> > See also the discussion with IANA and Benjamin. The idea is that if someone
> wants to add a curve or a hash function, but following the method in this
> draft.
> Do we really need an RFC or is the IANA registry the reference? The idea is
> that
> the IESG should be able to determine that the new curve provides benefits
> without the need for an RFC. Do we have it wrong?
>
> I see. Perhaps adding a few words to the document to indicate the
> circumstances under which each registration policy might be expected to be
> used would help.
>
I came to the same conclusion; please consider the text I added in v-19, cf.
my response to Roman, quoting below:
----------
This seems to match Benjamin's direction as well. Maybe:
"
Crypto-Type: 8-bit unsigned integer. The type of cryptographic
algorithm used in calculation Crypto-ID indexed by IANA in the
"Crypto-Type Subregistry" in the "Internet Control Message
Protocol version 6 (ICMPv6) Parameters" (see Section 8.3).
"
This discussion is related to the points Benjamin and Alissa made on "IESG
Approval" in the IANA section "
Assignment of new values for new Crypto-Type MUST be done through
IANA with either "Specification Required" or "IESG Approval" as
defined in BCP 26 [RFC8126].
"
The goal behind that is to enable the addition of Crypto-Types without the need
of an RFC as long as the IANA registry contains enough information to use this
specification as is for the new crypto type.
Suggestion to add after the text cited above:
"
The "Defining specification" column indicates the document that
defines the length and computation of the digital signature, which
could be this for values defined through "IESG Approval".
"
-------------------
Does that help?
Pascal
_______________________________________________
6lo mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6lo
