Considering the current practice and the working code of
non-pure-US-ASCII Internet and the rough (that is, non-US)
consensus of implementors, the following is the result of the
recent discussion related to MIME and its specification:
1) Both MIME charset label and ISO 2022 escape sequences label
character set properly, regardless of whether (font of) the
designated character set is supported or not. In the latter case,
the recipient has little chance to be able to understand the
language of the content, anyway, even if the recipient can,
somehow, display the content.
2) All the current encoding used on the Internet (including ASCII
and ISO-8859-1) follows the designation and invocation rules
of ISO 2022.
3) Suporting bi-lingual (Native and English) encoding is localization.
Supporting such localization of all the native languages of the
world is merely a set of US-centric localizations and is not
internationalization.
4) That MIME charset mechanism swiches several localizatons has
nothing to do with internationalization of the Internet.
5) 7bit representation of text according to ASCII with ISO 2022
extension is, by any means, comformant to ASCII and, thus, RFC 822,
while defaulting non-ASCII variant of ISO 646 designation may or
may not.
6) 7bit representation of text according to ASCII with ISO 2022
extension causes no interoperability problem, while defaulting
non-ASCII variant of ISO 646 designation may.
7) Inappropriate description about RFC 822 in the current MIME
draft MUST be modified.
Masataka Ohta
PS
I'll be absent till 1/3. Merry Christmas and a happy new year.