The Rise and Fall of VRML: Part 1

The Internet is filled with monumental success stories. But one of its most intriguing tales concerns not success, but one of the most expensive and colossal failures we’ve ever seen. The story of the rise and fall of a rich media technology known by the initials VRML has the makings of a tragedy of epic proportions.

The full story – a tale of hubris, sabotage, egos, and mismanagement – would take a book to cover properly (one I am planning on writing). So, under the assumption that we learn more from our failures than our successes, let me tell you about how VRML almost became a standard for Internet rich media advertising, but then ended up dying on the vine.

VRML (pronounced “vermel”) stands for virtual reality modeling language, and it represented the first real attempt to bring real-time, interactive 3D graphics to the Internet. As you can tell by the name, it was developed at a time when virtual reality was a very hot topic, so the name reflects both the times and a cute play on the term HTML (hypertext markup language), creating a moniker that is a marketing brand manager’s nightmare. We are now going to compress a lot of data here: VRML caught on, and the plug-in necessary to make it work was bundled into Netscape and Internet Explorer. Over 25 million VRML players were distributed, and it had, in one form or another, the backing of Silicon Graphics, Microsoft, IBM, and other major corporations. There was only one problem: Nobody knew what to do with it or how it was ultimately going to make enough money to justify the millions being spent on developing and promoting it.

It hadn’t started out that way. The original notion was that we would all be donning our 3D “avatars,” virtual representations of ourselves, as we headed into the vast virtual worlds of cyberspace to meet and greet each other. If you’ve read the book “Snow Crash,” you know what I mean. (When I worked for SGI’s Cosmo Software, developer of the leading VRML player, it was mandatory reading.) In fact, the business plans for VRML represent the first (but not the last) time that a revenue model was derived solely from a work of science fiction.

Again, compressing a lot of data: Millions were spent promoting the idea of virtual worlds. The idea failed. Which brings us to when advertising enters the picture.

Developing advertising banners using VRML was not a new idea. Linda Hahner, who owns a design studio in San Francisco called Out of the Blue Design, developed some of the first banner ads using VRML technology back in 1997. DDB Needham and Out of the Blue Design developed a VRML ad for Pepsi, which got a fair amount of press at the time. But these were one-offs: proofs of concept of what could be done. No one was really interested in VRML as an ad tool, especially the big companies who had set their sights, and everyone’s expectations, on large 3D virtual worlds populated with chattering avatars.

Warning! Technical Info Ahead

A brief tech break is in order so that you understand why VRML was suited to ads in the first place. At the core of what makes VRML work is something called vector graphics. Vector graphics is the same technology that Macromedia’s Flash is based on. Most images, like GIFs, are known as “rasterized” graphics: lots of tiny pixels that are sent down the Internet pipeline one at a time. That would be almost 37,000 pixels, in a typical uncompressed GIF banner, which is why it takes so long to download all those fancy graphics.

Vector graphics, on the other hand, are not images but really more like instructions that tell the computer how to draw an image. This tiny instruction file is shot down the pipe and is read and interpreted by the plug-in (which is why you need a plug-in to view Flash content). The plug-in interprets the instructions and draws the image in real time. As long as there is a fairly powerful computer at the other end, very tiny files can “draw” incredibly complex, interactive large images and animations.

Cosmo Software developed a series of demos that showed what could actually be done in a 12K banner using VRML and unleashed them on the advertising community. All of a sudden, things changed. Advertisers and agencies, instead of looking at a large VRML world and exclaiming, “That’s cool. How do you guys make money?” were looking at 30-second animations, with total interactivity, delivered in 10K or less, and saying, “That’s cool. Can I get that tomorrow?”

Next week: VRML finds the Holy Grail then crumbles into a pile of shale.

Related reading