From: "John A. Limpert" <johnl@radix.net>
Date: Tue, 12 Sep 95 07:32:41 PDT
To: "thad@hammerhead.com>
Subject: RE: 64 bit crypto
Message-ID: <01BA80F9.97A820A0@dialin18.annex1.radix.net>
MIME-Version: 1.0
Content-Type: text/plain


<code deleted>

>Then the prepare_key routine would take much much longer.
>
>The idea is that a 64 bit crypto routine can be arbitrarily
>secure against brute-forcing, if you are willing to pay a
>runtime penalty every time you use it.

My thought was that there might be shortcuts that the attacker
could use to compute the permutation of the state array in less time.

I wrote a small test program to look for cycles and weak keys in the
65536 iteration permutation. I assumed a 40 bit key. One weak key
was found, 0x0101010101, that produces a 255 iteration cycle.
375 randomly chosen keys were tested and no cycles were found.

There ought to be a faster way to compute the permutation with
a more analytical approach. My skepticism tells me that nothing
is a replacement for more key bits.

In a semi-related question, why don't we see Feistel systems with
larger numbers of rounds? Would a modified DES with 256 rounds be
any more secure than standard DES?