I have found a bug in the version of KCL available on CLI.Com: As was reported in a note from Sandra Loosemore on 16 Oct 87, the ETYPECASE macro tests the arms in reverse order. TYPECASE and CTYPECASE are both fine in this respect. A letter from Masami Hagiya on 20 Oct 87 said that this would be fixed soon, but perhaps it fell through the cracks. Pavel