The Diorama first prototype consists three main components: the Diorama Services Manager, the query mediation toolkit (subsection 4.3), and the wrappers (subsection 4.4). The Diorama Services Manager (DSM) is the primary service module which receives and processes client requests. The DSM services include metadata services (such as repository registration, consumer interface creation, and metadata browsing), query services (such as query installation, meta query retrieval, and distributed query processing), and object linking services, which link consumer interfaces with the relevant repositories which ''materialize" that interface. For each query request, the corresponding IDL interface definitions and the consumer's profile will be utilized, together with the metadata of relevant source registrations and the respective producers' profiles, to dynamically establish connections between a consumer's request and producers' information resources.
To process requests, the DSM communicates with an Official Metadata Catalog (OMC) server, which handles the storage of the information source repository and interface repository [17]). For example, when a DSM needs to load mediator information to process a consumer query, it instantiates the mediator by retrieving the mediator definition from an OMC.
Each different repository type needs a wrapper to connect with the system and ultimately to the consumers. To facilitate the incorporation of new information sources, each Diorama wrapper is specified with a suite of generic functions [16] which calls underlying local functions specific to the corresponding source repository. New sources are able to participate in Diorama by implementing these underlying local functions. Following this scenario, wrapper generation can be partially automated by having meta-wrapper functions produce skeleton wrapper code depending on the category of data repository (structured, semi-structured, unstructured). Section 4.4 will provide a further discussion.