OK. So it's an early morning and the coffee hasn't taken full effect
yet.
But alas, my eyes really did see message talking about determining
message structure/relation semantics based on the presence of an
ending blank line INSIDE a body part or the use of NULL body parts.
No. I don't think you folks can possibly be serious.
Nathaniel ends his note by suggesting we pursue clean handling of these
cases by defining appropriate content-types (in these cases, probably
all will be multipart variants.) To his suggestion, I can only
respond: Good heavens, yes.
D/