Path: illuminati.io.com!uunet!gatech!psuvax1!news.cc.swarthmore.edu!netnews. + upenn.edu!mail2.sas.upenn.edu!egendorf From: egendorf@mail2.sas.upenn.edu (Robert Egendorf) Newsgroups: sci.crypt Subject: Re: Algorithms Date: 13 Nov 1994 00:37:02 GMT Organization: University of Pennsylvania Lines: 31 Message-ID: <3a3n3e$ihn@netnews.upenn.edu> References: <199411120422.WAA04049@pentagon.io.com> NNTP-Posting-Host: mail2.sas.upenn.edu X-Newsreader: TIN [version 1.2 PL2-upenn1.1] Terry Ritter (ritter@io.com) wrote:: 2) Cloak2 -- a 992-bit key, plus 992-bit message key, file cipher. : The mechanism is a two-level nonlinear Dynamic Substitution : combiner stream cipher, with 16 second-level combiners, and huge : nonlinearized RNG's. The ciphertext is binary. : Both of the above are commercial secret-key ciphers currently : implemented for MSDOS (and which now function well under Microsoft : Windows). They are not exportable. But while the ciphers proper : are important, probably the major practical value of these systems : lies in extensive key-management by open alias, with the ability : to cleanly update keys without disturbing most users. Central key : management can be important to allow a business to retain access : to information for which it originally paid. And archived : ciphertext under old, replaced keys can be accessed easily. : For example, I used a 992-bit key (over 10 times as *long* as : an 80-bit "large enough" secret key) in my Cloak2 stream cipher, : because it was a reasonably clean design without too much overhead. : Since that design uses an RNG with about 37.8K (Bytes) of state, : I could instead "easily" equip it to use honest 310,048-bit keys, : provided we are willing to generate and keep such keys. (The user : keys would be kept inside enciphered alias files, of course, but a : 40K message key on a 2K message might seem a little much.) Has anyone else evaluated the Cloak2 cipher? What tests has it been subjected to? What is Mr. Ritter's background in cryptography?