At 12:06 PM -0500 12/30/03, Eastlake III Donald-LDE008 wrote:
I'm inclined to modify the draft to define "MAY+", meaning MAY but
likely to change to SHOULD, and to list AES-CCM as having that status.
Please don't make a new definition at this late date. We could easily
get bogged down into trying to say what "MAY+" means, and it would
almost certainly cause debate about whether particular
quasi-important encryption algorithms would be MAY+.
Instead, simply create a new section in this document that aligns
with section 3.2.3 of draft-ietf-ipsec-esp-v3-06.txt, say that
combined modes will require proper structuring of an ESP
implementation, say why combined modes are useful (speed
improvements, soon to be required in 802.11), and they say "there are
no suggested or required algorithms at this time, but AES-CCM is
expected to be of interest in the near future". That way,
implementers know that even though there isn't a MUST or SHOULD right
now, they still need to think about how their code should look if
there is one in the future.