IIRC, not all ECC stuff is patented, only curves over GF(q), q even,
which can be implemented efficiently using two-valued logic.
As I read it, this I-D includes these.
But it also includes Curves over GF(q), q prime or a power of some
odd prime. I wouldn't appreciate it much to exclude some stuff,
now that we have a definition including simply every finite field.
But my suggestion doesn't require an application to use binary
fields, and other curves are also given (and own for which noone
could claim any rights could be generated), so licence problems are
easy to avoid.
As I already mentioned to some people, I think it is the main topic
for an internet standard to enhance interoperability. So im still
convinced it is a good idea to provide a common way to store data
for ECC algorithms even if a particular implementation decide not to
use ECC (or not all suggested curves) for licence reasons.
I see the point that we shouldn't do the work for those who claimed
rights that should not be granted to anyone (like patenting a number).
But 1. most of the work is done, 2. a patent won't last forever and
if a standard is available before it's most likely that future
applications will conform to it.
Biodata Application Security AG