As the Internet has grown, there has never been any meaningful debate
against the need to support fully internationalized text throughout the
net. The Mime effort was delayed at least 6 months by virtue of its
attempts to satisfy this desire. The charset= parameter is the best the
working group could come up with.
"charset" is to distinguish multiple localizations and unrelated "to
support fully internationalized text".
To my knowledge, the problem has not gotten any easier. Hence, it is not
productive to discuss this topic in general, anymore. The only thing that
makes sense is for folks who want to suggest changes to create detailed
specifications and submit them for review.
The detailed specifications have been submitted long before first as
RFC1554. It has received several comments and revised to be
ISO-2022-INT-*.
So far, no technical errors have found. Implementations exist. The
encoding is used for the real internationalization.
So, your suggestion does not make sense at all.
But 'complete overhaul of text on the Internet' is not the purview of this
mailing list.
What is "complete overhaul"?
If you think "charset" or some other negotiation based approach is the
way to go, you are actually proposing the complete overhaul of text
on the Internet to modify all the exiting text-related protocols to
support "charset" parameter.
On the other hand, ISO-2022-JP have been proven to work on the current
Internet (including 7bit RC822) without modifying anything at all.
So is ISO-2022-INT-*.
AND this mailing list has quite-extensively discussed and
pursued the degree of overhaul it is capable of, several years ago.
That's a wrong way of discussion.
Just introduce a new encoding scheme and don't overhaul the rest of the
Internet.
Masataka Ohta