To conclude this presentation, a suggestion to people who start with handling unicode data: try to keep all information which is related together in objects, so the objects can handle the complexities caused by character and other encodings.