RE: Model Digest Algorithm

>The algorithm was merely indented to illustrate the approach.

Cool.

>
>Currently it is a XOR over all statement digests having advantage that
>model digest can be easily recomputed dynamically when statements are
>added/deleted.
>

I'm not a cryptographer, so I'm just going from general knowledge here.  We
are looking for a cryptographically strong digest function D of a set S, of
digests of statements.

So, by definition, D must be such that it is computationally infeasible to
find two sets S1 and S2 such that:

	D(S1) = D(S2)

The suggestion is that XOR is such a function. Effectively, the question is,
given a stream of effectively random numbers, how hard is it to find two
subsets which XOR'd together, result in the same value.

My guess is that it is not crypto hard, but I've asked one of our crypto
guys so hopefully will know for sure soon.

> ... Are you aware of other symmetrical algorithms that are
>more secure and have the same nice property? If security of the current
>approach is insufficient, this property can be dropped.

Aw hell, I'm the wrong person to ask.  I'm not a crypto guy.  Since you ask
the
question, I take it that the security of XOR as a digest aggregator function
is an open question in your mind too.

>I'm putting together another bunch of updates and will release another
>version of the API soon.

In that case, would it be timely to offer some feedback based on my
implementation experience using the current one?  I've been accumlating a
few comments.  Didn't want to send them to the list piecemeal and I'm not
finished yet.  But if now is a good time for you, I could write up what I've
got.

Brian McBride
HPLabs

Received on Sunday, 30 April 2000 13:12:59 UTC