A powerful wide-area information client

Wide-area information sharing is rapidly gaining momentum. However, current information systems provide inadequate functionality for users to efficiently and effectively search and use the information made available. Users are faced with a vast information space apparently in disarray. Assuming current client-server information system architecture, we argue in this paper that more functionality is needed on the information client side. We describe an information client prototype that supports (1) the use of object technology to model the information space, (2) declarative querying as a complement to navigational search, (3) uniform access and integration of data from diverse sources, and (4) user-specified and transparent caching that is integrated with permanent local storage of remote data. We also identify the needed support from the server side to better achieve these capabilities.