Mnemonic is NOT a general solution. It only handles alphabetic
languages and leaves the ideogrammatic languages out in the cold.
That limitation of mnemonic probably cannot be helped, but the use of
Real-* headers is the more general solution, is not Euro-centric, and
works world-wide. Let's try to think globally here and solve the
general problem for all users.
Oh, well, mnemonic is general, with support defined for the Chinese,
Japanese and Korean characters in the design, and I do have
the 5 CJK 16 bit character sets tabled, but these are not yet
in RFC-CHAR - and I do not support it in the current implementation.
So mnemonic could be used technically with reasonable results also
on CJK.
Keld