About the xVRML Project

Some History

The Early VRML Spec Process

Over the period 1994-7, an online group known as "the vrml list" (because we "met" on the www-vrml mailing list) worked together to create the original Virtual Reality Modeling Language specifications. By 1997, version 2 of the VRML spec was firmed up and became known as VRML⁹⁷. After review by the International Organization for Standardization, VRML⁹⁷ also became ISO/IEC 14772-1:1997.

The Emergence of the Consortium

During the period 1997-9, a new group formed which became known as the Web3D Consortium. As it charged money to join, the Consortium became primarily an industry-oriented group. The "spec process" was moved off of "the list" and into the hands of "the Consortium". The theory was that VRML had "matured" and that a smaller and industry-oriented group needed to control the future development of VRML. In 1999, the author of this document left his position at NASA Ames Research Center (Moffett Field, California) to join blaxxun technologies (a company which developed and marketed VRML plugins for web browsers and a multi-user server technology). I became their representative to the Consortium's "Contributor's Group". The charge of the Contributor's Group was to create the next generation of the VRML specification, and to base the new spec on XML, the eXtensible Markup Language. I wrote one of the first "straw man" early-draft DTDs for the Contributor's Group during this period.

The Emergence of XML Schema

Due to a number of factors, DTDs have (in an operational sense) been largely supplanted by XML Schemas as the primary working documents used to define valid XML. Due to a number of factors, Schema developement by the Consortium for the next generation of VRML has been based on converting from DTDs rather than "starting from first principles". One of the results of this and other factors is that the current Consortium Schema document does not appear to be valid XML, another is that the current Consortium Schema document is about 6000 lines long, and a third is that the current Consortium Schema document is difficult for humans to read and to understand (one of the reasons DTD has been supplanted by Schema in so many venues).

Project Goals

I began developing "from first principles" a Schema to describe VRML documents in the form of an XML Schema in July of 2003. My goal was to approach the task in an "evolutionary, not revolutionary" manner:

to develop a human-readable XML Schema describing what a valid and well-formed XML-based VRML instance document should look like, and based directly on the VRML⁹⁷ specifications

I have chosen to confine the initial Schema to those VRML⁹⁷ nodes which were actually implemented successfully in browser plugins (in other words, prior to the introduction of Geospatial and NURBS nodes into the specifications in 2002).

After the xVRML Schema & Specifications are developed, a Java^™ package will be created (probably using the Castor data-binding package), both stand-alone and browser-plugin viewers will be developed, and controller systems will be created.