Your compiler treats the source as whatever it likes (either latin1 or utf-8 I guess) but there your native chars are no available / wrong.
Don't use native chars in source code as you can never predict how (== which encoding) a specific compiler interprets the source.