|
its main purpose, i.e., providing the user with an interactive interface, bears the load of any job that might be delegated to hardware better suited to the specific task. In contrast Distributed RUBATO attempts to port the functionality of traditional RUBATO to a distributed and more generic environment. The original Rubette software modules are replaced by network-transparent server components (also called Rubettes) that may run locally or wherever in the world the Distributed RUBATO infrastructure is implemented. A good example for a distributed Rubette is the replacement of local mass storage with a Rubette providing access to large databases containing for example musical scores in electronic form. These databases typically reside in different locations, say libraries at IRCAM or at the University of Zurich. Any RUBATO system may connect to these servers using a common Rubette interface, and the RUBATO user may access and browse the databases in a transparent way. The generic nature of the Rubette interface enables the servers themselves to access other database Rubettes. Other examples of distributed Rubettes beside the library Rubette, or BiblioRubette as one may tentatively call it, include a MIDIRubette, driving an external sound studio, or a specific computational Rubettes, running on high-performance, parallel or clustered, number-crunching computer systems. Figure 1 illustrates the configuration just described. Thus the workstation is freed from tasks it is not specifically designed to run, and puts its entire power to the disposal of the user for the manipulation of a graphical interface. Needless to say that such a graphical user interface behaves itself like a Rubette and so makes the consistency of the environment complete.
|
