From: cme@ellisun.sw.stratus.com (Carl Ellison)
Date: Mon, 4 Oct 93 11:39:08 PDT
To: cdodhner@indirect.com
Subject: Re: POISON PILL
Message-ID: <9310041835.AA15292@ellisun.sw.stratus.com>
MIME-Version: 1.0
Content-Type: text/plain



>> Something else you can do is use a cipher which takes two input streams
>> and merges them into the one file, with one key extracting the 'harmless'
>> information and another extracting the 'harmfull' information. 
>> 
>> Matthew.
>> --
>> Matthew Gream, M.Gream@uts.edu.au. "... encryption is the ultimate means of
>> Consent Technologies, 02-821-2043.  protection against an Orwellian state."
>
>Has anybody writen one? Can we try for this as a feature in the next pgp?
>Should I write it myself? Of course I should... silly question... Ok does
>anybody have any suggestions or wishlists? Lemme know.
>

AFAIK, the only way to do this is with a Vernam OTP.  You have a key file (A)
the same length as your real data (B) -- encrypt the data by XOR to get (C).
Then you take an innocent text (D) and XOR with (C) to get an alleged key
file (E).  You hide (A) someplace, destroy (B) and (D).  Leave (C) around
and put up just enough resistence in letting folks have (E).

Does anyone know a simpler way?  I'm willing to bet that it can be proved that
the key would have to be at least 1/8 the length of the message in order for
this to work but I don't know of any schemes using less than the message
length to do it.

 - Carl