>Re: append() and nelem: >There are only two append() calls within sort(1). I believe the other >one should not be even encountered when nelem == 0. I'd rather have >sort bomb if the condition would happen. see, up to you. itojun