BOUNCE Non-member submission from [Rick Innis <>] (fwd)

Brian Behlendorf (
Wed, 22 Jun 1994 13:15:41 -0700 (PDT)

(Sorry about this - from now on all digest readers can post to www-vrml

Brian, who is on vacation, really, and would kill for something
better than a 14.4 modem and dialing through netcom....

---------- Forwarded message ----------
Date: Wed, 22 Jun 1994 10:28:25 -0400
From: Rick Innis <]%20(fwd)">
Subject: Re: LANG: OOGL [VRML Digest 21 June 1994]
To: ]%20(fwd)"
Organization: SoftQuad Inc
X-Mailer: Ream 4.16c - The Choice for a New Generation

> Sam writes:
> >Then base it on the same ISO standard that HTML is based on -- ISO 8879,
> >Standard Generalized Markup Language.
> Apples and Oranges.
Probably true.

> HTML works well as a subset of SGML because it is *mostly* doing
> exactly what SGML was intended for -- marking up text.

SGML is more like a meta-language, which you use to write a
description of the structure of a <class of> documents. In SGML terms
these are called Document Type Descriptions, or DTDs.

HTML is an SGML DTD which describes the class of documents used by WWW

SGML is about _structural_ markup. It says nothing about the
_appearance_ of a marked-up document.

> Our application is almost entirely different. I'm very nearly certain
> that attempting to use SGML as our basis would be a disaster, because
> it's just not intended for this sort of purpose.
Depends. From what little I've had time to look at on this list (I'm
a one-man QA department with four products to deal with this month...)
people seem to be spending more time talking about appearance than

You might be able to create a DTD (or even a suite of DTDs) which
would let you describe the structure of a VR "room", including what
rooms it's linked to (via HyTime links, for example).

How a VR browser chooses to present these is up to it, and should be
covered by a separate process. There are attempts within the SGML
community to develop presentation descriptions based on SGML, but
wouldn't hold my breath waiting for one....

> Using standards is a Good Thing, but you have to pick the right standard...
Very true.