Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!caip!nike!ucbcad!ucbvax!jade!ucbtopaz!newton2
From: newton2@ucbtopaz.berkeley.edu
Newsgroups: net.crypt
Subject: Re: Efficient identification and signature schemes
Message-ID: <1016@jade.BERKELEY.EDU>
Date: Tue, 29-Jul-86 03:45:50 EDT
Article-I.D.: jade.1016
Posted: Tue Jul 29 03:45:50 1986
Date-Received: Tue, 29-Jul-86 19:00:57 EDT
References: <8607201331.AA27476@ucbvax.Berkeley.EDU> <1474@well.UUCP> <990@jade.BERKELEY.EDU> <1508@well.UUCP>
Sender: usenet@jade.BERKELEY.EDU
Reply-To: newton2@ucbtopaz.UUCP ()
Organization: University of California, Berkeley
Lines: 15

Gee, if the *author* of Mailsafe's documentation is so easily confused,
God help the reader :>).

One more time: RSA is painful and relatively slow, DES can be fast.
Mailsafe (I think) uses RSA for key passing and authentication/signature.
To the extent that actual file encryption doesn't use RSA, it can be
fast (your 5Kbyte text file in ten seconds); even with fast and
clever schemes, RSA (I'm suggesting) takes tens of seconds to decrypt a
512-bit block (i.e. using a 512-bit modulus) via Mailsafe on a stock
PC. This only happens as semi-fixed overhead for a given transaction,
but the point is relevent to the earlier assertion that RSA Data Security's
products could do RSA orders of magnitude faster than was commonly believed
possible. This isn't true, and they don't claim it.

Doug Maisel