A Distributed LISP-STAT Environment

1 Motivation With the advent of networking and high-powered workstations, and the rise of end-user computing alongside the traditional centralised computing model, the heterogeneous network is emerging as the most signiicant platform for many computing activities. A heterogeneous network consists of a number of resources (e.g. workstations, leservers, database engines and computation nodes) interconnected by a fast data network; open systems standards facilitate interoperability in this (possibly multivendor) environment. This platform ooers signiicant beneets for statistical computing applications. Not only can programs access repositories of online data as required, they can also access other processors in order to exploit specialised features or just to harness idle resources. Hence computationally intensive tasks can be accelerated by farming out subtasks to multiple processors. Further, this platform supports groups of users and hence there is scope for the development of applications which facilitate collaborative working, such as a team of researchers working on a common dataset. 2 LISP-STAT LISP-STAT 5] is a statistical computing environment based on the Lisp language; it is available on a variety of processors and operating systems. Lisp is a dynamic language and as such lends itself to distributed applications| its exibility, interactive nature and reeective (`code as data') properties make it an eeective means of supporting multiple cooperative processes on a heterogeneous network. For these reasons, LISP-STAT is an ideal basis for the development of a statistical computing environment for heterogeneous networks. LISP-STAT provides dynamic, interactive graphics. Under the X-windows system, LISP-STAT processes running on one workstation can use another for their display and user interaction. These facilities alone are suucient 1