i think you mean with method signature the method name and the parameters, because implementing A and B at the same time would clash:
interface A{
void bla();
}
interface B{
String bla();
}
apart from that i think it is not quite correct to define interface methods solely by their signature. there is a lot of semantics behind it, e.g. what you input and what you get (see
design by contract). that means, though your interface methods look identical they offer different semantics.
so if i was you i would go for some rename-method refactorings to avoid such a problem. or if the semantics are the same, i would try use move method to form new interface groupings.
[ August 16, 2006: Message edited by: manuel aldana ]