On 1/2/2020 2:33 PM, John Levine wrote:
If a host doesn't get v4 connections, it doesn't need to recognize v4
address literals. Similarly, if it doesn't get v6 connections it
doesn't need to recognize v6 literals. Standards are about
interoperation, and cases that never happen don't matter for interop.
For defining a common core, the question is what is to be supported by
everyone, all the time. Supported is not the same as used, thereby
distinguishing between requirements for interoperability technical
specifications versus constraints of operational policy.
Is there any believe that any version of SMTP will be expected to
operate without support for domain names? (Again note, I said support,
not use.) Without support for IPv4? v6?
These questions need to be answered in terms of what is desired for
products being offered, not for operational choices in particular
For example, if there really is a view that some SMTP products
reasonably could operate only supporting IPv6 address, then after we get
through the discussion of pragmatics that justify recommending
standardization of that choice, we should consider how to create a
common SMTP that works for that choice as well as the more general, open
For this particular example, I'd say that the base protocol spec would
have to say something generic like 'text' and then leave the constraints
on the text to particular specifications that tailor things. mumble...
ietf-smtp mailing list