Thanks for the bug report! I need to test the new code, and check for other such cases, but I'll try to upload a new version today. (The optimizer assumed pair?, not null?, would be used where an improper list might occur, but it still shouldn't have crashed).