It is still too hard to build, but we are clever lads and gals.
As good as it is to build social games, a WikiPedia3D can also be a game.
1. There are only a handful of basic sensor types. As the core GUI, these are semiotic engines. They respond to events and emit signs where those signs are the events in 3D. An avatar pragma + scene events == situation. Situations have semantics. See SemWeb.
Because the avatar is under real time control of the user, it has pragmatics: the rules of intension.
2. Clica (gestures as sensor events) can entertain, attract, repel, restore, whatever. They can also teach.
3. The advantage to a WikiPedia3D is it can use situation semantics to conduct events, therefore, instruct in context.
A buildable world teaches the pragmatics of building.
Online worlds are as good as the personalities of the inhabitants are strong in the media. People build to their tastes so all the world adds is amplification if that is all one does with a world. The games should emerge out of the environment itself because in the virtual world, the environment is an ambient/semiotic server.
The key is to use the sensors to collect data that enables the cultural models to be harvestable by the server itself. It analyzes the models selected, the text descriptions of the world, the behaviors selected by metatypes and feedsback those dynamics to the scene itself as backgrounds, terrain, avatarBots, botHumans, and so on... express that metadata in a common ontological dataset. Standardize it.
Different terms such as virtual ambience or server-ubiquity or sensate world apply. Some just call it virtual reality, a term out of vogue but still applied.
Imagine a server site that instructs in practices because the building tools enable a simple triples engine for declaring semantics, and easy assembly of semantically enabled models where the ontological root is religion... or math.
No magic here. It is an ontology built into a real time 3D rendering so the use of it modifies its rendering.
Affordances are given to conformance with norms but affordances are the means to change the norms. In gamespeak, you get powers for figuring out the puzzles, but the puzzles include the kind of world you build.
I've been around web3D a long time. Vendor worlds are about systems. User worlds are about content because content *is* about people.
The problem of VRML is it was too early; the problem of X3D is that it is VRML.
OTW, this works well and if the authors stick to good authoring techniques, it works acceptably well. Geekdom aside, VRML and X3D are still the HTML of web3D. The bet of X3D is that ownable content is a better investment for the customer. By that I mean, fully ownable down to the syntax. X3D is a bet that standards always achieve that goal better than proprietary systems. Proprietary systems don't have to negotiate with competitors, and that my friend, is the rub.
But the BEST deal for the content builder on the web is still X3D. It is the best bet for your customer too. Why?
OWNING THE CONTENT WHEN YOU ARE DONE.
Let the software vendors debate open source, the profit in it for the content builder and the customer of the content builder is to OWN THE CONTENT.
Ownership is ultimate resale. Ultimate reapplication. Freedom to build an avatar that you can show off on any X3D server operated by ANY X3D company. Your potential set of worlds only gets bigger and bigger and will not go away if a server company folds and friends, LOTS OF THEM HAVE. Think of it as your blog disappearing one day, oopsie...
... but until the freakin' X3D browsers and conforming X3D are interoperable, and the servers are stable, it's just a bet that what you are building works with other 3D worlds, unless of course, you pay competent 3D artists to do their voodoo. Yes, customers build but professionals can build better and get it up faster and keep it up longer. The prize is there for the taking if X3D is adopted by a community of content builders.
The difference between a call girl and a hooker for the same product is quality.