Maybe I'm just impatient. Perhaps I'm becoming bitter. In Web Years (that is, One Web Week equals One Real Month), I've been involved in VRML for about six years. Like my parents who remember where they were the day Kennedy was assassinated, I remember the day I discovered VRML. I know. I need a life. A random browse to SGI's site and a download of WebSpace led me to Planet 9's Virtual SOMA. Being in Cleveland at the time, I didn't know what a SOMA was, but I knew this was revolutionary. Being a 3D artist and animator, I knew that I might be able to do this someday. Being an easily distracted person, I threw myself into it. I knew it would be difficult, and different from anything I'd ever done before, but I wanted to be a worldbuilder.

These days, I live in San Francisco. I work for a wonderful VRML company. I not only know what SOMA is, I have been there in person: a strange and wonderful experience, that - to see something virtually before seeing it in reality. Of course, I'm still easily distracted and VRML is still difficult. I watch and wait for and dream of creating the one world that will knock everyone's socks off and make VRML an overwhelming success. Instead, it seems I spend more time following discussions on things like avatar standards, scripting languages, and something called an external API. It makes me want to pull out my old e-mail .sig ("I'm an artist, not a programmer, dammit!"), but instead I think I'll use a new one: "It's The Content, Stupid!"

Now, I'm not writing this to tick anyone off, nor to revive the useless "Artiste vs. Programmer" debate. I have nothing but admiration for the folks who have given us this wonderful specification and the cutting edge tools to do something with it. They have changed my life and the lives of many others. But something is still missing: killer content for VRML. I lie awake at night wondering, "Was I crazy devoting my life to what amounts to a file format?" Okay, maybe that's all the coffee that keeps me awake, but still I find myself waiting for that One Great World.

I've lived on both sides of the fence now, first as a VRML architect and now as a VRML toolmaker. Hopefully, I can bring a perspective that is of interest to both sides. I recently talked to two well known VRML worldbuilders about their work, their tools and experiences with VRML. They don't do VRML for a living, but to call them amateurs would be an insult. They create wonderful worlds, and they're doing VRML simply because they love it.

So, what has been the most difficult thing for a worldbuilder to deal with? "Having to pick one browser and develop for it, knowing full well that there are other products that folks would want to work with," says Len Bullard. Len has been doing some fantastic worlds, most notably his "Kate Bush - HomeGround Towers" and "The Princess of Talos". His works are some of my favorites -- VRML for the sake of art, full of compelling visuals and audio. He calls VRML his "obsessive hobby" and sees it as "theater in a box". As a key player in SGML development, he knows standards all too well. And standards for VRML could be the biggest obstacle to creating compelling content. "A VRML standard was supposed to get around that (authoring to the browser), and it just ain't so. Look at all those variations for just one set of worlds and imagine how much upkeep it takes when combined with the HTML framework variations, other notations, scripting languages, etc. The WWW is heading for a train wreck, or has already wrecked. We are dead and just haven't noticed for all the noise being made by the rescuers."

Dennis McKenzie, a virtual reality enthusiast whose worlds are well known to readers of VRMLSite, has another problem. "It's a little frustrating to learn VRML and have to learn it all over again knowing it won't be all that long until it will change yet again. The fact that it has taken so long to catch on is also a problem. There really isn't any ready made market for VRML worlds out there." BR>

So, is there a killer application for VRML coming that will help create that market? Dennis doesn't think so. "I've heard this question before, about four years ago. Back then the subject was 'the killer app for VR'. I think they're still asking that one and I don't think it will happen that way. It will sneak in."

In fact, when it comes to the difference between VRML and traditional VR, Dennis would like to erase the distinction entirely. "Let's drop the ML from VRML. VRML is a computer language. VR is a concept. VR is what we are all talking about. I find it odd that the VRML community seems to find that so hard to accept. In a way, it almost seems like they are fearful of the same thing happening to VRML that happened to VR - that it won't live up to its hype. Be that as it may, what we are talking about is VR. Let's start calling it that."

Another concern, of course, is bandwidth. It's hard to get as creative as one would like when you're designing for an audience using 28.8 modems. Len doesn't let this stop him, though. "There are those who like to beat up on me for building "slow loaders". But I push it to the limit to make something compelling. I realize the problems of folks who can't wait for it to load, but I just don't care. VRML is an artform. Let the system catch up to the expression; not the other way around."

On the other hand, bandwidth is sometimes a symptom, and not the problem. When I was running WebWorlds, I was Chief Architect of Yahoo!3D (due to debut this month). I was given a set of requirements from Yahoo, and a team of brilliant Caligari 3D artists. To my dismay, though, I was getting worlds for review that were anywhere from 1 to 3 megabytes, uncompressed! And that would be for just one world, such as "Yahoo: Society and Culture". There were to be 15 worlds in all. "Guys," I said, "even if I compress this, I can't guarantee that anyone will sit through these downloads, let alone be able to walk through the worlds!" The real problem was in the requirements themselves. Yahoo insisted that every "3D icon" had to be labeled above with text descriptions, "just like our HTML version". That one phrase told me that the problem was in the concept itself.

So, is the process of creating VRML worlds all doom and gloom? Not at all. Len won IDS's V-Realm Builder contest earlier this year with a VRML site that he says his seven-year old son assured him would win. "He was sitting beside me when the news came in. The software, T-shirts and bucks were great, but impressing the boy was really fantastic." Both Len and Dennis love getting the positive feedback, but Dennis adds, "It would be nice to get some critical comments also. There just aren't any manuals out there that tell you how to build a good world. What works and what doesn't? It's hard to tell without a lot of both kinds of feedback."

Dennis and I also find another benefit from being early adapters of VRML... sharing the knowledge. "I really enjoy it when I can help someone out who is just getting started, " he says. "I remember working in a vacuum when I first started out, and won't ever forget the feeling. When someone you are in contact with gets hooked on building worlds, it kind of affirms the fact that you aren't over the edge with this stuff. Others feel the same way as you." In fact, Dennis is currently involved with a group of worldbuilders called ACTICON, building a universe of worlds called UNIVAR for the multi-user community. "Multi-user presents an entirely different challenge than single user worlds. In single user, you are trying to build a world that participants can enjoy by themselves. With multi-user, the main goal is to facilitate social interaction."

When asked what tools they were using, it seems that variety remains a key ingredient to building worlds. Len uses Live 3D, Community Place, CosmoPlayer and V-Realm as his browsers, and Caligari Pioneer, V-Realm Builder and ParaGraph's Home Space and Internet Space Builders for content creation. Dennis takes a different approach, using the VR freeware application Rend386, and a modeler called MODit to create low polygon-count objects. In the end, though, both rely on a text editor to finish their work. "WordPad is awfully useful," says Len.

So what do they want from the next generation of VRML worldbuilding tools? The list is long, but Dennis sums it up nicely. "A debugger, editor, modeler and viewer all in one. Cut and paste behaviors would be a definite plus, but there will always be hand editing involved with this, I think." Len adds, "We will need more editor power for behaviors and interpolators. Figuring out a path by hand is tedious. Eventually, as we begin to string worlds together with code in game form or interactive fiction, we'll need event graphs or scenario editors." I would add that, beyond worldbuilding tools,

In next month's article, I'll take a look at a few of the big studios that devote whole teams to major VRML projects. Finally, we'll look at some of the toolmakers, and find out what's in store for VRML artists in the near future. Three months, three articles... that's about one year in vrmlTime, so who knows what will happen by then? As Dennis says, "The best is always yet to come."

Robert W. Saint John is Administrator of VRML Development for IDS's V-Realm line of products. He used to run the VRML content company WebWorlds in Cleveland, Ohio. Fortunately, he woke up one day and found himself in San Francisco, in more ways than one.