Re: [rdiff-backup-users] Interesting write-up of 'compare-by-hash'.

[David S.]

> > 
> My impression was the write-up was talking about an inadvertant
> hash-collision, not hackers.
> 
> Since the hashing process is lossy (ie. non-reversable), then it is
> possible that two totally different data sets could generate the same
> hash, and in turn invalidate the backup checksum check.
> 
> I don't know what the odds are of this happening with rdiff-backup.

The presence or absence of collision is a property of the hash function
used - sha1, md5, whatever - and not 'rdiff-backup'.  Actually, any
hash function will have infinitely many collisions.  The proof is an
exercise for the reader ...

> 
> I assume that they are exceedingly small, but not zero.

Over the entire domain of a hash function, the probability is more
like one than zero.  Commonly used hashes are commonly used because they 
have infrequent collision over their practical domain.  That's an empirical
observation, however, not established, as far as I know, by any sort
of proof.

David S.

>