Rethinking last-revision

[Aaron Bentley]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Olaf Conradi wrote:
> On 31/03/06, Aaron Bentley <aaron.bentley at utoronto.ca> wrote:
> 
>>>And would simplify the check on standalone branches vs metadir
>>>branches (I currently check if  working_format == branch_format ==
>>>repo_format).
>>
>>That's not the right dichotomy.  A metadir branch may be standalone.
>>
>>A standalone branch is one whose repository, branch and working tree are
>>all stored in the same .bzr.  A metadir is a .bzr format that may or may
>>not contain a standalone branch.  Earlier formats could only contain
>>standalone branches.
> 
> 
> The format check was half of what I do.
> I first check the bzrdir locations, then check formats. When locations
> are the same and formats, then standalone.

This is necessarily true, because when locations are the same, then
standalone.

>  If flocations are the same
> and formats differ, bound branch. 

This is incorrect.  If the formats differ, it may be a standalone
branch.  A bound branch is a standalone branch for which
get_bound_location does not return None.  Try doing 'bzr upgrade
- --format metadir' in a standalone branch...

Aaron
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFELGUm0F+nu1YWqI0RAtBqAJ9MSU8MlCSPGGGNqAcc6CgwwfX7QwCfdoxb
9SmOg2UAH30D4HJ/eqq/Y1M=
=SJBt
-----END PGP SIGNATURE-----