Hallam-Baker, Phillip wrote:
Could we agree on a consensus point that:
'Any application developer who designs a protocol on the assumption it
will not be subject to NAT66 may be disappointed'
I think it would be far more constructive to tell application developers
what they _can_ assume... and to make sure that they have enough "safe"
assumptions to implement not only client-server apps but also multiparty
protocols with referrals.
(And trying to make it all hinge on DNS names is a non-starter)
Ietf mailing list