Re: Limitations of Emacs' vc when using modern backends

[Andre Spiegel <spiegel@gnu.org>]

Juliusz, [ RMS, please see comments at the end ]

> 1. It should be possible for a backend to override
> vc-previous-version.

Agreed, it surely has to be backend-dependent.

> (2) CVS versions files, while modern systems version trees; the effect
> is that the correct ``previous version'' depends on the file name.

That's also reasonable.  Out of curiosity, I wonder whether that is
really the only place you came across where "versioning files" and
"versioning trees" conflict?  I would expect that there are more places,
and vc-previous-version sounds like a pretty inocuous, although no doubt
relevant spot.

> I believe that the functionality of vc-darcs-rev-to-hash (getting a
> canonical revision identifier for a given revision identifier) should
> be moved into vc itself, and called whenever vc gets a revision from
> the user.  The default implementation would just be the identity, but
> it should be possible for a backend to override it.

Although the hash identifier seems rather specific to Darcs, I can see
how this could be useful for other backends -- you can also refer to CVS
and RCS revisions symbolically (both branches and individual revisions),
and it would be nice to canonicalize those too (turning them into real
version numbers), if only optionally.  If this is done, auto-completion
for symbolic revision names would be a logical next step (but certainly
not now).

So your suggested changes seem very low-impact to me (they would be
trivial from the perspective of other backends).  I don't see a reason
not to make them right now, but perhaps RMS should have the final word
on this.  Richard?