Path: utzoo!attcan!uunet!cs.utexas.edu!wuarchive!udel!rochester!kodak!doering
From: doering@kodak.UUCP (Paul F. Doering)
Newsgroups: comp.society.futures
Subject: Handwritten vs typed input
Message-ID: <2442@kodak.UUCP>
Date: 10 Apr 90 15:01:18 GMT
References: <54020@microsoft.UUCP>
Reply-To: doering@kodak.com (Paul F. Doering)
Distribution: na
Organization: Kodak Research, Rochester NY
Lines: 38

We're getting involved here in a debate on the merits of the handwriting-
recognition capabilities of the newly announced Sony Palmtop portable
computer. There seems to be a general admiration for the technology Sony
has mastered, leaving us -- as Gary Ericson has observed -- discussing the
relative efficiencies of the keyboard and the handwriting interface: a
keyboard on a tiny portable computer is too small for efficent use, so a
handwriting interface is preferable. (Have I been faithful to your
position, Gary?)

I find that two emerging technologies -- handwriting and document scanning
-- are counter-human. If there is one task at which humans are NOT good, it
is proof-reading. Character-recognizing interfaces turn me from a data-
enterer into a proof-reader, and the handwriting interface is by far the
worse from that perspective. Consider the process: I draw a character. While
I am drawing the next (or at least THINKING about drawing the next), the
system is trying to recognize my input. It then makes its best guess
about my intended input and displays that guess. At the very moment my mind
is looking forward, the system requires that I look backward to confirm the
system's recognition. Counter-human.

If you'd like a demonstration of the dilemma, try this. Using a pencil and
a pad, write out the characters of the reply you want to make to this
posting, while a colleague reads back to you the character you have most
recently written. You will soon discover that the future/past confusion in
the data-flow will slow you to a crawl. Now, would you really want to make
ALL your computer entries that way?

Gary, I haven't solved your problem. Itty-bitty keyboards are truly a
bad approach to input. I say that handwriting recognition is no better,
but for a different reason. Can we admit that we don't have a solution?
Opting for a new bad technology over an old bad technology is progress only
for the academicians and hackers. It won't solve the user's problem, and
it won't sell computers.
-- 
  =========================     ======================================
   Paul Doering (for self)         Man will never arrive,              
      doering@kodak.com               man will be always on the way.   
  =========================     =============== -Carl Sandburg =======