Re: Fwd: [DES-ANNOUNCE] New clients and 'spamming'

Justin Dolske (dolske@cis.ohio-state.edu)
Mon, 5 May 1997 23:46:18 -0400 (EDT)


On Mon, 5 May 1997, Dan Oetting wrote:

> You missed the point. I was following your process but instead of telling
> the client exactly what decryption result it was searching for you provide
> a limited number of bits of the result

Ohh... I get it. Interesting variation!

The only real problem that comes to mind is that you've got to carefully
tune how many bits of the expected result you give the client. Give too
many, and the odds are very good that there's only 1 result like that in
the key block. Give too few bits to the client, and the client find an
absurd number of matches in the keyblock. [I guess for matching N bits,
the client can expect to find (keys in block)/2**N matches]

This also somewhat complicates both the protocol and optimizing the
client, but those are probably fair tradeoffs. The extra work for the
server could also be a problem.

Please implement this, and send us the results. :-)

Justin Dolske <URL:http://www.cis.ohio-state.edu/~dolske/>
(dolske@cis.ohio-state.edu)
Graduate Fellow / Research Associate at The Ohio State University, CIS Dept.
-=-=-=-=-=-=-=-=-=-=-=-=-=- Random Sig-o-Matic (tm) -=-=-=-=-=-=-=-=-=-=-=-=-
I am glad I will not be young in a future without wilderness.
-- Aldo Leopold