5.4 Conclusion

This chapter offered an overview of the building blocks that are required for a me-centric computing architecture. As you will discover, the building blocks are at different development stages while the networking infrastructure is well developed; the appliances and interfaces are just about to be rolled out. On the Web services side, things look even more distant . The reason for these differences is the length of time that the different fields have been researched. While networking has been researched for decades, appliances have been researched for only a few years , as the cost for appliances dropped only in the 1980s, and the quest for a good Web services infrastructure started only at the beginning of the 1990s.

To make me-centric computing successful, it is necessary to have all three areas of development integrated into a single framework. But one of the strengths of the me-centric computing paradigm is that while a networked infrastructure is desired, the components can also work without any connectivity to the network or other intelligent components. Therefore, we will see point solutions appearing in the near future that will grow together naturally with the introduction of new and better solutions. Another advantage of this approach is that it will make it possible to connect intelligent components with differing development stages. That means that there is no need to start an infinite update spiral as with today's software on personal computers. If a certain intelligent component, device, network node, or service does not provide a certain feature, the other components will know how to handle it.

The architecture needs to make sure that the following features are enabled: compound applications, delivery mechanism, and loose coupling. The compound applications will enable the integration of different data sources, ranging from local files to databases on the Internet. The delivery mechanisms enable the data to be passed between compound applications easily, and the loose coupling makes it possible for the applications to run even though networks or other resources are not available.

This end-to-end architecture allows service providers and application programmers to develop and deploy network-based applications to address a range of devices with similar user interface requirements, while allowing content providers to create content in a device-independent format, insulating the content from device evolution.

In the next chapters, we will explore in more detail the technologies that are required to run a me-centric computing environment and see what is presently available. We will also show how these technologies are integrated today and how integration can be improved in the future.