On 27-Jun-20 15:06, Nico Williams wrote:
> On Fri, Jun 26, 2020 at 09:15:19PM -0400, Michael Richardson wrote:
>> 1) So, BCPs can point to multiple documents.
>> BCP14 actually is RFC2119 and RFC8174 now, so we can add a third document
>> if
>> desired. That is, it's cheaper to not spin 2119.
>
> Sure. Plus, RFC2119 is iconic, and used far and wide outside the IETF.
>
>> 2) I'm not claiming others won't use it, I just don't know if they will.
>> Crypto progresses... what other things do that in the same way?
>
> Any time we have a migration from one [sub]protocol to a newer one and
> need to retain support for the old one for a while, we will need
> something like SHOULD-. I agree this comes up most often in the context
> of cryptographic algorithms.
Interesting discussion but I want to see the ACP in the RFC Editor queue, can
we just do what is required for that and nothing else?
I am about a millimetre away from proposing to recall the GRASP spec from the
RFC Editor so that we can amputate its normative dependency on the ACP. I
really wish I'd done that a year ago.
Grumpy Brian
_______________________________________________
Anima mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/anima
