brisbane: initial cut at a mergeline cache

[John Arbash Meinel]

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

Andrew Bennetts wrote:
> John Arbash Meinel wrote:
> [...]
>>> Why would you be numbering revisions that aren't in your ancestry?  To
>>> describe a pending merge?
> [...]
>> Or stuff like "bzr qlog branch1 branch2", yes. Or if you wanted to
>> consider dead heads in the repository, or ...
> 
> How does qlog label a revision that has different dotted revnos in branch1
> vs. branch2?
> 

I believe they fake a merge point, and use the current numbering. Note
that this gives ~ the same values that they would have if the revisions
were actually merged, given what I mentioned before.

I probably am slightly in favor of numbered from merge rather than
numbered from source. I just wanted to mention a specific benefit of the
latter, so that we are genuinely evaluating them.

> I'd be happy to have non-ancestor revisions be displayed with some different
> revision number scheme.  annotate already does this with uncommitted changes
> (and maybe pending merges?), using a "?" in the revno.  Failing all else
> there's the revid, of course...
> 
> -Andrew.
> 
> 

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAknUGuMACgkQJdeBCYSNAANePwCfaqTE43hcfDJ/h9PrbLUrmKXX
PxEAniOIxnbTc/6OPpJIoMHdTRwnJMXJ
=b7fG
-----END PGP SIGNATURE-----