[Dshield] Crypto Question

[David Brodbeck]

On Mar 4, 2009, at 5:10 PM, John Hardin wrote:
> My point is *don't assume the algorithm is flawless*. If you design  
> your
> protocols with the assumption that the algorithms are imperfect,  
> then your
> protocol is robust when the flaws are actually found.
>
> If certificate signing, and PGP et. al., and other protocols had  
> included
> provision for signing with two hashes using different algorithms,  
> then we
> would not be seeing attacks like this:

To me this seems like basic "defense in depth."  Engineers would call  
it redundancy.

-- 

David Brodbeck
System Administrator, Linguistics
University of Washington