[net.crypt] Want fast public-key system

gary@arizona.UUCP (Gary Marc Levin) (04/20/84)

Can anyone offer a public-key crypto-system that can be efficiently
implemented, say in Pascal? It need not stand up to sophisticated
attacks.  I am willing to trade weakness to attack for time.  The
adversaries are beginning programmers with limited computer facilities,
so that even a key space of 1M would probably circumvent exhaustive
search.

Here is the situation spelled out in some detail.

    The VAX/VMS system used for academic computing (in contrast with
    the VAX/UNIX system I am currently working on) leaves something to
    be desired in the area of file protection.  The problem is not
    security, but rather access.

    As an instructor, it would be convenient to have access to my
    students' files.  With this access, I could collect assignments and
    use the machine for preliminary analysis.  Unfortunately, if I have
    access to someone else's files, EVERYONE has access to those
    files.

    One solution would be for students to encrypt their files in such a
    way that only I could decrypt them and then lower the protection so
    that everyone can read them.  This has the added advantage that I
    only have access to files that they wish for me to see (preserving
    privacy).

    My current solution is to use a data compression program.  It
    effectively scrambles the file, destroying byte boundaries, depends
    on the entire file, etc.  The problem is that if the expand program
    is made available to many instructors, it will eventually be
    compromised.  The data compression technique is sufficient for my
    needs in so far as encryption goes.  The fact that there is
    essentially only one key possible makes the system of little
    long term use.

I should be able to encrypt or decrypt at least as quickly as I could
compile the same file.  While code would be nice (the system supports
Pascal and FORTRAN), I would be pleased to receive pointers to
algorithms.

Thanks for any suggestions that may appear.  Either respond directly to
me or post to the net.  I will post a summary of responses.
-- 
Gary Levin / Dept of CS / U of AZ / Tucson, AZ 85721 / (602) 621-4231

bill@utastro.UUCP (William H. Jefferys) (04/22/84)

The last two issues of "Dr Dobb's Journal" has a complete RATFOR
implementation of a version of the RSA public-key algorithm.  I don't know
how fast it is, but the fact that the code is already written for
you may be of interest.
-- 

	Bill Jefferys  8-%
	Astronomy Dept, University of Texas, Austin TX 78712   (USnail)
	{ihnp4,kpno,ctvax}!ut-sally!utastro!bill   (uucp)
	utastro!bill@ut-ngp			   (ARPANET)