[00:05] <TDJACR> How does launchpad implement bazaar commit control?
[00:06] <lifeless> each user has their own file space
[00:06] <TDJACR> I mean, how does bzr+ssh:// only allow you to modify certain branches and stuff
[00:07] <wgrant> It's a custom SSH server.
[00:07] <TDJACR> Oh
[00:07] <TDJACR> Is is, like launchpad, F/OSS?
[00:07] <wgrant> It is
[00:07] <spiv> Yes.
[00:07] <TDJACR> Cool.
[00:07] <wgrant> See lib/lp/codehosting in the LP tree.
[00:08] <TDJACR> I was wondering how bzr+ssh limited what a user could do with ssh.
[00:08] <spiv> Although probably Launchpad could probably use OpenSSH if it didn't need to support sftp and using the Launchpad DB to get the public key(s) for a user.
[00:09] <spiv> Because apart from those things the bulk of the access control is done in a bzr plugin.
[00:09] <TDJACR> Which?
[00:10] <spiv> A custom one :)  It's in Launchpad's source somewhere.
[00:10] <TDJACR> For example, if I only wanted jim and bob to be able to commit to my server, but not execute any other command, how would I do that?
[00:11] <spiv> With OpenSSH, you can create an authorized_keys file on the server that restricts a user that auths with a particular key to only running a particular command.
[00:11] <TDJACR> spiv: So would I make that comman bzr?
[00:11] <spiv> There's even a helper script in bzr's source tree for that, just a sec.
[00:12] <spiv> http://bazaar.launchpad.net/~bzr-pqm/bzr/bzr.dev/annotate/head%3A/contrib/bzr_ssh_path_limiter
[00:12] <TDJACR> spiv: Thanks so much!
[00:12] <spiv> (there's also a much more complex bzr_access script in that contrib directory)
[00:13] <TDJACR> spiv: But that one would make a protected SSH bzr smart server, correcy?
[00:13] <TDJACR> t
[00:13] <spiv> Depending on what you mean by "protected", yes.
[00:14] <TDJACR> spiv: Users can't log on and execute irssi
[00:14] <spiv> Right.
[00:14] <spiv> Assuming they authenticate to that account with that key.
[00:14] <TDJACR> spiv: How would I control which branches they may  commit to
[00:14] <spiv> So you'd want to turn off password auth, or make the account have no password.
[00:15] <spiv> Use filesytem permissions.
[00:15] <TDJACR> spiv: So I ocould add the user to a group.
[00:15] <TDJACR> Sounds good.
[00:15] <TDJACR> Thanks so much
[00:15] <spiv> And/or use the argument that bzr_ssh_path_limiter takes
[00:16] <spiv> In the example it gives, command="bzr_ssh_path_limiter /home/foo/code" ..., the client won't be able to access any bzr data outside /home/foo/code
[00:17] <TDJACR> spiv: But I could make two branches in it, correct?
[00:17] <TDJACR> And limit them with fs permissions?
[00:18] <spiv> Right.
[00:18] <TDJACR> Sounds awesome, thank you.
[01:57] <bp0> hello
[01:58] <bp0> what is the difference between https://bugs.launchpad.net/ubuntu/+source/tvtime and https://bugs.launchpad.net/tvtime
[01:59] <micahg> bp0: one is the upstream project and one is the ubuntu package
[01:59] <bp0> how are they used differently
[02:00] <bp0> what is different about the bugs that are reported for each one
[02:01] <bp0> ?
[02:05] <micahg> bp0: bugs for the ubuntu package are reported against the package
[02:06] <micahg> tvtime actually uses sourceforge for bug tracking
[02:06] <micahg> but the launchpad project would list bugs linked to sourceforge
[02:08] <bp0> so where should that bug about ekiga killing sound in tvtime be reported?
[02:09] <micahg> it needs to be triaged in Ubuntu most likely
[02:10] <micahg> bp0: did you add all those tasks?
[02:10] <micahg> and are you experiencing the issue?
[02:11] <bp0> yeah
[02:11] <bp0> i was trying to get it listed with the others
[02:11] <micahg> bp0: are you using Ubuntu?
[02:11] <bp0> now, i think i've figured out how
[02:11] <bp0> not this moment
[02:13] <micahg> well, if you're not using Ubuntu, then sourceforge would be the place to file your bug
[02:13] <micahg> with tvtime
[02:14] <bp0> oh, yeah, i am using ubuntu
[02:14] <bp0> but not this moment
[02:14] <bp0> so i couldnt test anything right now if you asked
[02:14] <micahg> ok, then I suggest hoping in #ubuntu-bugs sometime tomorrow and hopefully someone can triage it for you
[02:14] <micahg> *hopping
[02:15] <micahg> I don't know enough about tvtime to do it right now
[03:16] <rasiq> bug notify - urgent - https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/460121
[03:18] <micahg> rasiq: the place for this would be #ubuntu-bugs
[03:19] <rasiq> thanks micahg..
[03:19] <rasiq> will post there.. bye..
[03:40] <lifeless> hmm
[03:41] <lifeless> LP apis seem to write to world readable dirs their credentials
[03:41] <lifeless> [sigh]
[04:14] <TDJACR> How does launchpad automatically add pushes to the working tree?
[04:19] <spiv> TDJACR: it doesn't; it runs loggerhead
[04:20] <spiv> (which is a web viewer for bzr branches)
[04:34] <meoblast001> hi
[04:34] <meoblast001> how long is the downtime expected to be for?
[04:35] <meoblast001> is this a major server change, or just a minor one
[04:35] <spm> meoblast001: heh. timing. Shortly hopefully.
[04:35] <meoblast001> ok, i have some code i was going to push, and i'm getting tired
[04:35] <meoblast001> i was wondering if i should wait or go to bed and push tomorrow
[04:36] <meoblast001> well, i do have a small amount of studying to do, so i will be awake for probably 15 more minutes anyways
[04:47] <meoblast001> spm: it's back up
[04:47] <meoblast001> at least the bazaar server is
[06:12] <al-maisan> Good morning
[06:12] <motionman> good morning :)
[07:53] <mdke> when trying to set ubuntu-doc@lists.ubuntu.com as the contact address for ~ubuntu-doc, I get this error: "ubuntu-doc@lists.ubuntu.com is already registered in Launchpad and is associated with Registry Administrators." Any ideas?
[07:54]  * spiv wonders if kfogel really is the help contact still...
[07:56] <wgrant> mdke: Sounds like it might have been registered to a team that was merged away.
[07:56] <mdke> wgrant: could be
[07:57] <mdke> can't you have more than one team with the same contact address?
[07:57] <wgrant> mdke: I guess your best bet is to ask a Question to have it removed or transferred.
[07:57] <lifeless> mdke: you'll need to file a request for the admins
[07:57] <wgrant> No.
[07:57] <lifeless> mdke: email addresses have to be unique
[07:57] <mdke> that's a bit odd, I think
[07:57] <lifeless> mdke: also note that having a contact address means that users preferences for contacting, bug mail etc, all stop working.
[07:57] <lifeless> generally, I think its a terrible idea to set an external list as the contact details for an LPteam
[07:58] <wgrant> I think the current implementation of the team contact address idea is hugely flawed.
[07:58] <wgrant> And should be avoided wherever possible.
[07:58] <mdke> a good alternative for us would be to disable the "Contact this team" button
[07:59] <mdke> it's a PITA for all team members to get those emails individually rather than through the mailing list
[07:59] <wgrant> Note that that button is only available to team members.
[07:59] <mdke> we don't use that team for bugmail or branch subscriptions
[07:59] <wgrant> But it's still crazy that it's not disablable.
[07:59] <mdke> well, we get quite a few email from individuals interested in participating
[08:00] <mdke> actually I don't think they are all members
[08:00] <wgrant> They can only contact the admins, IIRC.
[08:01] <mdke> we had an email recently from ~efsc70, who isn't a member, and I think the whole team got it
[08:02] <wgrant> I only get a "Contact this team's owner" link.
[08:02] <mdke> ok, that one must have only gone to me then, I guess. it didn't say that in the email
[08:02] <mdke> the team gets them too sometimes though, I suspect it is people who join the open subteam, and then use it
[08:03] <mdke> it really would make sense for such emails to go to the mailing list, I feel
[08:03] <mdke> Launchpad even suggests using a mailing list in the interface (although it's referring to a LP mailing list)
[08:04] <mdke> anyway, for the original problem, do I open a question on the launchpad project?
[08:05] <wgrant> You do.
[08:05] <mdke> ok
[08:42] <lifeless> mwhudson: you wouldn't happen to know when merge requests via mail will work again?
[08:55] <slangasek> I wonder if it would be possible to prevent users from changing the state of the tasks on bug #368497, or lock it to a group somehow.
[08:59] <wgrant> slangasek: There is no such functionality.
[08:59] <slangasek> wgrant: then I wonder if the accounts of the people who are playing bug volleyball could be locked
[09:00] <wgrant> Bug #73122
[09:00] <lifeless> wow
[09:02]  * wgrant returns to messing with ddeb domination logic.
[09:02] <wgrant> Odd that that bug has been hit so many times.
[09:02] <lifeless> I think there is a good reason to have final states
[09:03] <lifeless> and a less crowd-approach for bugs
[09:03] <slangasek> wgrant: odd, but clearly not coincidental.  The bug was also given a nickname at one point, which I removed, hoping that would stem the flow of pointless state changes; someone, somewhere on the Internet, is telling other people this bug should be reopened
[09:04] <lifeless> have you googled for it ?
[09:04] <slangasek> yes
[09:04] <slangasek> didn't find anything except LP itself
[09:04] <wgrant> slangasek: But they weren't directly reopened.
[09:04] <slangasek> wgrant: how do you mean/
[09:04] <wgrant> Oh. Different tasks.
[09:04] <wgrant> I just saw lots of transitions in each block, which tended to indicate idiocy rather than malice.
[09:05] <wgrant> But I see now that they were for different tasks.
[09:05] <slangasek> I guess alternatively, I could unsubscribe ubuntu-sru from the bug and make it Not My Problem, but I figured I should at least /ask/ before abandoning the bug to chaos
[09:06] <slangasek> wgrant: I'm pretty sure it is idiocy rather than malice; if it were malice, why would they all be focused on a single bug instead of reopening bugs at random :)
[09:07] <wgrant> slangasek: By idiocy I meant accidental state changing, which does happen sometimes.
[09:07] <wgrant> But in that case it should hit other bugs.
[09:12] <micahg> wgrant: how often are merged branches hidden?
[09:12] <micahg> is it supposed to be instantaneous?
[09:13] <wgrant> micahg: As soon as the branch is scanned (that is within a couple of minutes, when the new revisions show up in LP)
[09:18] <micahg> wgrant: this branch isn;t showing as merged: https://code.edge.launchpad.net/~micahg/prism/prism-1.0b2-karmic
[09:20] <micahg> or rather, it's not being hidden
[09:20] <wgrant> micahg: Ah, this mess...
[09:21] <wgrant> micahg: A branch's status is set to Merged automatically only when it is merged into a series branch.
[09:21] <wgrant> micahg: There are no series branches in that project.
[09:21] <micahg> ah
[09:21] <micahg> so I need to manually set it
[09:21] <wgrant> Right.
[09:21] <wgrant> This is unfortunate.
[09:22] <micahg> ok, I'll talk to someone about creating a series for it
[09:22] <micahg> should I file a bug?
[09:22] <micahg> for it not being marked merged
[09:22] <wgrant> It's not a bug.
[09:22] <wgrant> It wasn't merged, in that sense.
[09:22] <wgrant> So it should not be marked merge.
[09:22] <wgrant> +d
[09:22] <micahg> if you say so :)
[09:23] <wgrant> I believe people are rethinking how all that works.
[09:27] <micahg> does lp linking branches only work for official branches with bug #s?
[09:28] <wgrant> LP will automatically link any branch to bugs that are referenced properly (that is, a 'bzr ci --fixes lp:123456', not just a textual reference in a commit message).
[09:28] <micahg> ah
[09:28] <micahg> that's what I need to do
[09:29] <micahg> and if it's multiple bugs?
[09:30] <TDJACR> How does launchpad add all pushes to the mainline?
[09:30] <wgrant> micahg: I think something like '--fixes lp:123456 --fixes lp:234567'
[09:30] <TDJACR> Er, working tree
[09:30] <wgrant> TDJACR: How do you mean?
[09:30] <wgrant> LP doesn't keep working trees.
[09:30] <wgrant> As was explained this morning, it uses Loggerhead to provide a web UI to the branches.
[09:31] <TDJACR> Ah. How does it keep all pushes as the "working tree"
[09:32] <wgrant> I still don't understand what you mean.
[09:32] <wgrant> Lp doesn't deal with working trees.
[09:32] <TDJACR> When I bzr push ./ lp:myproj How does it become the files, in the directory.
[09:33] <spiv> TDJACR: you mean how is it that you can view the files in the branch via the web site?
[09:33] <TDJACR> If I push to my own server, I need to ssh into it and bzr update it
[09:33] <spiv> TDJACR: it uses loggerhead (https://launchpad.net/loggerhead0
[09:33] <wgrant> TDJACR: Loggerhead is a web application which looks at the branch (not the working tree) and displays the file listings and gives you files. They don't exist in a working tree on disk.
[09:33] <spiv> TDJACR: launchpad doesn't host any working trees
[09:34] <TDJACR> spiv: So, If I branched off of my server, I'd still get a working tree, but my server wouldn't have one?
[09:35] <spiv> TDJACR: yes
[09:36] <spiv> TDJACR: a working tree is where you can use a text editor etc to make changes that you can then "bzr commit".
[09:36] <TDJACR> When I push to my server I get This transport does not update the working tree of: bzr+ssh://bzr@....
[09:36] <TDJACR> But not when I push to LP
[09:37] <spiv> Launchpad is configured to not create working trees
[09:37] <lifeless> spiv: there is also that bug
[09:37] <wgrant> It will give you that message only if there is an existing working tree on the other end, IIRC.
[09:37] <lifeless> with a spurious message
[09:37] <spiv> You can create a shared repository on your server that does the same thing (bzr init-repo --no-trees)
[09:37] <TDJACR> spiv: I'll try it.
[09:37] <TDJACR> Thanks.
[09:38] <TDJACR> Now, why doesn't branch remember the location when I push?
[09:38] <spiv> And you can remove an existing, unwanted working tree with "bzr remove-tree" (run on the server in this case)
[09:38] <spiv> It should, unless it has already remembered a different location.
[09:39] <spiv> If you need to override an existing remembered location you can do "bzr push --remember your_location"
[09:40] <TDJACR> Okay, cool.
[09:42] <TDJACR> What is the purpose of having a tree?
[09:47] <spiv> It's the place where you edit files.
[09:47] <spiv> And "bzr commit" those edits.
[09:48] <spiv> Also called a checkout or a working copy.
[09:51] <TDJACR> Let me refine that, why would I have a tree on my server? The pushes wouldn't appear as files.
[10:24] <oSoMoN> hi all
[10:25] <oSoMoN> I've requested a one time import of translation templates in rosetta, and the files are markes as uploaded by someone else in the import queue
[10:25] <oSoMoN> and this someone else (who used to be responsible for the translations on this project) received mails to confirm the upload
[10:25] <henninge_> oSoMoN: the branch owner, I think
[10:26] <oSoMoN> doesn't seem so, the branch owner is a team
[10:26] <oSoMoN> and this person is not part of the team anymore
[10:27] <oSoMoN> could that be a bug in rosetta, or some setting I need to change somewhere?
[10:28] <henninge_> oSoMoN: let me check
[10:28] <oSoMoN> thx
[10:31] <henninge_> oSoMoN: see the code ;) http://paste.ubuntu.com/301941/
[10:32] <oSoMoN> pretty clear indeed
[10:32] <henninge_> oSoMoN: so templates is the series driver or the series owner
[10:32] <henninge_> *for templates*
[10:33] <oSoMoN> but where do I set the series driver?
[10:33] <henninge_> oSoMoN: on the series home page, I am pretty sure
[10:34] <oSoMoN> I don't seem to be allowed to set it (although I'm admin of the project, and it's already set to the team anyway, not this person)
[10:34] <oSoMoN> see https://launchpad.net/elisa/1.0
[10:40] <henninge_> oSoMoN: this *is* a bit strange, I don't see how Philippe Normand comes into play at all ...
[10:41] <oSoMoN> me neither!
[10:41] <henninge_> oSoMoN: can you please file a bug about it so we investigate the matter further?
[10:41] <oSoMoN> he used to be part of the elisa-developers team and in charge of translations
[10:41] <oSoMoN> sure
[10:41] <oSoMoN> what project on LP should I file it against?
[10:41] <henninge_> oSoMoN: Please add that information to the bug report
[10:42] <henninge_> oSoMoN: rosetta
[10:51] <slangasek> nice, bug #368497 reopened again
[10:52] <wgrant> Wow.
[10:55] <slangasek> uhm, interesting correlation
[10:55] <slangasek> each of the last three accounts that reopened the bug was created on the day they did so
[10:56] <oSoMoN> henninge: I just checked using launchpadlib, and Philippe Normand is the series owner
[10:56] <oSoMoN> the thing is, I cannot find where to change this
[10:56] <henninge> oSoMoN: right, so the owner does not show in the UI any more?
[10:57] <oSoMoN> no, it doesn't
[10:57] <henninge> oSoMoN: can you search if there is a bug this? it should be in launchpad-registry
[10:57] <oSoMoN> sure
[10:58] <henninge> oSoMoN: I always thought Maintainer was another word for Owner ...
[10:59]  * slangasek finds three more accounts listed on there that were created within a week of meddling with that bug state, nice
[11:01] <oSoMoN> henninge: I can't find a bug for it, will file one
[11:02] <henninge> oSoMoN: please do.
[11:03] <henninge> oSoMoN: Maybe the owner is meant to be obosoleted but than the translation import code should not use it, either.
[11:03] <oSoMoN> I'll include this info in the bug report
[11:03] <henninge> oSoMoN: thanks
[11:04] <henninge> oSoMoN: please subscribe me to the bug
[11:04] <oSoMoN> sure
[11:13] <oSoMoN> henninge: https://bugs.edge.launchpad.net/launchpad-registry/+bug/460980
[11:14] <oSoMoN> note that I manage to set myself as the series' owner using launchpadlib
[11:15] <henninge> oSoMoN: thank you. good job!
[11:16] <oSoMoN> thanks to you
[12:11] <Laney> Can I get a PPA which builds on armel/sparc please?
[12:11] <Laney> I need to test some FTBFS fixes
[14:57] <simon-o> Hi, is this error known? http://bazaar.launchpad.net/~simono/ubuntu/karmic/gnoemoe/fixes-bug-459164/revision/6
[15:01] <PhilDick> I've reported a few bugs in Launchpad for ubuntu.  Is there an easy/convenient way to get a list of bugs I've created or commented on?
[15:05] <thekorn> PhilDick: https://bugs.edge.launchpad.net/people/+me/+reportedbugs
[15:06] <thekorn> and .../+commentedbugs
[15:27] <kfogel> launchpadlib experts: what am I doing wrong in this?  project.getBranches(status="any")
[15:28] <kfogel> it gets a HTTP 400 "Bad Request"
[15:28] <kfogel> I've tried it without the status= too.
[15:28] <kfogel> (that is, with just "any")
[15:31] <bodhi_zazen> What is the best way to contact a LP admin ?
[15:32] <jml> bodhi_zazen, ask a question on answers.launchpad.net/launchpad
[15:32] <jml> bodhi_zazen, or just ask your question here.
[15:33] <bodhi_zazen> jml: done, but I did not get an answer
[15:33] <bodhi_zazen> https://answers.launchpad.net/launchpad/+question/85249
[15:34] <geser> kfogel: I just noticed the API doc got now the toc, finally :)
[15:34] <jml> bodhi_zazen, our admins aren't always able to answer questions as quickly as they'd like.
[15:34] <jml> bodhi_zazen, this week they are unusually busy helping with the Ubuntu release.
[15:35] <bodhi_zazen> I understand
[15:36] <bodhi_zazen> my request is over 2 weeks old =)
[15:37] <kfogel> geser: yeah, saw that.  \o/
[15:37] <jml> kfogel, o hai
[15:38] <jml> kfogel, does it work with, say, status='merged'?
[15:39] <kfogel> jml: it's working now with no args (which means "any active status"), and that's fine for my purposes.
[15:40] <jml> kfogel, ok.
[15:40] <jml> kfogel, I think that the API docs are terrible wrt enums.
[15:41] <kfogel> jml: I haven't used them enough yet to form an opinion, but I'll take that as a warning.
[15:41] <jml> kfogel, well, you just bumped into that as a problem. :)
[15:42] <thekorn> kfogel: you can also catch the HTTPError object and print error.content, this should give you all possible values
[15:42] <thekorn> in this case: status: Invalid value "any". Acceptable values are: Experimental, Development, Mature, Merged, Abandoned
[15:43] <kfogel> thekorn: thank you for that tip!
[15:47] <jml> also (and I'm sure I've asked this before), but why isn't that included in the str() of the HTTPError?
[16:07] <Chipzz> Hey people, my apologies for the spam, but for those that didn't see it yet, may I steal a couple of minutes of your time and draw your attention to http://www.fosdem.org/2010/distrominiconf ? :)
[16:08] <Chipzz> IMO the launchpad ppl could contribute a lot here :)
[16:22] <nekohayo> anyone else having problems with apport popping up on each coming back from sleep/suspend in karmic, and then launchpad not being able to parse the stuff it sends?
[16:22] <nekohayo> been like that for a week
[17:34] <bond> hey
[17:36] <bond> How do you run a receipt describing the bug and its status through the bot commands?
[17:50] <rowinggolfer> really love the new changes to the site.
[17:57] <Pici> bond: What bot are you referring to?
[18:00] <bond> Pici: !bug #NNNN
[18:02] <Pici> bond: if you mean ubottu....
[18:02] <Pici> !usage | bond
[18:03] <bond> Pici: I am interested in plugin that takes information about the bug on launchpad
[18:04] <bond> a plug encyclopedia I have already installed and functioning well. Or have you laid there in the encyclopedia? :)
[18:05] <Pici> bond: That would be the bugtracker plugin, code is here: https://code.edge.launchpad.net/~tsimpson/ubuntu-bots/tweak .  Help can be found in #ubuntu-bots
[21:07] <tgm4883> Should users always be able to ask questions in the Answers section of a project, even if the project unchecks the "users can ask questions" checkbox?
[21:07] <tgm4883> specifically, we disabled this function a while ago, and a user just asked a question though it https://answers.edge.launchpad.net/mythbuntu
[21:13] <beuno> tgm4883, I think the shouldn't
[21:13] <beuno> there's a bug around for it
[21:14] <beuno> there are reasons why it's enabled, but it needs some thought
[21:19] <tgm4883> beuno, ok, thanks for the info
[22:52] <zorael> What does it mean when it says "Entity-body was not a well-formed JSON document" when I try to modify the description of a bug?
[22:59] <intellectronica> zorael: it's a bug
[22:59] <malept> hi, I think my PPA build needs to be kicked on bohrium
[23:26] <malept> the PPA build in question seems to be stuck on unpacking the gettext package...
[23:29] <wgrant> malept: I've seen bohrium get stuck several times lately.
[23:33] <malept> wgrant: oh joy.
[23:50] <vadi2> Are launchpad PPAs affected by the slowdown? I uploaded a package, but I never got a confirmation email and it's not showing up on my PPA page.
[23:58] <wgrant> vadi2: PPA uploads are happening fine. You probably didn't sign the package correctly.
[23:59] <vadi2> but there is no error about it: http://paste.pocoo.org/show/147193/