In the meantime I'll go ahead and reject this until we can get some more definitive feedback in regards to what the pgsql team expects here dealing with "enum udt" types from jdbc...