PHIL: VRML Visions

Mark D. Pesce (
Sun, 12 Jun 1994 21:38:57 -0700 (PDT)

VRML List Members:

I have been working on the development of a VR interface to WorldWideWeb
for several months. During this period of time I have been inspired by what
will be possible when _our_ work is complete. I'd like to share some
of these visions with you, in the hope that it will help to formulate the
needs for VRML in the short term, and into the future.

We're all approaching this design from different points of view; my training
has greatest depth in networking and visualization, others on this list are
architects and designers. Despite differing needs, our goals are quite often
congruent, and we must strive to develop a technology which meets the
needs of the vast majority of its users without sacrificing elegance or

Mark Pesce
June 1994


Vision One: The United States Holocaust Museum
When: Now
The Project:

The USHM project is a joint development between Husky Labs (Washington, D.C.)
and the Labyrinth Group to produce a WWW site which has both documentary and
spatial exhibits. The USHM is a unique museum; the architecture of the space
plays an important role in the story being told. Parts of the Museum look
like villages in Eastern Europe, a concentration camp, an oven. Visitors
to the Museum are overwhelmed by the gestalt that is created; that a space
can be so evocative is one of the strengths of the Museum, and, without it,
the Museum loses some of its impact upon the imagination and the soul.

When completed, the Museum will be accessible, as a space, from anywhere on
Internet, to any WWW client which is VRML-capable. It will be possible to
"tour" the Museum, just as is done in real space, with many links from the
virtual space of the Museum into Web pages which describe, in greater
detail, the content of the Museum's exhibits.

This project requires very little technology which has not already been
developed and demonstrated at WWW '94. Our very simple (moronic) VRML
language (the specification for which is given at the VRML Forum WWW site)
is capable, with very few extensions, of handling the expression of the
entire space, with extensive linkages to WWW pages to be viewed from
Mosaic or another Web/MIME viewer.

Vision Two: The United States Library of Congress
When: October 1994
The Project:

My initial inspiration for the creation of Labyrinth and VRML was to develop
a way to model the complexity and richness of the Web in such a manner that
it is intrinsically navigable, without any instruction required on anyone's
part. I use, as an example, the US Library of Congress GOPHER site. It is
rich, it is deep, but it is difficult to navigate or browse, especially if
one is unacquainted with the site, or one doesn't know precisely what one
is looking for.

I assert that it is possible to model the Web in three dimensions in such
a way that it is _intrinsically_ navigable, such that the form of an object
within a particular space is derived from the content of that object. It
is possible, then, to build a "library", or data repository, which looks
only coincidentally like a library in the real world; in fact, we are better
served if this virtual library looks like its content. If we succeed in
this, it should be possible for anyone, of any age, to browse the collected
content of the world's greatest library (or as much of it as they care to
place on-line), without having to be taught hermetic classification systems,
searching methodologies, and the like..

At the very least, it is possible to take the GOPHER site
and "wrap" a VRML interface around it. In so doing, we make the Library
accessible to people who would otherwise be intimidated by a GOPHER interface,
and can organize the space so as to facilitate browsing and examination of
the resources at the site.

This, again, requires very little that VRML does not already have.

Vision Three: Earth
When: January 1995
The Project:

One of the most inspirational parts of the Web, for me, are the pages at the
Michigan State University where the Current Weather Maps/Movies are kept and
maintained. This is a collaborational effort of many people at many
university sites across the country, and is a trully brilliant example of
what is already possible with the Web.

Extending this into the virtual world, I'd like to be able to create a real,
live "Earth" on my desktop, not unlike the version described by Neal
Stephenson in _Snow Crash_ (a book full of good ideas), which would take the
sattelite data, massage it just slightly, and wrap it onto a sphere which
represented the planet. This is certainly possible _right now_, and would
be an extremely effective demonstration of the combined capabilities of VR
and WorldWideWeb.

Further, one can imagine that this "Earth" simulation can become one
of the interfaces for NASA's EOS project, the "Mission To Planet Earth",
which begins later this year, and will result in exabytes (10^15 bytes) of
cartographic and other ecological mappings of the planet. Thus, VRML can be
used to create a rich, dense, yet easy-to-use interface to what will be the
single largest data collection task ever undertaken by humanity.

It has been a long-held believe of mine that VR/Cyberspace represent an
effective, necessary technology for planetary management. In an age where
we must always be conscious of our effect upon the surrounding environment,
we need powerful tools to visualize it, and to track our effect upon it.

Technically, this project is only a little more advanced than those which
have been described already. It would be useful for VRML to incorporate the
concept of a "data stream" (something in the works for the WWW libraries) so
that "real-time" data flows could be incorporated into the planetary

Vision Four: The WorldWide Marketplace
When: June 1995
The Project:

[There is a small, but vocal, minority of people who find the idea of
commerce on Internet anathema. To these I would ask that they skip to the
next and final proposal.]

Many commercial organizations are excited by the possibility of being able
to offer goods and services through Internet-based channels. In fact, it
should be possible to create a "space" with a very wide array of goods, both
physical and more ephemeral, without advertising, (or with it, if desired),
so that a "marketplace" can exist for trade. Moreover, this market need not
be (should not be, in my own opinion) monolithic, no single Macy's or HSC or
what have you, but more like a bazzar, with a dense, organic ecology of
businesses which work together (and are co-located) in order to support each

It might look like a suburban shopping mall, although that will not have a
great appeal for everyone. It might look more like Marrekesh than
Minneapolis. It does need to be flexible, though, so that there can be a
wide availability of goods/services, and so that organizations can come and
go quickly, as a market need arises, is satisfied, and disappears. While
cooperation is necessary, any monolithic entity which attempts to "own" the
entire retail space would end only in stifling the diversity of the market.
Therefore, it has to be possible to create a concatenated "space" which can
draw from sources (Web sites) all across Internet, and yet can create a
continuous, regular perception of space. Finally, it all has to be integrated
seamlessly with other Web-based markets.

This project is somewhat more complicated than those outlined above, as it
requires an overall protocol architecture to support a distributed
description of a virtual space. I have outlined our approach to this
problem in the "Cyberspace" paper at the VRML Forum Web site, and with the
innovations described in that work, it is possible to create, maintain and
evolve a marketplace on the Web.

Vision Five: The Agora and the Senate
When: March 1996
The Project

"The net, the very network itself, you see, is merely a means to an end...
The end is to reverse-engineer government, to hack politics down to its
component parts, and fix it." - Joshua Quittner, _WIRED_ 2.06

One of the great benefits of communications technology is that it can create
and nurture communities which exist outside of demographic bounds. However,
demographic communities need to be brought together as well, for the
purposes of self-governance. VRML can be instrumental in the implementation
of these systems of self-governance.

Democracy is founded on the conceptualization of "informed consent". This
means that you, as a citizen, have a right and a responsibility to exercise
your franchise based upon the best possible information. While the history
of politics is the history of disputes about which information is "true",
"right" or "moral", underlying all of it is a understanding that
information, in whatever from, from whomever it comes, is necessary for
informed consent, and informed consent is necessary for democracy.

In America, our founding fathers did not give us a direct democracy, at the
national level, because no communications infrastructure of the eighteenth
century could support such a proposition. However, in New England, and
in several of the other colonies, towns governed themselves directly, and
rather anarchically, through a system of "town meetings", in which all the
citizenry could play a role. This itself was modeled after systems which
evolved in England, and before that, in Athens and Rome, where all of the
citizens of the state were free to participate in the debate on the issues
of the day. To them, government was a place as much as a process; the area
around the Senate in Rome contained the Library, a public area for
discussion and debate, the baths (where more discussion, debate, and
deal-making went on), and an ampitheatre for speeches.

If we attempt to augment our own mechanisms of self-governance, we will need
to look at this model, and adapt its most functional aspects to our own
methodologies, and create a Senate of our own. This can be done with what
we have now, and with what we are developing, although VRML will have to
move toward _complete_ interactivity before it is possible to stage a
"street debate" complete with an attentive (or heckling) audience. Already
it is possible to peruse the budget of the United States; and the town of
San Carlos, California, this week announced its own WWW site. We can use
VRML to put an attractive interface on these gold mines of useful data;
that alone is worth doing. But further, we can use the Web and VRML to
create a "place" where democracy can happen.

That, I believe, is a noble goal, and one worth working towards.


I have shared my visions with you; these projects are possible _immediately_
in the future, not in some distant, barely imagined cyberpunk evocation.
Together we can sculpt the tools which will define these worlds, define
their behaviors, and allows us all to share in each other's creations.

Thank you.


|| * That's AL * WorldWideWeb