Antti,

     Ken is organizing a co-operative group to collect and analyze
computer viruses, starting with the people in the anti-virus industry
(like us) who are already doing this informally.  Ken is concerned (and
rightly so!)  with transmitting virus samples over the Internet in
plaintext.  So, we're looking for a DES implementation that we can use.
Since most of the viruses we're concerned about right now are DOS
viruses, we're looking for a DOS-based DES program.  We need to have
a DOS DES program available to all the participants in the co-op
before we can get started.  So we're *very* interested in your programs!

     We wanted to contact you before spending a lot of time
understanding what needs to be changed in your code for DOS.  (We didn't
want to do the work and write it up if you weren't going to be
interested, after all!)  Here are the things we've run into so far:

- DOS filenames must be 8 or fewer characters long.  So, many
  of the filenames you use need to be abbreviated, and references
  to them in INCLUDE statements must also be abbreviated.

- Most DOS MAKE facilities only implement a subset of the usual
  Unix MAKE language.  We're using the IBM C/2 compiler, and
  had to write our own MAKE file for your programs.  I think
  we could just provide that to you directly?

- Our compiler didn't like one of the constant definitions /
  initializations that you used (I don't have my notes with
  me right now, so I can't tell you which one).  That's
  easily fixed.

     You're right that endian stuff needs to be looked at.  What are the
function return types that you're worried about?

     Anyway, we would be happy to help you make a DOS version available.
If you are interested, we can start writing a list of the changes that
would be needed.  Should we go ahead?

                          Thanks,

                           Steve

P.S.  What convention did you use for the "last block rule" in CBC mode?
      Did you use the ANSI X9 convention?

