[MERGE] revert deletes files when it is safe

[Martin Pool]

On Wed, 2006-09-27 at 16:58 -0400, Aaron Bentley wrote:

> > We already have a common idiom of using
> > 'self.build_tree([])' for creating a tree with arbitrary contents. 
> 
> Right, but
> - - build_tree is not a good name when you're updating an existing tree
> - - using lists is overkill when you're just trying to update one or two
>   files.
> - - interpreting the list as tuples is not discoverable, and hard to
>   finger-type correctly.

Good points.  How about prepare_tree()?
> 
> > I *think* Robert is wanting to switch the tests over to MemoryTree where
> > possible. Which means he can change build_tree_contents() to work on a
> > Transport, and not have to actually write to disk for lots of tests.
> 
> I'm not entirely comfortable with this kind of simulation, but I wasn't
> trying to block it.  I think we should get Martin to write a sequel for
> distcc: distunittest.

:-) 

Well, we should have a situation now where the test are all independent
of each other, which is a good foundation.  So we could, at least, run
multiple threads locally, which might help substantially for things
blocked on file IO.  Parallelizing them is an almost necessary first
step for useful distribution.

-- 
Martin Pool <mbp at sourcefrog.net>