[core] Reading -13

[Jim Schaad <[e-mail]>]

* Section 4.1.3.1 - I am unclear why an OSCORE error response can be cached
by an OSCORE unaware intermediate would be cached, but a success message is
never going to be cached.  Based on this, I don't plan to set an outer
Max-Age option.

* In section 5.4 you have the text

request_piv: contains the value of the 'Partial IV' in the COSE object of
the request (see Section 5), with one exception: in case of protection or
verification of Observe cancellations, the request_piv contains the value of
the 'Partial IV' in the COSE object of the corresponding registration (see
Section 4.1.3.5.1).

I am unclear how/why this is different for observations.   A cancelation is
a message, so the IV is the request.  A re-registration is a request
message, any response would correspond to that request.  The only
interesting question has to do with updating the MID on a re-registration
but not how PIVs work for this field.

* Section 6.1 - Is a registry needed for the leading byte of compression?
Behavior if bits 0, 1, or 2 is set in the flags byte on decode?

* Section C - given the way that my system is implemented, it would be nice
if the outputs included the first full IV to be used for both the sender and
the recipient.  That would allow for a test that the combination of ids and
common ivs is done correctly.  In my case I do not have the shared IV
available for testing as I immediately or in the id.