🚀 go-pugleaf

In article <17pj4fINNm8i@early-bird.think.com> barmar@think.com (Barry Margolin) writes:
> In article <FRIEDMAN.92Aug29214928@nutrimat.gnu.ai.mit.edu> friedman@gnu.ai.mit.edu (Noah Friedman) writes:
> >   Axman's message suggests that the data format or the conversion
> >algorithm may be patented as well as (in its details) trade secret.  This
> >means that even if we figure out how to do the job (which may not be
> >impossible), we'd be sued for implementing it.
> 
> How can something be "patented as well as trade secret", when the two are
> mutually exclusive?  A patent by definition makes the details public, while
> a trade secret requires that you guard the details (anyone you disclose
> them to must agree to keep your secret).

Perhaps he meant that the algorithm would be patented but the design
details would remain a trade secret; in other words, the "how" would
be public, but the "why" would remain private.  I believe the Data
Encryption Standard is set up this way: the algorithm is public
knowledge, but its design is kept secret by the NSA (or whichever
government agency formulated it).

Jim
--
   _    Jim Thompson           |  Wanted:
 _| ~-  thompson@wg2.waii.com  |	Nick Park's "Grand Day Out"
 \,  _} Western Geophysical    |	on VHS-format videotape
   \(   Houston, Texas         |  Email me if you know where I can buy it.