Merge progress

[Aaron Bentley]

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

Martin Pool wrote:
| I see you still partially work in terms of generating a changeset to
| apply.  Do you think we should do that rather than just operating on the
| trees?

Using changesets seems to reuse code quite effectively.  I bet
libbazing/merge.py would be almost as long as libbazing/changeset.py if
we weren't using changesets.  This way, optimizations of changeset
generation are also merge optimizations.  Plus, hairy stuff like the
renames/adds/deletes code should only exist be implemented in one place
(or less!), if at all possible.

It does mean we do a diff on each changed file, then throw it away.  But
that's O(changed files), so probably not a big concern.

| So you like the @ syntax?  I was thinking about actually making it ./@12
| rather than just . at 12 to make it a bit less likely to collide.

Sure, I'm happy to do it that way.

Aaron
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFCZzgV0F+nu1YWqI0RAkpWAJ90YjTD+Sl6IH5TLweXnlOK9mDMWgCffY5c
CT+S2U9Qz4zrTMlewZKv4+g=
=vrUe
-----END PGP SIGNATURE-----