-by Justin Couch
When the development of VRML first began, deliberate decisions were taken to design the language a step at a time. Version 1.0 was purely for static scenes. Version 2.0, which Part IV examined in detail, added behaviors that could be programmed. The future certainly has other surprises in store.
This chapter looks at some of VRML's possibilities, as well as some interesting current developments. This chapter includes the following topics:
As time for the release of VRML 2.0 grew closer and closer, the VRML development community realized that many of the hoped-for parts were not going to be finished in time. Instead of pushing back the release dates, the specification writers dropped a number of areas they felt were not yet complete. Among these was the VRML binary file format and the method of talking to a VRML world from another application.
Despite what has been said over the past years, virtual reality is still in its infancy. True VR requires the heavy use of 3D graphics and some serious hardware. Until the release of the Pentium CPU, there was never enough horsepower on the home pc to participate in a VR world-at least, not with any degree of speed. With the release of Microsoft's Direct3D standard, however, there has been a new interest in 3D graphics on the desktop. Now almost every major video card manufacturer has video cards offering 3D acceleration in the pipeline. The combination of the two has led to a rush of 3D applications on the desktop, with VRML leading the way.
Even VRML lacks some of the basic qualities required of a virtual reality system. Although it's possible to create a lot of the functionality for multiuser systems within VRML 2.0, there are no built-in capabilities within the language itself. Even some of that capability was left out in the effort to get VRML 2.0 ready by its release date.
Some of the unfinished portions of the 2.0 specification are slated for an update of the standard in version 2.1. This next version should be finished some time before the beginning of 1997. Don't worry; the plan is to not change the file format you have learned in this book and not to make incompatibilities between the two versions. You could say that 2.1 is more a maintenance release of VRML, with a few extra features thrown in.
Probably the first goal for 2.1 is to get the binary format sorted out, which will provide a customized format for distributing large VRML worlds across the Internet. Until now, users have had to rely on gzip, a compression program that compresses the text file before sending it to the end user. A binary format will provide an even better solution because extra space won't need to be wasted on text strings to define the world. Instead of, say, eight bytes to contain a node description, you will use only one.
The binary format is expected to be based on Apple's 3DMF format, which is the basis of many of its current offerings. 3DMF is more a method of producing these files than the actual format itself. Apple already has a QuickTime Movie and VR products based on it. This will be a welcome edition to VRML. If you learn by looking at other people's code, don't despair-you will have the ability to convert between text and binary formats. This was one of the implied requirements of the binary file format.
Next on the to-do list of 2.1 is the external API definition. This API will enable you to control the world from an outside application. As shown later in this chapter, it's possible to get the VRML world to control the HTML page in a limited way, but there's absolutely no way of doing the opposite. The external API contains all the functionality of the internal scripting browser interface and adds a little more. So if you have a Java or JavaScript program, you will be able to control what happens in your VRML world from a Web page or even a complete standalone program.
There are a few other issues related to 2.1. One of these is exactly which language will be supported in the Script nodes. When this issue was first raised, it resulted in a language war on the development mailing list. At the moment, there's no one required language that all browsers must support. To be completely cross-platform means you will need to write your behaviors in both Java and JavaScript.
My crystal ball (shown in Figure 22.1) is last season's model, so I hope it still works. Anyway, I'll give it a shot…
Figure 22.1: Looking into the future of VRML, the wizard is unable to see a clear picture.
Based on the discussion undercurrents happening on the VRML development list (www-vrml), it seems there are several areas to be addressed in the next major version of VRML.
This is a tricky topic-everybody has a favorite protocol he or she wants to use. However, with VRML 2.0, world writers have the ability to experiment freely and easily with protocol. It's likely that by the time VRML 3.0 is being discussed, a consensus on how to achieve this will be in place. This may not be an exact definition of a protocol, but maybe an agreement from the browser writers can be made that VRML 3.0 worlds will be inherently multiuser capable.
There are already several proposals concerning the creation of seamless worlds in VRML 2.0. At the moment, VRML worlds are static in size. At some point, you reach the boundary and need to click on an object to go to the next part of the world. The current methods of providing seamless worlds are based on extending VRML 2.0 with Java so that this can be done. Where this could lead nobody knows. One possibility is that it ends up creating the equivalent of William Gibson's matrix on the Internet-a global 3D world accessible by everyone.
Another interesting topic that will probably be addressed is the issue of heads-up displays (HUD) within the VR world. If you want to display text related to some particular part of the world, then you need to run a separate HTML window. VRML enables you to add head-mounted displays and data gloves so that you can have an immersive VR environment. When you are in an immersive environment, you lose the ability to use the flat screen, text-type pages you get on a screen. HUD capabilities have come up a number of times in the discussions of www-vrml, and it's likely some form of them will be in VRML 3.0.
By now, you should have a pretty good idea of what VRML can and cannot do. Many of these restrictions depend on your own abilities and requirements. It wouldn't take much for the average programmer to code some solutions, but you probably aren't experienced enough in handling the complexities of TCP/IP communications to whip up your own multiuser scripts in a couple of days. After all, this book shows you how to enhance an existing site with VRML rather than show you the insides of VRML. How to write scripts in VRML can fill a book by itself.
VRML is still, by definition, a single-user world. Unless you go out of your way to include multiuser capabilities, you will be enjoying your own solitude in your virtual environment. To add multiuser capabilities, you will either have to write your own system for both the VRML end and the server to handle it or use somebody else's code and server. Later in this chapter, some of the systems that VRML browser companies have written using their own extensions are examined. In VRML 1.0, this meant you had to use that company's browser to access the world. In VRML 2.0, you just include the VRML extension mechanisms examined in the next chapter, so you no longer need to rely on one browser.
To a certain extent, seamless worlds can be created by placing proximity sensors around the edge of the world, then loading in the neighboring section when the user gets near the edge. This is a lot more complex than you might first imagine. It's a research topic all by itself. A number of companies and individuals (including myself) are actively trying to solve this problem in a useful way. This probably isn't a problem you will try to tackle yourself.
One problem not yet solved is interactions between objects already in the scene. Chapter 16, "Adding a Dash of Reality," looked at the Collision node. This node provides collision detection between the user and objects, but it doesn't cover what happens when your animated robot walks through a wall. Currently, the robot can walk straight through, but you aren't able to do so. The problem lies in the amount of CPU power it takes to compute the collisions. With the speed of current home computers, everything would be way too slow.
VRML also lacks the ability to type text into the world directly. Although you should be able to type text directly in the future, at the moment you must use the multiframed approach presented later in this chapter. You can't directly affect the world contents (at least not with VRML 2.0-though you will with the external API and VRML 2.1). The text documents need to present information in a way that the VRML world cannot. Jeff Sonstein's VRMLYahoo! front-end experiment combines the VRML part to provide the search capabilities with the HTML document to show the search results. The combination works effectively and probably means that direct input abilities to the VRML worlds will be a while in coming.
The popular press always seems to refer to the whole Internet/WWW
thing as cyberspace. If you are a Gibson fan, you probably
cringe every time you read about references to the whole Internet
thing being cyberspace. However, there are companies out there
trying to make cyberspace happen. Many of these companies use
their own proprietary format for describing the worlds, so this
chapter looks only at the ones that use VRML as the basic system.
| Coution |
OnLive and Worlds Chat work only on MS Windows 95 or 3.1. If you're using NT, you will have some problems with the system crashing. OnLive won't even install if it finds you have the incorrect hardware and software set up. |
OnLive has a unique 3D chat environment where you can use your sound card with speakers and a microphone to talk and hear conversations, rather than typing messages as you do in other chat environments. OnLive can be found on the Web at
http://www.onlive.com/.
OnLive's main product is called Traveler. It's the front-end that enables you to travel to other worlds and chat with online visitors. The main meeting world is called Utopia, but new worlds are showing up all the time, including some movie-related theme-based worlds.
So, ever wanted to go to Utopia? You need the right program to visit, and OnLive's Traveler works great (unless you're on Windows NT, as noted previously).
| Note |
You need to fill out a form to download the program. The program, currently in beta form, is still free. |
| Note |
When running the setup program, Traveler becomes the default VRML browser for your browser. To change this, go to Options | General Preferences in the Netscape browser. |
Figure 22.2: Selecting an avatar using OnLive's Traveler.
Figure 22.3: Using OnLive's Traveler program to visit the Utlpia 3D chat world.
The examples in this book have used version 2 of CyberPassage, so you should already be familiar with its interface. With the move to VRML 2.0, Sony dropped the support for their proprietary format. At the time this chapter was written, you needed to have two copies of CyberPassage: Version 1 for Sony's chat and version 2 for testing examples. The move to the VRML 2.0-based worlds should be complete by October 1996, and you will probably need only one browser-version 2.
CyberPassage takes a different view of the world from most of the other browsers. You may choose from only two Avatars. When you're wandering around in the world, it becomes difficult to spot who you're talking to because everybody looks the same.
Participating in Sony's worlds is similar to the Onlive method:
This is probably the most popular of all the multiuser VRML browsers, and the one I spend too much time in. The browser is based on the principles outlined in the book Snow Crash. You can define personal and business cards for yourself, introducing what you do for a living, as well as your interests, hobbies, and so on, which can then be exchanged with other participants. (See Figure 22.4.) It also enables you to completely control how much information you want to reveal to others.
Figure 22.4: A snapshot of PointWorld, the entry world for BlackSun users.
CyberGate is a bit different from the other browsers because all the areas for chatting and information are contained in one window.
Like most of the other browsers, you can use either a pre-built avatar or construct your own. Many people just go with the default avatars, but some regular users eventually create their own. You see some quite interesting ones out there. (Some of the interesting ones that I've seen recently include the U.S.S. Enterprise 1701-D, a toilet, a camera, and a pod of dolphins.) BlackSun avatars are any legal VRML 1.0 file, so whatever you create you can use-but it does have to follow the normal rules outlined later in this chapter.
Our final browser comes from Worlds Inc. The Worlds Chat environment was one of the first. At their Web site (http://www.worlds.net/), you'll find a demo of Worlds Chat, but a Gold version is also available that offers a lot more features. The demo version is a chopped down version of the commercial version, but you can still go online and chat with others.
Using the demo, your selection of pre-built avatars is limited to 15 and you can't assign your own name or custom avatar. The interface is nice to use, but the limitations of a 28.8K modem soon start showing in a busy world. The avatars are different as well-they are made of a series of images that change depending on the viewpoint. Figure 22.5 shows the avatar gallery, where you choose your persona.
Figure 22.5: They entry of the space station of Worlds Chat.
Once you have an avatar, you can proceed to Worlds Chat by clicking on the Go To Worlds Chat. You appear in the Lobby along with several other somewhat lost-looking avatars. Your chat box shows the conversations. By clicking on the colored spheres in the lower-right corner (see Figure 22.6), you can port to different worlds.
Figure 22.6: The Worlds Chat environment.
One of the fun ways to roam about cyberspace is as an avatar you've created yourself. The problem is that what you create depends on which browser you're intending to use. At the middle of 1996, all the multiuser browsers were still VRML 1.0, which meant that what you used was completely dependent on the browser. Some used just a collection of pictures from different angles (like the creatures in DOOM), and others took any VRML 1.0 file you created. With the arrival of 2.0-based browsers, it's expected that all browsers will start using a VRML file for the body.
Many browsers offer a collection of basic avatars to choose from, as discussed in the preceding sections. Some worlds, like OnLive, let you make simple modifications like changing the avatar's dimensions or colors. It's even more interesting when the worlds support custom-built avatars.
When you create an avatar, there are a few rules of thumb to follow:
One of the more interesting uses of VRML is to combine VRML with other Web technologies. When the specification was being written, the writers realized that VRML won't always be deployed the way you think it will. The parameters field in the Anchor node was the result of this realization. Besides using just the URL of the link, you may also want to communicate other values.
No doubt one of the most common ways of using VRML will be to use multiframe documents, with VRML in one frame and HTML in the others.
Remember the texture-mapped columns from Chapter 15, "Sprucing Up Models with Textures and Materials"? This example puts those plain VRML primitives representing the VRML logo in the middle of the room. Each one of them is linked to a different thing:
Listing 22.1. The basic frames layout of the page.
<HTML>
<HEAD>
<TITLE>An example HTML and VRML World</TITLE>
</HEAD>
<!--
An example of combining VRML and HTML together to provide
interesting content for your world.
-->
<FRAMESET COLS="*, 120">
<FRAMESET ROWS="75%, 25%">
<FRAME SRC="column_world.wrl" NAME="vrmlFrame">
<FRAME SRC="blank.html" NAME="docFrame">
</FRAMESET>
<FRAMESET ROWS="*, 120">
<FRAME SRC="index_file.html" NAME="indexFrame">
<FRAME SRC="vrml2.jpg">
</FRAMESET>
</FRAMESET>
</HTML>
This is a good example of how combining different Web technologies can add more functionality to your Web page. The preceding example is shown in Figure 22.7.
Figure 22.7: Combining frames and VRML can add another dimesnsion to your VRML worlds.
Where do you go from here? With the ability to put in programmable behaviors, one of the most obvious uses for VRML is in front-ends for pre-existing games. There has been a DOOM WAD-to-VRML file converter around for some time, and without a doubt there will be a Quake front-end.
VRML is designed to create platform-independent content, so it will never run as fast or as smoothly as a custom-built game. VRML games are more likely to be of the not-so-high performance types, like mysteries and adventure games. However, some time in the not-too-distant future, there will be Internet games based on VRML worlds. Like all new technologies, this one is waiting to be explored.
A good preview of what's possible can be found at Keith and Margo's Murder Mystery, found at www.murdermystery.com. This mystery is composed of several VRML rooms that you explore to find clues. By solving the mystery, you qualify to win trips and prizes. This site, shown in Figure 22.8, is a good example of how VRML and frames can work well together.
Figure 22.8: A scene from Keith and Margo's Murder Mystery VRML site.
Although VRML is and will be behind the gaming scene for a while, the gaming world isn't waiting around. The entire gaming industry is moving toward multiplayer games across the Internet.
The way to get around the bandwidth problems is for each player to have a copy of the game stored locally on their machine. That way, only the game updates, such as the positions of other players, need to be transmitted. Several hosting sites have begun to appear that enable you to play commercial games over the Web. One such site is the Total Entertainment Network, or TEN for short. At TEN you can play Duke Nukem 3D and Terminal Velocity, and they aren't stopping there.
Another good example is Internet Monopoly by Virgin games, and Looking Glass Software is planning to make multiplayer Internet games a regular feature on all their future games.
You have almost reached the end. The tour of VRML has nearly ended, and you should be confident enough now to start making your own worlds. In the last VRML chapter, the finishing touches are added to your world, creating the full multimedia VR experience. In the following chapter, the world developed in Chapter 17, "Real-Life Examples: The VRML Art Gallery: A VRML World by Hand," has everything you learned in this part of the book put into practice. Hold on-it's one heck of a ride!
| Will we ever see Gibson's matrix happening in the future? Is the Internet the start of it? | |
| Tricky question. Certainly VRML has the makings to be the foundation of the matrix, but there other competing technologies out there. At the moment, VRML has the backing of just about every major
computer company except Microsoft, which has decided to do its own thing. Whether Microsoft can steal the march from VRML remains to be seen.
Before users can go all out with immersive VR systems, a lot of work needs to be done. Remember that there are no defined multiuser protocols, and the current head-mounted displays are still cumbersome for extended periods of time. Besides, they aren't a common item you pick up at your local computer discount store. Bandwidth on the Internet is the other major problem. Cable modems and other solutions will be the start of developing the required high-capacity links. | |
| Can I incorporate other technologies, like Shockwave Director and RealAudio, into my worlds? | |
| Currently, there's no real way to integrate Shockwave in a VRML world, but some sites use Shockwave as a front-end for a VRML world (the two technologies exist as separate elements on the HTML page). | |
| Where can I find out more about creating my own VRML avatar? | |
| There are several sources on the Web, but take a look at Gerry Paquette's Avatar guide at http://www.magmacom.com/~gerryp/howtoav.html. This guide takes you through the process of building your own avatar by using Pioneer. Gerry is well known for his avatars in the CyberGate community
and in several worlds he has built, one of which features a pool with sharks and a bartender you can order drinks from.
RealAudio is a real mystery. The definition of the sound system is examined in the next chapter, but for the moment, RealAudio inclusion will probably happen fairly quickly |