Confronting the myth of rapid obsolescence in computing research