On 8/6/11 7:43 AM, "Leif Johansson" <[email protected]> wrote:

> So in that case I still support having the spec explain (very carefully)
> how you do "raw keys" - i.e keys with no intrinsic semantics - and then
> also adding the cryptographers caveat to that.

To be clear, when I was talking about raw keys, I didn't mean keys with no
intrinsic semantics.  I just meant something along the lines of PKCS1; a
modulus, an exponent, some algorithm info, and that's about it (I'd be ok
with defining private keys in the same format as well).  It would have
defined semantic, it just wouldn't be tied to an identity.

A definite goal for me that is NOT met by PKCS1 however, is to *limit* the
choices and extensibility in certain directions to reduce the overall
complexity.

Of course, once you've got primitives for sign and encrypt and you've got a
key format, doing something like PKIX is possible.  Perhaps we could sketch
that out as potential follow-on work in the charter, so we can make progress
on some of the shorter-term stuff first?

-- 
Joe Hildebrand

_______________________________________________
woes mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/woes

Reply via email to