moore(_at_)cs(_dot_)utk(_dot_)edu (Keith Moore) wrote on 03.01.03 in
<200301031753(_dot_)h03Hqvj12371(_at_)astro(_dot_)cs(_dot_)utk(_dot_)edu>:
2. But in fact, if you look at the headers we have defined so far, you
will see that the only places 8bit freedom is allowed is in comments,
phrases, parameters and newsgroup-names.
you obviously don't get it. the definition of message/rfc822 is for
the content-type, not for the use of that content-type in email.
content-type definitions are independent of where they are used.
(the only exception is text/plain where the default charset is
"iso-8859-1" in http and "us-ascii" in email - but this was only a
hack for backward compatibility with pre-standards http.)
it's absolutely unacceptable for usefor to redefine the meaning
of message/rfc822 - even if they only pretend to do so for usenet.
Frankly, from past discussions on both USEFOR and this list, it appears
that using message/rfc822 as-is has unacceptabable constraints(1),
allowing a workaround is unacceptable(2), using another media type does
not work(3), and sending news via mail is not going to stop(4).
Now what?
From what I can see, (2) is *by far* the easiest "problem" to ignore.
Of course, if this debate continues to stall at every option, I can
guarantee you that what you get is what we have today, which is
essentially (4) stomping on *all* these solutions - i.e. sending a naked
news article as-is with only an added To: header. USEFOR nearly had a
revolt of the Usenet moderators just for daring to suggest that that had
problems. (See the archives.)
And frankly, the holy restriction you're arguing for with (2) is already a
very silly idea in the first place.
MfG Kai