Improving Visual Representations of Code

E.L. Burd, P.S. Chan, I.M.M. Duncan, M. Munro, P. Young

Liz.Burd@durham.ac.uk

Visualisation Research Group
Centre for Software Maintenance
University of Durham
South Road, Durham, DH1 3LE, UK

Network Health Warning! The images contained within this paper are GIF encoded bitmaps grabbed from the original encapsulated postscript (EPS) files. These gifs are both considerably large and of substantially lower quality than the original images. All the images in this paper occur in chapter 3 only and we strongly recommend you turn image loading OFF when loading this chapter. For hard-copies of this paper please use the available postscript version.


Keywords: code visualisation, call graphs, control flow, graph simplification, graph presentation

Abstract

The contents of this paper describe the work carried out by the Visualisation Research Group in the Centre of Software Maintenance at the University of Durham. For obtaining a high level understanding of the code systems graphical representations are more useful than purely textual representations. However, graphical representations still have a tendency to provide the maintainer with too much information. This information overloading is detrimental to program comprehension therefore ways of reducing graphical complexity, preferably without information loss are required. Thus the work of the Visualisation Group has been to investigate strategies and heuristics for graphical information representation. Strategies considered include clustering and hiding nodes, layout, the use of colour, and the use of 3D technologies. This paper describes the results of this work and provides examples from a detailed case study that has been performed.

Contents

Table of Figures


Forward to Section 1, Introduction
To References

This page is maintained by Peter Young, please send any comments, jokes, insults or general abuse to (peter.young@durham.ac.uk).
Last updated: Monday 4 March, 1996.