Traditional mouse-based desktop interfaces support a number of fundamental capabilities that frequently become problem areas for novel interaction devices with new capabilities. The same interaction and device design problems come up with these devices over and over again, with researchers and designers repeating the same mistakes. This note describes these fundamental capabilities, how they are supported on mouse-based interfaces, and identifies fundamental problems in extending them to novel interaction devices. My hope is that this can help designers identify basic mismatches between input device and interaction technique in the early stages of a design, as well as to consider design solutions (or to be more precise, various design compromises) that have been proposed.