Precisely because these characters are user defined, administrators must consider carefully whether to include them in a database. If user-defined characters occur in a database, all systems that access the database must define them consistently. Also, there are no standard algorithms for converting user-defined characters from one code page to another. For example, OpenEdge cannot convert user-defined characters from SHIFT-JIS to EUCJIS, since a user-defined character appearing in multiple code pages does not necessarily occupy the corresponding position in each code page.