VRML 97, the revolutionary new specification for the Virtual Reality Modeling Language, has arrived, and everyone interested in web development or 3-D graphics is asking the same question. Is it for real? Has the shining promise of interactive 3-D environments on the World Wide Web materialized into a practical solution? Should web developers and 3-D artists be rushing to hone their VRML development skills in anticipation of major commercial demand for this technology?

This column, the first in a series covering this watershed development in 3D graphics for the web, will address these questions specifically. Subsequent lessons will delve into the details of the new specification, its creative possibilities and the new navigation and authoring tools. But first things first. Should we all be interested and excited about this new development, and if so, why?

We can't begin without an honest acknowledgment of VRML's past. When, a couple of years ago (a million internet years ago) the first specification of a language purporting to provide web users with an interactive, immersive 3-D experience was presented to the public, interest was enormous. Everyone took a look, and the vast majority were disappointed and never looked again.

The problem was not with the technology itself. It was actually quite impressive for a first generation swipe at such an extraordinary and ambitious goal. The problem was rather that the average user at the time had nothing close to the hardware support necessary to render out even the simplest 3-D scenes in even a rough approximation of real-time movement. Another problem was that the technology was hyped and oversold. Books and articles in the press spoke of immersive "worlds" and exploited the then-lively buzzword of "virtual reality." Visionaries blurred the concepts of cyberspace and 3-D space in vague, enticing terms. User and developer expectations were raised to unrealistic levels and the result was disappointment. Most people who took a serious look a VRML in the past have ceased to keep up with developments.

When I went over the massive Silicon Graphics Inc. (SGI) campus last week to interview the official VRML Evangelist at the company's Cosmo Software division, I was naturally skeptical. I was prepared for exaggeration that required considerable discount before it could be responsibly communicated to this audience. But, to my surprise, the Evangelist displayed the sober confidence of someone holding a hand full of aces. Bill McCloskey is a fascinating man, just the kind of extraordinary blend of talents and experience that we see everywhere at the top levels of both web development and 3-D computer graphics. A fine artist, a painter established on the New York scene, Bill encountered 3-D animation some years ago and fell in love. After stints with major players in 3-D and establishing himself as an expert in VRML design, he was asked to evangelize the technology for Cosmo. SGI is undoubtedly the leading commercial force in VRML, having provided its proprietary Open Inventor file format to jump start the original version of VRML. (By the way, Bill confirms that it's perfectly acceptably to pronounce it "vermil," even though he concedes that it sounds like the name of a rodent.)

I came to Bill to ask the hard question. Is VRML now for real? Is now the time to jump in and catch the wave? His answer was, unsurprisingly, "yes," but his reasons were detailed and, to me, persuasive.

In short, he argued that all of the foundation has now been laid for serious, large-scale VRML development. A new and powerful specification for the VRML language is now in place, browsers are being written to it and being rapidly distributed, and the client hardware base is rapidly moving to the level necessary to support this technology.

Let's consider these issues in detail.