[COSE] Issue #37 - Use the JOSE key

[COSE] Issue #37 - Use the JOSE key_ops terms

"Jim Schaad" <ietf@augustcellars.com> Sat, 07 November 2015 06:27 UTC

From: Jim Schaad <ietf@augustcellars.com>
To: cose@ietf.org
Date: Sat, 07 Nov 2015 15:24:51 +0900
Message-ID: <047b01d11925$0444d800$0cce8800$@augustcellars.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Thread-Index: AdEZHgG5HAXCBmEuRzaei5RysxNH0A==
Content-Language: en-us
Archived-At: <http://mailarchive.ietf.org/arch/msg/cose/rgGSTk3VbQTnHzYGu5e_uFG-QaI>
Subject: [COSE] Issue #37 - Use the JOSE key_ops terms
Precedence: list

The key_ops were added for the use by WebCrypto group, it is not clear that
they did a good job in defining them.

Some question that I would like responses to are:

1.  I there ever a case when one would do derive_bits but not do derive_key,
and vice versa.

2.  If one is using a polyfill in the WebCrypto group in order to have a
non-suported key, one needs to have the derive_bits option set.  However,
this would most likely not be set if one thinks it is going to only be used
for deriving keys.  Should this be an options which is not support?   Derive
key can only be used to generate keys that are supported in native code but
not for new polyfill algorithms.

3.  Not clear that this is a place where size is an issue.  It would mostly
be an issue if there is a statement that key_ops should be placed in the
ephemeral key structure.   I would note that it is not clear that it is
legal in WebCrypto to have key_ops on a public key.  Do people believe that
we should make a statement about providing a key_ops or ephemeral keys?

4.  Do people believe that we should worry about the size of keys in other
locations.  Specifically, key_ops is currently an array but potentially
could be an array or a single entry.  Is this something that needs to be
done?

Jim

[COSE] Issue #37 - Use the JOSE key_ops terms Jim Schaad