Charles Lindsey wrote:
In <3E244CD9(_dot_)60209(_at_)Sonietta(_dot_)blilly(_dot_)com> Bruce Lilly
<blilly(_at_)erols(_dot_)com> writes:
Any raw untagged utf-8 in any header field is an example (that is exactly
what one would get if the properly tagged and encoded form which is legal
in header fields were decoded).
"Encoded-word" is a technical term defined in RFC 2047. It encompasses
objects of the general form
=?charset?encoding?data?=
No the *general* form is =?charset*language?encoding?text?=.
Whilst an encoded-word might indeed be decoded into raw untagged utf-8
under some circumstance, it is totally illogical to claim that every
instance of "raw untagged utf-8" that one might encounter must therefore
have resulted from "decoding some encoded-word".
We're not discussing "every instance [...] that one might encounter";
we're discussing header fields such as To, Cc, Bcc, Reply-To, Subject,
etc., and the *only* way that non-ASCII content can appear in header
fields is via the 2047/2231 mechanisms.
I have repeatedly invited you to indicate exactly where the Usefor draft
requires (or even allows) any encoded-word to be decoded (other than for
final display in a UA).
That invitation is still open.
You have yourself claimed that encoded-words in header fields can be
decoded for a followup, and the Usefor draft clearly permits such text
resulting from decoding those encoded-words to subsequently be sent in
raw untagged utf-8 to a gateway.