[Cvsnt] Next release features

[Tony Hoyle]

Jonathan M. Gilligan wrote:
  Such a check could help address the greatest problem I see with CVS: there
> is no good way to check the integrity of a repository. This means that
> corruption in repositories can go undetected for years until you try to
> check out an old revision and discover that you cannot reconstruct it
> because one or more of the difftexts between the head revision and the
> target is corrupt.

It does actually do something like this as part of the client-server
protocol, so the data is already part of the commit somewhere, just not
written to RCS.  The problem is putting it into the rcs code without
breaking stuff - the rcs backend is the worst part of the cvs code.

This would tend to break the ability to hand fix the rcs file.  Although
it's rare for people to do that, it's a lifesaver if something does go
wrong and you have to recover.

Tony

_______________________________________________
Cvsnt mailing list
Cvsnt at cvsnt.org
http://www.cvsnt.org/cgi-bin/mailman/listinfo/cvsnt