Innovation

Correspondingly, innovation on the Web seems to have slowed from an end user perspective as these entry level requirements for a public site have been raised. Any site hoping to service a large number of users needs high bandwidth, the high availability provided by multiple servers, multi-homed networks, and load balancing systems. These are all out of reach for a small developer or business starting with just a good technical or business idea, discouraging such people from having a go.

For any application, the development vision is likely to be compromised by the business reality of Web deployment. The need for coordinating the development process with the launch publicity results in very hard deadlines, which in turn discourage the use of higher risk technologies, resulting in a tendency to go with tried and true systems and techniques, because they minimise risk, eliminate the need to solve basic deployment details, and can be outsourced.

Indeed, much of the risk associated with new technologies is not associated with the application itself, but in the re-solution of these details. And a truly innovative application may end up stretching an existing platform to or beyond its limits.

If ``tried and true'' technology existed at a lower, more general level, it could be used to support existing high level abstractions, and still be compatible with new ways of doing things.

The Jtrix model allows functions to be outsourced from the application development process at a much lower level, and in a much less contraining way than with high level Content Management or Application server systems. It allows a developers ``tool kit'' to provide the basic functions with less constraining baggage associated with them.

Jim Chapman 2001-08-16