Path: utzoo!utgpu!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!rutgers!gatech!purdue!decwrl!sun!pitstop!sundc!seismo!uunet!mcvax!cernvax!cui!bertrand
From: bertrand@cui.UUCP (IBRAHIM Bertrand)
Newsgroups: comp.cog-eng
Subject: Re: Human Factors: Paper-Like Interface
Message-ID: <314@cui.UUCP>
Date: 19 Dec 88 14:33:37 GMT
References: <2690003@hpdsla.HP.COM> <88Dec10.134912est.10521@ephemeral.ai.toronto.edu> <2442@ficc.uu.net> <88Dec14.210656est.10862@ephemeral.ai.toronto.edu>
Reply-To: bertrand@cui.UUCP (IBRAHIM Bertrand)
Organization: University of Geneva/Switzerland
Lines: 57

>>> RE IBM's announcement of a "paper-like" interface:
>>
>>> Q:  How many people would really want an interface like this?  ...
>>> I don't really think, however, that a stylus would be better for
>>> text input ...
>>
>>An obvious answer is to draw a typewriter keyboard on your LCD display
>>and use its touch sensitivity to determine what "keys" you type.
>
>That's not a very good answer -- have you ever tried to type on a membrane
>keyboard (one that has no tactile feedback) before?  It's not easy or
>accurate.  Perhaps the keyboard will start to go away in machines like
>executive workstations and shop-floor computers, where it only gets in 
>the way anyway, but in the case of people who have to work with text, 
>be it words or program code, I think the keyboard is the better way to
>go, perhaps *aided* by touch-sensitive input devices.  Now if we could 
>only make a _better_ keyboard so everyone would be happy...

Here, at the University of Geneva, we have built more than 15 years ago such 
an interface based on a plasma display and an infrared "touch panel". The 
display was square (about 8 inches sides) with 512x512 pixels and the input
device had a resolution of 160x160. Any opaque object could be used to point
to the screen (including a finger) and the position could be sampled 50 times
per second, allowing a complete tracking of the movements of the finger.

The equipment had also a regular keyboard attached to it. So, for text 
intensive applications like text editors, we used the keyboard for character
input and the "touch panel" for cursor positioning and text selection with the
finger in a very natural way. I think that having to use a stylus makes it much
less comfortable since you have to pick it to make a selection and drop it to
type on the keyboard.

For graphic intensive applications like graphic editors where you had very 
little text to enter but mainly free hand drawing or rubberbanding, we used a 
simulated keyboard drawn at the bottom of the screen. Typing was not very fast 
(about two characters per second) but acceptable.

An interresting consequence of the tight coupling of the input and output 
devices was the disappearence of the notion of cursor (for graphics). The 
finger WAS the cursor. Another interesting point was also the use we made
of the finger up <-> finger down transitions to build a more user friendly
interface.

For those who are interested in getting more information about it, I can send
hardcopies of papers published on this.

Bertrand Ibrahim

email:              to BITNET                             to EAN
from
BITNET IBRAHIM@CGEUGE51.BITNET                 bertrand%cui.unige.ch@CERNVAX
ARPA   IBRAHIM%CGEUGE51.BITNET@CUNYVM.CUNY.EDU bertrand%cui.unige.ch@ubc.csnet
UUCP                                           cernvax!cui!bertrand.uucp
UUCP                                           mcvax!cernvax!cui!bertrand.uucp
UUCP                                           cui!bertrand@cernvax.UUCP
EAN                                            bertrand@cui.unige.ch
JANET                                      bertrand%cui.unige.ch@uk.ac.ean-relay