[01:13] <dato> jelmer: oops, you're right. I guess I forgot to commit, and I can't find the branch now
[01:14] <dato> jelmer: so, I guess I'll create a commit and merge it tomorrow. I don't think you should've pushed, though, but never mind.
[01:18] <jelmer> dato: Well, it should be trivial to merge your commit into the current branch. I haven't uploaded anything yet
[01:20] <jelmer> dato: Mind if I just import-dsc the 1.4.1 release?
[01:21] <dato> so, a couple things
[01:21] <dato> are you interested in dm-upload-allowed: yes maintain it?
[01:22] <jelmer> dato: I'm interested in contributing to the maintainance of the bzr package
[01:23] <jelmer> i.e. doing uploads now and then if I have the time and nobody else is doing it
[01:26] <jelmer> not sure if that's what you're asking :-)
[01:28] <dato> ok
[01:30] <dato> so due to my fuck-up, I really think I should push --overwrite you... I would really dislike the idea of having a 1.4 in the branch that does not include the 1.3.1 changelog, what do you think?
[01:30] <jelmer> dato: oh, yes - I agree
[01:30] <dato> so it seems I've lost the commit, if it ever existed
[01:31] <dato> so I'll create a 1.3.1 commit, and push --overwrite
[01:43] <jelmer> k
[02:00] <dato> jelmer: done, going to bed.
[02:11] <jelmer> dato: thanks, goodnight
[09:59] <dato> jelmer: so, I'll upload bzr 1.4; I'll add the dm-upload bit
[10:06] <emgent> morning
[10:19] <matkor> Hmm, recent repo olive-gtk:
[10:19] <matkor>     from bzrlib.plugins.gtk import icon_path
[10:19] <matkor> ImportError: cannot import name icon_path
[10:20] <matkor> ?
[12:50] <jelmer> dato: COol, thanks
[12:54] <dato> jelmer: I'm in a conference, just need to find a spot to do it :)
[12:54] <fullermd> jelmer: Are you working up to a bzr-gtk release?
[12:56] <jelmer> dato: k :-)
[13:28] <jelmer> fullermd, yep
[13:47] <fullermd> jelmer: May want to look at https://bugs.launchpad.net/bzr-gtk/+bug/224914 for it.  Kinda sucks   :|
[13:48] <fullermd> Oh, hithere.
[13:48]  * fullermd pets ubottu.
[13:48] <fullermd> (it's a little more complicated than the desc; it seems to have weird interactions with the sort of children, whether direct or merge revs...   but anyway)
[13:57] <jelmer> oh crap
[13:57] <jelmer> we also have to other issues to fix before the release
[13:57] <jelmer> so I guess we should delay it another week
[13:57] <fullermd> Just doing my part to keep you from being bored   ;)
[13:59] <jelmer> see, this is why I prefer working on vaporware. it doesn't have users and so it's a lot easier to pretend there are no bugs
[13:59] <jelmer> :-P
[13:59] <fullermd> Oooh, I didn't know you were working on bzr-sccs...
[14:00] <bob2> haha
[14:10] <hersonls> poolie: you be here?
[14:13] <bob2> 2313 aest
[14:13] <rexbron> Hello, will branches that have been created with a plugin, ie bzr-svn, use the same methods as regular bzr branches?
[14:13] <rexbron> I am trying to integrate that functionality into a script one the branch has already been created
[14:14] <rexbron> once rather
[14:15] <jelmer> rexbron: yes
[14:16] <rexbron> jelmer: ok, and will the .push() method just work for pushing to a remote bzr branch, right?
[14:41] <hersonls> now the bazaar slackbuild has updated :D
[16:52] <jelmer> rexbron: yep
[17:07] <Kamping_Kaiser> hi all. i diffed two files, and one was (apparently) created 1970-01-01.  is this because it never existed?
[17:08] <LeoNerd> diff uses that to say it's a new file
[17:08] <Kamping_Kaiser> cool. thanks for confirming that
[17:40]  * Peng gasps.
[17:40] <Peng> bzr-svn uses old-style classes?
[17:41] <jelmer> oldstyle being "class foo:" rather than "class foo(object):" ?
[17:45] <Peng> Yeah.
[17:46] <jelmer> I've fixed a bunch of them a couple of weeks ago
[17:46] <jelmer> If you can still find any old-style classes in the 0.4 branch, please let me know :-)
[17:47] <Peng> Connection and ConnectionPool in transport.py.
[17:49] <jelmer> thanks. It's a bad habit..
[17:49] <jelmer> (these were two classes I had added recently)
[17:50] <Peng> ConnectionPool.connections could also be a dict/defaultdict. Probably not worth the effort though.
[18:09] <Peng> Woah, pylint really doesn't like bzr-svn.
[18:10] <jelmer> patches welcome :-)
[18:11] <Peng> It's 3500 lines long!
[18:14] <Peng> Heh, pylint triggers a DeprecationWarning of its own.
[18:15] <Peng> jelmer: "Your code has been rated at 4.20/10" :D
[18:16] <jelmer> :-/
[18:16] <Peng> I've never used it before and didn't configure it.
[18:16] <Peng> It's very whiny. Did you know one of your classes has "too many" instance attributes?
[18:22] <jelmer> Peng: yeah - have you tried running it on bzr itself?
[18:22] <Peng> Haha, no way.
[18:22] <jelmer> It complains about any class that has more than 20 methods I think
[18:22] <Peng> It's slow too.
[18:22] <Peng> Something like that, yeah.
[18:22] <Peng> It complains about lots of things like that. Too many args, etc.
[18:29] <MattCampbell> Suppose I want to make a bzr branch from a given release of an open-source project that uses svn, e.g. Trac.  If I do "bzr branch http://svn.edgewall.org/repos/trac/tags/trac-0.10.4/", then make modifications to that branch, will someone then be able to merge those changes with a bzr branch made from svn trunk?
[18:30] <MattCampbell> I wonder because I know that svn doesn't have the explicit concepts of branches and tags as bzr does; it uses cheap copies for both.
[18:31] <Peng> You could try it, of course. :P bzr-svn should be smart enough to make the branches related, though.
[18:32] <Peng> jelmer: When are you planning to release the next version of bzr-svn?
[18:34] <MattCampbell> My main fear with version control systems is always that I'll do something wrong at the beginning of a project, with regard to tree organization and the like, and never be able to cleanly recover from the mistake.  That's why I asked that question.
[18:53] <dato> jelmer: uploaded
[18:59] <jelmer> dato: thanks!
[18:59] <jelmer> Peng, tomorrow hopefully
[19:02] <Peng> jelmer: Ok, cool.
[19:44] <libwilliam> I am working on a Bazaar Plugin for Anjuta and I am wondering if there is C bindings for Bazaar. I haven't seen anything online but thought if anyone knew if someone was working on it you all would know.
[19:46] <MattCampbell> A better idea would probably be to embed Python in Anjuta, write Python bindings for Anjuta's API's, and do the bzr integration from Python code running inside Anjuta.
[19:46] <thatch> libwilliam: you can load libpython and call the apis directly
[19:48] <MattCampbell> I suppose this will be a perceived disadvantage of bzr compared to svn for a while; since svn is a bunch of C libraries, it's probably more attractive to third-party IDE/tool developers who aren't using Python already.
[19:50] <libwilliam> Alright I will look into that. I have never done something like that so unsure how easy it is, but gives me a good reason to learn.
[19:58] <libwilliam> thatch: is this similar to what you were talking about? http://www.linuxjournal.com/article/8497
[19:59] <thatch> libwilliam: yeah, that's what MattCaand I were suggesting
[19:59] <libwilliam> alright thanks, I will mess around with it
[22:48] <awmcclain> Anyone have any success compiling pycurl for leopard?