Oh yea, that's right. 8859-1 is a subset of Unicode character encoding, but not UTF-8. I mixed them up.

So one can find existing xml files with various encodings? Does that mean a parser needs a conformance statement as to what encodings it can understand?