[MERGE] New option --author for commit

[Aaron Bentley]

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

John Arbash Meinel wrote:
> John Arbash Meinel has voted tweak.
> Status is now: Conditionally approved
> Comment:
> I'm not sure you should remove the test entirely.Instead "assert author
> not in revprops" you could:
> 
> if author in revprops:
>   raise AssertionError()
> 
> which will always be raised.

Well, it seems a long way to go for a programming error.  So I was fine
with an assert.  Really, if we can't use assert for programming errors,
we shouldn't use them at all.

> The only other question is whether the Author should be displayed in
> short logs, rather than the committer.

I vote yes.  I'm forever crediting people in my commit messages so that
they'll appear in logs.

> If we change the pqm to use
> --author as the person requesting the submission (since they, in effect,
> authored the merge commit), it would make 'bzr log --short bzr.dev' a
> bit more interesting. Though really the primary author should be the
> person who does the patch, and then we would have submitter. (Either way
> makes bzr log --short more interesting for bzr.dev.)

Most PQM commits credit the author.  I think PQM should use --author,
but should use the "apparent author" of the head revision being merged:
 - "author" if present
 - "committer" otherwise.

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

iD8DBQFGwwHJ0F+nu1YWqI0RAlwDAKCIbBfRyRtT/hOL2HY8GZ/DCEwRzwCgh6Ye
+H+QJM562NYvcISxsQcodyA=
=Fu/9
-----END PGP SIGNATURE-----