A distributed system is one where all the major components may run on different machines and communicate with one another over a local- or wide-area network. According to that definition, ENQUIRE is not yet a fully-distributed system, although some form of client-server relationship exists between its components.
It would be feasible to further separate the various system components, e.g. by setting up an RPC connection between the query layer and the BSS, and by adapting the interface software to run on users' local machines. Some of these possibilities will be explored in future Okapi projects. In principle distributed systems are intrinsically more flexible, permit load-sharing and parallel database searching, and enforce component modularity such that errors can be more readily identified.
In the practical context of the ENQUIRE implementation, however, achieving co-ordination between the various machines, and providing a full GUI-based retrieval service over a wide-area network (from the London Business School to City University) proved less than straightforward. The research staff were able to keep the system in operation over the period of time necessary to generate evaluation data for the project, but more reliable network connections and systems software would be essential to maintain the long-term operation of the service without dedicated staff.