From: mpj@netcom.com (Michael Paul Johnson)
Date: Mon, 4 Apr 94 22:58:08 PDT
To: cypherpunks@toad.com
Subject: Earn $271.82 if you break new crypto freeware.
Message-ID: <199404050559.WAA14989@mail.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain


-----BEGIN PGP SIGNED MESSAGE-----


    __
   /  \     New Cryptographic Freeware Available:  Data Lock
  /____\
  \    /    * Uses the Diamond Encryption Algorithm (slight variation on MPJ2)
   \  /     * Includes complete source code
    \/      * No patent infringement problems
            * Includes source code library for Diamond & Diamond Lite
            * Can be strong enough to protect very valuable data (see below)
            * Can be weakened enough to be exportable in executable form only
            * Ciphertext doesn't advertise its algorithm or key (stealth)

Documentation only (exportable):
ftp:csn.org//mpj/public/dlockdoc.zip or dlockdoc.tar.gz
ftp:ftp.netcom.com//pub/mpj/public/dlockdoc.zip or dlockdoc.tar.gz

Full package (including all source code & executable file):
ftp:csn.org//mpj/I_will_not_export/crypto_???????/mpj/dlock.zip or
 dlock.tar.gz
?????? and anti-export warning given in ftp:csn.org//mpj/README.MPJ

ftp:ftp.netcom.com//pub/mpj/I_will_not_export/crypto_???????/mpj/dlock.zip
 or dlock.tar.gz
?????? and anti-export plea given in ftp:ftp.netcom.com//pub/mpj/README.MPJ

Colorado Catacombs BBS 303-938-9654  DLOCK.ZIP

Data Lock itself may be useful, but its greater value lies in the fact that it
is written more to be used as a function library for people who want to
incorporate encryption into other applications.

The only change from MPJ2 to the Diamond Encryption Algorithm is that the key
expansion mechanism now distinguishes between keys of different lengths.  For
example, the keys "aaaaaaaa" and "aaaaaaaaaaaa" would have been equivalent in
MPJ2, but are not in Diamond.  Diamond Lite is just the logical contraction of
Diamond from a 16 byte (128 bit) block to an 8 byte (64 bit) block.  Even
though "Lite" is in the name, it yields better security for small numbers of
rounds because of the faster avalanche effect (1 bit to 64 in just 2 rounds
instead of 1 bit to 128 in 5 rounds).  MPJ2 and Diamond are derived from the
MPJ encryption algorithm, invented in 1989, but use an improved key scheduling
algorithm that eliminates the slight bias in the way the substitution arrays
were filled.

The following is a comparison of some of the symmetrical key ciphers available
to the general public today.  Included is a (somewhat subjective) strength
comparison to give a general idea of how Diamond and Diamond Lite compare:

ALGORITHM    BLOCK KEY      ROYALTY ECB   RELATIVE  APPROXIMATE WORK FACTOR
NAME         SIZE  SIZE     FREE?   MODE  SPEED     TO BREAK log base 2 of
             BITS  BITS             OK?             number of operations

DES          64    56       YES     YES   MEDIUM    43
3DES         64    112-168  YES     YES   SLOW      64-168
DIAMOND      128   variable YES     YES   MEDIUM    128-key size
DIAMOND LITE 64    variable YES     YES   VERY FAST 64-key size
BLOWFISH     64    variable YES     YES   VERY FAST 64-key size?
SHA-CFB      160   variable YES     NO    VERY FAST 80-160
MD5-CFB      128   variable YES     NO    VERY FAST 64-128
REDOC II     80    160      NO      YES   FAST      80-key size
REDOC III    64    variable NO      YES   VERY FAST 64-key size
KHUFU        64    512      NO      YES   ?         64-512
IDEA         64    128      NO      YES   FAST      64-128
MMB          128   128      ?       YES   FAST      128?

The "work factor to break" column is somewhat subjective, and is based on the
best attacks that I have knowledge of (hardly an exhaustive list), combined
with a GUESS at what attacks might succeed.  They also assume that the key
length is at least as long as the numbers indicated when the key length is
variable.  You should decide for yourself what you will believe. Don't put all
your eggs in one basket.

Besides those things listed above, some distinguishing characteristics of
Diamond and Diamond Lite include:

      * They can easily and securely take a pass phrase directly as a key.
      * They can be extremely fast when implemented in hardware.
      * They allow you to perform some security vs. speed and size tradeoffs.
      * Key setup is designed to make brute force attacks very costly.

OK, this is where I put my money where my keyboard is.  If you are the first
one to correctly decipher my challenge text (CHALLENG.ENC in DLOCK.ZIP),
enciphered with DLOCK.EXE, before midnight UTC, 29 April 1997, and follow the
redemption instructions contained within the challenge text, then I will send
you US$271.82 of my hard-earned money.  You must (1) reveal to me how you did
this, and (2) not break any laws in the process to collect the prize.  I know
that this isn't enough to justify much serious cryptanalysis, but it should
demonstrate that such a challenge is beyond the reach of the average hacker.

                  ___________________________________________________________
 |\  /| |        |                                                           |
 | \/ |o|        | Michael Paul Johnson  Colorado Catacombs BBS 303-938-9654 |
 |    | | /  _   | mpj@csn.org  ftp:csn.org//mpj/README.MPJ for crypto stuff |
 |    |||/  /_\  | aka mpj@netcom.com mpjohnson@ieee.org mikej@exabyte.com   |
 |    |||\  (    | m.p.johnso@nyx.cs.du.edu CIS 71331,2332 PGP key by finger |
 |    ||| \ \_/  |___________________________________________________________|

-----BEGIN PGP SIGNATURE-----
Version: 2.4

iQCVAgUBLaDz1j9nBjyFM+vFAQHJ8QP/UgnrRX0u5AAnEoOIuNPi1Y8yRPrY7U3R
BWTb04eyi1hqSuWnVQaAkINp84R5d/PhyS7wa5xEEoq+UmhISEoGHoSVc6e2QWr+
xsSR5vjvUQpc5zkPIdkFOpVb94aCUCDHh5Zv4bU6WsVoKI+zAXSrRDL7o4zhwfxp
+H6ov+NPI1M=
=9Ul1
-----END PGP SIGNATURE-----