Recording where branches that are absent are

[Jan Hudec]

On Sat, Feb 04, 2006 at 12:35:52 -0600, John A Meinel wrote:
> Robert Collins wrote:
> > On Wed, 2006-02-01 at 21:16 -0500, Aaron Bentley wrote:
> > 
> > 
> >> | The .bzr/branch/format file would be different yes, and the fact
> > they
> >> |  both occupy .bzr/branch/* would have that effect.
> > 
> >> I think all working trees should specify the location of their branch,
> >> even if it's just '.'.  This breaks that regularity.
> > 
> > Another way of thinking about what I am proposing is that all working
> > trees will have their branch object available via
> > self.bzrdir.open_branch(). So its extremely regular, requires no working
> > tree specific logic, and will work identically with any working tree
> > implementation.
> > 
> 
> ...
> 
> > Meh, I see it as increasing code regularity, preserving separation and
> > leveraging the filesystem structure.
> > 
> > Impasse - I agree on that.
> > 
> > Martin, John, your thoughts?
> > 
> > Rob
> > 
> 
> I think having Branch be able to determine where it is located, rather
> than having WorkingTree do it is a better separation.
> So I prefer having ".bzr/branch/branch-location" to having
> ".bzr/checkout/branch-location".

IMHO .bzr/branch should *BE* the branch. So it either is there, or it is not.
And checkout should know where it's branch is.

> Robert and I had a big go-around about the meaning of
> Branch.open_containing() on IRC, though it settled out at least the
> meaning of BzrDir.open_containing().

Indeed, this makes sense. BzrDir knows what it contains and knows whom to ask
for the other bits.

-- 
						 Jan 'Bulb' Hudec <bulb at ucw.cz>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20060205/ba80ff9a/attachment.pgp