Ben Wing <ben(a)666.com> writes:
in the process of getting mule under windows working, i'm being
forced to put unicode encoding natively into mule. as soon as this
is in a bit better shape, i will look into adding support for the
"language tags" recently introduced in unicode, and then we can
convert all those goddamn files to utf-8.
That would be a good first step. But for new code, I would be most
happy if we could avoid I18N literals in chars altogether (strings
should be OK under UTF8), and always use a variant of `make-char'.
That way even a non-Mule XEmacs will be able to compile the Mule code.