There are a couple of things I've found out about recently which I think are significant developments in the 3D web.
Second Life Architecture Working Group
The Second Life Architecture Working Group are publicly working on a defining a set of protocols which will open up the next generation of the Second Life "grid" to allow others to host Second Life style worlds. Second Life is a popular (the most popular?) online virtual world and currently has an open source client and closed source server. All servers are currently run by Linden Labs, but the company recognises that if Second Life is to become as ubiquitous as the World Wide Web, they have to open up the technology.
What I find interesting about this standardisation effort is the willigness of the group to investigate the use of existing open Internet standards where possible. Examples include HTTP, REST APIs, XMPP, FOAF, XFN and OpenID. I think this is a much more sensible approach than trying to define new protocols for every part of the technology.
The basic approach of the group currently appears to be to take each feature of Second Life and either match it to an existing open standard, or if none applies then define a new one. Meetings happen online, even inside Second Life itself, and the chat logs are available to view on their wiki.
Vivaty
Vivaty were previously known as MediaMachines who created the FluxPlayer for X3D viewing and FluxStudio for X3D authoring. Under their new brand they have recently launched a beta of a new 3D social networking service which uses X3D technology to provide online virtual worlds similar to Second Life, but with a greater emphasis on social networking. This is really a flagship for the X3D standard and it will be interesting to see how well it performs.
This new direction for the company appears to be an evolution from creating developer tools and implementations of Web3D standards, to providing end user web services which use those standards. I think this speaks volumes about the maturity of the technology.
3D Internet vs. 3D Web
What's interesting about contrasting these two developments is that they are both working towards building distributed online virtual worlds, but going about it in different ways. One is creating a 3D Internet and one is creating a 3D Web. Also, one is taking a commercial service and turning it into an open technology, the other is taking an open technology and turning it into a commercial service.
Here's my question. Are virtual worlds and the web different uses of the Internet in the same way that email clients and web browsers are different, or are virtual worlds one possible application of the 3D Web? It could be that both are true, similar to the fact that both email clients and webmail exist. In which case, the 3D Web is a web interface to virtual worlds. This then strays into the much wider debate of desktop vs. web browser as a software platform.
The Second Life client includes not only the real time rendering of interactive 3D vector graphics, but also a huge array of other technologies including authentication, instant messaging, prescense, friends lists and even currency. I would call this a rich Internet client, because it involves much more than just a web browser. It's something which implements many different protocols over the Internet and is designed to be installed on a desktop PC, separate from a web browser.
The pure 3D Web approach I envisage would be slightly different. The client (web user agent) would include only the downloading/uploading (over HTTP) and rendering of 3D scenes (in X3D), with a client-side scripting engine (ECMAScript). HTTP authenticaion might also be included, as it is currently included in web browsers, but application specific protocols like instant messaging, exchange of currency and friends lists would be left to server-side web services. An instant messenger client may well use the Jabber protocol (XMPP), but would not require the user to download an IM client, it would simply be used via a web interface much like Meebo. Currency? 3D PayPal perhaps.
The pure 3D Web option fits very neatly in the context of other web standards – we already have XHTML and the start of SVG implementations in web browsers, X3D could be the next step, with CSS and ECMAScript playing their parts.
XMPP for the 3D Web?
As an aside, one interesting idea which has been mentioned in the Second Life Architecture Working Group discussions is the idea of using XMPP not only as an instant messaging protocol but as a general purpose point-to-point protocol. I've thought about this before – the idea of using XMPP in place of HTTP to overcome the limitations and synchronous nature of HTTP. I know XMPP is extensible, but I don't know enough about it to know whether this would work. The 3D Web is surely going to be a big stretch for the hack that is AJAX and XMPP could hold the answer to truly interactive 3D scenes.