🚀 go-pugleaf

>From Moran@PARC-MAXC Mon Dec  7 18:47:48 1981
As a first-order approximation, meta-control-d is THREE keystrokes, ie, each
keypress is counts as a keystroke, a la the Keystroke-Level Model (CACM,
1980).  I believe this simple additive model is approximately right.  While
the ESC is "conveniently located", it is outside of the touch typing region
(for most people) and so takes longer to reach.  I guess I believe that the
above example is less than three keystrokes according to a more sophisticated
model, but not much less.  We have not run any experiments on these special
shift-key combinations; and I'm not interested in doing so, since I think the
effects really are second-order.  You might try applying the Rumelhart/Norman
typing model to see if the effects are more substantial.

Tom Moran

[ Tom should know.  He's one of the few people around who is actually doing
  research in this area!  I have asked Dave Rumelhart for his opinion, and
  will publish it here if he has any interesting observations.
							JQJ
  ]