1995-01-12 - Re: Multiple symetric cyphers

Header Data

From: grmorgan@freenet.vcu.edu (Greg Morgan)
To: cypherpunks@toad.com
Message Hash: bc41f918ccccb603a78e4aed8bbc63aa2755658c8d824dc9749eefc23d33385d
Message ID: <9501122245.AA18097@freenet.vcu.edu>
Reply To: N/A
UTC Datetime: 1995-01-12 22:46:08 UTC
Raw Date: Thu, 12 Jan 95 14:46:08 PST

Raw message

From: grmorgan@freenet.vcu.edu (Greg Morgan)
Date: Thu, 12 Jan 95 14:46:08 PST
To: cypherpunks@toad.com
Subject: Re: Multiple symetric cyphers
Message-ID: <9501122245.AA18097@freenet.vcu.edu>
MIME-Version: 1.0
Content-Type: text/plain




>
>> But selecting a single cipher is just as much a fixed policy as a
>> randomly selected one is.  Far better to let the user pick a policy,
>> both about sent and accepted ciphers.
>
>If you do give the user control, what is an acceptable mechanical
>implementation? Let's say I have a file encryptor which allows the
>user to choose between DES, 3DES, IDEA, Diamond, and RC5. Must I
>require the user to tell that program what cypher was used to encrypt
>the file she wishes to decrypt?
>
>Is storing the cypher type as part of the encrypted file a weakness?

Perhaps an MD5 of the password could be encrypted along with the
plaintext using the method(s) of choice.  At the decryption
phase, the password would be MD5 hashed again and the block of
bytes the size of an MD5 hash would be "decrypted" with each of
the methods... which ever one matches the original MD5 hash
would be used to decrypt the rest of the cyphertext.

This method wouldn't leave a known plaintext in the file to
attack.

--
-----------------------------------------------------------------------------

Greg Morgan <grmorgan@freenet.vcu.edu> | "I dunno Brain, me and
Pipi





Thread