4.5.4. A Cypherpunks outlook: taking the abstractions of academic conferences and making them concrete
- One thing Eric Hughes and I discussed at length (for 3 days of nearly nonstop talk, in May, 1992) was the glacial rate of progress in converting the cryptographic primitive operations of the academic crypto conferences into actual, workable code. The basic RSA algorithm was by then barely available, more than 15 years after invention. (This was before PGP 2.0, and PGP 1.0 was barely available and was disappointing, with RSA Data Security’s various products in limited niches.) All the neat stuff on digital cash, DC- Nets, bit commitment, olivioius transfer, digital mixes, and so on, was completely absent, in terms of avialable code or “crypto ICs” (to borrow Brad Cox’s phrase). If it took 10-15 years for RSA to really appear in the real world, how long would it take some of the exciting stuff to get out?
- We thought it would be a neat idea to find ways to reify these things, to get actual running code. As it happened, PGP 2.0 appeared the week of our very first meeting, and both the Kleinpaste/Julf and Cypherpunks remailers were quick, if incomplete, implementations of David Chaum’s 1981 “digital mixes.” (Right on schedule, 11 years later.)
- Sadly, most of the abstractions of cryptology remain residents of academic space, with no (available) implementations in the real world. (To be sure, I suspect many people have cobbled-together versions of many of these things, in C code, whatever. But their work is more like building sand castles, to be lost when they graduate or move on to other projects. This is of course not a problem unique to cryptology.)
- Today, various toolkits and libraries are under development. Henry Strickland (Strick) is working on a toolkit based on John Ousterhout’s “TCL” system (for Unix), and of course RSADSI provides RSAREF. Pr0duct Cypher has “PGP Tools.” Other projects are underway. (My own longterm interest here is in building objects which act as the cryptography papers would have them act…building block objects. For this, I’m looking at Smalltalk of some flavor.)
- It is still the case that most of the modern crypto papers discuss theoretical abstractions that are not even close to being implemented as reusable, robust objects or routines. Closing the gap between theoretical papers and practical realization is a major Cypherpunk emphasis.
