Cyphernomicon Top
Cyphernomicon 17.10

The Future:
Crypto Research


  17.10.1. Academic research continues to increase
  17.10.2. "What's the future of crypto?"
           - Predicting the future is notoriously difficult. IBM didn't
              think many computers would ever be sold, Western Union
              passed on the chance to buy Bell's telephone patents. And
              so on. The future is always cloudy, the past is always
              clear and obvious.
           - We'll know in 30 years which of our cypherpunkish and
              cryptoanarchist predictions came to pass--and which didn't.
  17.10.3. Ciphers are somewhat like knots...the right sequence of moves
            unties them, the wrong sequence only makes them more tangled.
            ("Knot theory" is becoming a hot topic in math and physics
            (work of Vaughn Jones, string theory, etc.) and I suspect
            there are some links between knot theory and crypto.)
  17.10.4. Game theory, reputations, crypto -- a lot to be done here
           - a missing link, an area not covered in academic cryptology
              research
           - distributed trust models, collusion, cooperation,
              evolutionary game theory, ecologies, systems
  17.10.5. More advanced areas, newer approaches
           + some have suggested quasigroups, Latin squares, finite
              automata, etc. Quasigroups are important in the IDEA
              cipher, and in some DES work. (I won't speculate furher
              about an area I no almost nothing about....I'd heard of
              semigroups, but not quasigroups.)
             - "The "Block Mixing Transform" technology which I have
                been promoting on sci.crypt for much of this spring and
                summer is a Latin square technology.  (This was part of
                my "Large Block DES" project, which eventually produced
                the "Fenced DES" cipher as a possible DES
                upgrade.)....Each of the equations in a Block Mixing
                Transform is the equation for a Latin square.  The
                multiple equations in such a transform together represent
                orthogonal Latin squares. [Terry Ritter, sci.crypt, 1994-
                08-15]
           + But what about for public key uses? Here's something Perry
              Metzger ran across:
             - ""Finte Automata, Latin arrays, and Cryptography" by Tao
                Renji, Institute of Software, Academia Sinica, Beijing.
                This (as yet unpublished) paper covers several
                fascinating topics, including some very fast public key
                methods -- unfortunately in too little detail. Hopefully
                a published version will appear soon..." [P.M.,
                sci.crypt, 1994-08-14]
  17.10.6. Comments on crypto state of the art today vs. what is likely
            to be coming
           - Perry Metzger comments on today's practical difficulties:
              "...can the difference between "crypto can be transforming
              when the technology matures" and "crypto is mature now" be
              that unobvious?....One of the reasons I'm involved with the
              IETF IPSP effort is because the crypto stuff has to be
              transparent and ubiquitous before it is going to be truly
              useful -- in its current form its just junk. Hopefully,
              later versions of PGP will also interface well with the new
              standards being developed for an integrated secure message
              body type in MIME. (PGP also requires some sort of scalable
              and reverse mapable keyid system -- the current keyids are
              not going to allow key servers to scale in a distributed
              manner.) Yes, I've seen the shell scripts and the rest, and
              they really require too much effort for most people -- and
              at best, once you have things set up, you can now securely
              read some email at some sites. I know that for myself,
              given that I read a large fraction of my mail while working
              at clients, where I emphatically do not trust the hardware,
              every encrypted message means great inconvenience,
              regardless." [Perry Metzger, 1994-08-25]


Next Page: 17.11 Crypto Armageddon? Cryptageddon?
Previous Page: 17.9 Crypto Standards

By Tim May, see README

HTML by Jonathan Rochkind