We use properties file (whose encoding is defined) to contain our UTF-8 characters. Yup...but it's defined as ISO-8859-1, not UTF-8 ( http://docs.oracle.com/ja

We use properties file (whose encoding is defined) to contain our UTF-8 characters. Yup...but it's defined as ISO-8859-1, not UTF-8 ( http://docs.oracle.com/javase/6/docs/api/java/util/Properties.html#load%28java.io.InputStream%29 http://docs.oracle.com/javase/6/docs/api/java/util/Propertie... ). You need to escape everything else. Also the escapes are only for BMP characters, so supplementary characters need multiple unicode escapes to represent a single 'character'. There's too many rough edges in interoperability. You do still need to know this stuff if you work outside of Latin-1.