Persistent Spatial Relations - a Systems Design Objective

This paper approaches the selection and development of appropriate computational geometry algorithms from a system design viewpoint: for one particular but important application. This is the domain of GIS (Geographic Information Systems) or Geomatics. We start with an outline of the subject, followed by a brief description of current commercial practice, then set specifications for a “next generation” system that attempts to overcome current limitations. Appropriate methods are then selected on the basis of this analysis, and the status of the current implementation of such a system is described. We conclude with a discussion of remaining questions whether the specifications are appropriate, the efficiency and practicality of the algorithms, and large-scale implementation issues.