[00:28]  * kfogel is away: Smithwicks w/ Omar
[00:40] <lifeless> Noldorin: hi
[00:40] <lifeless> Noldorin: I put some other ideas in the bug
[01:17] <poolie> emmajane: thanks for the nice update
[01:17] <lifeless> spiv: ping
[01:18] <spiv> lifeless: pong
[01:18] <lifeless> I have an odd failure
[01:19] <lifeless> which may collide with your patch
[01:19] <lifeless> branch(Pdb) wt_a.branch
[01:19] <lifeless> RemoteBranch(bzr-v2://127.0.0.1:59801/extra/a/)
[01:19] <lifeless> (Pdb) wt_a.branch
[01:19] <lifeless> RemoteBranch(bzr-v2://127.0.0.1:59801/extra/a/)
[01:20] <lifeless> -> branch_b = wt_a.branch.bzrdir.sprout('b', revision_id='1').open_branch()
[01:20] <lifeless> (Pdb) branch_b
[01:20] <lifeless> BzrBranch7('file:///tmp/testbzr-Hs3ARm.tmp/bzrlib.tests.per_branch.test_branch.TestBranch.test_clone_branch_parent%28RemoteBranchFormat-v2%29/work/b/')
[01:20] <lifeless> (Pdb) branch_b.repository.chk_bytes.keys()
[01:20] <lifeless> set([])
[01:20] <lifeless> wt_a.branch.repository is a 2a repo
[01:21] <poolie> hi lifeless, spiv
[01:25] <sque> hello! I am running ubuntu 8.04.3 server with bzr ppa and installed loggerhead
[01:25] <sque> when I try to view file contents from the webbrowser I get an exception from loggerhead
[01:25] <sque> An unexpected error occurred whileproccesing the request:
[01:25] <sque> AttributeError: 'module' object has no attribute 'ProgressBarStack'
[01:26] <sque> What is the problem?
[01:26] <lifeless> sque: you either need a newer loggerhead or an older bzr.
[01:26] <spiv> lifeless: perhaps!  I'm not sure.
[01:26] <lifeless> spiv: can I call?
[01:26] <spiv> Sure.
[01:27] <sque> ii  bzr                                     1.17-1~bazaar1~hardy1                   easy to use distributed version control system
[01:27] <sque> ii  loggerhead                              1.10-1                                  Web viewer for Bazaar
[01:27] <sque> those are the version that are currently installed and they both come from this official ppa https://launchpad.net/~bzr/+archive/ppa
[01:28] <sque> you mean that they are not compatible?
[01:32] <spiv> Yes; that version of Loggerhead is too old to work that with version of Bazaar.
[01:33] <lifeless> spiv: found it I think:
[01:33] <lifeless> 1759 ->         source = repo._get_source(self.to_format)
[01:33] <lifeless> 1760            if isinstance(source, RemoteStreamSource):
[01:33] <lifeless> 1761                return repository.StreamSource.get_stream(source, search)
[01:33] <lifeless> 1762            return source.get_stream(search)
[01:34] <lifeless> spiv: have you changed that?
[01:42] <spiv> I don't think so, let me check.
[01:42] <lifeless> spiv: I've put up a merge proposal
[01:43] <spiv> No, I didn't change that.  It does seem like an obvious problem, though!
[01:55] <emmajane> poolie, you're welcome for the update. :)
[01:57] <sque> beuno: ping
[02:00] <igc> morning all
[02:34] <mozmck> on ubuntu 9.04, I installed the latest bzr-gtk from source as described on the website.  but when I try to run olive-gtk is says "ImportError: No module named gtk.ui"
[02:34] <mozmck> am I missing something?
[02:43] <AfC> mozmck: a good technique in these cases is to install a -dev package of what you are trying to build first, to make sure you have all the dependencies
[02:46] <mozmck> I don't see a -dev package for bzr-gtk.  I have the bzr ppa in my repositories list.
[03:00] <meoblast001> is it hard to set up a bazaar server?
[03:00] <meoblast001> i'm thinking about doing it on my server for backup of my source files
[03:12] <igorgue> hi, to install a dependency I use: pip install -e 'bzr+http://bazaar.launchpad.net/~txamqpteam/txamqp/trunk/#egg=txamqp'
[03:12] <igorgue> do you know how to add that to my setup.py?
[05:46] <_habnabit> So, I'm not clear exactly what a 'branch' is or how branches can be subdivided. I'm trying to migrate from svn to bzr, and I've created a repository with 'bzr svn-import'. I made a new repository to work in, and I tried to make a branch with just one project from the svn repository, but I get an error saying that its directory is not a branch. I can only make a new branch from the entire svn trunk.
[05:47] <_habnabit> Should each separate project have its own branch?
[05:52] <spm> _habnabit: i hit something similar when I first migrated. this may help: http://www.stedee.id.au/2008/11-06/manually_migrating_subversion_repository_launchpadbzr ??
[05:53] <spm> _habnabit: the key part of that for you would be the: svn2bzr.py --scheme=trunk --exclude=branches --exclude=tags <== to only get what you want
[05:54] <_habnabit> Well, what's the difference between using svn2bzr and bzr-svn? Aside from that it seems like it's more one-way.
[05:54] <_habnabit> I'm not sure if it would solve the issue of trying to split my svn trunk into separate bzr branches.
[05:55] <spm> I tried bzr-svn and it didn't work well for me - then. svn2bzr did. ymmv.
[05:55] <bob2> branches in bzr and svn are pretty similar in concept, but with the proviso that in bzr you can't check out arbitrary dirs like you can in svn
[05:56] <_habnabit> So with bzr, I would want separate branches for each project.
[05:56] <bob2> well, separate branch for each branch of each project
[05:56] <lifeless> _habnabit: a branch is one line of development for a project
[05:56] <_habnabit> Right.
[05:56] <lifeless> _habnabit: and its not subdividable to subdirectories
[05:56] <_habnabit> Okay.
[05:57] <_habnabit> How do I deal with this, then? They are in subdirectories at the moment.
[05:57] <_habnabit> It looks like svn2bzr might help, hm.
[06:00] <bob2> bzr-svn import lets you explain how the svn repository is layed out, too
[06:24] <lifeless> spiv: ping
[06:32] <spiv> lifeless: pon
[06:32] <spiv> pong, even.
[06:32] <lifeless> did you do anything along the lines of getting xml inventories from CHK repos?
[06:33] <lifeless> turns out we have to implement it
[06:33] <spiv> No, but I believe John did for the bundle code.
[06:33] <spiv> IMBW.
[06:33] <lifeless> ah yes, thats what was tickling my memory
[06:58] <vila> hi all
[07:02] <lifeless> EODing
[07:15] <poolie> hey vila
[07:15] <vila> hi poolie
[07:16] <poolie> vila, did you mention the existence of the build farm yet?
[07:16] <poolie> it might be god
[07:16] <poolie> good*
[07:16] <poolie> liw just asked about whether we had one
[07:16] <vila> ok,
[07:17] <vila> Hey guys, we have a build farm running the test suite on a couple of slaves !
[07:17] <poolie> :)
[07:18] <vila> well, the truth is karmic fails and the windows slave setup is broken :-/
[07:22] <vila> poolie: more seriously, what do you have in mind ? An announce on bazaar@lists.c.c ?
[07:23] <poolie> yes :)
[07:23] <poolie> that's no good about karmic
[07:23] <poolie> liw mentioned that too
[07:23] <vila> Did you read the discussion with jam about not allowing forcing builds from the public server ? Do you agree with my proposal ?
[07:23] <poolie> you should file a bug?
[07:23] <poolie> i did read it
[07:23] <poolie> um
[07:24] <vila> the karmic failures may be minor, I just haven't dig them yet
[07:28] <vila> jelmer: ping, BB is down, care to review my patch for bug #403340 anyway ?
[07:29] <vila> cough bug #403430
[08:54]  * igc dinner
[09:01] <poolie> hm
[09:04] <lifeless> poolie: how was the conference?
[09:04] <poolie> hasn't happened yet
[09:04] <lifeless> ah
[09:04] <poolie> it's on the weekend
[09:05] <poolie> am just working from here, mostly on trying to get my presentation to feel more coherent
[09:05] <lifeless> laserpres!
[09:05] <poolie> the short story is
[09:06] <poolie> laserpres?
[09:06] <poolie> anyhow the short story is: open source development is fun and fulfilling,
[09:06] <poolie> being done entirely over the net it's very much mediated by and shaped by the tools
[09:07] <poolie> so the point of our work is to make tools that fit how people work and help them work better
[09:08] <poolie> specifically: it's pretty amazing when people translate your software or documentation into languages you can't speak yourself
[09:12] <lifeless> laserpres<- coherent presentation
[09:12] <lifeless> and that story sounds compelling
[09:15] <poolie> 2- open source projects are so much more open to bug reporting that it's easy to get swamped, and bugs can be very fragmented
[09:18] <vila> poolie: regarding bugs: I think it's part of the open thing, in the long term they *will* be fixed, because the underlying assumption is that FOSS *cares* about software without bugs and dislike the tendency to make users think that buggy systems/tools are to be *expected*
[09:18] <poolie> 3- if you invite people to help, you're obliged to accept that help - don't drop patches but put them through some kind of coordinated merge review
[09:19] <poolie> mm
[09:19] <poolie> well
[09:19] <poolie> neither open nor closed projects fix all their bugs
[09:19] <poolie> and i'm not convinced closed projects care any less
[09:19] <poolie> at least, the best projects of either type probably care the same amount
[09:23] <vila> poolie: not IME, but don't let me disturb you
[09:23] <poolie> well
[09:23] <poolie> so the question is, what should i say about either how to handle bugs better, or what launchpad/bazaar do to help you handle bugs better
[09:24] <lifeless> anecdotally
[09:25] <lifeless> I just found that a patch I put up for cppunit 4?maybe more years ago - still in the bug tracker. 'postponed'
[09:26] <vila> lifeless: at least it's there and others can use it
[09:26] <poolie> well
[09:26] <poolie> kinda
[09:27] <vila> lifeless: it's not as good as if it has been integrated, but it's better than not available at all
[09:27] <poolie> though actually stephane did dig a patch out of the svn bug tracker a while ago
[09:27] <lifeless> I posted a dup of my patch - it was so long ago I'd forgotten that I'd posted it there ;)
[09:27] <lifeless> anyhow
[09:28] <vila> :-D
[09:28] <lifeless> linking bugs to code, management of the data is Good Stuff
[09:28] <awilkins> I do like the "proposed for merging" branches thing
[09:29] <lifeless> code review tied into scm and bugs is good
[09:29] <awilkins> We have a meeting about dev practices this morning and I fully intend to wave Launchpad around given the opportunity
[09:29] <awilkins> I just wish the network wasn't so crapulous here or I'd have an instance running by 1100
[09:31] <lifeless> :)
[09:31] <awilkins> Have all the sources just need the deps...
[10:13] <lvh> hi
[10:13] <lvh> I have two ssh keys, one with password and one without. I'd like to use the one without password as a signing key. Is that a bad idea?
[10:14] <lvh> also, am I supposed to get ssh to use the different key, or is there some way to get bzr to ask for a different key? (the former is how git does it)
[10:34] <krisives> How can I branch a repo into a target directroy?
[10:35] <krisives> I want to do `bzr branch /from/here /into/here` without it creating a new sub directory *inside* "here"
[10:37] <RAOF> krisives: Is there anything wrong with "bzr branch /from/here /into"?
[10:37] <krisives> I am branching user directories
[10:37] <krisives> So what I am trying to do basically is `bzr branch /home/developer /home/me`
[10:38] <krisives> Get it?
[10:41] <lifeless> krisives: you can do 'bzr init; bzr pull URL' as an alternative
[10:41] <krisives> Will it still be a repo?
[10:42] <krisives> ty, that seems to work well
[14:40] <vila> jam: ping, what did you do ? windows builds works again 8-/ (and yes, I half expect that you will answer: nothing)
[14:41] <corporate_cookie> it looks like the default version of bzr in Ubuntu Server LTS 8.04 is 1.3.1; is there a more recent supported version of bzr ?
[14:41] <beuno> corporate_cookie, you have the official PPAs, yes
[14:42] <vila> corporate_cookie: https://edge.launchpad.net/~bzr/+archive/ppa ?
[14:42]  * beuno waves at vila 
[14:42] <corporate_cookie> are they 'supported' as in can I tell my boss I can call canonical for 'help'  ?
[14:42]  * vila waves back :)
[14:43] <vila> corporate_cookie: AIUI yes
[14:43] <corporate_cookie> thanks : )
[14:43] <corporate_cookie> hooray : )
[14:43] <vila> pfew, that one was easy :-)
[14:44] <corporate_cookie> hehe thanks vila
[15:37] <lvh_> hm
[16:03] <Colonel-Rosa> What's the syntax for commit a spefic file?
[16:04] <Colonel-Rosa> bzr commit -m "message" /file/path
[16:04] <Colonel-Rosa> right?
[16:04] <luks> yes
[16:04] <Colonel-Rosa> Cheers
[16:14] <davidstrauss> Does anyone know how to take old-format branches that were upgraded to get rich root support to properly support "join"?
[16:40] <kfogel> I need to convert an older branch to be a branch in a new 2a repository.  Details here; see question at end: http://paste.ubuntu.com/252586/
[16:41] <kfogel> james_w: (am I on the right track above?)
[16:42] <james_w> kfogel: yeah
[16:42] <james_w> but I can't answer your question fully, sorry
[16:42] <kfogel> james_w: it's that easy, huh?  Great! :-)
[16:42] <kfogel> james_w: np, thanks for looking.
[16:42] <james_w> I thought there was a "migration guide" knocking around somewhere or something
[16:42] <kfogel> let me see
[16:43] <beuno> kfogel, hold on
[16:43] <beuno> why aren't you doing bzr upgrade?
[16:43] <beuno> you should reconcile *before* upgrading
[16:43] <kfogel> beuno: thank you.  (Er, that's why I'm *asking*.)
[16:44] <kfogel> beuno: IMNSHO bzr ought to notice and just DTRT, but since it apparently won't, I'll dance the upgrade dance.
[16:44] <beuno> kfogel, please nag the bzr list about this
[16:44] <kfogel> beuno: so, reconcile the old tree first; then bzr upgrade it; then branch it into the new shared repository
[16:45] <beuno> I've been tryingo the the upgrade story straight
[16:45] <beuno> kfogel, you can just upgrade the repo
[16:45] <kfogel> beuno: orig tree is not in a shared repo
[16:45] <kfogel> beuno: I mean, I could put it in one, just for this, but if that step can be avoided, then even better
[16:45] <beuno> kfogel, https://lists.ubuntu.com/archives/bazaar/2009q3/061057.html
[16:45] <beuno> kfogel, gotcha
[16:45] <beuno> yes then
[16:45] <kfogel> beuno: so if this works, I'll just write a bzr wiki page on it and post that
[16:46] <kfogel> beuno: thanks for the link!
[16:46] <beuno> kfogel, also: http://people.canonical.com/~ianc/doc/en/upgrade-guide/
[16:47] <kfogel> beuno: aaaaah
[16:47] <beuno> kfogel, also, https://lists.ubuntu.com/archives/bazaar/2009q2/058377.html
[16:47] <beuno> actually
[16:47] <beuno> no
[16:47] <beuno> ignore
[16:47] <beuno> that
[16:48] <kfogel> beuno: no, that one does have a section on data migration
[16:48] <beuno> Peng_ wrote something to the list
[17:01] <jam> vila: for windows builds working... it may be that I had created a hard link, which I subsequently deleted
[17:01] <jam> that, at least, is my best guess
[17:01] <jam> I also submitted some patches to the builder
[17:08] <vila> jam: k, thks for telling, anyway, I should have an XP setup locally soon (without cygwin) and will see from there
[17:12] <sque> Hello
[17:13] <sque> Can someone plz explain the main concept of how to create a multi-user bzr server?
[17:13] <sque> with different credentials for each user
[17:23] <kfogel> beuno-afk: I'm thinking maybe this is the page that needs to be updated: http://bazaar-vcs.org/FormatUpgrades  :-).
[17:28] <Tak> sque: super easy with ssh
[17:31] <sque> Tak: I have setup bzr+ssh at the moment and have created one user
[17:32] <sque> Tak: and the repository exists in /bzr. till now I can give this one user to all commiters to commit changes. I thought to create different users but how it will manage the file permissions?
[17:34] <Tak> it just uses the native file permissions, doesn't it?
[17:35] <sque> Tak yes it does. But if user A commits changes to repository user B will be unable to change some files. is this normal?
[17:36] <sque> Tak Or I bazaar repository can work with different file owners?
[17:37] <Tak> you can set the permissions however you want
[17:38] <Tak> world-writable, writable only by a developers group, project-specific write permissions, ...
[17:38] <sque> I think I am missing something here...
[17:47] <vila> jam: just see revno 4603 (your changes to the builder), the build worked before that (4602 at least, but I doubt that's related), so that hard link you're talking about sounds like a better explanation (which was it by the way ?)
[17:47] <sque> Tak I created a test branched did many commits from a root user and normal user... to see if it would have permission problems, but as I see bazaar does not change files it creates files for each new commit so alla the commits are readable from any user
[17:48] <jam> vila: I had hardlinked the build-win32 directory to try to avoid having 10 copies of all the download files everywhere
[17:48] <jam> but it looks like windows + hard links + acls just doesn't get along very well
[17:48] <vila> haaa, and you tried to use the one in the buildbot hierachy ?
[17:48] <sque> Tak I though it was working as svn that the repository should not be written by different user otherwise it would be fucked up.
[17:49] <jam> vila: something like that, yeah
[17:49] <vila> jam: right, that matches the fact that I had to delete a couple of files around,
[17:49] <vila> and why the failure was transient
[17:51] <vila> jam: by the way, the test farm now run selftest with/without locale so we should be able to remove that for pqm, shouldn't we define an explicit target for pqm by the way...
[17:51] <jam> vila: you mean something other than "make check" ?
[17:52] <jam> And yes, we probably could remove the second pass for PQM as long as we update documentation and general workflow
[17:52] <jam> to make people aware that some tests will be run asynchronously
[17:52] <jam> i'm guessing we aren't 100% ready to make the switch yet
[17:52] <jam> but we're probably getting close.
[17:54] <vila> jam: yeah, something like 'make pqm' with pqm: check to describe today behavior but make it clearer that it's the pqm rule
[17:55] <jam> vila: just call it "pqm-check"
[17:55] <vila> right, the missing bit is removing --no-plugins :-) See bug #412930
[17:55] <vila> jam: yup
[17:55]  * vila EODing
[17:56] <jam> so... they *should* be running with plugins
[17:56] <jam> and I'm not sure I agree about bug #412930
[17:56] <jam> but that is what discussion is for :)
[19:46] <elmo> how do I turn on http debugging?
[19:59] <kfogel> elmo: 'bzr help debug-flags' ?
[19:59] <kfogel> elmo: looks like 'bzr -Dhttp'
[20:00] <elmo> kfogel: thanks, I id RTFM but it doesn't even mention the word 'debug' :(
[20:00] <elmo> also -Dhttp doesn't appear to do anything
[20:01] <kfogel> elmo: Well, I didn't say it worked :-).
[20:01] <kfogel> elmo: wireshark?
[20:02] <fullermd> It worked last time I tried it (which admittedly, was a year or two ago, but...)
[20:19] <vila> elmo: -Dhttp will write all HTTP headers to .bzr.log
[20:20] <vila> can someone confirms that there are problems with bazaar.launchpad.net since ~1h ?
[20:22] <kfogel> vila: try in #launchpad too?
[20:22] <vila> kfogel: was just doing that :-)
[20:33] <divokz> I have a question about the "Advanced shared repository layouts" (as found in the User Guide).  How are these created on the server?  Do I just init-repo and mkdir directories like "branches", "tags", etc?  I was under the impression that this could be done in a space efficient way, but I'm having a hard time finding anything.  Is there not a command just to pull down the entire tree, including "trunk", "tags", "releases", etc?
[20:33] <divokz> I've heard of stacked branches, but if I make stacked branches on the server, do they pull down the branch and the stacked-on repository?
[20:34] <divokz> We have a lot of data, so space can be a concern
[20:39] <garyvdm> divokz: If you have a shared repository, that will share most of the data amongst different branches (that are in the shared repo.)
[20:40] <garyvdm> So creating a new branch won't consume alot of space.
[20:40]  * garyvdm looks at "Advanced shared repository layouts"
[20:40] <divokz> garyvdm: so if I do `bzr init-repo` and then make branches inside there, I'll get the behavior I'm looking for?
[20:41] <divokz> `bzr init-repo project_name`
[20:41] <divokz> `cd project_name`
[20:41] <garyvdm> bzr branch
[20:41] <garyvdm> yse
[20:41] <garyvdm> yes
[20:41] <divokz> okay
[20:42] <divokz> and then, if I want a features dir, I just `mkdir features`, `cd features`, `bzr branch ../trunk some_feature` ?
[20:42] <divokz> Won't I have to check out each branch on a local machine separately?
[20:43] <LarstiQ> divokz: yes, although I'd start out with less hierarchy and just 'bzr branch trunk some_feature'
[20:43] <garyvdm> You can just do  bzr branch trunk some_feature project_name
[20:43] <garyvdm> You can just do  bzr branch trunk some_feature *in* project_name
[20:43] <LarstiQ> divokz: ehm, I'd do the branching locally
[20:43] <divokz> LarstiQ: but what if the branches need to be shared?
[20:44] <garyvdm> divokz: Have a shared repo on both the server and your computer
[20:44] <LarstiQ> divokz: then you share them when you reach that point. You are right that there is no 'repository cloning'
[20:45] <divokz> LarstiQ: do you mean without storing them on the server?  (decentralized)
[20:45] <LarstiQ> divokz: if you feel more comfortable with storing them on the server, then do that :)
[20:45] <divokz> LarstiQ: :)
[20:46] <LarstiQ> divokz: but if it is a short lived feature branch only hacked on at a sprint, why bother, it will be merged soon anyway
[20:46] <divokz> Yeah, I'd use a local branch in that case
[20:46] <divokz> I'm talking about something longer
[20:47] <LarstiQ> divokz: sure, then it can make sense
[20:47] <divokz> But to check them out, I have to get each branch individually, correct?  (what you meant by 'no repository cloning'?)
[20:47] <LarstiQ> divokz: you have to get each branch you want individually, yes
[20:48] <divokz> okay, great -- I've been flipping through documentation for too long without finding this stuff out
[20:48] <LarstiQ> divokz: which is what I mean with no 'repository cloning'
[20:48] <LarstiQ> divokz: a repository is only a common revision store
[20:48] <divokz> Thanks, LarstiQ and garyvdm -- very helpful
[20:49] <garyvdm> divokz: Maybe be you misunderstand the "shared" in "shared repository". The data is shared by the different branches in side the shared repo. A shared repo is not nesserly shared as in accessible by other people.
[20:49] <divokz> garyvdm: yes, I did misunderstand that
[20:50] <divokz> garyvdm: so what does `bzr init-repo --no-trees repo_name` do in relation to that?
[20:50] <garyvdm> A branch normally contains a Working tree, a list of tags, and a pointer to the latest revision.
[20:51] <garyvdm> You can also have a branch with no working tree.
[20:51] <garyvdm> if you do bzr init-repo --no-trees, by default branches won't have a working tree
[20:52] <divokz> so, I would want that if I'm storing branches inside 'repo_name'?
[20:53] <garyvdm> A working tree is not the same thing as a repository
[20:53] <garyvdm> A working tree is the copy of the files on the disk, that you can edit, and then commit to the repository.
[20:54] <divokz> yes...
[20:54] <divokz> I guess I don't understand why else you wouldn't want to have a working tree in a branch
[20:54] <garyvdm> A branch with no working tree and be pulled to a branch with a working tree
[20:54] <garyvdm> cases where it is usefull:
[20:55] <garyvdm> on a server, we no one will be editing the working tree(wt), only pulling from their branches.
[20:55] <garyvdm> or
[20:56] <garyvdm> If you have one branch the you work in, and many other branches that you switch between
[20:56] <garyvdm> You bind the one that you work in to another on that has no working tree.
[20:57] <garyvdm> That way, you can have lots of branches, one working tree
[20:57] <divokz> could you point me to an example of this workflow?
[20:57]  * garyvdm looks at the docs
[20:58] <divokz> Or tutorial, etc -- I'm used to using more of the svn-oriented workflow
[20:58] <divokz> (thanks)
[20:59] <garyvdm> Can't find something in the docs, let me type up and example
[20:59] <garyvdm> *an
[21:07] <garyvdm> divokz:http://paste.ubuntu.com/252737/
[21:07] <divokz> garyvdm: I just *finally* found http://bazaar-vcs.org/SharedRepositoryTutorial -- it's helpful
[21:07] <divokz> garyvdm: thanks
[21:07] <garyvdm> I actual use that example I gave you alot.
[21:08] <divokz> garyvdm: I think that makes sense
[21:08] <divokz> I'm saving it at any rate :)
[21:10] <garyvdm> divokz: If you are not branching from somewhere else, I would recommend using the 2a format. It uses much less space.
[21:11] <garyvdm> if you are branching from some where else, then stick with the format of the branch you are branching.
[21:11] <divokz> somewhere else meaning a 3rd party?  Or just a separate machine?
[21:12] <garyvdm> 3rd party
[21:12] <divokz> k
[21:12] <divokz> I'll look into that -- could be helpful for some large datasets
[21:13] <garyvdm> If it's on a separate machine, you can just upgrade that branch too.
[21:13] <garyvdm> The reason why I say not to do it from a 3rd party, as that the conversion is one way.
[21:14] <divokz> makes sense
[21:14] <garyvdm> s/as/is
[21:18] <Noldorin> lifeless: pingf
[21:18] <Noldorin> ping*
[21:23] <lifeless> moin
[21:23] <lifeless> hi Noldorin
[21:24] <Noldorin> hi lifeless
[21:24] <Noldorin> how are things?
[21:24] <lifeless> ok
[21:24] <Noldorin> :)
[21:24] <Noldorin> you got a few free minutes now?
[21:25] <lifeless> I will in 20 or so
[21:25] <Noldorin> ok cool
[21:25] <Noldorin> i'll message you then
[21:49] <Noldorin> lifeless: so i noticed your updates to my bug report
[21:49] <Noldorin> looks like a pretty completely summary now :)
[21:51] <Noldorin> lifeless: was wondering where you wanted to proceed from here though
[21:57] <_thumper_> how can I find out which was the last mainline revision that a particular file was modified in?
[22:08] <jam> _thumper_: I seem to be getting a timeout when accessing http://bazaar.launchpad.net/~jameinel/bzr/1.18-lock-warnings/.bzr/branch/format
[22:08] <jam> is something going on with LP right now?
[22:08] <thumper> jam: yes
[22:08] <thumper> it's minorly fucked
[22:08] <thumper> we're working on it right now
[22:08] <jam> k
[22:10] <Noldorin> lifeless: you there?
[22:16] <lifeless> Noldorin: hi yes, been doing reviews; more than I expected :)
[22:16] <lifeless> Noldorin: someone needs to follow through on the tests I've proposed in the bug report
[22:16] <lifeless> hi jam, thumper
[22:23] <Noldorin> lifeless: right, ok
[22:23] <Noldorin> lifeless: so no more suggestions for the moment?
[22:23] <Noldorin> i have the source on this PC
[22:23] <Noldorin> if you want to do any more tests
[22:24] <lifeless> not interactively; got a _lot_ to do to get 2.0 out the door; I'll monitor the bug, and if I can find the time put a patch up to do the read-back test
[22:24] <lifeless> you could do that yourself fairly easily I suspect
[22:24] <lifeless> there is a function in that file that the lock object uses to check that its in the right place on disk and has the right nonce
[22:25] <lifeless> calling that function right after the rename of held to the tmp-for-unlocking would be useful
[22:25] <lifeless> with a try;
[22:25] <lifeless>     ...rename...
[22:25] <lifeless> except errors.BzrError:
[22:25] <lifeless>     pass
[22:26] <lifeless> else:
[22:26] <lifeless>     print "lock still in place after rename-to-unlock"
[22:26] <Noldorin> lifeless: ok, i'll give that a go.
[22:26] <Noldorin> thanks
[22:27] <lifeless> I would like to spend more time myself on fixing this, but - 2.0 calls. Sorry!
[22:29] <Noldorin> lifeless: that's fair enough. i look forward to trying v2.0 anyway :)
[22:29] <Noldorin> maybe if we figure this out in time, you can encorperate the fix?
[22:32] <lifeless> Noldorin: I'd be delighted to
[22:47] <garyvdm> Is it ok (provided all test pass) to change a public method from returns [] to a generator (yield x)? Or should I create a new method iter_xxx?
[22:48] <garyvdm> The method in question is Repository.find_branches
[23:15] <james_w> "bzr: ERROR: Cannot commit from a lightweight checkout to a stacked branch. See https://bugs.launchpad.net/bzr/+bug/375013 for details."
[23:15] <james_w> but I'm not in a checkout!
[23:15] <james_w> is that an overzealous check, or is the message unclear?
[23:16] <lifeless> james_w: its a little unclear. You're in a stacked branch locally, right?
[23:16] <james_w> stacked against a remote branch
[23:16] <lifeless> thats been causing corruption for quite some time
[23:17] <james_w> Standalone tree (format: unnamed)
[23:17] <james_w> Location:
[23:17] <james_w>   branch root: .
[23:17] <james_w> Related branches:
[23:17] <james_w>   parent branch: bzr+ssh://bazaar.launchpad.net/~dmitrij.ledkov/xiphos/main/
[23:17] <james_w>      stacked on: bzr+ssh://bazaar.launchpad.net/~dmitrij.ledkov/xiphos/main/
[23:17] <lifeless> it doesn't work, and will give ACF errors
[23:17] <james_w> I'm actually testing due to bug 413291
[23:18] <james_w> back in a bit
[23:20] <garyvdm> Never mind about my question. There are test fails, so I'm going to add iter_find_branches
[23:20] <garyvdm> Or should that be iter_branches?
[23:46] <igc> morning
[23:49] <lifeless> hi igc
[23:49] <igc> hi lifeless
[23:49] <garyvdm> Hi igc, lifeless
[23:50] <igc> hi garyvdm
[23:51] <lifeless> hi garyvdm
[23:51] <lifeless> need more people to say hi to