Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site allegra.UUCP
Path: utzoo!watmath!clyde!floyd!harpo!ulysses!allegra!jdd
From: jdd@allegra.UUCP (John DeTreville)
Newsgroups: net.crypt
Subject: Re: Random Pad is Correct
Message-ID: <2368@allegra.UUCP>
Date: Wed, 28-Mar-84 13:45:41 EST
Article-I.D.: allegra.2368
Posted: Wed Mar 28 13:45:41 1984
Date-Received: Fri, 30-Mar-84 00:15:46 EST
References: <2363@allegra.UUCP>, <1306@sdccs6.UUCP>
Organization: AT&T Bell Laboratories, Murray Hill
Lines: 24

    > Someone quoted (or misquoted) Divida as saying true random bits
    > sequences could not be used for one time pads.  This was because one
    > could get a long run of zeros as long as the message and thus wind up
    > sending the text plain.
    > 
    > That is an invalid argument.  If the enemy sees a message that says
    > "launch all missles", he does not know that the key was zero!  The
    > chances are exactly that same the the key was some pattern of bits that
    > converted any arbitrary message into "launch all missles".  It is
    >  completely correct to use totally random bits in a one time pad.

    It is not!  The chances that a message in english crypt into another
    message in english seem much less likely than just a string of zeros.
    Plus, the enemy could say, "Gee, they must have forgotten to crypt it."

This reply shows how intuition and mathematics can give differing answers.
In such cases, I would suggest that mathematics should win out.'

This reply also illustrates how foolish it is to depend on netnews for one's
technical information.

Cheers,
John ("Low Probability") DeTreville
Bell Labs, Murray Hill