I don't understand the response to formal comment #75. In particular, the proposed (make-condition type alist) procedure seems to have a problem when several fields of the type have the same name, as can occur through inheritance and would otherwise not be problematic. Will