[00:25] <poolie> what do i need to do to create a team with 'canonical' in the name?
[00:25] <poolie> ask a question i guess
[00:25] <thumper> yep
[00:28] <spm> poolie: create a dummy name; I'll rename in the db
[00:29] <poolie> oh jolly good
[00:29] <poolie> thanks!
[00:29] <spm> well... it's for my use to so... :-)
[00:30] <poolie> https://edge.launchpad.net/~nonameyet-australia-staff
[00:30] <poolie> spm^^
[00:30] <spm> ta
[00:37] <spm> poolie: is done. https://edge.launchpad.net/~canonical-australia-staff
[00:37] <poolie> thanks
[00:40] <poolie> sinzui: duping my own bugs, how embarrassing :)
[00:49] <mrooney> poolie: haha, everyone surely does it at some point
[00:54] <thumper> spm: canonical-south-pacific?
[00:55] <wgrant> thumper: It's all Australia!
[00:55] <thumper> wgrant: but I'm not in Australia!
[00:55] <wgrant> thumper: Conform!
[00:56] <poolie> beuno: two things briefly on the user page: it's a bit too long now, and maybe the badges should be in the rhs portal
[00:56] <poolie> "when we say australia we mean nz too" :-)
[00:56] <mwhudson> thumper: canonical-nz would be rather boring
[00:56] <thumper> mwhudson: that it would
[00:57] <spm> mwhudson: thumper: I'm failing to see a problem there? (with canonical-nz being boring) :-P
[01:05] <jml> if you can have linux.conf.au in NZ, you can put up with canonical-australia-staff
[01:06] <jml> you can't have your cake and eat it too
[01:06] <jml> or should I say...
[01:06] <jml> pavlova!
[01:07] <thumper> oi
[01:08] <poolie> spm: can you rename it again to -australasia
[01:08] <spm> poolie: how about 'canonical-staff'? :-P
[01:09] <poolie> jml, reckon wikipedia has "List of New Zealander things claimed by Australia"?
[01:09] <jml> poolie: not yet!
[01:10] <poolie> "Some sources say the recipe originated in New Zealand, while others claim it originated in Australia." classic :-)
[01:12] <beuno> poolie, interesting. they could be moved there...
[01:12] <beuno> what would you do to make it shorter?
[01:13] <Usuario_Compaq> hi
[01:13] <YuCaTeRcO> alguien habla español?
[01:14] <YuCaTeRcO> help
[01:17] <spm> poolie: like so? https://edge.launchpad.net/~canonical-australasia-staff
[01:17] <YuCaTeRcO> ??
[01:18] <thumper> YuCaTeRcO: just ask (in english)
[01:18] <YuCaTeRcO> thumper:  español?
[01:19] <poolie> beuno: can you help is spanish?
[01:19] <poolie> s/is/in
[01:19] <YuCaTeRcO> algun:)
[01:19] <YuCaTeRcO> :(
[01:19] <YuCaTeRcO> alun canal en español?
[01:19] <beuno> YuCaTeRcO, si, en que te podemos ayudar?
[01:19]  * thumper looks at beuno
[01:19]  * beuno feels observated
[01:20] <YuCaTeRcO> beuno:  como adquiero mi disco de instlacion de servidor ubuntu?
[01:20] <beuno> YuCaTeRcO, eso que tiene que ver con Launchpad?
[01:20] <beuno> de aca: http://www.ubuntu.com/getubuntu/download-server
[01:20] <beuno> pero este no es un canal para ayuda con Ubuntu
[01:21] <YuCaTeRcO> beuno:  y esta canal de que trata'
[01:21] <beuno> YuCaTeRcO, Launchpad!
[01:21] <YuCaTeRcO> :o
[01:21] <YuCaTeRcO> okis
[01:21] <YuCaTeRcO> sorry
[01:36] <sinzui> poolie: I didn't want to mention it
[01:38] <sinzui> poolie: I believe you can delete this milestone if it really is an error
[01:38] <sinzui> https://edge.launchpad.net/bzr/+milestone/1.14rc1-error
[01:39] <poolie> thanks
[01:43] <poolie> lifeless: can you click https://edge.launchpad.net/~canonical-australasia-staff for me?
[01:43] <poolie> do you get permission denied?
[01:44] <lifeless> I'm a bad test
[01:44] <lifeless> and no
[01:45] <mwhudson> poolie: i get 404
[01:45] <lifeless> I think we want private, not private membership
[01:45] <lifeless> dont we?
[01:46] <poolie> do we?
[01:46] <poolie> i don't mind people knowing that it exists
[01:46] <poolie> like other canonical people for example
[01:46] <lifeless> ok
[01:50] <lifeless> poolie: I think it should be moderated not restricted
[01:51] <poolie> spanish CD requests...
[01:51] <lifeless> poolie: or if restricted, make sure the sysadmins will be managing it and putting hires into it
[01:51] <poolie> but anyhow, i think i'll make it public and moderated
[01:51] <poolie> i found a lp bug though :)
[01:51] <lifeless> au naturale
[01:55] <poolie> bug 373039
[01:58] <lifeless> it should get an upside down logo, like my personal logo
[01:59] <thumper> no
[01:59] <thumper> it shouldn't
[01:59] <lifeless> we're in the land down under here
[02:02] <thumper> it should be an upside down map of nz and au
[02:02] <lifeless> that would be good too
[02:02] <lifeless> or should we say
[02:02] <lifeless> a rightside up map
[02:04] <thumper> right, who says north is up
[02:04]  * thumper is on top of the world
[02:05]  * mwhudson has a south-up world map in his hallway
[02:32] <poolie> sinzui: if i make a private-membership team will its list archives be public or private?
[02:32] <sinzui> private
[02:32] <sinzui> true private teams will be landing in a few weeks
[02:33] <poolie> thanks
[02:33] <poolie> what is the difference between true private and private-membership meant to be?
[02:34] <sinzui> Everyone knows about a private membership team, but they do not know who is in it
[02:34] <sinzui> no one knows about a private team, and it can subscribe to bugs, have branches and PPAs
[02:34] <poolie> modulo bug 373039
[02:34] <poolie> ok thanks, that's what i thought :)
[02:35] <poolie> spm: oh sheesh, once it has a list i can't make it private?
[02:35] <sinzui> PS, WONTFIX, as designed
[02:36] <sinzui> Even if you were told the name, you get a 404
[02:36] <sinzui> Since developers are not super users, I cannot tell as real 404 form a private one
[02:36] <poolie> sorry i don't see how that's consistent with "everyone knows about a private membership team"?
[02:37] <lifeless> sinzui: surely you should be able to visit the page for a private-membership team
[02:37] <lifeless> sinzui: how else will you apply to join it?
[02:37] <sinzui> we are removing private membership teams in two cycles
[02:37] <poolie> oh ok
[02:38] <sinzui> Private membership teams were a hack, and we spend a lot of time dealing with issues where they cannot do anything accept revel into their own privacy. It is better it subscribe to bugs and branches
[02:39] <poolie> that's fine, i don't see a case where they're really needed
[02:39] <poolie> can you tell me what to do to make this team private after it has a list?
[02:39] <poolie> can i delete the list and then change it?
[02:39] <sinzui> poolie: yes, that is what you need to do
[02:40] <poolie> is just "deactivate list" enough?
[02:40] <sinzui> barry: updates a page elaboration on the steps last week
[02:40]  * sinzui looks
[02:45] <spm> poolie: I can remove and purge the list. (if you haven't already...)
[02:47] <sinzui> poolie: I don't know if that is required since you are keeping the team name, but you may need to take up spm's offer
[02:47] <lifeless> spm: please do
[02:47] <lifeless> spm: better safe
[02:47] <poolie> apparently i do
[02:47] <spm> argh. the name justchanged underneath me. prefixed private-
[02:47] <sinzui> The deactivate feature should do that automatically.
[02:47] <poolie> thanks spm, sorry for the noise
[02:47] <spm> poolie: actually is good - I've not had a lot to do with private teams, so is a learning experience. albeit the hard way :-)
[02:51] <sinzui> spm: The instructions to rename a team with a list are strikingly similar to the step we took to import launchpad-users' archive
[02:51] <sinzui> oh, the page even notes that
[02:52] <spm> sinzui: is probably where they came from :-)
[02:54] <spm> sinzui: when I made the team private, "something" prefixed "private-" onto the team name. Is that prefix relevant elsewhere as a pattern match? I assume not, but....
[02:54] <sinzui> spm: it a namespace/blacklist prefix
[02:55] <spm> ahh. to get around the blacklist? right.
[02:55] <sinzui> just like I cannot create a team with canonical or launchpad in it's name, the same is true for private. Only you have the power
[02:56]  * spm does a HeMan impersonation "I HAVE THE POWER!!!!!" zing boom flashing lights etc
[02:56] <sinzui> Fight the power
[02:56] <spm> ha!
[02:57] <Hellow> I GOT THE POWA
[02:58] <sinzui> CC Music Factory have lost the power
[02:59] <spm> I have a Rubber Ducky, and am not afraid to use it /super_silly_mode
[03:02] <poolie> my router power supply went 'phut' :/
[04:49] <tomas__> hola
[04:49] <tomas__> alguien m puede ayudar
[05:22] <wgrant> poolie: How do you propose to do permissions checking in the librarian? It doesn't know who you are.
[05:23] <wgrant> spm: Fix lpnet kthxbye
[05:24] <poolie> wgrant: i don't know, not my problem
[05:25] <wgrant> poolie: Aha, yes, that's always a useful way out!
[05:25] <poolie> there seems to be a tough tradeoff with the cookies being on a different domain but there's probably a way
[05:27] <wgrant> You can't have the cookies on the librarian domain.
[05:27] <wgrant> So librarian can't possibly know who you are.
[05:27] <wgrant> So we have a very difficult situation.
[05:28] <poolie> right
[05:29] <poolie> wgrant: thus my suggestion that we mostly try to avoid putting the real librarian urls in email etc
[05:29] <poolie> make the url the secret not the user's key
[05:30] <wgrant> poolie: Right, I suppose that's reasonable as long as the key in the URL isn't as pathetic as it is now.
[05:30] <wgrant> That has been the source of a few vulnerabilities before.
[05:34] <poolie> so there's an arguable problem that once you know the url you always know it, even if you lose access to the bug
[05:34] <poolie> otoh the content never changes
[06:18] <jamesh> poolie: we don't want any authentication cookies sent to the librarian.
[06:19] <jamesh> if there is stuff that needs to be protected by authentication, it should be segregated from the librarian and make sure no active formats can be uploaded by the user.
[06:20] <jamesh> no html, svg, swf, probably pdf, etc.
[06:21] <jamesh> I know adobe's viewer can execute javascript in a PDF, but I don't know what it can access, so you'd probably want to ban it.
[06:25] <poolie> yeah that's what i was thinking
[06:26] <wgrant> jamesh: Adobe Reader's JavaScript interpreter lets you execute arbitrary code, most of the time...
[06:26] <wgrant> It hasn't got the best security record.
[06:27] <jamesh> wgrant: I was more wondering about what it lets you do that isn't considered a bug ...
[06:27] <jamesh> e.g. read cookies
[06:27] <wgrant> Right.
[06:59]  * maxb wonders what the time-on-queue build score adjustment is supposed to achieve
[07:00] <maxb> given that is doesn't boost score enough to preempt anything other than changelog urgency
[07:01] <maxb> Which doesn't really get used in ubuntu anyway
[08:28] <lifeless> spm: do you know if soyuz will see a new key on my keyring automatically?
[08:28] <spm> lifeless: nope. I'd assume "yes" but...;  cprov ^^ ?
[08:30] <wgrant> lifeless: By keyring do you mean LP account?
[08:30] <cprov> maxb: it will preempt similar builds (same component, urgency, pocket)
[08:31] <soren> lifeless: It did for me when I added a new one.
[08:31] <wgrant> cprov: But surely the actual time is taken into account too, rather than just the score given by the time?
[08:31] <maxb> cprov: Indeed, but if they're similar in those ways they'll be getting time-boosts too
[08:31] <wgrant> Because most PPA builds will have identical scores, yet they build somewhat in-order.
[08:31] <cprov> maxb: but the older will be prioritized
[08:32] <maxb> But isn't the queue processed in order anyway?
[08:32] <cprov> maxb: but yes, the time-based scoring is to timid to be effective.
[08:32] <maxb> (subject to scores)
[08:32]  * maxb attempts to be more clear
[08:33] <cprov> maxb: the queue order is the score (nothing else is considered)
[08:33] <wgrant> If the ordering is on (score, queued date), then why is the waiting time part of the score?
[08:33] <wgrant> Oh.
[08:33] <maxb> queued-date is not used to provide ordering for builds of the same score?
[08:33] <wgrant> Well, other stuff *is* considered, but only in as much as PostgreSQL has to order them somehow.
[08:34] <cprov> maxb: the build.id is the tie-breaker in the current implementation.
[08:34] <maxb> The other thing that I'm wondering: does time-based rescoring actually become visible in the web ui, or is the time-based boost a sort of "virtual" score on top of what's actually in the build record?
[08:35] <lifeless> soren: thanks; new machine, new signing key
[08:35] <lifeless> wgrant: no, I mean keyring; used for validating mails and package uploads
[08:35] <maxb> cprov: Ah! Which approximates to order of enqueuing in the common case?
[08:36] <wgrant> lifeless: But you mean the one attached to your LP account?
[08:36] <cprov> maxb: yes, if builds have the same score (which is very likely for PPAs, due to the component smash)
[08:36] <lifeless> wgrant: naturally; I wouldn't expect lp to see new subkeys on keys I haven't told it about
[08:37] <wgrant> lifeless: Right, that change will take effect immediately. But binding an OpenPGP key to a machine is a little strange.
[08:38] <lifeless> wgrant: its pretty normal if you keep your primary key offline
[08:38] <soren> I don't know if it's actually normal, but I think it's a very sensible thing to do.
[08:38] <wgrant> lifeless: Ah, I see.
[08:39] <maxb> On the basis that if you have to revoke it, the signatures you made on other machines remain valid?
[08:40] <lifeless> maxb: right, machines get compromised, keys on the machine get compromised
[08:40] <lifeless> http://fortytwo.ch/gpg/subkeys <- has info
[09:04]  * wgrant kicks the appservers again.
[09:07] <wgrant> lifeless: Why would branches for a sourcepackage appear on the project page?
[09:07] <wgrant> They're on completely separate objects.
[09:07] <lifeless> wgrant: they are branches of the same code
[09:08] <lifeless> so why wouldn't they at least get a nod
[09:08] <wgrant> lifeless: It would possibly make sense when the package is linked to the project, but the package doesn't exist (it's a bug that you could create that branch), so it can't be linked.
[09:09] <lifeless> wgrant: I don't think its a bug that I could create that branch; I can create the package too
[09:09] <lifeless> wgrant: and will when I push to the ppa
[09:09] <lifeless> wgrant: if ppa's are to build from branches we need this exact workflow
[09:09] <lifeless> as for the link being missing; that will get sorted
[09:11] <wgrant> lifeless: /ubuntu/+source/subunit is meant to 404, because it doesn't actually exist in Ubuntu. It just happens that Soyuz doesn't currently do that existence check, and the SourcePackageName exists because somebody uploaded subunit to a PPA.
[09:11] <lifeless> wgrant: link created
[09:11] <lifeless> wgrant: everything except the 404 sounds reasonable tome
[09:12] <wgrant> lifeless: Why shouldn't it 404?
[09:12] <wgrant> It doesn't exist.
[09:12] <wgrant> That link is also wrong, now, and you shouldn't have been allowed to create it.
[09:12] <lifeless> wgrant: think through the use case of uploading a new package
[09:12] <wgrant> There is no subunit package in Ubuntu 9.04, so that link is impossible.
[09:13] <wgrant> lifeless: That page will 404 until somebody uploads the package to some Launchpad archive, so you'd have to upload it to a PPA before you can create a branch for it.
[09:13] <lifeless> wgrant: you're repeating what happens today. I want you to imagine things
[09:13] <lifeless> wgrant: I want you to imagine that you can push instead of dput
[09:14] <wgrant> I don't imagine that NMSP is going to happen in my lifetime.
[09:14] <wgrant> But yes, let's imagine.
[09:14] <lifeless> wgrant: now, go through the workflow of getting a new package in a ppa uploaded
[09:14] <lifeless> and you'll see exactly why I'm doing what I'm doing in the order I'm doing it
[09:15] <lifeless> the upstream link is arguably right, because for *the ppa the binary is in, in jaunty*, it is correct.
[09:16] <wgrant> No, that link clearly says 'in Ubuntu Jaunty'
[09:16] <wgrant> A PPA is not Ubuntu Jaunty.
[09:16] <wgrant> That there is no way to link to a PPA is a bug, but that upstream link is certainly not right.
[09:17] <lifeless> I'm pretty sure ppas will inherit it
[09:17] <lifeless> OTOH if you can demonstrate some harm by it being there I will remove it
[09:17] <wgrant> It's wrong, but not harmful.
[09:17] <wgrant> PPAs won't inherit it, because it carries no meaning.
[09:18] <wgrant> I don't think anything uses it except the 'Also affects project' bug link and perhaps rosetta.
[09:18] <lifeless> and perhaps showing links to package branches
[09:18] <lifeless> which is where this discussion started
[09:18] <wgrant> I'm trying to think how a package branch for a PPA can be done cleanly.
[09:19] <lifeless> good! I'm trying to do it :)
[09:19] <wgrant> Package branches are already very strange, because they use quite a different model from project branches.
[09:19] <wgrant> (Source package branches are always associated with a series, and may be officially associated with a pocket; only official project branches are linked to project series)
[09:21] <wgrant> I don't recall any discussion about this on the spec...
[09:23] <wgrant> To do things properly, you need to be able to associate the branch with a PPA, not just a distro. It should be a different target. I thought this was a bit odd, as it would mean merging branches between targets, but that already happen's with Debian->Ubuntu merges.
[09:23] <wgrant> Wow.
[09:23] <wgrant> Apostrophe fail.
[09:24] <maxb> I didn't even know you could have package branches without a project
[09:24] <wgrant> It's a very new, unadvertised, and broken feature.
[09:25] <maxb> ah :-)
[09:25] <wgrant> maxb: https://code.edge.launchpad.net/~subunit/ubuntu/jaunty/subunit/releases, for example.
[09:28] <wgrant> spm: You wouldn't happen to still be around to kick a dodgy lpnet appserver, would you?
[09:29] <mpt> "Sorry, there was a problem connecting to the Launchpad server. "
[09:29] <wgrant> Yep.
[09:29] <wgrant> That Storm bug again.
[09:29] <wgrant> As usual.
[09:30] <wgrant> Juuuust as my boss creates a Launchpad account, too :(
[09:30] <Spads> wgrant: kicked.
[09:31] <wgrant> Spads: Thanks.
[09:32] <wgrant> Do you know why that has just started happening now?
[09:32] <wgrant> s/now/a couple of weeks ago/
[09:36] <mwhudson> the fix for the bug landed, so hopefully it'll stop happening soon
[09:37] <mwhudson> (when there are LOSAs awake again i guess)
[09:37] <wgrant> A nice simple fix, too.
[09:37] <mwhudson> yeah
[09:40] <wgrant> What time does the staging restore happen these days?
[09:41] <mwhudson> it's never been very predictable
[09:41] <mwhudson> and i certainly have no idea at the moment
[09:41] <wgrant> It would be useful to know, as new users can't use it until the next restore.
[09:52] <Kulithalai> open Office 3 Installed but the printer adminstration wizard therein spoils the party it, we can not print  if any one can help
[09:52] <wgrant> Kulithalai: You might have better luck in #ubuntu.
[09:53] <Kulithalai> Launchpad failed to load
[09:53] <Kulithalai> |help
[09:53] <wgrant> Kulithalai: Try refreshing the page - that generally helps.
[10:04] <mnemo> "see bug 360846"
[10:04] <mnemo> that's a good one :P
[10:04] <turutosiya> is login.launchpad.net/+openid down?
[10:05] <mnemo> turutosiya: all pages refuse to load pretty much every other time I try here...
[10:05] <rowinggolfer> turutosiya: see topic
[10:06] <turutosiya> ok thanks
[10:07] <wgrant> Spads: It's gone again.
[10:21] <apachelogger> cprov: ping
[10:21] <cprov> apachelogger: pong
[10:22] <apachelogger> cprov: hey, I just had kind of a PPA idea ... did you ever consider letting the PPA owner create and manage pockets/sub-collections
[10:23] <cprov> apachelogger: do you mean arbitrary pockets ?
[10:23] <apachelogger> I guess
[10:23] <cprov> apachelogger: no, we didn't. But tell me more about the use-case you have in mind.
[10:23] <apachelogger> for example: Kubuntu has a PPA for experimental software and packages, a user might not want to be exposed to the full load of packages, but say only those needed for amarok
[10:25] <apachelogger> so the kubuntu peeps create a pocket which only contains amarok-related packages making users happier as not their whole system falls apart when the kubuntu people publish a new KDE version to the PPA
[10:25] <cprov> apachelogger: isn't that a separate archive/ppa ?
[10:25] <apachelogger> well, the idea is to still have them all-in-one
[10:26] <cprov> apachelogger: also, you realize that splitting the needed packages is more complicated that it looks, right ?
[10:27] <apachelogger> cprov: didn't really think about it, the idea just hit me :)
[10:27] <mnemo> im getting LP errors that dont even have OOPS numbers on them... you you guys aware of those??
[10:27] <mnemo> i.e. "Sorry, there was a problem connecting to the Launchpad server."
[10:27] <cprov> apachelogger: I understand your point, but at the end named-ppa is exactly the same thing (uses/share the same resources)
[10:27] <apachelogger> yep
[10:28] <apachelogger> and $user can still get all of the packages in there
[10:28] <apachelogger> the pocket would eventually just a virtual filter on top of the actual packages list
[10:28] <cprov> apachelogger: it's a real problem, you are right. Sometimes you have to split/combine packages according to their audience
[10:28] <apachelogger> I think there is a "be" missing somewhere
[10:29] <cprov> apachelogger: and that's the problem that named-ppa is supposed to solve.
[10:30] <cprov> apachelogger: they share the maintainer group, the signing key, etc ... the only difference is the way you specify it in the client sources_list, really.
[10:31] <apachelogger> well, that takes away the user's possibility to get all Kubuntu experimental updates with one repo line and the developer's to just put everything in one PPA and present it to the user in a pre-defined way
[10:32] <apachelogger> like when you need a specific lib in two experimental PPAs you either create a common PPA shared amongst those or copy the package around, while a post-build virtual pocket management woud allow it to be in both pockets at the same time as long as both of them are part of the same PPA
[10:34] <apachelogger> I suppose it could be done the other way round i.e. create an overlay on top of all present PPAs to manage multiple appearances of one package in the various PPAs
[10:35] <apachelogger> cprov: if you want I can write a complete description of what I have in mind with use cases in a bug report ... since I only got up 10 minutes ago I might not make any sense :D
[10:36] <cprov> apachelogger: yes, that would be ideal
[10:37] <apachelogger> ok :)
[10:38] <wgrant> mnemo: It's the bug mentioned in the topic.
[10:38] <cprov> apachelogger: thanks for thinking about this problem, though.
[10:38] <wgrant> cprov: Don't package sets also solve this problem?
[10:38] <wgrant> But that would require PPA overrides, I suppose.
[10:39] <cprov> wgrant: yes, I suppose, but it would require pinning.
[10:40] <wgrant> cprov: Which the buildds and some clients are going to need anyway, I'd imagine.
[10:40] <cprov> wgrant: buildd yes, client I don't think so.
[10:41] <wgrant> cprov: Hrm, OK.
[10:41] <cprov> wgrant: well, i'm not sure it will be *required*
[10:41] <wgrant> I wish it had all been thought out better.
[10:41] <cprov> things will discovered on-course, it's a complex problem.
[10:49] <jordi> I think its the first time I go to work after having no sleep and no alcohol involved.
[10:50] <jordi> it's still 11:49, this is a disaster
[12:07] <rconan> win close
[12:07] <rconan> woops
[12:11] <apachelogger> cprov: bug 373197
[12:12]  * apachelogger notes that he used the wrong browser again :D
[12:24] <Hellow> *sigh* still getting "Please try again" errors
[12:36] <mpt> and so is ubottu, apparently
[12:53]  * wgrant is a little confused as to why this is being allowed to continue.
[12:53] <wgrant> Shouldn't one in every few requests failing being enough to get a swift cherrypick?
[13:47] <kiko> wgrant, a cherry-pick of what?
[13:47] <wgrant> kiko: The fix for the Storm bug that keeps killing the appservers.
[13:48] <kiko> wgrant, who says it's a storm bug? :)
[13:48] <wgrant> kiko: The diff.
[13:48] <wgrant> And flacoste.
[13:48] <kiko> it's not a storm bug.
[13:49] <kiko> the storm bug just makes the actual problem more serious
[13:49] <kiko> but it won't fix the core issue
[13:49] <wgrant> But it will let the appservers recover.
[13:50] <kiko> it won't
[13:50] <kiko> you're confused, trust me :)
[13:50] <wgrant> Oh.
[13:50] <kiko> it will make them degrade more slowly
[13:50] <kiko> but they will still die a horrible death
[13:50] <wgrant> Lovely.
[13:50] <wgrant> Have you identified the real problem yet?
[13:51] <kiko> yes, am shopping for an admin
[15:34] <fgiraldeau_> Hello, I would need more space for a PPA, is there somebody who can help me?
[15:40] <beuno> cprov, ^
[15:41] <cprov> fgiraldeau_: yes, file a question via Launchpad on the soyuz product specifying how much extra space you need (and why)
[15:42] <fgiraldeau_> cprov, thanks!
[15:42] <cprov> fgiraldeau_: np
[16:50] <james_w> is there any chance to have "debian/etch" as a series on lp?
[17:16] <SamB> james_w: for PPA, you mean?
[17:16] <james_w> nope, I mean the object
[17:16] <SamB> object?
[17:17] <james_w> just like https://launchpad.net/debian/lenny
[17:17] <mrevell> james_w: Probably best to ping sinzui about that
[17:17] <james_w> thanks
[17:17] <mrevell> of course, that's precisely what I've just done :)
[17:17] <mrevell> in writing that
[17:17] <mrevell> but he has no context, heh
[17:18] <sinzui> james, mrevell someone need to contact the owner/driver of debian to do that
[17:18] <mrevell> sinzui: Registry Admins, I think
[17:19] <sinzui> and as this is owned by our admins, a question is needed to ask them to do it
[17:19] <james_w> thanks
[17:19] <sinzui> mrevell: I am developer, or an admin, the difference is like:
[17:19] <sinzui> Hey, I just make the bomb, I don't drop it
[17:20] <mrevell> :)
[17:21] <james_w> rockstar: hey. Does the scanner have to pick up a branch pushed via the smart server before I can access that branch through the API? If so, is there a maximum latency that will be involved?
[17:22] <james_w> I'd rather not have to sleep for a few minutes every time a branch is pushed
[17:23] <rowinggolfer> who dropped the bomb?
[17:24] <emsenn> The dogs, shortly after they were let out.
[17:24] <rockstar> james_w, no.  When you push, it creates a db record.
[17:24] <rockstar> james_w, out of curiosity, what are you doing through the API?
[17:24] <james_w> rockstar: hmm, I'm getting 404s on lp.load(branch_api_url) after doing the push
[17:24] <james_w> setting the official branch for a source package
[17:25] <rockstar> james_w, ah.  Hm.
[17:26] <rockstar> james_w, if you wait five minutes, does the lp.load work?
[17:26] <rowinggolfer> is that the launchpad equivalent of "have you tried switching it off..."
[17:27] <rowinggolfer> james_w: BTW - did you push the whole of debian etch?
[17:27] <james_w> rowinggolfer: nope
[17:27] <james_w> rockstar: trying that now
[17:28] <rockstar> james_w, I just tested it with `bzr push; bzr lp-open` and it's working.  I suspect something fishy in the API cache.
[17:44] <james_w> rockstar: sorry, my fault
[17:44]  * james_w hangs his head in shame
[17:44] <rockstar> james_w, :)
[17:44] <james_w> trying to load the branch from staging when I pushed to edge
[17:45] <rockstar> Ha!  :)
[17:45] <sinzui> I've done that
[17:51] <james_w> (<lp.code.model.seriessourcepackagebranch.SeriesSourcePackageBranchSet object at 0x6fba650>, 'delete', 'launchpad.Edit')
[17:52] <james_w> ^that was the "content" of the 401 error I'm getting, does it make sense to anyone?
[19:30] <dlacombe> hello all
[19:31] <dlacombe> e
[19:36] <TresEquis> lol, "see bug 360846"
[19:36] <TresEquis> the intermittent problem I'm seeing is that I can't view bugs ;)
[19:39] <TresEquis> or, when I do see them, I can't update them
[19:42] <_Jeff_> Hi guys
[20:21] <colonelqubit> How do I add a comment to an ubuntu blueprint on launchpad?  Someone in the #ubuntu-bugs channel mentioned that every blueprint should have a corresponding wiki page, but I don't see a link to a wiki page on this particular blueprint page.
[20:23] <timlinux> hi
[20:23] <timlinux> I need to rename a ppa for our project
[20:24] <timlinux> is there a way to do that? I dont see one from from the ui
[20:24] <Ursinha> colonelqubit, there's a whiteboard in the blueprint, it's the way of making comments
[20:25] <Ursinha> timlinux, you can edit your ppa and there's the Displayname
[20:25] <Ursinha> timlinux, in Change details
[20:26] <timlinux> yes I need to change not the display name but the name itself
[20:26] <timlinux> otherwise the url still looks odd
[20:26] <colonelqubit> Ursinha: Ah, okay
[20:26] <timlinux> let me show you
[20:26] <Ursinha> timlinux, so I guess you need to open a question and request that
[20:27] <Ursinha> timlinux, go ahead
[20:27] <timlinux> https://launchpad.net/~qgis/+archive/stable <-- tried to rename to unstable
[20:27] <timlinux> so when you open the page it shows as unstable
[20:27] <timlinux> but the url is inconsistent
[20:27] <colonelqubit> Ursinha: The bit about "Any notes on the _status_ of this spec" was what confused me. I was going to add notes on how to implement the feature.
[20:28] <timlinux> Ursinha, where do I open a ticket for manual intervention? Is there a ticket tracker somewhere for this?
[20:29] <Ursinha> timlinux, sure, answers.launchpad.net/ :)
[20:29] <Ursinha> let me hand you the link
[20:29] <Ursinha> timlinux, https://answers.edge.launchpad.net/launchpad-project/+addquestion
[20:29] <Ursinha> colonelqubit, let me look closer, I recall a url field
[20:29] <Ursinha> a closer look, even
[20:30] <Ursinha> timlinux, or https://answers.launchpad.net/launchpad-project/+addquestion
[20:30] <colonelqubit> Ursinha: yeah, bdmurray on #ubuntu-bugs said that there should be a wiki page, but there's not one set for this blueprint
[20:30] <Ursinha> colonelqubit, let me see, do you have the blueprint link?
[20:30] <colonelqubit> https://blueprints.launchpad.net/ubuntu/+spec/niversal-updater
[20:31] <Ursinha> niversal :)
[20:31] <timlinux> Ursinha, many thanks I will take it there
[20:31] <Ursinha> timlinux, no problem :)
[20:32] <colonelqubit> Ursinha: a "niversal ubuntu trough", to be specific
[20:34] <colonelqubit> Ursinha: I mean, I do see my machines as repositories of information -- heaps of data, even -- but I usually don't go so far as to consider them barnyard food boxes.
[20:39] <Ursinha> colonelqubit, here, you have to edit the blueprint to add the url specification
[20:40] <Ursinha> colonelqubit, but only allowed people can do that
[20:40]  * colonelqubit hopes he gets to level-up.
[20:41] <Ursinha> colonelqubit, so you can comment using the whiteboard and nag people for them to add the url in there
[20:41] <colonelqubit> Ursinha: ok
[20:43] <colonelqubit> Ursinha: If commenting on a wiki page is the preferred method of commenting on a blueprint, why doesn't the creation of a blueprint just create the wiki page at the same time? (automagically)
[20:45] <Ursinha> colonelqubit, I guess each team decides how's the best way to exchange ideas over a blueprint
[20:45] <Ursinha> but that's an idea
[20:46] <colonelqubit> Ursinha: If I wanted to file a feature request, it would be against launchpad?
[20:47] <colonelqubit> or the ubuntu launchpad team?
[20:47] <Ursinha> colonelqubit, against launchpad blueprints
[20:47] <colonelqubit> ok
[21:18] <james_w> http://github.com/rbmntjs/open-baskerville/commit/c5b3322fb49baa3104363ca98546c684d940c0ef
[21:50] <colonelqubit> Is there a way to mark Questions (under the Questions/Answers launchpad system) as dupes?
[22:16] <kiko> colonelqubit, no, there isn't -- you should just invalidate and store a URL in the comment.
[22:20] <colonelqubit> kiko: thanks
[22:42] <colonelqubit> kiko: Do I need to be an answer contact for a give project in order to be able to mark a Question as invalid?
[22:43] <kiko> colonelqubit, I don't think so -- can you not change status?
[22:43] <colonelqubit> kiko: what does the interface look like?
[22:44] <kiko> there's a link to the right that says "change status"
[22:44] <kiko> sinzui, do you know the answer to colonelqubit's question?
[22:44] <colonelqubit> kiko: the links I have are Edit question, Subscribe, Link existing bug, Create bug report, This is a FAQ
[22:44]  * sinzui reads code
[22:48] <sinzui> colonelqubit: yes, you must be an answer contact to make a question as invalid
[22:48] <colonelqubit> sinzui: thanks
[22:48] <sinzui> colonelqubit: it must also be open or needsinfo for the status to be presented
[22:49] <colonelqubit> sinzui: is any of this doc'd somewhere?
[22:49] <colonelqubit> sinzui: it's been kind of a steep learning curve trying to navigate launchpad
[22:50] <kiko> colonelqubit, may not be documented; sinzui could email matt revell to sort that out
[22:50] <colonelqubit> sinzui: and you guys probably have been things to do than tell me via IRC.. :-)
[22:50] <colonelqubit> s/been/better/
[22:50] <sinzui> I do not think so. I'm reading a unit test to remember these rules. They are complex because we take into acounnt the user and email vs web
[22:51]  * sinzui looks at help.launchpad.net
[22:51] <colonelqubit> yeah, I was looking at https://help.launchpad.net/Answers/OfferingHelp, but didn't see much there
[22:52] <sinzui> yes, that needs some elaboration
[22:53] <sinzui> what a tease! The link to question status just takes you to the top-level answers page
[23:03] <sinzui> colonelqubit: I updated the definition of the invalid question status, and partially fixed the link on the answer contact page to the status defintions
[23:03] <sinzui> I don't think I can get the anchor to work without apply extreme violence to moin
[23:04] <colonelqubit> sinzui: heh
[23:05] <colonelqubit> sinzui: thanks for updating the page
[23:05] <colonelqubit> I'll go mark some dupes
[23:38] <ablert> I'm trying to learn how to create a meta package - whose directives are to install a bunch of other packages
[23:38] <ablert> anyone know of a good learning resource?
[23:39] <wgrant> ablert: apt-get source ubuntu-meta
[23:42] <ablert> thanks
[23:44] <poolie> sinzui: is <https://edge.launchpad.net/bzr/+milestones> new? it looks really good
[23:45] <sinzui> It landed two days ago
[23:45] <poolie> it does have a regression that milestones without dates sort to the top
[23:45] <sinzui> I have not bragged about it because I think the milestone/release page is more prone to timeout. I need to add batching
[23:46] <sinzui> poolie: that too
[23:46] <poolie> which kind of sucks for https://edge.launchpad.net/bzr/+milestones
[23:46] <sinzui> We will fix that next week
[23:46] <sinzui> s/we/me/
[23:46] <poolie> it may not strictly be a regression in that page but it occurred elsewhere before
[23:46] <poolie> i'd love to see on that page a Bugs column saying "%d/%d" for open and total
[23:46] <poolie> though it's only a click away
[23:55] <poolie> lifeless: maybe launchpad should have an expiry mechanism for bug importance?
[23:55] <poolie> "<janitor> you haven't done anything on this for 15 days, it's obviously not critical..."
[23:57] <exarkun> maybe it's only critical relative to all the other bugs filed against a project
[23:57] <exarkun> and there's only 6 minutes of developer time available every three months
[23:57] <exarkun> but they should be spent on the critical things first