[00:00] <RAOF> Jordan_U: Got a read-only $HOME?
[00:00] <Jordan_U> RAOF: No
[00:01] <RAOF> Sure? :)  Lucid's been resuming with a readonly home for me on occasion ;)
[00:01] <RAOF> ...and a read-only home is one trigger of that behaviour that I know.
[00:02] <Jordan_U> RAOF: Yes, I have been working with files in $HOME without problems
[00:02] <fullermd> Check ~/.bzr.log, make sure it's owned by you.
[00:03] <Jordan_U> fullermd: Ahh, it's owned by root
[00:03] <Jordan_U> Probably from trying to use etckeeper with bzr
[00:04]  * Jordan_U Thinks etckeeper in Ubuntu should not default to using bzr untill bzr can properly deal with being run with sudo
[00:05] <Jordan_U> fullermd: That fixed it, thank you
[00:50] <Noldorin> hi
[00:50] <Noldorin> i'm trying to install bzr-git on windows
[00:50] <Noldorin> but am having no luck
[00:50] <Noldorin> bzr plugins does not list it.
[01:11] <Supertanker> Is there any reason I should not access a bzr branch stored on my Samba drive?
[01:12] <bob2> do you have filenames with : in them or mixed caase?
[01:13] <Supertanker> Shouldn't be for the first; for my projects, I use all lowercase names
[01:13] <Supertanker> Er wait no.
[01:13] <Supertanker> I don't use all lowrcase names
[01:13] <Supertanker> But the Samba share is hosted on a Liknux box and I don't have files with the same names but different cases in the same directory
[01:13] <Supertanker> (I do jhave a login.php in one directory and Login.php in another...)
[01:46] <Supertanker> Yay, bzr for windows finally downloaded
[01:58] <Noldorin> hi
[01:58] <Noldorin> i'm trying to use bzr dpush to push to a GIT repo, but am having the following problem:
[01:58] <Noldorin> C:\Users\Alex\Documents\Visual Studio 2010\Projects\IRC.NET\devel>bzr dpush git+
[01:58] <Noldorin> ssh://git@github.com/Noldorin/IRC.NET.git
[01:58] <Noldorin> bzr: ERROR: [Error 2] The system cannot find the file specified
[03:10] <Noldorin> hi
[03:11] <Noldorin> i'm trying to dpush using bzr-git
[03:11] <Noldorin> getting the following error though:
[03:11] <Noldorin>  http://pastebin.com/m78e07c36
[03:34] <fullermd> As a guess, version mismatch.
[03:35] <fullermd> That's a reasonably recent bzr-git, and a several version back bzr.
[05:15] <GuyFromHell> anyone on gentoo know where olive-gtk is hiding?
[05:16] <Supertanker> What is this "treeless branches" feature I keep reading about, and why is it important?
[05:18] <fullermd> I dunno what you keep reading about, but a branch may or may not have a tree, depending on...  well, whether it has a tree.
[05:18] <mneptok> GuyFromHell: https://edge.launchpad.net/bzr-gentoo-overlay/
[05:19] <mneptok> GuyFromHell: bzr-gtk is part of that overlay
[05:19] <GuyFromHell> mneptok, oo, shiny. I could not find it anywhere
[05:19] <GuyFromHell> thank you
[05:19] <mneptok> np np
[05:27] <Supertanker> fullermd, what exactly is "tree" referring to?
[05:29] <fullermd> The working tree, where you edit and read and touch and caress-by-the-fireside your files.
[05:29] <Supertanker> Ah, okay
[05:31]  * SamB_XP can't help but remember something about some windows pack-in called "wang" ...
[05:31] <fullermd> If you have a branch with a [colocated] tree, you can get rid of the tree with 'bzr remove-tree', and you can get it back with 'bzr checkout'
[05:32] <fullermd> (and some args to reconfigure also, I believe)
[08:31] <pattern> i just did a "bzr add myfile", but then noticed that i'd misspelled the filename
[08:31] <pattern> i haven't committed yet, so is there some way to "unadd" the file?
[08:31] <fullermd> Well, if you handed it a filename that doesn't exist, it shouldn't have done anything...
[08:32] <pattern> it does exist
[08:32] <pattern> but there's a misspelling in the filename
[08:32] <pattern> so i need to rename the filename
[08:32] <pattern> and "unadd" it from bzr, then add it again with the right name
[08:32] <pattern> then commit
[08:32] <pattern> i just don't know how to "unadd" it
[08:32] <fullermd> Why unadd/readd, instead of just 'bzr mv'ing?
[08:32] <fullermd> Anyway, you'd use 'rm' to unadd.
[08:33] <pattern> i tried rm and it complained about not being able to safely remove the file
[08:33]  * fullermd stabbies the !@&$ DWIM crap in there.
[08:33] <fullermd> rm --keep
[08:33] <fullermd> (and then add a bzr alias for it so it stops trying to outsmart you ever afreakingain  :p)
[08:34] <pattern> thanks, fullermd
[11:46] <Noldorin> hello
[11:47] <Noldorin> i'm trying to use bzr-git to push a bzr branch to a git repo
[11:47] <Noldorin> however, i'm geting some horrible errors...
[11:47] <Noldorin> bzr: ERROR: exceptions.AttributeError: 'InterRepository' object has no attribute
[11:47] <Noldorin>  'fetch_objects'
[11:47] <Noldorin> the full log is here: http://pastebin.com/m78e07c36
[11:48] <bob2> are you using dulwich,bzr-git and bzr trunk?
[11:48] <Noldorin> bob2: indeed i am
[11:48] <Noldorin> i believe everything is installed correctly, though i had some trouble yesterday...
[11:49] <fullermd> Not on the latter you aren't.
[11:49] <fullermd> bzr 1.17 on python 2.5.2 (win32)
[11:49] <Noldorin> oh, that's true
[11:49] <Noldorin> odd
[11:49] <Noldorin> i'm sure i had bzr 2.0 installed
[11:50] <bob2> you may well need actual bzr bzr
[11:52] <Noldorin> bob2: how do you mean?
[11:52] <bob2> as in bzr trunk
[11:52] <Noldorin> ah right
[11:53] <Noldorin> hmm
[11:53] <Noldorin> brb, restarting
[12:15] <jelmer> fullermd: What's the status of DWIM?
[12:15] <fullermd> vila landed it a few weeks ago.
[14:08] <maxb> Is there anything along the lines of architecture documentation for Bazaar? e.g. anything that helps understanding the internals?
[14:30] <jml> maxb, maybe in the HACKING guide or in the developer doc section
[14:31] <jml> maxb, in the tree, that is.
[14:31] <jml> maxb, I've mostly got a sense of things from looking at commands in bzrlib/builtins.py and drilling down.
[16:32] <SabreWolfy> Using Ubuntu Karmic and trying to install bzr-explorer, but conflicts with qbzr version preventing this; latest qbzr version in Karmic is 0.14; bzr-explorer needs 0.17
[16:39] <SabreWolfy> any suggestions ?
[16:43] <SabreWolfy> Adding QBZR PPA solved
[16:52] <Noldorin> hi
[16:52] <Noldorin> i hqave bzr 2.0 installed
[16:52] <Noldorin> but for some reason, when i run bzr from the command line, it uses version 1.17
[16:52] <Noldorin> looks like i have bzr 1.17 in my path somewhere, but i can't figure out where :S
[17:04] <Noldorin> anyone?
[17:08] <verterok> Noldorin: 'bzr version' should show the path to the executable and bzrlib
[17:08] <Noldorin> hmm ok
[17:08] <verterok> Noldorin: ops, only the bzrlib path
[17:09] <Noldorin> verterok: c:\Python25\lib\site-packages\bzrlib
[17:09] <Noldorin> yeah
[17:10] <Noldorin> surely i want the one on program files though?
[17:10] <Noldorin> if i do...
[17:10] <Noldorin> "C:\Program Files\Bazaar\bzr.exe" version
[17:10] <Noldorin> i get v 2.0.3
[17:11]  * Noldorin gets out procmon
[17:13] <verterok> Noldorin: check in c:\Python25\Scripts
[17:13] <Noldorin> ok
[17:13]  * verterok goes to the grocery store, brb
[17:13] <Noldorin> verterok: seems to be a bzr in there
[17:13] <Noldorin> should i remove it?
[17:14] <Noldorin> ah, removing it does the trick :)
[17:14] <Noldorin> thanks
[17:17] <Noldorin> not bzr says it can't find dulwich :S
[17:17] <Noldorin> yet it's in the python path
[17:44] <verterok> Noldorin: if you'r using the standalone bzr.exe I think you need to install plugins in a different place as bzr.exe use it's own python interepreter
[17:44] <Noldorin> verterok: yeah, so it seems
[17:45] <Noldorin> any idea what i need to do with dulwich?
[17:45] <Noldorin> since i want the bzr binary to use it
[17:45] <Noldorin> well, bzr-git actually
[17:51] <Noldorin> verterok: i've pasted the git and dulwich folders into C:\Program Files\Bazaar\plugins
[17:51] <Noldorin> but it's not recognising dulwich still
[17:51] <Noldorin> No module named dulwich.errors
[17:51] <Noldorin> Unable to load plugin u'dulwich' from u'C:/Program Files/Bazaar/plugins'
[17:51] <Noldorin> bzr: ERROR: Unable to import library "dulwich": bzr-git: Please install dulwich,
[17:51] <Noldorin>  https://launchpad.net/dulwich
[17:52] <fullermd> dulwich isn't a bzr plugin, it's a python module/library/whatevertheycallit.
[17:53] <fullermd> I don't think you can use such things with the prebuilt bzr.exe; you'd have to use the python version.
[17:53] <Noldorin> fullermd: yeah, that's what i was thinking
[17:53] <fullermd> (Or I could be full of bullcaca; I don't really know windows)
[17:53] <Noldorin> was wondering why bzr-git was looking for it in plugins there
[17:53] <Noldorin> heh
[17:53] <Noldorin> fullermd: i think you may well be rigtht - i may need the python version
[17:54] <Noldorin> fullermd: hrm, can't i just edit the PYTHONPATH for bzr.exe somehow?
[17:54] <Noldorin> http://osdir.com/ml/bazaar/2009-11/msg00361.html
[17:55] <fullermd> I'm not sure you can; bzr.exe bundles it all up internally.
[17:55] <Noldorin> hrm
[17:57] <Noldorin> fullermd: so you recommend i just set up the python version of bzr 2.0, etc.? will it happily coexist along side the exe version?
[17:57] <fullermd> I have no idea.  It's been 13 years since I had a system running Windows.
[17:57] <Noldorin> hah ok
[17:58] <fullermd> I'd tend to think having both would be sticky.
[17:58] <fullermd> But it may not be.  I'm well into "guessing" territory there.
[17:59] <Noldorin> mm, as am i
[18:00] <Noldorin> fullermd: if i only have either in my PATH, then things might be ok
[18:01] <Noldorin> brb
[18:11] <Guest46284> hi. Is there a way of finding out what "bzr update" will do without actually running it
[18:58] <Noldorin> hello. do i need the python installation of bzr to run bzr-git/dullwich?
[19:46] <maxb> Is Bundle Buggy genuinely still alive, or does HACKING.txt lie?
[20:00] <pattern> i have a question regarding how best to do something with bzr...
[20:01] <pattern> it's a little long, though... so rather than flooding the channel, i've pasted the question here:  http://paste.pocoo.org/show/161426/
[20:07] <gutworth> pattern: you can just commit specific files
[20:08] <gutworth> or use a makefile which will only rebuild things if there is a change
[20:08] <gutworth> or make an ignored file where you list files to ignore
[20:09] <pattern> well, i could use a makefile that ignores files that aren't changed... however, LaTeX will still read them all to create the new pdf... so it doesn't really help
[20:11] <pattern> the only way to have those files not read by LaTeX is either to comment out the code that includes them, or to delete that code
[20:11] <gutworth> commit specific files then
[20:12] <pattern> i do commit each individual file in bzr
[20:12] <pattern> but the problem is more with the files that have the code that includes the individual files
[20:13] <pattern> they're like #include directives in c
[20:13] <pattern> and i have those lines commented out while i'm working
[20:14] <pattern> each of those include directives is just \include{foo}, where "foo" is the filename
[20:14] <pattern> so it's not like the file that has these include directives actually contains the whole file it's "including"
[20:14] <pattern> it's just a direction to the compiler to include the file
[20:15] <pattern> the contents of those files are completely seperate from the file they're included in
[20:15] <pattern> and both the file that's doing the inclusion and the included file are seperately versioned in bzr
[20:16] <gutworth> well, that's tricky
[20:16] <pattern> yep :)
[20:16] <gutworth> there must be some way to avoid rebuilding deps if they're not changed
[20:16] <pattern> that's why i come to the experts :)
[20:18] <pattern> i'm pretty sure there's no way to avoid rebuilding the files that haven't changed other than either commenting out or deleting the include directives
[20:21] <pattern> actually.. there might be another way
[20:22] <pattern> i think LaTeX has some control flow commands, which i haven't been using yet
[20:22] <pattern> but i should look in to them
[20:22] <pattern> that way it could maybe skip a part of the file if a variable is set
[20:22] <pattern> that would be idea
[20:22] <pattern> ideal
[20:23] <pattern> then the variable could be set from the makefile
[20:23] <pattern> and make could be told to set this variable from the command line
[20:23] <pattern> that would be perfect
[20:25] <pattern> hmm... but now that i think about it... it might not solve the problem after all
[20:26] <pattern> as, say i bracketed off files 1-95 as being compiled or not depending on a variable... and check that version of the include file in while i worked on files 95-100
[20:27] <pattern> at some point i'll want to bracket off files 1-100 while i work on files 100-105 instead
[20:27] <pattern> and that will entail rewriting the code in the include file
[20:28] <pattern> unless maybe i think of some way to reduce the granularity such that i can somehow include each individual file or not, from the command line
[20:28] <Noldorin> hi. do i need the python installation of bzr to get bzr-git/dulwich to work on windows?
[20:28] <pattern> anyway... this gives me some leads...
[20:28] <pattern> thanks for your help, gutworth!
[20:29] <Noldorin> getting the following error at present: http://pastebin.com/m3a6a2e4e
[21:13] <maxb> How alive/dead is the BundleBuggy project now that bzr itself no longer uses it?
[21:15] <Noldorin> >	hi. do i need the python installation of bzr to get bzr-git/dulwich to work on windows?
[22:04] <Noldorin> bob2: ping?
[22:04] <bob2> I don't have an answer
[22:04] <bob2> I was just wondering if you'd asked on the list
[22:05] <bob2> lots of people are on holidays atm so might not see discussion on irc
[22:05] <Noldorin> yeah, good point
[22:06] <Noldorin> probably worth a shot
[22:06] <Noldorin> bob2: it's a pain to set some of this beta stuff up when you don't know much about python (as i don't)...
[22:06] <Noldorin> it seems it's just dulwich that's causing the problem
[22:06] <Noldorin> i want bzr.exe to recognise it in the path
[22:07] <Noldorin> hrm
[22:13] <pattern> gutworth: i finally figured out a solution for my problem... what i'll do is keep all the include directives uncommented in files A and B, and then have my makefile run them through a filter that will comment out what i need and save the output under a new name, say A2 and B2
[22:13] <pattern> A and B will be versioned, but A2 and B2 will not
[22:13] <pattern> and A2 and B2 will be the files that latex uses to compile
[22:14] <pattern> of course, what the script run by the makefile comments out will depend on what options i pass to make on the command line