I note that R. Koster is unveiling his Areae product after a year of speculative hype build up. It turns out he is selling a framework for world builders.
Think DotNetNuke for a world platform. Maybe.
DNN's claim to fame is being a 'no-programming required' portal builder. The reality is it is a market framework (and a good one) into which those who can program on ASP.Net can add and sell modules to be dragged and dropped, skins for new looks, etc. There is a healthy market here for DotNetNuke developers. Being a real open source and free distro for the package with purchased add-ons, as long as the MS components have been purchased, installed, managed and maintained, these layered systems can do the job.
In short, there is no free lunch. Building and hosting virtual worlds or any hypermedia content for that matter always carry platform expenses. The more standard the components, in theory, the better the costs for the developers at the top or bottom depending on perspective, of that layer cake. That is the sale Koster is trying to make. His claim is he is creating a standard for this.
He is certainly creating his own specification. So is Linden Labs. The W3DC already has a standard for scene graphs, Collada for object interchange, H-anim for avatar bodies, and more coming for that family of technologies. Assessing Koster's chances here: I'd say poor for creating an accepted standard. For building a layer cake framework for non-pros to build worlds: pretty good. This isn't new territory technically. Getting a community of developers to create components for it will be the next hard sale. Then there is the problem of use. Stand-out 3D worlds tend to involve a considerable amount of handcraft. Drag and drop worlds can be a place to start, but eventually one is vertex twiddling and creating textures, so it still means learning what the average world user won't learn, but the market can fill those gaps quickly as long as there is 'natch on the transaction.
The question will be who is the customer for these layer cake worlds?
Answering a question on Tao Takashi's blog:
Standards are like roadsters, Baba. They are one part performance, one part replaceable parts, and no parts that only work on the car you happen to be driving. Specifications, the kind say the W3C or IETF make, for new technology can be blown glass from top to bottom with no hope of every replacing any broken parts. They can fail or succeed but they don’t have a legacy so they don’t have any promises to keep.
Koster complains today’s VR is based on 20 year old technology. He’s right. What else was that true of: HTML, DHTML, XHTML, SVG, XSLT, XML and more. The trick is that by the time something is ready to be standard, it is usually approaching or ending the middle cycle of age. That means it is roadworthy and there are lots of users who can benefit from the smoothing effect of standardization. It also means that initial burst of money from being the new kid in town is already earned and possibly spent. VC firms don’t like standards and they like monopolies even less. On the other hand, the market has been convinced to only buy standards-based technology.
This is where the dogs of the open sourcers unleashed early in the web history come back to bite the hands that held the leash. A VC wants open source because it is a cheap startup AND they want IP because if they can get a market lock, it is gravy money for years to come. They want standards to help them do that. That these are mutually incompatible positions bothers them not at all because they believe and have been shown that with the right combination of buzz, personality and paid placements, they can convince the public that their new entry in the market is *ta da!!*: standard. The truth is few people are actually professionally qualified to do real standards work. Thus ISO. They hate ISO.
So it is a good idea to evaluate Koster’s offering in terms of what it seems to be technically: the VR version of DotNukeNet except probably not BSD licensed. The first part of that will be a good thing for the types who build private business sites and don’t need to-the-metal handcrafted worlds, or for the newbs who just want a party house on the web.
Ultimately, however, the VR business is yet-another-data-center. The profits are in server space and services. The 3D stuff is just ‘onset cue simulated presence’ and while fun, not a big change. The right way to analyse offerings such as Areae once past the hype is to look at the services mix and determine which customers will pay for that and what they will use it for, and will they take over more IT responsibilities over time.