>>>>> On Fri, 11 Apr 2003 00:18:41 +0900,
	Bang Jun-Young <junyoung@netbsd.org> said:

> I'm not sure about what is the best naming scheme to take. Another
> candidate is Content-Type: field as seen in the mail header.

The problem of MIME charset name (that you refered as Content-Type)
is that there are a lot of (terrible) variants in the specification.

I don't think we should support names like
"Extended_UNIX_Code_Packed_Format_for_Japanese" (which is the canonical
name of the MIME charset "EUC-JP").

> And case should be insensitive.

I prefer simplest one. (i.e. locale codeset name only, case sensitive.)
Note that case insensitive comparison may require loading codeset
database.
--
soda