More questions.

Randy Stiles (stiles@aic.lockheed.com)
Tue, 6 Dec 94 00:48:12 PST


Hi,

Since we all seem to be weighing in on this "VRML
is not VR topic", I thought I would too.

It is true that VRML has been focused on
specification of a 3D scene, and not saying too
much on details about immersion. So what. Those
can be added in.

I think it is wise at this point to focus on
specifying 3D scenes. This does not lessen its
possible impact with regard to immersive VR.

For Performer fans, I have been relentlessly
pestering the performer development group to build
a loader capable of loading OpenInventor 2.0
scenes. So have others, and I believe they will
be done soon - Michael Jones has made a couple of
announcements about this inventor loader on the
Performer newsgroup.

So if it takes Performer to convince you that it
is possible to use VRML immersively, then that may
be possible soon, given that a person could write
their own routines for selecting objects, and
loading them inside Performer.

And I don't think the VR should be removed from
VRML. I think they are trying to do something
useful for VR - a uniform, well-thought out scene
definition language that can be used across a
world network. I don't think anyone is pretending.

Date: Mon, 5 Dec 1994 23:32:40 -0800
From: hitchner@netcom.com (Lew Hitchner)
Sender: owner-www-vrml@wired.com
Precedence: bulk

>> Date: Mon, 5 Dec 1994 10:39:23 -0800
>> From: globus@nas.nasa.gov (Al Globus)
>> To: softpres@crl.com
>> CC: jch@nell.oki.com, www-vrml@wired.com
>> Subject: Re: More questions.
>>
>> .vrml is 4 characters, incompatible with DOS, .wrl doesn't sound like
>> .vrml, so how about .vr? There is no law that says there must be 3
>> letters in a dos file extension (c language files are .c by convention).
>> 'Dot VR files' kinda has a ring to it, also to a user a .vr is more
>> obviously 'virtual reality'.
>>
>> A note, vrml is really a 3D specification language with URLs thrown in.
>> It doesn't have much to do with VR. In fact, people who do real
>> VR with Silicon Graphics machines don't usually use Open Inventor, they
>> use Performer --- another SGI product designed for speed -- or roll their
>> own So why don't
>> we quit pretending that VRML really says anything about immersion,
>> admit that what it really describes is 3D graphics, and name the
>> thing appropriately? wrl is fine.
>>
>> Al Globus, CSC
>> NASA Ames Research Center
>>

I second Al's opinion. The consensus in research and popular journal
publications, even among the hyped-up media, is that a 'Virtual Reality
application' MUST include Interaction, and in it's 'true' form also
should include Immersion, though there are many commercial,
non-immersive, desktop VR software systems. That doesn't require
high-end SGI's to qualify, either. Even the lowest-end, inexpensive
desktop VR systems capable of running on a bare bones Windows 386 PC
system support some form of interaction. The VRML proprosal is, in my
opinion, essentially an Open Inventor Viewer with extensions to handle
WWW hypertext-like links -- nothing much VRish about that since it
supports neither interaction nor immersion. I think in choosing the
file naming extension, the emphasis should be focused more on 3D and on
markup language (ml) to be consistent with common usage among WWW
terminology. Also, since HTML file names aren't restricted by DOS's 3
character file extension limit, why should VRML (which should probably
be renamed to remove the VR)? I think .3dml would be more appropriate.

-Randy

# Randy Stiles Lockheed AI Center
# stiles@aic.lockheed.com Orgn 9620 Bldg 254F
# Office: 415.354.5256 3251 Hanover Street
# Fax: 415.354.5235 Palo Alto, CA 94304-1191