>>>>> On Sun, 5 Sep 2004 14:53:26 +0900 (JST),
	Curt Sampson <cjs@cynic.net> said:

>> If we will conform to MIME charset names, the names
>> must be case sensitive, that isn't what people usually
>> expect in UNIX environment.

> I am sure you mean case-insenstive above.

Yeah, you are right.

> You're right that people don't usually expect that, but I fail to
> see how matching in a case-insensitive way cause problems with
> people who use a specific case.
> 
>     User 1: Always uses the same case. Case-insensitive matching works.
> 
>     User 2: Uses the wrong case. Case-insensitive matching works.
> 
> Same result, either way.

The problem is that the way User 2 uses doesn't always work.
As I already wrote, 
- It won't work with some third party applications which directly
  parse locale names.
- It won't work on older NetBSD installations.
- It won't work on older Linux installations.
- It won't work on commercial UNIX variants.

As you see, we should not promote to use case insensitive names.
It just spreads confusion, and doesn't give any real benefit to
users.
--
soda