VRML issues

Randy Stiles (stiles@aic.lockheed.com)
Mon, 13 Jun 94 00:54:09 PDT


I am new to the VRML mailing list. I am familiar
in concept with HTML and WWW, and I like the idea
you folks are putting forward about VRML very much.
I have worked with visual simulation/ virtual environment
systems, for education and design for the USAF and
ARPA, and can bring something from that perspective.

Date: Mon, 13 Jun 94 08:31:36 GMT
From: roger@is.co.za (Roger Layton - KMC)
Content-Length: 3228
Sender: www-vrml-owner@wired.com

Subject: WEB : Mapping out communal cyberspace
From: Roger Layton roger@apollo.is.co.za
To: www-vrml@wired.com

Date: Sunday 12 June 1994

The evolution of the Internet, and as reflected in the navigation tools
which support it, has led to a set of loosely connected, geometry-free
worlds, islands of information.

Whereas this approach is suitable for hypertext, and in fact is highly
desirable, with the introduction of VR as a medium we need to expose the
underlying geometry, and to treat coordinate space as a shared resource.
In other words, the coordinates of the worlds we defined, no matter
whether we choose to use local systems or not, are all mapped onto a
globally agreed coordination standard.

Yep, since the difference between VRML and HTML is in large
part spatial.

I strongly believe that it will be essential to CREATE a coordinate system
which defines WHERE things are, on a world-wide, globally agreed basis.
In effect to create a world coordination system, so that everything not only
has addressability as per current Internet standards, it also has relative
positional addressability in some world - it can be placed onto a map.

This is a very good point, in the sense that people will have to have
a common substrate on which to write their VRML, and showing one
collection with respect to another.

How many in this group are familiar with Distributed Interactive
Simulation? They adopt a geocentric frame of reference,
meters and radians for yaw,pitch,roll. I think we could
just use the same thing. DIS also addresses the network efficient
update of many active entities. What DIS does not address, and
what VRML could be very strong in, is 1) abstract systems,
2) preserved state, 3) collective world-building.

Adopting some parts of DIS could be a big help if
you ever want to see more than one person in the VRML
areas you create.

There is some work going on at the Naval Postgraduate School
concerned with hyper-objects in a virtual world, and this
could have some relevance to VRML. What is more, they
provide examples of using DIS and SIMNET code in a virtual

The map does not need to co-incide with the physical world, although
it will be useful to have some model of the planet somewhere. This is
the model for some new inner world, our own evolving cyberspace.

No, strictly it doesn't, but you will find it very useful if it does,
or is capable of doing so.

I recommend the RingWorld approach (a la Niven), with a ring 1,000,000 units
wide, circling at a distance of 100 million units around a central point.

I would reccommend subscribing to a coordinate set, one of possibly many:

1) matrix ( as in "the matrix" big, flat abstract
area requiring no advanced mapping)
2) earth (DIS world, many of us live here, and often display information
with respect to earth)
3) galactic space
4) ring-world?

This is a good point - maybe we want some standard symbology for types
of links.