The paper describes an empirical study O/software maintenance that was carried out in a system software department in 1989 and 1990. The study Jocused on error occurrence andJault detection. Over 400 problem reports were studied. The study showed some unexpected results. It showed, for example, no relation between the phase of error occurrence and the solution time. An explanation is the gap between the methods as they are supposed to be applied and reality. Assessment O/the size o[the gap is one O/'the contributions ~[ this kind Of empirical stu~tv. mahltenanee, d~l~'cts, nwtrics
Maintenance takes up a considerable amount of software engineer's time. Conte et al. ~ claim it takes up to 60 per cent of effort, and Lehman 2 states that 70 per cent of the expenditure on a program is incurred after initial installation. This paper describes an empirical study of software maintenance. The goal of the study was to gain insight into the origin of maintenance. This insight should enable software engineers and managers to take improvement measures that should reduce future maintenance efforts.
The study is an example of analysis of the software engineering process. Analysis of a software process should lead to improvement of that process. Data on the development process are required to be able to analyse it. The need for measurement and data collection of software development and maintenance is often stressed. Many organizations do not, however, practise these activities in software development. For instance, one survey indicated that 50 per cent of software development organizations in the Netherlands do not collect any data on the software process 3. The study described in this paper can be perceived as an empirical study on the one hand. On the other, it is an example of the fact that data collection in software development can provide an organization with important insights and that data collection does not have to be complicated. It is hoped that this example inspires others to improve their software process by analysis that is based on facts and figures.
The paper consists of four sections. The first section discusses software maintenance --the topic under study. The second section addresses analysis of software deve-lopment. The third section details the design of the study, with the results and interpretation appearing in the fourth section. Finally, the paper is rounded off with some conclusions.