CDR: RE: New OLD cryptograph patent for NSA

[David Honig]

At 10:55 PM 10/12/00 -0700, Bill Stewart wrote:
>
>It's often hard to tell whether a physical object violates
>a given patent or not - bitspace is often pretty subtle stuff,
>especially if it's manufacturing methods rather than end results
>that are the subject of the patent.
>
>But increasingly, the interesting patents are (gak) software,
>(gak gak) algorithms, and (gak phfft) business methods,
>all of which are basically bits that are potentially easy to make
untraceable.
>Sure, if you actually have to ship somebody the infringing code
>on a CDROM or DVD, then there's some traceability,
>but that's decreasingly interesting as a distribution method.

Before anyone else starts, don't take my hypothesis that patents 
will survive in a crypto-abundant world as endorsement for the
USP&TO lunacy we've all seen.  You can limit the context to 
physical-object or manufacturing patents.  It is has been pretty
well argued that bits will be very hard to regulate in any sense
of that word; and also that USPTO has been doing too much PCP during
work hours.