Re: [LTC] ECC <= 112


Nils Durner wrote:
Hi,

as I am not aware of a mailing list for LibTomCrypt, I post my patches here.

There is no list.

<snip>

While I think it's great you are working [experimenting?] with LTC I
highly recommend against users using non-standard ECC curves for
fielded applications which is why I rejected your patch in the first
place [well that and your ECC-112 curve params didn't look right].

As for the raw sign/verify functions, EC-DSA specifies the output is a
ASN.1 SEQUENCE of the form {r, s INTEGER}.

I don't see any security problems with your function, I just don't want
to support it as it's not a standard. I support the raw DSA [over
integers] since DSS actually specifies the output to be the two
integers concatenated to one another. My dsa_sign() function is NOT
standard as I encapsulate the two signature values in an ASN.1
SEQUENCE.

Eitherway, it's cool that you're experimenting with it. Just be
careful with the params and make sure your ECC curves are actually
working properly.

Tom

.