc/members [08:10] - teammemberships (which are not the memberships IN this, but OF this, even though the data model does it the other way around) [08:10] salgado: yes, because 90% of the details are the same [08:10] where there are forms and funcitonality that are UNIQUE to a team, have that in a different +foo page [08:11] i'll fix it up and land it next week, i just wanted you to know it's going to be a big patch [08:11] but it should look good [08:11] and its going quickly, because youve given me a pretty solid base [08:12] that's all so far [08:12] nice to know the work I did is at least easy to be refactored. I always work with that in mind [08:12] i can see, and appreciate it [08:13] I've been refusing to play with the problems between person/teams pages because we had some discussion about moving teams to another path (/teams) for example [08:13] and I was affraid that I could have to refactor it all over again [08:14] but anyway, thanks for helping me with this. :) [08:21] Merge to rocketfuel@canonical.com/pybaz--devel--0: [r=jamesh] raise SourceTreeError instead of Exception in ArchSourceTree constructors (patch-36: ddaa@ddaa.net) [08:26] salgado: ok, another thing [08:26] person.subscriptionPolicyDescription [08:26] is going [08:27] because you can just context/subscriptionpolicy/title [08:27] and /description [08:28] indeed. it used to return a description that was hardcoded instead of using the values from dbschema [08:32] it's goooone :-) [08:32] salgado: did you deliberately want to show the preferred email of a team, even to anonymous users? [08:33] I don't think that's necessary [08:33] we can show them only to logged in users, as we do with people email [08:33] i wondered if that was just accidental :-) [08:34] actually, there is some argument to say that a team's email is less sensitive than a person's [08:34] but a team email address is usually something public [08:35] it isn't hard to join launchpad [08:35] so, i don't see a problem with people having to join launchpad in order to see a team email address in the clear [08:36] mpt is on such crack if he did these pages! [08:37] hey SteveA [08:37] evening [08:37] how's the debconf? [08:40] sabdfl: did you see the news about mandrake / connectiva's debian derivative ? [08:44] SteveA: yes [08:44] SteveA: doomed, i think [08:45] if they make that their primary distro base, they will each lose big chunks of their user base to us :-) [08:45] they could become an ubutu derivative, of course... [08:46] salgado: is merging accounts working? if so, i'm going to do a quick usability test with jbailey on it right now. [08:46] bradb, please DO IT. :) [08:46] and yes, it should be working === bradb DOES IT [08:46] salgado: The bug is still open. =) [08:46] jbailey, which one? [08:47] 572 [08:47] up to 275 lines of jbailey brain dump so far... [08:47] I think I'm amusing him [08:47] OMG, it's assigned to malone. :( [08:48] Sure, I was in malone and trying to do the work. [08:48] Is that bad? [08:48] I always do some triaging in foaf/launchpad bugs [08:49] but never in malone bugs. I wasn't aware of it [08:50] Ah, sorry for the hassle. [08:51] no worries [08:51] but dude, you have three accounts in launchpad [08:51] Do I? === _--Luciph3r--_ [~Luciph3r@217.221.80.207] has joined #launchpad [08:51] you must be a rock star. all our scripts are creating accounts for you. :) [08:51] I'm aware of two of them, I guess. [08:51] https://launchpad.ubuntu.com/people?searchfor=all&name=jeff+bailey [08:51] salgado: sssh! [08:51] i'm doing a usability test [08:52] oh, sorry [08:52] :P [08:52] *lol* [08:52] One of the accounts has no email address associated with it, apparently. That should be fun to figure out. =) [08:53] It'll be fun when this thing starts merging in all my upstream accounts too. =) [08:54] maybe it has email addresses associated. the problem is just that that email is not validated and thus we can't make sure it's yours [08:54] I think I use 3 different ones. [08:55] I could tell you exactly what is going on with your accounts (why you have them and the reason of the problems), but they closed the backdoor I used to use to access the production database. :P === debonzi [~debonzi@200-171-140-32.dsl.telesp.net.br] has joined #launchpad [08:57] salgado: Apathy tends to overcome anyway. As long as it's all working, I'm not fussed about how it works. I spent enough time staring at bugzilla's guts and gnats' guts. If I never have to hack another bug trackig system, I won't feel unfulfilled. =) [08:58] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=jamesh] Fix buttress test coverage. (patch-2061: david.allouche@canonical.com) [09:00] hmmm. I just realized I can do an evil hack that would be of great help for people merging accounts [09:01] bradb, btw, what do you think of make our vocabs show something like "username (No validated email address)" in the cases where a user doesn't have a preferred email? [09:06] salgado: one sec [09:07] hm, i like the intent [09:07] but i confess that i don't think a normal person will know the implications of "No validated email address" without having to do some reflection [09:09] that might be an mpt question [09:10] indeed === salgado watches bugmails popping up here [09:16] right, we're wrapping up at jeff's, i'm going to dump the braindump to the list when i get home (so people can see what it sounded like right from the users mouth), and then i'll write a proper summary in the next few days [09:26] Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial] Bug 130: Multiple paragraphs in DOAP project description field (patch-2062: morgan.collett@canonical.com) === Mez [~mez@cpc2-lich4-3-0-cust115.brhm.cable.ntl.com] has joined #launchpad [09:30] how do i get access to malone? [09:30] like i have editbugs on bugzilla === _--Luciph3r--_ [~Luciph3r@217.221.80.236] has joined #launchpad [09:38] anyone? [09:39] Merge to rocketfuel@canonical.com/cscvs--devel--1.0: [r=spiv] test suite style fixes, better reaping of cvs server processes (patch-100: david.allouche@canonical.com) [09:52] salgado: do you know the branch name that will become production on tuesday? [09:53] lifeless: ^? [09:59] sabdfl, no, I don't know. === bradb [~bradb@modemcable082.64-130-66.mc.videotron.ca] has joined #launchpad === Luciph3r [~Luciph3r@217.221.80.244] has joined #launchpad [10:15] Merge to rocketfuel@canonical.com/launchpad--devel--0: [trivial] Bug 1289: Make product text less bold (patch-2063: morgan.collett@canonical.com) === _--Luciph3r--_ [~Luciph3r@217.221.80.248] has joined #launchpad [10:43] Merge to thelove@canonical.com/dists--bazaar--1.5: new build (patch-36) [10:44] Merge to thelove@canonical.com/bazaar--devo--1.5: Much simpler and better fix for archive-mirror source choosing (patch-29: Matthieu.Moy@imag.fr) === lamont is now known as lamont-away [10:56] New Malone bug 1315 filed on product Malone by Brad Bollenbach: Usability braindump needs to be summarized into a report [10:56] https://launchpad.ubuntu.com/malone/bugs/1315 [10:57] kiko-afk: interestingly enough, the EoS seemed pretty much dead obvious to jbailey [10:58] there was no indication from observing his actions of him wondering things like "does this get added as a comment?", etc. [10:59] i have a feeling that's because he's used the whiteboard before in bugzilla. someone not familiar with bugzilla (and even people that are, that never bother with the whiteboard) may be a little bit less comfortable with that field [11:13] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=spiv,jamesh] Auto Buildd System implementation. (patch-2064: celso.providelo@canonical.com) [11:16] dilys: knock knock [11:17] growing impatient? [11:17] wondering why she didn't spew out the bug report i just filed (i got the bugmail) [11:17] (and she's in the recipient list) [11:17] dilys seems to drop packets sometimes [11:18] she did not log one my merges that go through [11:18] bradb: hi :-) when will the ubuntu memebrs section be ready? I get system error on it [11:18] hi sivang. to what URL are you referring? [11:20] bradb: sec, I'll paste it here [11:21] bradb: it works now! [11:21] what the hell happend [11:21] anyone here who can give me the equivalent of editbugs on malone? [11:22] Mez: we don't yet have the functionality to edit more than one bug at a time in malone [11:22] bradb ... so who assigns bugs to someone ? [11:23] bradb: how can I get my name into ubuntu memebrs? I'm an approved one [11:23] who closes bugs etc etc [11:23] sivang - speak to mako [11:23] Mez: on a public distro bug, anyone. on a public upstream bug, only the maintainer/assignee can edit the task [11:23] bradb... so why Cant I do anything to any bug except comment on them [11:24] Mez: can you give me a URL? [11:24] https://launchpad.ubuntu.com/malone/bugs/1201 [11:25] Mez: what do you want to do from this screen? [11:25] well, prefereably mark the bug as fixed / similar [11:25] Mez: what have you tried so far? [11:26] I cant see anywhere to change the status [11:26] Mez: can you tell me a bit about where you're attempting to do this? what are you expecting to be possible at this point? [11:26] ok [11:27] I've just found where to change it [11:27] ok :) [11:27] you click on the bit under needs fixing in [11:27] grr [11:27] thats not very self explanatory [11:28] Mez: how would you have preferred or expected that to have worked? [11:28] oh god, I'm only now reading the news about the foundation [11:28] somewhere more intutive... or labelled "change the status" or somethign [11:29] Mez: right. i'm glad you mentioned that this confused you. there's been enough users that have brought that to my attention now to indicate that it's a high-priority usability fix we need to make. [11:29] s/there's/there have/ [11:29] lol - I'm glad it's not just me then :D [11:32] jamesh: ping [11:32] bradb: https://launchpad.ubuntu.com/distros/ubuntu/breezy/+members <== again gives an error === SteveA finds it unlikely jamesh will be around at this hour [11:33] SteveA: true, I already tried earlier :-) [11:40] sivang: sorry dude, that's outside my area of lp knowledge. i'll file a bug for you though so that the right person sees it. [11:42] Merge to rocketfuel@canonical.com/launchpad--devel--0: [r=kiko] simple clean up and little fixes (patch-2065: daniel.debonzi@canonical.com)