home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.crypt
- Path: sparky!uunet!europa.asd.contel.com!darwin.sura.net!wupost!cs.utexas.edu!milano!cactus.org!ritter
- From: ritter@cactus.org (Terry Ritter)
- Subject: Re: Growth of mcrypt due to packets
- Message-ID: <1992Aug16.171913.13994@cactus.org>
- Summary: Corrections and Comments
- Keywords: mcrypt
- Organization: Capital Area Central Texas UNIX Society, Austin, Tx
- References: <1992Aug15.162947.688@mintaka.lcs.mit.edu>
- Date: Sun, 16 Aug 1992 17:19:13 GMT
- Lines: 49
-
-
-
- My thanks to those who remind me that each byte has eight bits.
- The values should be 95^1024 keys, controlling 2^8192 possible
- cipherings. But I'm not sure it matters much.
-
- Now that I think about it, Mass Encryption is odd in that the
- User Key is so closely connected to the internal key. Of course,
- any cipher can be insecure with a short key, but the difference
- here is that the User Key is so directly related to the ciphering
- process itself. Whereas a short key in other ciphers may select
- in a well-distributed fashion from a small number of choices
- hidden among all the possible cipherings (and even this is weak),
- in this cipher a short key means that only a small portion of the
- cipher is really unique. The rest of the cipher is functionally
- derived from the unique portion which is exposed to attack.
-
- Whereas in other ciphers a User Key selects a particular
- ciphering as a unit, in this cipher *each character* in the User
- Key *independently* selects a portion of the cipher, a known
- particular byte. This means that individual characters of the
- User Key can be isolated and attacked separately. This is unusual
- and very weak.
-
- Indeed, each User Key character may control *several* particular
- bytes, due to the key expansion formulas; such bytes could be
- anagrammed in an attempt to find a single solution for the
- associated key character, independent of the rest of the key.
-
- When selections are poorly distributed, the number of possible
- cipherings tells the wrong story. For example, Simple Substitution
- on 26 letters should select from 26! (that's about 2^88) possible
- cipherings. Nevertheless, people somehow manage to solve these
- ciphers virtually by examination. Trial-and-error occurs, but it
- is not a *random* sort of trial-and-error; it is instead directed
- by probable words and probable adjacent characters. For solution
- purposes, each letter of the substitution key can be worked on
- separately, most of the time.
-
- Evaluating the number of possible cipherings is useful only to the
- extent that other attacks have been eliminated, and brute force
- remains as the only available attack. But that is not the case in
- Simple Substitution, and that is not the case in Mass Encryption
- either.
-
- ---
- Terry Ritter ritter@cactus.org
-
-
-