8.14.1 Development of the Information Personae took place during the last two years of my PhD thesis. A full year was spent on expanding, researching and debating philosophical, social and aesthetic issues with colleagues from many disciplines before embarking on actual software development. The core philosophy behind the Information personae is that content creates the architecture, counter to the idea of 'containers' that 'content providers' fill after the architecture has been completed.

8.14.2 An Information Persona (IP) agent system is best described as a set of tools and interfaces for the dynamic construction, distribution, querying, and rendering of an "embodied" collection of information. Information is at all times linked to a person, searches result in links to people who carry the information we are looking for.

8.14.3 The IP can be broken down into several components, which collectively provide the sets of tools and interfaces for community members. These components will handle three tasks: 1) content management; 2) mobile agent brokering; and 3) dynamic view generation. The IP bypasses traditional notions of client and servers by containing the capabilities of both, allowing for decentralisation of computing resources via mobile or transportable agents, a convenient paradigm for distributed computing since they hide the communication channels but not the computational location. Through this architecture, powerful content-centred communities can form, providing a dynamic infrastructure which facilitates collaboration and communication.

8.14.4 Participants' interaction with their IP occurs through a variety of modes. For example, the most basic mode of interaction is via email messaging. One begins by submitting a request to create a new IP, and attaches a listing of all web sites, documents, imagery, etc. that are to be incorporated into the IP as content. The IP responds as needed, prompting more input about configuration parameters and preferences on issues ranging from rendering styles to access constraints. Once the basic IP is in place, "search engine" style requests can be submitted, notification about all access patterns in relation to content get logged, and additional content can be added as necessary-all via an ASCII message-based interface. [top]