Unfortunately, it isn't the only case in which symbol collections should
be rendered differently depending on the language with which they are
used. As I understand it, we have at least one more in the case of
Devanagari.
That's all. Isn't it?
I
wouldn't have a serious problem, except that I'd be real interested in
what recommendation we would make to people who are using 10646 with
*neither* Japanese or Chinese.
You don't have to recommend any, because any is OK.
Or, if you really want to have a bottom charset, by some unknown reason,
you can have, say,
charset=iso-10646-euro
which contains code points lower than 2048 only, just as RFC1345 defines
charset=ISO_646.basic:1983
And keep in mind that most messages sent intra-Japan and between
mutually known Japanese speakers are going to use Kanji, not Chinese
Han, so the necessary decoding information is externally available
without any of this extra tagging
How can the "externally available" information available?
I'm in Japan. I daily send and receive many intra-Japan messages using
Kanji, not Chinese Han. Still, on my workstations, the information is
not available at all.
Can you give me a specification how the information be available?
(whether or not use of the external
information is desirable is another issue).
Desirable? It is NECESSARY for correctness.
Masataka Ohta