An assertion failure in CommentType led to the discovery that some observations (e.g. try recent eps aur and chi cyg obs from database) contain a '0' (zero) comment code. The submitter may have meant '0' (capital 'o'). This should be looked into. For now, we treat the errant comment code as '0', i.e. "OTHER". What is the most appropriate thin to do? We probably should not use an assertion here.