> if we're trying to properly display a Japanese name, it's probably easier to put the appropriate language context in a UTF-8 document than it is to embed Shif

> if we're trying to properly display a Japanese name, it's probably easier to put the appropriate language context in a UTF-8 document than it is to embed Shift JIS text into a UTF-8 document. You'd think that, but in practice I've found the opposite. Applications that use encodings managed to display things properly. Applications that hardcode UTF-8 don't.