However, as I read RFC 2026, _any_ substantive change from the
Draft Standard 5321 that has not gone through Proposed Standard
in a separate document first takes the 5321bis document.
AFAICT, there is no exception there that allows introducing a
substantive change that would cause some existing conforming
implementations to become non-conforming and still have the
document move to Internet Standard, even if that change has been
widely implemented by large email providers.
Hmm.
But doesn't a document also require implementations? I admit I may not have
paid attention to that part of the lesson, but I vaguely remember a field
on the shepherding document that said "list conformant implementations
here", or words to that effect.
The current limits of the localpart (character set, case sensitivity and
length) seem a little lacking in implementations, compared to other parts
of the document. So it might be both permissible and sensible to have the
next document give a brief overview of current practice, note that some
receivers do enforce <description blah bla description> and suggest rather
than specify that senders stay within these limits.
Arnt
_______________________________________________
ietf-smtp mailing list
ietf-smtp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf-smtp