Process Improvement and Software Reuse and Good Documentation

The purpose of this paper is to describe a maintenance documentation system with a number of features contributing significantly towards process improvement and software reuse. The system makes extensive use of techniques of logic programming [1] to enable effective software change management, requirements traceability and continuous creation of templates for cross project use. A repository of such templates is essential to the attainment of level 5 in the CMM model [2]. This is because with such a repository in place, a manager can switch a project from a waterfall process model to a prototyping model when there is such a need. Furthermore, these templates are useful towards procedure reuse [3] in the sense that they can be interconnected and then instantiated to give rise to new and more complex process [4]. With regard to software reuse, the system is also successful largely due to the extensive use of techniques of logic programming. An example of such scenarios is given by the fact that programs are stored alongside with their specifications and designs. This database of programs permits search by specification written in the Horn syntax [5] [6] [7]. If the system is coupled with suitable code generation techniques, it is then able to support generative reuse, source code reuse, design reuse and specification reuse [3]. Towards the end of the paper, we argue that it is possible to devise a metric to measure the “goodness” of any maintenance documentation system. Subject to detailed statistical studies with project history data and in this first instance, the metric ρ may be envisaged to have a linear form, ρ = Σaβχβ, where β enumerates all the attributes of the documentation system that are positive towards process improvement and software reuse. For example, if we agree that software reuse comprises of idea reuse, artifact reuse, procedure reuse, vertical reuse, horizontal reuse, etc, then in its generic form, ρ will have a χ term for each of these modes of reuse, χideal reuse, χartifact reuse and so on. These terms take on a value of 1 if the corresponding attributes are supported; otherwise, they take a value of zero. The coefficients aβ measure the relative importance of these attributes and are dependent on a number of factors including the nature of the project, the organizational management style etc. The use of metrics is intimately related to process improvement [8]. Corporate savings achieved through reuse and process improvement are well documented in the literature [9] [10]. Therefore implementing a system such as the one described above is an issue that needs justified attention.