"bzr merge" and "bzr log -n0"

[Eli Zaretskii]

> Date: Sun, 29 Aug 2010 04:57:33 -0500
> From: "Matthew D. Fuller" <fullermd at over-yonder.net>
> Cc: nils-gmane at ackermath.info, bazaar at lists.canonical.com
> 
> History is all accessed by dereferencing parent pointers on the
> revisions, so the presence of one state implies the presence of all
> its predecessors.  If we don't want to have 2, we can't claim to
> have 4.

Right, thanks.