On Fri, 19 Sep 2003 20:55:46 +0200
Bjoern Hoehrmann <derhoermi(_at_)gmx(_dot_)net> wrote:
UTF-8 everywhere is a reasonable principle and much simpler to
understand and implement than any means to specify use of legacy
encoding schemes.
UTF-8 has its own technical problems (the Unicode signature, representation
of non-BMP characters, etc.). Moreover, people do not throw away legacy
encodings but stick to them. For example, although I think that
Unicode is better than Shift-JIS and I do have Unicode-aware text editors, I
still use Shift-JIS, which is so convenient at present.
I am not saying UTF-8 is bad. I'm just saying UTF-8 everywhere is even
more unrealistic than any other options at hand.
For inbound encoding declarations, generic syntax
does not work. Whatever syntax you choose, it will look odd in many
formats and many authors won't use it anyway.
Please see my mail to Martin.
Cheers,
--
MURATA Makoto <murata(_at_)hokkaido(_dot_)email(_dot_)ne(_dot_)jp>