Representational programming for design analysis

Abstract

We propose an approach to programming we feel is very powerful for design analysis in the context of design modelling applications. The approach has some similarity to dataflow programming, especially in its visual form. Where the operational nodes in a dataflow program network operate on the data directly, in our approach the nodes operate instead on the representational structures containing the data. For this reason, we term the approach representational programming. In this paper, we define representational programming and compare it to dataflow programming and the Grasshopper® graphical algorithm editor. We illustrate representational programming using a constructive approach to representational structures, termed sorts, and we exemplify its application to building design analysis. Finally, we show an implementation of a small design analysis tool extending the Autodesk® Maya® modelling environment using visual programming with sorts.