Re: Concretizing the theorizing

Chris Holt (Chris.Holt@newcastle.ac.uk)
Mon, 24 Apr 1995 16:09:25 +0100 (BST)

From: "Andrew C. Esh" <andrewes@cnt.com>
> On Sat, 22 Apr 1995, Brandon Van every wrote:
>
> > I've noticed that after the initial flurry of ideas on this list, we
> > seem to have lost steam. I think it's because we need to ground our
> > discussion in something more tangible, so that ideas can be directed,
> > stored up, and built upon.
>
> I could start and argument, if you like. :)

Uh-oh.

> How about this. Let's completely divorce the VR idea from any connection
> to the Internet, except for distribution of spaces (worlds) in stand-alone
> (single disconnected file) format. You could put a space on a CD-ROM and
> sell it. You could distribute spaces via satellite link. I imagine
> products like that which would result from a combination of Myst and Doom,
> where the quality is higher, and you can walk around in realtime.

Well, of course people will put spaces on CD-ROMs (both static and
eventually modifiable); but these will have links out into the rest
of the world. It just happens if you try to go through such a link
and you're not connected, you'll bang your nose on the door. The
question is whether the distribution of spaces on such a disk will
really be any different in nature or structure from those on the
main web. After all, when someone is designing a room, they aren't
going to want to know exactly where the corridors leading off it are;
they'll just want to have books of doors available. [Presumably if
a link is working, the book will open at that page, and if it isn't,
you can't open to it.] But I don't see that the issues involved in
designing VRML differ significantly in either context.

> This approach opens the disconnected home PC market to us, which is much
> bigger than the Internet X-Windows community. I also completely sidesteps
> the "Server" and "Multi User Synchronicity" problems. Some inter-user
> interaction can take place over point-to-point (P2P) links, like any
> Head-to-Head game on the market today. P2P links are going to get much
> faster and more reliable than present day modem connections. ISDN is
> being installed quickly, it carries 128Kbps, and it connects in less than
> a second.
>
> Let's not depend on the Internet. It hasn't helped us at all so far. The
> more we sit around and wait for the eggheads over in WWW-VRML to figure
> something out, the longer we will have to wait.

I'm not sure what you mean here. Are you going to design a VRML all
on your own? Are you going to design spaces and then translate them
(or not) into VRML when it's suitable? Are you going to just dream
up a wish-list of properties, hoping that someone will implement them?
Do you want to design something just for PCs? If so, will it be
VRML-compatible? Or are you talking about transport protocols?

------------------------------------------------------------------------------
<a href="http://www.cs.ncl.ac.uk/chris.holt/">Chris.Holt@newcastle.ac.uk</a>
------------------------------------------------------------------------------
Long is the web / And hard, that out of hell leads up to light.