Mališa Vučinić writes:
> Thanks Tero for this feedback! Could you check if this commit takes care of
> it:
> 
> https://bitbucket.org/6tisch/draft-ietf-6tisch-minimal-security/commits/dee6cf8074f2
> 
> The algorithm identifier is added, it is optional and if it is not present the
> IEEE802154-AES-CCM-128 algorithm is assumed. Apart from the key length, I also
> added the nonce length in the description of the algorithm in the registry.

Looks good. Formatting the algorithm ids as negative numbers is bit
wierd, but I assume it allows making the field optional as you can
detect from the the nint that it is algorithm identifier not key
usage...

Other option could be to combine the key_usage and algorithm to same
field, i.e., add algorithm to key_usage tables, and when AES-CCM-256
is added then double the key_usage entries to contain both possible
algorithms. This might have the same problems TLS have with cipher
suites, that we end up with quite large table with all possible
combinations.

Or we could rename key_usage to key_usage_and_algorithm and split it
so that (key_usage_and_algorithm & 0x1f) is the actual key_usage, and
(key_usage_and_algorithm >> 5) is the algorithm id, which would still
encode key_usage_and_algorithm as one octect for first few algorithms.
-- 
[email protected]

_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch

Reply via email to