We have some russian, macedonian, czech and chinese on i2geo.net We had several issues at the beginning which were mostly related to the DB. The default configuration was still using an iso-8859-15 encoding (iso-8859-1 + euro). Changing the default encoding of the MySQL was not enough, we had to change the default collation (all unicode something). Supplementary to this everything web should, of course, be utf-8. Some defaults still remain here and there to use the java default charset... changing the sys-prop for charset is a good practice I think. The storage issue was the real problem because you do not realize it right away: the hibernate cache is fully unicode aware. It's only after you return from the DB back into java. But we had no Korean around. paul Le 1 juil. 2011 à 15:55, Gerritjan Koekkoek a écrit :