From: cme@ellisun.sw.stratus.com (Carl Ellison)
Date: Mon, 18 Oct 93 07:47:16 PDT
To: blaster@rd.relcom.msk.su
Subject: Re: Monitor radiation overlooking.
Message-ID: <9310181439.AA21879@ellisun.sw.stratus.com>
MIME-Version: 1.0
Content-Type: text/plain


>Once more, sorry for my poor English.

Your English is much better than my Russian. :-)

I've cleaned up spellings and punctuation in the quote below, to reflect my
understanding.  Was I correct?

>I want to say that DES isn`t group. We can`t say
>that:
>A k1, k2; E k0; ST des(k2, des(k1, text))=des(k0, text)
>        (A means "for all", E means "exist", ST means "such that");
>
>or: E k1, k2; A k0; ST des(k2, des(k1, text))!=des(k0, text)
>
>I say that:
>E f(k, text); A k1, k2; E k0 ST len(k0)=len(k1)=len(k2) and
>des(k2, des(k1, text))=f(k0, text)
>
>f() is like des() but f() is not des()!!!
>My friend says that the intruder must try only 2^56 keys for des(des()).


What's missing in the last statement is the qualification of <text>.
Is it

	A text
or
	E text

?

Meanwhile, some method of learning f would be nice :-).

Is it also claimed that:

E g(k,text); A k1,k2,text; E k0  ST len(k0)=len(k1)=len(k2) and
des(k2, f(k1, text)) = g(k0,text)

for f(,) as defined above?  (That is, does this method continue to
apply for an indefinite number of DES copies?)

 - Carl