Dmitry Eremin-Solenikov dbaryshkov@gmail.com writes:
Would it be ok to change ecc_point size to become a per curve option?
If needed, yes.
(I've also been considering switching to using extended X,Y,Z,T coordinates as internal representation for twisted edwards curves, but unclear if it's worth the effort. See the section on cooordinate choice on https://www.lysator.liu.se/~nisse/misc/ed25519-msp430.html for some details).
A close rival to ecdsa sign/verify (see rfc 7091 and 5832).
Ok, I'll have to look that up (but not at the moment).
Regards, /Niels