Path: utzoo!utgpu!attcan!uunet!lll-winken!lll-lcc!ames!ncar!noao!arizona!gln
From: gln@arizona.edu (Gary L. Newell)
Newsgroups: comp.cog-eng
Subject: Re: Human Factors: Paper-Like Interface
Message-ID: <8710@megaron.arizona.edu>
Date: 14 Jan 89 03:53:27 GMT
References: <316@cui.UUCP> <2690009@hpdsla.HP.COM>
Organization: U of Arizona CS Dept, Tucson
Lines: 33

In article <2690009@hpdsla.HP.COM>, garye@hpdsla.HP.COM (Gary Ericson) writes:
> Assuming that you use a technology that doesn't require you to keep your arm
> out of the way (i.e., infra-red lines would be interrupted by your arm if
> you set it down on the screen), I was talking with someone about how it
> > typing when necessary, then make it go away when not needed.  
> For me, if it was a small amount of text and I already had a pen in hand for
> pointing, I think I'd rather write out the text in longhand on the screen
> with the pen.  Typing on a non-tactile-feedback screen would be more tedious 
> to me than writing the words out by hand.
> Gary Ericson - Hewlett-Packard, Workstation Technology Division

There has been quite a bit of work at IBM T.J. Watson Research in the last
two years on Gestural Interfaces. The use of a transparent tablet over a
flat display seems to be optimal for such a paper/pen like interface. It
eliminates the problems of determining precise points being indicated on
the screen. Applications in editting and spreadsheets or any info.
processing problem are being looked at. I have a number of references for
anyone interested.

Also, note that a keyboard is only preferable to stylus when entering 
text, if you are dealing with a small alphabet language (English, 
proof reading symbols, etc.) not for languages like Japanese or Chinese,
for these languages a good gestural interface would be superior and
many foreign companies are doing work in this area.

Much more work is needed in this area, especially in the areas of
hardware design (tablet technology is not quite up to the task yet)
and recognition algorithms (gestural recognition has subtle differences
from 'normal' character recognition problems and new algorithms are
likely to be needed if reasonable recognition results are to be 
gained). 

gary newell