[MERGE] Enable setting and getting make_working_trees for all repositories

[Aaron Bentley]

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

Robert Collins wrote:
> On Wed, 2008-04-09 at 22:41 -0400, Aaron Bentley wrote:
>>> For that matter, I don't really
>>> understand how this is a significant improvement over raising
>>> NotImplementedError.
>> NotImplementedError usually means someone forgot to implement
>> something.
>>       I think that kind of error should trigger a test suite failure.
>>
>> ValueError means that the supplied value was unsuitable, and for
>> AllInOne repositories, False is not a usable value.  (True is, because
>> that's the value they always use.)
> 
> We use NotImplementedError pervasively in bzrlib for 'optional method
> that this class does not implement'.

Really?

Because the use I'm familiar with is pseudo-pure-virtual-function, like
Branch.pull and Tree.get_file.  I have never had any reason to believe
that those methods were optional.  I strongly suspect that any
implementation of Branch or Tree which lacked those functions would fail
the test suite, and rightly so, in my opinion.

Aaron
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFH/YNc0F+nu1YWqI0RAnTZAKCBe9Vd5W2zHWrik8bEmaPflBnrgACfeR9N
pYfrIrsd1/JMg3zLb/5uzSQ=
=jitb
-----END PGP SIGNATURE-----