You've probably been wondering where Brian & I have been
over the last few weeks; you know we've been moderating the
list, putting out the digest, and so forth, but you haven't
seen us directly. We've been very busy; Brian is
participating heavily in the design of HOTWired, Wired's on-
line alter ego, and we've been busy ourselves (Labyrinth
Group, that is). I'd like to spend some time telling you
about the work we've been doing, and what we hope to return
to you, the VRML community, because of it.
*****
I'll be making three postings on the following subjects:
1) Annouincing a public-domain VRML browser (*this* posting)
2) Why Open Inventor should be the VRML base language
(A commentary on the survey)
3) VRML projects underway or planned
I have an unusual position; I am a list moderator and I also
have a firm (Labyrinth Group) which is designing *products*
that build upon the work we perform here. I have tried,
insofar as I have been able, to execute my duties as list
moderator fairly, and faithfully; now I want to weigh in
with my personal opinions, something which I have stayed
away from previously. In this I can't help but speak both
as a founder of the VRML effort *and* as a businessperson
who has certain commercial goals. I believe that the
highest commercial goals are also the best long-term goals
for the VRML community, and I believe that will become clear
over this series of postings.
*****
A Public Domain VRML Browser
The response from the public and press to the US Holocaust
Memorial Museum tour at SIGKIDS was phenomenal. We are
gathering interest from all corners of the Internet
community and the virtual reality (I prefer "interactive
simulation") community. People are starting to understand
the enormous potential both of the World Wide Web and
visualization tools like VRML browsers, both of which make
an inhospitable environment (Internet) friendly and
comfortable for anyone, anywhere.
This interest has led to an invitation from the Program
Committee for the Second International World Wide Web
Conference (Fall '94 Conference on Mosaic and the Web), to
put together a panel presentation on "Visualization and the
World Wide Web". I would very much like, as the centerpiece
of this panel, to present a baseline VRML specification to
the Web community for their examination, comment, and
approval. It's the right time to do it; the next conference
will be at least six months after that, much too far away in
this time of accelerating acceleration, especially around
Internet.
This means that we need to come to some decisions about how
to produce a workable VRML specification, and in a fairly
small amount of time; my next posting addresses this issue
extensively.
This conference opens on 17 October 1994, in Chicago,
Illinois, USA. At that time we (Labyrinth Group, that is,
although, with your permission, the announcement will be
made with the blessings and approval of this list) will be
announcing the availablity, via FTP "superdistribution", of
public domain (*free*, NOT shareware) VRML browsers for the
following platforms:
* Microsoft Windows 3.1 (Win32s, WINSOCK)
* Macintosh System 7.x (MacTCP)
* OSF Motif (SunOS 4.1.x, Solaris, IRIX)
We have secured a *royalty-free* license to a scaled-down
version of the RenderMorphics RealityLab rendering engine to
use the VRML browser. While it will only do flat shading,
it will render approximately 120,000 polygons/sec on either
Pentium(TM) or PowerPC systems, which is *very* nice.
We committed to a release of a public domain VRML browser at
SIGGRAPH, but this is the first announcement which we've
made to the list about this. We feel it's a good idea
because it achieves the following goals:
* Puts VRML browsers into *everyone's* hands
* Encourages the development of "worlds"
* Gives VRML the same accessiblilty that HTML now has
Millions of Web users (there are millions, aren't there?)
will be able to access VRML documents as easily as they get
to HTML documents today. That's something that benefits
everyone.
We will also be providing "reference source" to the
community, so that interested hackers and students and
professors, etc., can use our knowledge base to create
specific VRML applications. In this way VRML can grow even
faster. We want to see an "ecology" of tools develop, as it
has with GNU, and a marketplace where the commercial tools
have to be very good indeed to compete against the free
ones. (One thing we'd like to see *right away* is an OOGL
to VRML conversion utility. Hint. Hint.)
Before we are accused of a devious altruism, we want to make
something clear: we believe that, "giving things away is a
really good way to make a living". We would cite as
examples Cygnus Support, which makes a living supporting
free software, and Husky Labs, which provides "hand-holding"
services for clients new to Internet. There's plenty of
money to be made, by everyone, without actually *owning*
anything. Check out Kevin Kelly's book _Out of Control: The
Rise of Neo-Biological Civilization_, Chapter 12, "E-Money",
for a comprehensive explanation.
The Labyrinth VRML Developer's Kit
Labyrinth Group will be shipping the VRML browser with a
"Hotspace" (Hotlist in 3D) of interesting places to go. The
third posting will talk about some of the "places" which are
already in preparation. We would like to solicit
contributions from the VRML community, so that *our* work,
not just the work of commercial or academic clients, is
available at the time the application is released.
We have just put the finishing touches on the Alpha version
(1.0a10) of the Labyrinth VRML Developer's Kit, which
includes *everything* you need to start creating and
publishing virtual worlds in the Web. In the absence of any
other standard (see posting #2), we have extended Labyrinth
VRML to support some of the features mentioned on this list,
but it is still quite inadequate for the long term. (I am
quite embarassed that Labyrinth VRML scored as many votes as
it did on the survey - it was whipped up in a few days. Had
I ever thought it would do so well, we'd probably have spent
more time on it.)
We have released the Kit to a very few developers, who are
working to develop "content", which is the most important
thing to VRML right now; a good specification, without
compelling content, is much less useful than a poor
specification with a few amazing implementations. We have
committed to providing conversion tools to the real VRML
specification, once *we* agree to one, so that their hard
work doesn't get lost.
We will be releasing a more comprehensive version of the
developer's kit on 16 September 1994. As we are interested
in having as much compelling content as possible available
on 17 October, we'd very much like to have VRML List members
become VRML developers. While the kit is *free* (again, our
philosophy is to bootstrap the technology, and not to worry
about the money), we don't have the support resources to
give out thousands, or even hundred of copies. We'd like to
get a maximum of 50 more developers working on VRML content,
so we need to pre-qualify all interested parties. (Assuming
there's that much interest, of course.)
If you are interested, please send email ASAP to
[email protected]
Subject: VRML Developer's Kit
Within the message body you should address the following
issues:
* Who you are, what organization
* What you plan to do with the kit
* If you plan to be done by 17 October
* What environment you use for development
CAVEATS: The current development kit is designed to run
under MS-Windows 3.1/Chicago/WindowsNT. While the Macintosh
and Motif versions of the kit are under development, we're
not committing to having them complete before 17 October, so
they may not be ready for 16 September. Furthermore,
Autodesk's .ASC format (generated by Autodesk 3D Studio) is
the only 3D import format currently supported by the tools.
Unless you can generate .ASC files, you won't be able to do
very much, unless you enjoy building polymeshes by hand.
Finally, if there are those of you in the community who are
interested in developing VRML tools, we're here to help.
Please contact us (at the same email address), and tell us
what you'd like to do. If it's appropriate, we can release
various parts of the source code (VRML parsers, etc.) to you
for your own projects. We will be making all of that
publicly available after 17 October, but if you want a head
start, let us know ASAP.
Thanks for your time; we've found this process incredibly
exciting, and look forward to working closely with all of
you in the future.
Look for posting #2 in a few hours.
Mark Pesce
VRML List Moderator - http://www.wired.com/vrml/
President,
Labyrinth Group - http://www.butterfly.net/lab/lg/
[email protected] - http://www.hyperreal.com/~mpesce/