[Hipsec] Re: The exact method used to generate a HIT from a HI

From: Tom Henderson <tomh(at)ponymail.com>
Date: Tue Mar 18 2003 - 19:05:35 EST

> On Fri, Mar 14, 2003 at 02:27:26PM +0200, Pekka Nikander wrote:

Yes, you are right, I did not read 2536 carefully enough. T can be inferred from the length of the key, I guess, plus has better alignment to leave it out. I therefore would support removing it from both the HI and the hash of the HI.

> My personal opinion would be to use the RFC2536 DSA format
I am not strongly for RFC2535 format, since original motivation was to make it storable in DNS (for which we have learned we need another RR definition anyway). But I do think that either an octet or four is needed as an algorithm field, to identify the underlying key structure, so that a HI could stand on its own and be interpretable.

Tom

---

Hipsec mailing list
Hipsec@lists.freeswan.org
http://lists.freeswan.org/mailman/listinfo/hipsec Received on Tue Mar 18 19:47:57 2003