David Daney wrote: > Per is correct, you are mistaken. >> The JLS v2 section 2.6.10 specifies what should happen in the case of > numeric operands. In the case of two char operands they both get > promoted to int. I didn't critisize him, because he cites the JLS. I critisized the JLS. So Per and the GCJ-Team are, since my last mail, out of the target line. Bojan