2008/04/14/#launchpad.txt

=== fta_ is now known as fta
=== jamesh_ is now known as jamesh
=== jamesh_ is now known as jamesh
=== mwhudson_ is now known as mwhudson
[06:41] <ubotu> New bug: #217115 in launchpad-bazaar "OOPS requesting a review of a proposal to merge branch" [High,Confirmed] https://launchpad.net/bugs/217115
=== asac_ is now known as asac
[07:13] <emgent> morning
[08:26] <carlos> morning
[08:32] <mpt> Goooooooooooooood morning Launchpadders!
[08:33] <emgent> morning mpt :)
=== cjwatson_ is now known as cjwatson
[09:16] <emgent> hi sabdfl 
[09:55] <ubotu> New bug: #217155 in launchpad "pgp signatures get added as attachements" [Undecided,New] https://launchpad.net/bugs/217155
[10:12] <sabdfl> hi emgent
[12:16] <ubotu> New bug: #217207 in launchpad "Wishlist: links to related packages" [Undecided,New] https://launchpad.net/bugs/217207
=== mrevell is now known as mrevell-lunch
[12:35] <ubotu> New bug: #217218 in launchpad-bazaar ""Change registrant" changes the branch owner not the registrant" [High,Confirmed] https://launchpad.net/bugs/217218
=== mrevell-lunch is now known as mrevell
=== vednis is now known as mars
[14:02] <ubotu> New bug: #217249 in soyuz "PPA uploads should preserve the original component" [High,In progress] https://launchpad.net/bugs/217249
=== mdz_ is now known as mdz
=== mrevell changed the topic of #launchpad to: https://launchpad.net/ || PPA builds slow 09.00 - 18.00 UTC April 15 | Next meeting (all welcome): Thu 17 April 2008, 1800UTC #launchpad-meeting | Help: https://help.launchpad.net | Questions: https://answers.launchpad.net/launchpad | Channel logs: http://irclogs.ubuntu.com
=== mrevell changed the topic of #launchpad to: https://launchpad.net/ || PPA buildds at half capacity 08.00 - 17.00 UTC April 15 | Next meeting (all welcome): Thu 17 April 2008, 1800UTC #launchpad-meeting | Help: https://help.launchpad.net | Questions: https://answers.launchpad.net/launchpad | Channel logs: http://irclogs.ubuntu.com
[15:32] <ronny> yo
[15:33] <ronny> is there any way to import hg branches ?
[16:17] <kiko> ronny, not currently but we're looking for somebody interested in doing that work under contract
[16:19] <ronny> hmmk - we switched pida to hg
[16:20] <ronny> wanted to keep using the translation stuff
[16:21] <kiko> ronny, you can -- there's nothing that ties those two together right now in LP
[16:22] <kiko> you can continue uploading templates and translating them
[16:22] <ronny> ok
[16:23] <ronny> then i just missunderstood something
[16:26] <statik> ronny: you can use the bzr fastimport plugin to import from hg, I've done it for a few projects
[16:27] <statik> ronny: there is a hg-fast-export.py included with that plugin, then you run bzr fast-import, and boom! you have bzr branches
[16:27] <statik> works for git too
[16:27] <statik> ronny: what was pida using before hg? (just curious)
[16:27] <ronny> hmmk
[16:28] <statik> https://launchpad.net/bzr-fastimport
[16:28] <ronny> well - we had a bzr, then we moved back to svn for a reimplementation, now we switched to hg
[16:28] <ronny> imho bzr just sucks
[16:29] <ronny> tonns of modes but lack of speed
[16:29] <statik> ronny: which things did you find slow, which version? (I'm interested because I'd like to know if they are things we have already fixed or that still need fixing)
[16:30] <statik> I'm working with some pretty large projects on migrating to bazaar, and there have been amazing improvements in the last few months, but I'm always wanting to know if there are more things lurking
[16:31] <ronny> well for older bzr'S everything was just plain slow, for newers i only know that pull/push sucks, since i use it only for tracking
[16:33] <ronny> from my point of view bzr is just plain wrong
[16:36] <statik> ronny: ok :) if you have concrete complaints,  I'm very interested in helping get them addressed, but if it's just philosophy differences I tend to approve of the design goals that the bazaar team has made
[16:37] <statik> but either way I'm a huge advocate of interop between bzr, hg, and git
[16:39] <ronny> my complaint are simpely that it has different modes and each of those imposes different working models, git and hg got that right
[16:45] <synic> if I disable the blueprints for a project, shouldn't the "blueprints" tab go away?
[16:46] <gmb> synic: Ideally, yes. At the moment, though, it doesn't. I believe there's a bug about it somewhere.
[16:46]  * gmb looks
=== thekorn_ is now known as thekorn
[17:01] <MacYET> hi, is there a way to configure a LP bugtracker in a way that only members of the core team can change the workflow state?
[17:01] <gmb> MacYET: How do you mean? You want to lock the bug status field to be changed only by members of a given team?
[17:01] <MacYET> jep
[17:03] <gmb> MacYET: No, there's no functionality for that. How do you think it would help you?
[17:04] <MacYET> well, there is some guy changing tons of zope 2 tickets...and nobody knows this guy
[17:04] <MacYET> as release manager I am interested that only members of the zope 2 team are allowed to change the state
[17:19] <gmb> MacYET: Right, I can see your problem.
[17:20] <gmb> MacYET: It might be worth talking to a Launchpad admin about this to see if they can help you out any.
[17:20] <gmb> Let me see if I can find one for you.
[17:20] <MacYET> ok, tnx
[17:26] <gmb> MacYET: All the admins appear to be busy at the moment. I think the best way to proceed would be to file a question on http://answers.launchpad.net/launchpad
[17:26] <gmb> And then someone will get back to you.
[17:29] <MacYET> no need for the hurry, trying to ask the guy first, but in general I was surprised that this option is available to anyone
=== salgado_ is now known as salgado
=== matsubara is now known as matsubara-lunch
=== salgado_ is now known as salgado
[17:50] <ubotu> New bug: #217337 in malone "Bug supervisors shouldn't be compulsorily subscribed to all bug reports" [Undecided,New] https://launchpad.net/bugs/217337
=== kiko is now known as kiko-fud
=== salgado is now known as salgado-lunch
[17:56] <mc__> How can I get a specific bug as RDF?
=== matsubara-lunch is now known as matsubara
=== salgado-lunch is now known as salgado
[18:47] <mc__> What would be the best way to measure the memory consumption of a programm? I need to track down what uses how much memory
[18:57] <mc__> nevermind, wrong channel
=== kiko-fud is now known as kiko
[19:12] <mohi> hi :)
[19:12] <mohi> I have a question abour PPA in launchpad. can I ask here?
[19:12] <mohi> about*
=== mc__ is now known as pummeluf
=== pummeluf is now known as mc__
[19:24] <beuno> mohi, sure, this is the place. What's the problem?
[19:26] <nand> hi!
[19:26] <mohi> beuno: I use an application that is not in the ubuntu repositories. can I make a .deb from this and publlish it to my PPA in launchpad despite I hadn't made this software?
[19:27] <beuno> mohi, yeap, as long as it's free software, sure
[19:28] <beuno> PPA actually makes the deb for you, so you just upload the source code
[19:28] <mohi> i have the rpm! I've builded deb from rpm.
[19:28] <mohi> beuno: ^
[19:30] <nand> I have asked for a LP mailing list. Unfortunately, since a few days, I get "This team's mailing list will be available within a few minutes." Not sure if this is related to bug 215118 ?
[19:30] <ubotu> Launchpad bug 215118 in launchpad "Odd mailing list resync behavior" [High,Fix released] https://launchpad.net/bugs/215118 - Assigned to Barry Warsaw (barry)
[19:31] <ubotu> New bug: #217374 in launchpad "OOPS re-activating an inactive milestone" [Undecided,New] https://launchpad.net/bugs/217374
[19:31] <beuno> mohi, well, you can't currently upload binaries to PPAs, so you really need the source code and get the packaging done
[19:35] <mohi> beuno: so I can't upload only the .deb without source!?
[19:35] <beuno> mohi, not to PPA, no
[19:35] <mohi> ok. ty beuno :)
[19:35] <beuno> you can upload it to the projects home page
[19:35] <beuno> LP let's you upload files
[19:35] <beuno> but to be able to use apt, you need to upload the source code, and it compiles server-side
[19:36] <mohi> aha... 
[19:43] <LaserJock> BjornT_: ping
[19:46] <BjornT_> LaserJock: hi
[19:55] <mc__> How can I get a specific bug as RDF?
[19:56] <ubotu> New bug: #217390 in launchpad-bazaar "A branch owner should be able to delete the branch also if someone is subscribed" [Undecided,New] https://launchpad.net/bugs/217390
[19:57] <BjornT_> mc__: you can't. you can get a text representation of the bug by appending +text to the url. e.g. https://launchpad.net/bugs/217390/+text
[19:57] <ubotu> Launchpad bug 217390 in launchpad-bazaar "A branch owner should be able to delete the branch also if someone is subscribed" [Undecided,New] 
[19:58] <LaserJock> BjornT_: regarding bug #217337
[19:58] <ubotu> Launchpad bug 217337 in malone "Bug supervisors shouldn't be compulsorily subscribed to all bug reports" [Undecided,New] https://launchpad.net/bugs/217337
[19:58] <LaserJock> BjornT_: I'm kinda confused about what you just said there
[19:58] <LaserJock> Bug supervisors aren't implicitly subscribed to bugs?
[19:59] <BjornT_> LaserJock: right, the aren't implicitly subscribed. when setting the bug supervisor, we automatically create an subscription. that subscription can be removed, though.
[19:59] <LaserJock> and how do I tell the difference between bug supervisors an package subscribers?
[19:59] <LaserJock> because right now people are implicity subscribed
[20:00] <BjornT_> LaserJock: in which context?
[20:01] <LaserJock> when people sign up under "Subscribe to bug mail" for a package
[20:01] <LaserJock> then they are implicitly subscribed to new bugs, correct?
[20:01] <BjornT_> right
[20:02] <LaserJock> then what the heck is this bug supervisor stuff
[20:02] <LaserJock> I can't find it anywhere
[20:03] <LaserJock> do bug supervisors actually exist yet?
[20:04] <BjornT_> LaserJock: well, we don't have bug supervisors for packages yet. by being a bug supervisor you get extra permission, you can set importance, and certain statuses.
[20:05] <BjornT_> LaserJock: we used to call package subscribers for 'bug contacts', but that was wrong. they were only subscribed to bug mail, they didn't have the permissions like all the other 'bug contacts' had.
[20:07] <LaserJock> ok
[20:07] <LaserJock> hmm
[20:08] <LaserJock> any ETA on bug supervisors for packages?
[20:08] <LaserJock> that would fix a rather annoying bug for us
[20:08] <kiko> LaserJock, I spoke to BjornT_ about this -- we have a plan
[20:09] <BjornT_> kiko: did we talk about a time-frame, though?
[20:10] <kiko> BjornT_, no, we didn't -- we need to be opportunistic about doing it, and if we are able to, great
[20:11] <LaserJock> cool
[20:11] <BjornT_> yeah. it won't be done for this milestone. we'll see how the next milestone looks like.
[20:12] <LaserJock> how would people/teams become bug supervisors then?
[20:16] <BjornT_> only certain people would have permission to add a bug supervisor. you'd have to ask those people. i'm not yet sure who those people will be.
[20:17] <LaserJock> hmm, that *could* get a little messy
[20:17] <LaserJock> but I'm not sure how else you'd do it
[20:18] <LaserJock> first thing I'm gonna want to do is set up teams as bug supervisor on like 500 packages
[20:19] <LaserJock> I can't imagine just one or two people would want to have to deal with that
[20:20] <LaserJock> and I'm not particularly looking forward to spending a couple hours clicking in LP to sign up either
[20:20] <BjornT_> well, if it's that many, maybe it makes sense to try to join the ubuntu-bug-control team, so you'd be bug supervisor for all packages?
[20:20] <LaserJock> I'm in that team
[20:20] <LaserJock> we want subscription
[20:21] <LaserJock> not permissions
[20:21] <LaserJock> per se
[20:21] <BjornT_> well, you can already subscribe to a package's bugs
[20:21] <LaserJock> that's what we already have
[20:21] <LaserJock> but we end up implicitly subscribed
[20:22] <LaserJock> when I want explicit subscriptions that Bug Supervisor provides
[20:22] <kiko> no
[20:22] <LaserJock> and really Bug Supervisor is the more accurate term
[20:22] <kiko> LaserJock, you're confused. bug supervisor doesn't provide explicit subscriptions.
[20:22] <LaserJock> kiko: BjornT_ said it does
[20:22] <kiko> bug supervisor doesn't imply any subscription, really
[20:23] <kiko> LaserJock, bug supervisor has permissions to change protected fields of a bugtask
[20:23] <BjornT_> no i didn't. at least i didn't mean to say it :)
[20:23] <LaserJock> bah
[20:23] <LaserJock> so what is bug #217337 about then?
[20:23] <ubotu> Launchpad bug 217337 in malone "Bug supervisors shouldn't be compulsorily subscribed to all bug reports" [Undecided,New] https://launchpad.net/bugs/217337
[20:23] <kiko> LaserJock, it's just as a convenience that we default bug supervisors to also being pillar subscribers
[20:24] <kiko> but that can be disabled
[20:24] <kiko> BjornT_, we need to use this term "pillar subscriber" or something like that to differentiate that to "bug subscriber", I think
[20:24] <LaserJock> I see
[20:24] <kiko> otherwise it's also very confusing
[20:24] <LaserJock> well, how does somebody become a pillar subscriber?
[20:24] <kiko> LaserJock, just "subscribe to bugs" for a project, distro or package.
[20:25] <LaserJock> but that gives me implicit subscription for packages
[20:25] <kiko> correct.
[20:25] <LaserJock> :s
[20:25] <kiko> LaserJock, I suspect the only reason you're chasing explicit subscription is because you want to unsubscribe to something.
[20:26] <LaserJock> yeah
[20:26] <LaserJock> is that wrong to want? :-)
[20:26] <kiko> why don't you say that instead of going through this very convoluted path? :)
[20:26] <LaserJock> because that what bug #217337 looked like to me
[20:26] <ubotu> Launchpad bug 217337 in malone "Bug supervisors shouldn't be compulsorily subscribed to all bug reports" [Undecided,New] https://launchpad.net/bugs/217337
[20:27] <LaserJock> looked perfect
[20:27] <kiko> that summary is confusing, and is that even a valid bug?
[20:27] <LaserJock> I want my teams to be Bug Supervisors and have explicit subscriptions or implicit subscription I can get out of
[20:27] <LaserJock> that's all I'm asking for :-)
[20:28] <kiko> it's more the latter
[20:28] <LaserJock> either way works for me
[20:29] <LaserJock> I can't tell the difference between explicit and implicit subscriptions other than one I can unsubscribe
[20:29] <LaserJock> obviously for you guys there is a difference
[20:29] <kiko> LaserJock, the difference has to do with what we do when bug data changes.
[20:29] <kiko> i.e.
[20:29] <kiko> I'm subscribed to bugs in alsa-utils
[20:30] <kiko> a bug is reported against the kernel
[20:30] <kiko> the bugtask is changed from linux to alsa-utils
[20:30] <kiko> if subscription is explicit, the kernel team stays subscribed and we need to specially subscribe alsa-utils
[20:30] <mc__> anyone op in here? there is a typo in the topic "buildds" two "d"'s
[20:30] <kiko> or else complicate the UI to make you decide
[20:30] <kiko> mc__, it's actually "build daemons"
[20:31] <mc__> oh, alright :)
[20:33] <LaserJock> wait a sec, you can change the package on a bug task?
[20:33]  * LaserJock goes to look
[20:34] <LaserJock> what the heck, why do people keep creating tons of new bug tasks for?
[20:36] <LaserJock> kiko: alright, well I guess I'll crawl back under my rock again
[20:37] <LaserJock> I just got excited that maybe the bug supervisor thing would fix the "can't unsub from implicit subscriptions" thing
[20:37] <kiko> LaserJock, how many bugs would you unsubscribe from?
[20:38] <LaserJock> oh, not all that many
[20:38] <kiko> LaserJock, how many?
[20:38] <LaserJock> it's sort of one of those, you don't need it often but when you need it you really need it things
[20:38] <LaserJock> oh probably 10 or so a release, for me personally
[20:39] <kiko> LaserJock, just echo "bug-id" >> ~/.ignored-bugs
[20:39] <kiko> and use that in your procmail filter
[20:39] <LaserJock> it's not about me
[20:39] <LaserJock> I don't mind it too much
[20:39] <LaserJock> it's about team members and upstream leaving
[20:39] <LaserJock> which I'd like to avoid if I can
[20:41] <LaserJock> people seem to not like getting automatic email that is irrelevent to them
[20:41] <kiko> leaving because of bugmail. sounds like a red herring. :)
[20:41] <LaserJock> and when they can't get out, that's even worse
[20:41] <LaserJock> no, I'm serious
[20:41] <LaserJock> I had people get like 100 emails in a day so they left the team, they'll probably come back now that it's over
[20:42] <kiko> why did they get 100 emails in a day?
[20:42] <LaserJock> because we have a bug with I think 50+ tasks
[20:42] <LaserJock> and each time anybody did anything it sent an email
[20:42] <kiko> really?
[20:42] <kiko> LaserJock, but people shouldn't receive multiple emails for that bug
[20:42] <LaserJock> https://bugs.edge.launchpad.net/ubuntu/+source/mopac7/+bug/201962
[20:42] <ubotu> Launchpad bug 201962 in lapack3 "gfortran transition" [Low,Fix released] 
[20:43] <kiko> it's just one bugmail per 5 minutes of changes
[20:43] <kiko> LaserJock, hang on, I need to reboot, will be 4 minutes
[20:49] <LaserJock> kiko: back?
[20:50] <kiko> yep
[20:50] <kiko> what's the story?
[20:50] <LaserJock> ok, so on that bug there are 54 tasks and 90+ comments
[20:50] <LaserJock> that adds up to a lot of bugmail
[20:51] <LaserJock> I guess I'm struggling with whether we should move process bugs off of Launchpad or not
[20:51] <LaserJock> it seems like Launchpad isn't designed around these kinds of bugs
[20:53] <kiko> LaserJock, I'm not sure how those two thoughts equate. unsubscribing implicit subscribers isn't possible, granted, but there's an easy workaround (filtering or the D key) and regular bugs and process bugs are similar in bug volume.
[20:53] <kiko> LaserJock, I also don't quite understand why the number of bugtasks would generate more bugmail
[20:53] <kiko> unfortunately I need to reboot again, so bear with me.
[20:53] <LaserJock> np
=== matsubara is now known as matsubara-afk
[21:04] <LaserJock> kiko: well, more bugtasks means more irrelevent email for a lot of bug contacts
[21:04] <kiko> LaserJock, they'd still have to unsubscribe manually, which isn't fun
[21:05] <kiko> LaserJock, maybe we should be asking ourselves why we're sending them bugmail at all?
[21:05] <LaserJock> well, actually I think that'd be pretty common
[21:05] <kiko> I mean, could we avoid somehow emailing them?
[21:05] <LaserJock> people want to get some bug mail, then be able to turn them off
[21:05] <kiko> what is the nature of the change being bugmailed?
[21:05] <LaserJock> well, people want relevent email, not irrelevent email :-)
[21:06] <kiko> so maybe we should start by analyzing why the email is irrelevant
[21:06] <kiko> is it me-tooing?
[21:06] <LaserJock> no
[21:06] <LaserJock> just involving many different packages
[21:06] <LaserJock> people don't mind getting the email for the package they are a bug contact for
[21:06] <LaserJock> but if that's like 1/10 of the bugmail they get from the bug report, they start complaining
[21:07] <kiko> maybe the problem is multiple bugtasks indeed
[21:07] <kiko> I'm not sure that's a feature
[21:08] <LaserJock> that's why I'm wondering if we should move those kinds of bugs off LP
[21:08] <LaserJock> if it's sort of a corner case
[21:08] <LaserJock> the other issue is that there are  different types of bugs
[21:08] <LaserJock> basically code bugs and process bugs
[21:09] <LaserJock> most subscribers are intersted in the code bugs
[21:09] <LaserJock> but process bugs can produce a lot of traffic and they have no way to get out
[21:11] <LaserJock> one suggestion that has the most potential, IMO is to ubsubscribe implicit subscribers if the task they are subscribed through is marked Invalid or Fix Released
[21:11] <LaserJock> however some people have said that they only would want that for Invalid
[21:12] <LaserJock> so I'm not sure if it would work without have some option, which I know you wouldn't want to do
[21:13] <kiko> at that point I think it's more generally useful to "block bugmail from this bug"
[21:14] <LaserJock> so do you see the issue or do you still think I'm on crack? ;-)
[21:35] <ubotu> New bug: #217427 in soyuz "Please support arbitrary arch/buildd affinity for arch:all builds" [Undecided,Confirmed] https://launchpad.net/bugs/217427
[21:47] <chx> hi. https://launchpad.net/drupal/main Last run:	Failure eight hours ago
[21:47] <chx> thanks if you can fix :)
[21:52] <kiko> chx, let me check what the error was
[21:52] <chx> kiko: <3 you
[21:54] <kiko> bzrlib.errors.SocketConnectionError: Unable to connect to SSH host escudero; EOF during negotiation
[21:55] <kiko> mwhudson, what do you think of that?
[21:56] <chx> kiko: can this be a temporary error? that server not answering..? canyou try rerun?
[21:58] <kiko> chx, I've retried, but I dunno
[22:02] <chx> mwhudson_: hey
[22:02] <mwhudson_> chx: hi?
=== mwhudson_ is now known as mwhudson
[22:03] <kiko> mwhudson_, drupal isn't syncing, and I'm seeing a lot of weird problems in the buildbot logs 
[22:03] <kiko> <kiko> bzrlib.errors.SocketConnectionError: Unable to connect to SSH host escudero; EOF during negotiation
[22:03] <mwhudson> kiko: hm
[22:03] <mwhudson> kiko: load on one of the buildslaves is 100+, this seems to be unlikely to be a coincidence
[22:04] <mwhudson> well, was
[22:04] <kiko> hmph
=== matsubara-afk is now known as matsubara
[22:15] <chx> mwhudson: so what can be done so that the Drupal import synces?
[22:15] <mwhudson> chx: it should be fixed now, let me give it a manual kick
[22:17] <mwhudson> chx: it should 'just happen' now, the system is working through a backlog and i don't have any control over the order it does things in :/
[22:17] <chx> mwhudson: my problem is that http://drupal.org/cvs?commit=110720 went in and I would prefer something better than the merge abilities of CVS ...
[22:17] <mwhudson> chx: it should be synced in a few hours
[22:17] <chx> mwhudson: great, thanks
[22:18] <mwhudson> i hope that's not too inconvenient
[22:18] <mwhudson> (when we've finished rewriting the innards of the system, i'd be able to prioritize it...)
[22:21] <chx> mwhudson: it is a bit, but i guess it'll be fine
[22:22] <chx> mwhudson: Murphy's law says that when a 850K patch lands then bzr mirror fails :)
[22:22] <mwhudson> chx: well, at least we know of the problems
[22:23] <mwhudson> and i've spent the last two weeks doing nothing other than working on the new system...
[22:23] <chx> mwhudson: awesome, thanks
=== fta_ is now known as fta
[23:15] <ubotu> New bug: #217472 in launchpad "MMUnknownListError when deactivating a missing list" [Medium,Confirmed] https://launchpad.net/bugs/217472
[23:48] <fta> on https://edge.launchpad.net/distros/ubuntu/hardy/+source/xorg-server/2:1.4.1~git20080131-1ubuntu8
[23:48] <fta> Redirect Loop
[23:48] <fta> Firefox has detected that the server is redirecting the request for this address in a way that will never complete.