textfiles/programming/CRYPTOGRAPHY/mutation.txt

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

             MUTATi0N: Cryptography for Hackers/Phreakers & Others
             =====================================================

                       Version 1.00, Released 7/27/91

                           Written By Professor Falken
                               Copyright (c) 1991

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


INTRODUCTION:
------------
   Mutation was written with the Hacker, Phreaker, Pirateer and Almost-Average
Joe in mind.  Its purpose is to extend highly secure cryptographic systems
to the end user of the program.  This program uses a modified National
Bureau of Standards DES (Data Encryption Standard) algorithm to achieve
this purpose.  Along with the normal Electronic Code Book cipher there is
also the option of creating a Block-Chained cipher.

   I decided to write THIS program because I felt there was an inadequate
supply of encryption utilities out there. Let me explain, the encryption
utilities that were in the public domain (Shareware & Copyrighted)
either employed insecure encryption algorithms or the key (password)
length was much too short to make cracking it hard.  Take Central Point
Software's "PC-Secure" for example.  PC-Secure employs NBS's DES algorithm
but the key length is limited to 32 characters.  While the DES standard
is safe, a key length of 32 is not all that secure.  Not to mention that
I'm sure Central Point Software put some type of backdoor in the program
just in case SOMEONE made a ciphertext that NEEDED to be deciphered via
abnormal means.  Another SHAREWARE utility is "The Private Line" the only
thing private about the ciphers this program makes is they won't be able
to be read your cat.  The Private Line employs DES, however, their key
length is a miserable 16 characters long.  I think my 486-33 could bust
these ciphers in a matter of weeks.  Unlike these previous programs, and
the hordes of other programs which follow in their footsteps, I am proud
to release a highly secure, command-lined, low memory overhead, and
portable cryptographic tool.


BRIEFING:
---------

   Mutation has a possible key length of 118 characters for Electronic
Code Book mode, and 56 characters per KEY for Cipher Block Chaining's
2 key requirement. Not to mention there are NO known trapdoors written
into this program. As for the DES algorithm, thats another story
(See:TRAPDOOR THEORY).

   Mutation can be used in Electronic Code Book mode, which is the
default mode if no options are specified, and a Cipher Block Chaining
mode.  The ECB(Electronic Code Book) mode uses only ONE key.  A key
is the password that you want to use to encrypt & thusly also decrypt
the file with.  While ECB only uses ONE password, CBC(Cipher Block
Chaining) uses TWO passwords to encrypt the file with.

   The files you wish to encrypt can be really anything.  A list of
telephone numbers, friends names, passwords for various systems,
technical notes, where you hid the dead body etc.  You can really
encrypt anything. However, I would recommend at this time to stay
away from encrypting EXE & COM files.  There is a small bug in the
program which will add a few bytes to the end of each file.  While
this will not effect the output of an ASCII file, it may mess up an
EXE/COM file.  I hope to fix this bug in upcoming versions.


USING THE ECB OPTION:
-----------------------

   In order to invoke the ECB option (default) of Mutation, all that
is needed is knowing whether you wish to ENcrypt(+) or DEcrypt(-) the
file, what password(KEY) you wish to use, and the INCOMING filename
and what the OUTGOING filename should be. For instance, I wish to ENCRYPT
a file called TELEFONE.DOC, I will use "fatcuntwhorecumbelchinggutterslut"
as my password.  I suppose I will output the ENCRYPTED text to a file called
TELEFONE.DOC . You see, this makes sure that the original text is
OVERWRITTEN and never seen again. Here's how its done:

C:> m + fatcuntwhorecumbelchinggutterslut <telephone.doc> telephone.doc

C:>
   OK, so now you have this nice file called TELEPHONE.DOC which looks
so messed up, if you didn't know better, you'd swear its the face of
your neighbors crack-smoking child.

   So, after an amount of time elapses you decide you need a telephone
number out of that file.  No problem. Enter:

C:> M - fatcuntwhorecumbelchinggutterslut <telephone.doc> telephone.dec

C:>
   Now, you can read TELEPHONE.DEC and get the phone number you need.
But, if you noticed, I used a DIFFERENT filename for DECRYPTING the
file. The reason for this is just in case I fucked up the KEY and
it OVERWROTE my encrypted file with the incorrect decrypting KEY.
So when I check out telephone.dec and see that it is FINE! I can then
rename telephone.doc to something else, copy some bullshit file over
its place, then erase it.  See that way no one will be able to match
your enciphered text with your deciphered text and figure out your key.
See how easy it was?!

USING THE CBC OPTION:
---------------------
< NOTE: CBC IS NOT AVAILABLE IN VERSION 1.00 OF MUTATi0N >
< COMING VERSIONS WILL HAVE CIPHER BLOCK CHAINING ENABLED >


   The Cipher Block Chaining option is a little bit different from the
ECB option, not to mention a bit more secure(See TECHNICAL SPECS.& NOTES).
CBC requires you to enter TWO passwords(KEYS).  Lets say we wish to use
CBC on a file called TECHNOTE.DOC, we will use
"everyonewantstobeabigshottheres" & "toomanychiefsandnotenoughindians"
as the two passwords.  We will output our ciphered-text to a file called
TECHNOTE.DOC

C:> M + <cbc> everyonewantstobeabigshottheres toomanychiefsandnotenoughindians
<technote.doc> technote.doc

C:>
   Notice that all that text overran the current line and splashed down
onto the next.  This is NO PROBLEM.  Mutation will use up the WHOLE
command line for incoming keys/filename etc... So if you have a pretty
long password/s don't worry about it wrapping over, just be sure to
leave enough room for the filenames.  NOTE: Names without extensions
(IE ibmbios. ) ARE valid in Mutation.

   Ok, so you have this NICE enciphered text.  I bet you would like to
read it later huh?! Well, remember your KEYS, else your screwed.
Ok, LETS DECRYPT!

C:> M - <cbc> everyonewantstobeabigshottheres toomanychiefsandnotenoughindians
<technote.doc> technote.dec

C:>
   Thats how its done.  No problem huh?! Once you verify that technote.dec
came out correctly be sure to overwrite/rename/etc. the encrypted
ciphertextfile before going on.

TECHNICAL SPECIFICATIONS & NOTES:
---------------------------------

   The Data Encryption Standard Algorithm was conceived by IBM
in 1977 and was adopted by the National Bureau of Standards as
the nations Encryption Standard. The Data Encryption Standard
Algorithm is a block cipher that enciphers identical eight-byte
plaintext blocks into identical eight-byte ciphertext blocks.
Or more technically: it transforms an input block (file) into a
string of output bits of fixed length(output file).  The
enciphering/deciphering functions are such that every bit in the
output block depends jointly on every bit in the input block and
on every bit in the password. This DES algorithm is also known
as the Electronic Code Book (ECB).

   Certain files exhibit great regularity, with many identical
eight-byte blocks (for example, a memorandum always begins with
TO: ).  Normally under ECB encipherment, such identical blocks
of plaintext yield identical blocks of ciphertext (when using
the same key).  Repetitions of some other blocks may also be
visible; for example, telephone numbers in the plain text when
aligned with the eight-byte DES blocks, can be recognized as
identical in the ciphertext.  It is doubtful whether this
phenomenon is a serious weakness: to date no technique has been
found which uses it to determine the password or obtain usable
plain text of alphabetic/numeric files. However, if the existence
of identical blocks can be concealed, a cryptographic system is
strengthened at least intuitively, because the amount of information
available to the cracker is reduced. This is where Cipher Block
Chaining comes in handy.

   By using Mutation's Cipher Block Chaining option it is possible
to use DES in such a manner that identical eight-byte plaintext
blocks virtually always yield different ciphertext blocks, without
altering the security of the existing algorithm. Thusly if you are
encrypting a file that has a lot of repetition (IE- Indented source
code, Spreadsheet, Listing) please use CBC to strengthen the code.
However, as in life for every GOOD there is an EVIL, the CBC is
easily cryptanalyzed on the basis of a moderate amount of ciphertext.
You see, from enough enciphered short blocks, a cryptanalyst could
recover the key, and then decipher each individual short blocks.
But, I am not here to analyze various cipher systems, I'm here to
teach the use of this product.

CRYPTANALYSIS OF THE DES ALGORITHM:
-----------------------------------

   So you ask, "Is DES crackable?"  Well from a time standpoint, YES.
But before you say "WHAT? I thought NO ONE could read it!" lets ask these
questions:

   1. Is it reasonable to assume that an opponent can obtain the
      required information to decipher the ciphertext. (IE-Finding them
      written down, extracting them from your head, etc.)
   2. What is the computational COST of the attack?
      (IE- How much manpower, fast enough computing facilities etc..)
   3. How much time does it require?
      (Can my opponent wait 30,000 years for the cracking of my text?)

   So it while it is TRUE that someone will be able to read your
ciphertext, I really don't think they will care 10,000 years from now
anyhow.  But, just for the hell of it lets make a worst case hypothetical
situation(I hate them).  First, lets assume you have made a CBC cipher
with two keys 55 characters per key.  Second off, we MUST assume, that
for some reason you have pissed off the WRONG people in the government
(namely the people that effectively OWN the modified Cray Y-MP that
the National Security Agency has) and they deem it necessary that they
devote 100% of the computing resources to cracking your file.  Third,
we must assume your DEAD else, they would have used sodium-penathol
a long time ago to get the KEY from you.  Fourth- These government
people plan on having their jobs AFTER the current administration leaves
the White House.  If all these pre-conditions are met, WELL, your up
chunky-shit creek without a paddle.  Because probably in 3-4 years
they will have deciphered it.

   If DES keys were tested at the rate of one per 60 microseconds, the
key trial would require in excess of 68,000 years.  However, as critics
are quick to point out: Massive parallelism and pipelining, the time to
recover a key can be reduced to any value.  In a pre-calculated scenario
properly combining precomputation, a machine can be constructed that
would recover the key with high probability in one day.  The estimates
of the cost are $5,000,000 with a precomputation time of 2 years.
Even if these estimates are accepted, the way DES would normally be
used, with Cipher-Block Chaining & Initial chaining value, would negate
the advantage gained in the time-memory tradeoff.  Thusly, it is VERY
questionable if any organization would risk an investment of 5 million
dollars and two years of precomputation under these circumstances.
So really, your ciphers are very safe.


TRAPDOOR THEORY:
----------------

   Some critics of DES are leery about the methods of analysis used
by IBM to validate the DES, and all such methods used by the National
Security Agency, have been [CLASSIFIED] by the U.S. Government.
Therefore, some critics have inferred that one cannot be sure that the
statements by IBM and NSA about DES are as claimed.  You see, the
National Security Agency advised the designers that certain design
criteria were considered by the agency to be sensitive.  IBM was
requested not to publish the selection criteria.  Critics have
suggested that special properties might have been incorporated into
the design of the substitutions and permutations (aka Trap Doors)
which would give a cryptanalytic advantage to a knowledgeable
party.  No proof can be offered to refute the existence of trap
doors.  Even if IBM, the National Bureau of Standards, or the
National Security Agency were to reveal the criteria, there would
undoubtedly be critics who would suggest that not all were revealed.
The BOTTOM LINE is that after nearly 11 years of scrutiny, no
viable method of cryptanalyzing DES has been discovered and published.
Bell Telephone Laboratories & the Lexar Corporation examined the
properties of the substitutions and permutations.  While the Lexar
Corporation reported some "peculiar" properties of the substitution
boxes, to date, no feasible cryptanalytic for DES has been found.
To keep the public happy, the Senate Committee on Intelligence
conducted an investigation into the matter. The following summarizes
the conclusions that were reached:
                                  1. DES is more than adequate for its
                                     intended applications.
                                  2. IBM invented and designed DES.
                                  3. NSA did not tamper with the design.
                                  4. NSA certified that the DES was free
                                     of any known statistical/mathematical
                                     weakness.
                                  5. NSA recommended that the Federal
                                     Reserve Board use DES for electronic
                                     funds transfer applications.

   So if there is a TRAPDOOR or not, I really doubt they would risk
releasing of their trapdoor to the public & intelligence community
at large just to decipher some piss on the wall hackers file (Their
point of view).

PASSWORD LENGTH:
----------------

   In the past there has been a tendency among users of cryptographic
systems to use a word or a name as a key.  Today this procedure is
VERY unsafe.  There are a fewer than a million words in an unabridged,
and fewer than a million names in the largest telephone directory.
That number of potential keys could be tested in a few minutes on a
powerful computer of today.  Even the use of eight randomly chosen
ASCII characters is unsafe.  If a DES key is introduced directly,
it should consist of 32 or MORE ASCII characters
(IE-A-Z,a-z,0-9,and punctuation, and ALT-key combinations).  You see
by having Mutation being able to use 118 bytes of password, the user
might, for example, enter several words chosen randomly and
independently from an unabridged dictionary.  Or the user may enter
some phrase from a song or TV commercial. It is essential that these
choices be made in such a way that the key cannot practically be found
by an opponent, either by guesswork or by enumerating some plausible set
of keys.

HINTS & EXTRAS:
---------------
< NOTE: CIPHER BLOCK CHAINING (CBC) IS NOT AVAILABLE FOR >
< VERSION 1.00 OF MUTATi0N, FURTHER RELEASES WILL FIX THIS >

   Mutation in ECB mode can use KEYS up to 118 characters in length. In
CBC mode, it can use TWO KEYS, with a maximum of 56 characters per key.
That is assuming your using one letter filenames. Example:
C:> m + <cbc> passwordOne passwordTwo <A> B
This would give you A. as your input file and B. as your output file.

   The LARGER the password you use, the STRONGER the ciphertext becomes.
So a 118 character password is MUCH better than a 70 character password.
However, in Block Chaining, having two 56 character passwords are even
STRONGER than one 118 character password.

   For the STRONGEST ciphertext possible, I suggest using Cipher-Block
Chaining, with 56-char. passwords. Note, that you could use a
100-character password for KEY1 and a 18 character password for KEY2.
Or a 34-character password for KEY1 and an 84-character password for KEY2.
Just realize that the numbers can be swapped around, and just make sure
its as long as possible for the STRONGEST text.

   Sometimes making the LONG password is about as much bother as it
will be for someone trying to break it.  Here are some ideas for
making up these LONG keys:
                         1. Sentences from books.
                         2. Phrases from obscure songs.
                         3. TV/Radio Commercials.
                         4. A column from an old newspaper.
                         5. Make up a short story.
                         6. Make faces with the ASCII chars. :0 :) ;) etc.

   I'm sure you can think of other things to write about, like maybe
your first screw, but that may too personal, incase someone breaks it.

   ALSO make sure you PROPERLY dispose of the ENCRYPTED filename, that
you just decrypted. First copy a file over, then another, then to be
safe, copy another, then rename the fucker, THEN erase the SOB.

   It would also be nice to have this program setup in your PATH.
So that you can encipher stuff all over your harddrive! When the nice
men in jackets come to borrow your computer don't let them have all
your data that easily.

CONCLUSION:
-----------

   This documentation is A LOT longer than I intended it to be. In fact
I shouldn't have written all that, I probably confused a lot of the
readers.  Oh well, you should be able do discern how to run the thing
from the previous text and the small example built into the program.

   I would now like to thank Legions of Lucifer-Phuck for releasing
this product on their intercontinental distribution system.  I really
helps a program when it has access to a WIDE dispersal network. Again,
thanks LoL-PHUCK.

   Earlier, I mentioned 'portability' of this code. The code is in 'C'
and is quite portable, but I am not releasing the code for various
reasons. First, there is the possibility that someone will be able to
figure out how the DES code was modified by myself, and secondly
I plan to release this on AMIGA & MAC computers.  I plan on having the
AMIGA version out sometime in the summer of 91'.  The MAC will have
to wait a few more months.

   Anyhow, greets go out to: All X Legion of Doom members hanging out
and around, P/H/A & Doctor Dissector (KC3.0 is the BEST!), Prometheus,
Sir Hackalot, Ex-Phortune 500s UNITE!, Shawn-Da-Lay Productions -
Wheres the NEW version of Telix Hacker?!!, Brew Associates Where RU?,
L.E. Pirate(What happened to Dragonfire?!oh no)(What Cult of the Dead
Cow?), The Rebel(716), 2600 Magazine(About time I get that damn
subscription information-1.5 months later), Phrack Magazine, Mondo 2000
Magazine (Wheres my subscription fuckers?), Twisted Sector, The Ranger,
The Viper, Slavelord, and all the others who give me phree file points...
Not that I am partial to people who give me free points or anything.(haha)

   And for EVERYONE who I left out- YOU TOO!

   Anyhow, this ends this SUMMERs(1991) IBM program release schedule for
me. Hopefully, you will see something from me sometime soon, other than
updates to this and CNA Inquiry.

   Programs released to date:
                             1. Automated CNA Finder (1989)
                             2. Phreak Tools v1.x    (1989)
                             3. CNA Inquiry v2.x     (1991)
                             4. MUTATi0N v0.x Beta   (1991)

- Professor Falken
= Ex-Legion of Doom!
= Ex-Phortune 500



MUTATi0N - Version 1.00
Conceived & Written by Professor Falken 6/91
Copyright (c) 1991 Released by Legions of Lucifer-Phuck