From: cyberoid@milton.u.washington.edu (Bob Jacobson)
Subject: Re: Computerized Reality: Better than VR
Date: Wed, 17 Jul 1991 21:57:47 GMT
Organization: HIT Lab, Seattle, Washington



I hereby exercise my moderator's right of first reply to Pierre Wellner 
of Xerox Europarc, who writes:

	Virtual worlds are lots of fun, but the real world is often better for 
	real work.  A virtual world can create environments that are complete-
	ly different from anything users experience in the real world.  This 
	can have advantages, but if the environments and objects are too 
	unreal, users can't rely on their knowledge and experience to use the 
	system.

	Users want to learn the *minimum* possible.  The best systems are 
	designed to be familiar to users, and it is real world objects and 
	environments that are most familiar.  The more a system is like the 
	real world, the easier it is to learn, and the easier it will be to 
	use (hence the "desktop metaphor,"  "direct manipulation," etc.).   
	If it's good to resemble the real world, then it's even better to 
	*be* the real world!

	Most people use computers to get work done.  Most computer applica-
	tions are synthetic worlds that users interact with to accomplish a 
	task.  Even the conventional workstation/PC "desktop" is a kind of 
	virtual world, completely seperate from the physical world.  But 
	why force users to abandon the real world? ...

In fact, most people do NOT use computers to get work done.  Most people
can't type, let alone interact intelligently with an onscreen presentation
of data.  Extrapolating from the experience of current computer users is
not the wisest design method, unless one is satisfied to market solutions
to 10% of the population (and then usually one application at a time).

The "desktop" metaphor is a good case in point just how much learning is
required to use a computer.  First, one has to work at a desk, to under-
stand the functionality of different types and forms of conventional
information.  Then, one has to be able to log on to the computer and be
able to input data using (1) a keyboard or (2) a persnickety pen-pad,
whose performance is still a great unknown.  (And the person's handwrit-
ing had better be legible.)  One has to understand that the "desktop"
that is represented vertically on the computer screen is actually a
rotated model of the real desktop, although it is unrealistically orderly
and the icons do not have any natural relation to the things they repre-
sent.  The desktop's functions are another whole school of learning, for
which people like myself read lots of books and do lots of experimenting,
on our own free time (easier if one is a professional and without kids),
to even begin to penetrate.

Virtual worlds are not a substitute for the onscreen (usually visual,
non-aural) interface.  They are a replacement for that inadequate inter-
face.  They are a design paradigm that comes at information collection,
manipulation, and management from the standpoint of the person, not the
machine.  Of course this is an ideal, which is why the current round of
virtual interface technology feels so disappointing:  it is still operat-
ing on the margins of acceptability (although the new VPL Eyephones HRX
are pretty smooth, and 3D sound works now).  In the future, virtual worlds
will increasingly come to resemble the real world, which computer terminals
never can.  

Moreover, virtual worlds need not resemble today's "whole hog" approach
of total encapsulation; that's just a response to weak existing trans-
ducers, which cannot function well in competition with physical sources
of light and sound.  They can and will be far more subtle.  The idea is
to make it possible to use machines with less learning than today.

But the main point of my little dialogue with Pierre is that virtual
worlds are more than a better way to use computers.  They are conceived
as a better way to do many things, in which computers are involved but
not necessarily the central issue.  Computo-centrism is a lot like
auto-centrism:  it occludes alternative solutions to basic human 
problems, like how to learn, to commmunicate, and to do the everyday
things that provide for a good life.

This is a common debate between those who are more comfortable working
onscreen ("it works NOW" is their bottom line) and those who are taking
a fairly high risk to do things off and away from the computer itself.
Let's here more, from both and other sides!

Bob Jacobson
Moderator