[sci.philosophy.tech] Touch Sensitive Screens

mrh@cybvax0.UUCP (Mike Huybensz) (05/21/87)

In article <314@phoenix.PRINCETON.EDU> (Daniel Yaron Kimberg) writes:
> I have reservations about TSSs though.  I've seen
> them in a few places, not in industrial control, but in retail sales.  They're
> nice, but I wouldn't call them wonderful.

I've seen TSS's in industrial control: about 5 years ago, Gould built a
very pretty system complete with a touch-driven programming environment.
I'm very glad I never had to program in it: it would have been incredibly
frustrating.

I worked with touch panel driven programs on the PLATO system for about 9
years.  They are rather friendly, but have alot of user interface problems.
The major problem is that people don't distinguish between pointing with
their finger and touching: thus the actual touch can be almost an inch from
the target when the fingertip is just below the target and the ball of the
finger makes the actual contact.  The PLATO touch panels had low resolution
and produced a single sample per touch: together these greatly limited
the practical number of choices on the screen.

One thing that could be done to improve touch panel interfaces would be to
have an indicator (like the mouse arrow) that would appear above the fingertip
to show where exactly you were pointing.  This would permit much greater
touch resolution, but you would still need some way of clicking (since it
doesn't make much sense to to be shown only after you completed the touch.)
Once you have that sort of pointing and clicking interface, it may not
much matter whether you point with your finger on the screen, mouse,
trackball, digitizer, or scratch pad.
-- 

Mike Huybensz		...decvax!genrad!mit-eddie!cybvax0!mrh