Re: ECC point compression trick


"Tom St Denis" <tomstdenis@xxxxxxxxx> wrote in message
news:1151511777.232321.25330@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Pubkeybreaker wrote:
Tom St Denis wrote:
Pubkeybreaker wrote:
The square roots of 2 mod 7 are 3 and 4.

And -3 == 4

OMG WOW PONIES

And -4 == 3.

You still don't know which one is negative.


The negative root would be the one you find by negating the
other root
you find.


And still the question stands - is 3 or 4 the negative square
root of 2 mod 7?

I don't know exactly how you are saying to use the one that you
choose - but if it requires that two processes be able to
independently come up with the same positive or negative root,
then you can't just declare that the positive root is which ever
one you find first, because the other process may find the other
one first. Just because both are deterministic doesn't mean that
they will both produce the same results as each other.

How about just saying that you will always use the smaller of the
two (when both are represented as integers 0 <= i < modulus)
where ever you are presently suggesting to use the "positive
root"? That has some basis in that the larger number would be
negative (and the negative of the smaller number) in a p's
complement (where p is the modulus) interpretation. If that
causes problems, then it would seem to be the same for your
"positive" vs "negative" root choice.




.