The strength of the web comes from unified protocols for network access (HTTP, HTML, RSS, RDF) and from hyperlinking.

Web 2.0 -- the social networking generation -- improved upon Web 1.0 by putting user-generated content first.

The promise of Web 3.0 is the demolition of information silos previously built. The end of Web 2.0 centered on service architecture: websites as services. However, the services were still silos.

For example, take perhaps the most successful of the Web 2.0 services: twitter. Twitter presents a simple, perhaps even revolutionary, way of communicating with a large set of people. However, there is only one twitter service provider: . If I wanted to set up my own twitter service and customize it to my ends, I couldn't, because the twitter software is not open-source or freely available. Even for web services and APIs that do have open-source backends, there is generally poor integration between parallel services.

I wrote about this in my blog earlier this year:

A big part of this service architecture is content conversion. That is, presenting interaction with a backend in multiple different formats. This flexibility allows users to interact with the web through whatever tools they find best to manage their information (and also the ability for tools to be designed to allow for different information management strategies). As an example, HTML can be converted to RDF: .

Like most other technologies, RDF will become useful when it starts to get adopted. Firefox uses some RDF files, and there is a smattering of RDF on the web. But it will grow fast once it becomes prolific enough for people to start using it. OpenCalais is really cool as they enable people to go there now with existing content. Maybe it will be enough to be a catalyst.

Text search has been the backbone of the web. But there are other ways of interacting with information, like tags, hierarchies, and directed graphs. RDF provides a good baseline of associations with which to make such graphs. So in addition to providing the building blocks for better information presentation, the encouragement of alernate ways of thinking of information democritizes the web. Not that Google is the most evil of overlords, but I'd look forward to alternatives.