J.M. Spencer
University of Newcastle upon Tyne. 1992
One of the issues which must be resolved in order to maximise the potential of advanced object oriented distributed systems is that of dynamically deciding the best placement for mobile projects. This position paper discusses work in progress on this problem and presents a new approach which proposes that the best object placement decisions will be made by taking into account the properties and attributes of each of the objects on an individual basis (i.e. no two objects are the same). Thus, for example, when considering the best locations for interacting objects, in addition to considering environmental factors such as CPU loadings and communication costs, an examination of the properties of the objects involved ought to reveal the best course of action. The contribution of this paper is to present a framework to help organise thinking about which object properties to use in deciding the placement of mobile objects. This framework is based upon logical genera analogous to those used in the biological sciences.