Re: Comments on draft-ietf-kitten-krb5-gssapi-prf-03.txt

As with a similar discussion on krb-ietf list how a KDC cert should
be verified in PKINIT, we should not ignore the fact that the affected
spec (here) is going to be normative for both, the gssapi mechanism
implementor and the application gssapi caller, and those two
will have an entirely different background and needs.

So I would suggest to actually quantify
 (a) what the mechanism implementor must provide as a minimum
     to be at all useful
 (b) what the application caller can rely on to be safely available.

I think that a number in the range of 1000-2000 should be good.

I would indicate to the gssapi implementor that a secure PRF is
necessary, and it would indicate to the application caller
that this function is not designed to produce a Stream-cipher like
pseudo random pad which it can (ab)use to XOR large piles
of application data.

We should add a reference to the document draft-eastlake-randomness2-10.txt
that is sitting on the Editor's queue:

  Date: Mon, 07 Feb 2005 17:31:14 -0500
  Subject: Protocol Action: 'Randomness Requirements for Security' to BCP
  Sender: ietf-announce-bounces@ietf.org

  The IESG has approved the following document:

  - 'Randomness Requirements for Security '
     <draft-eastlake-randomness2-10.txt> as a BCP

This document contains lots of useful information about cryptographic
randomness.

-Martin

_______________________________________________
Kitten mailing list
Kitten@lists.ietf.org
https://www1.ietf.org/mailman/listinfo/kitten