As software becomes more crucial to organisational success, there is a tendency to turn towards third-party software houses for cheaper, faster and more effective solutions. This paper discusses the problems encountered by small suppliers, offering innovative software solutions to industry. It suggests that there is a common pattern to the development of such suppliers; and proposes a model of their evolution and development. This model, in turn, identifies a fairly consistent set of issues which should be addressed by organisations purchasing such innovative, strategic software products. The paper provides an appropriate criterion for judging the likely success of products purchased from such a software house, additional to the standard criteria of product quality and suitability.

Growing competition in the investment banking industry has given rise to increasing demand for high functionality software applications that can be developed in a short period of time. Yet de-livering such applications creates a bottleneck in software development activities. This dilemma can be addressed when firms shift to develop-ment methods that emphasize software (eusabili-ty. This article examines the productivity implications of object and repository-based in-tegrated computer-aided software engineering (ICASE) software development in the context of a major investment bank's information systems strategy. The strategy emphasizes software

This position paper describes the reuse space in five dimensions which are more or less orthogonal. With this description in mind we discuss when and how to learn reusable components and how to apply them. From this discussion we will conclude an operational environment for reuse. Keywords: description of kinds of reuse, components-based reuse, to learn components, application of components Workshop Goals: getting and giving feedback; networking; Working Groups: reuse process models, reuse education. quotation from Larry Latour, reuse workshop Pittsburgh 1989 Dusink- 1 Background The interest for reuse dates from the time that an Ada program library had to be built by our group. This means that we started from a component-oriented view on reuse. In this view reuse has two sides: the making of reusable components and the reuse of them. Of course, they both influence one another. With these ideas in mind a two-track research was started. One track working on the form and represen...

This paper proposes a new vision for the measurement and management of development productivity related to computer aided software engineering (CASE) technology. We propose that they be monitored and controlled via the application of dynamic software development "life cycle trajectory metrics. " This view develops out of management accounting approaches for process control and recent advances in CASE technology that make automated measurement possible. We suggest that current approaches involve the use of "static metricst ' for estimation and evaluation, with the result that the depth of the insights they can provide to management is necessarily limited. They only provide "point estimatesn of output or productivity at the beginning and end of the project. Yet to manage software development proactively for improved efficiency and effectiveness, management needs to track

Reuse of software components emerged as an industry goal subsequent to a NATO conference in 1969 where Doug McIlroy first introduced the concept. Over the two decades since, reuse remained a topic of much discussion and some research; and, although different views exist on the degree of success enjoyed by software developers in today's industry, most students of the state of software development practice agree that McIlroy's original vision has yet to be achieved and that increased reuse of software components is possible and remains a goal worth pursuing.

Our research centers around exploring methodologies for developing reusable software, and developing methods and tools for building with reusable software. Roughly speaking, developing with reusable components involves three steps: 1) searching and retrieving reusable components based on partial specifications, 2) assessing the reuse worth of the retrieved components, and, possibly, 3) tailoring the reusable components to the specifics of the problem at hand [3, 16, 19]. In this paper, we focus on a number of component retrieval methods that are developed, or being explored, in the context of SoftClass, an experimental CASE tool developed at the University of Quebec at Montreal to explore issues in software reuse [12, 15]. The methods discussed in this paper fall into two general categories: 1) "traditional" keyword-based retrieval algorithms, which aim at retrieving a single component that satisfies the user's needs, and 2) "component composers", which explore combinations of componen...

Software reuse is concerned with capturing software components in some form and then applying them to the construction of another application. The ultimate goals of software reuse are: to improve the quality of software produced; reduce the costs of software development; and increase the productivity of software developers. However, the present situation, and the concern amongst researchers, is that software reuse is not living up to its original expectations [FBPD + 91]. Although there have not been very many scientific studies that validate the claim, it is nonetheless a strongly held belief among researchers and developers that objectoriented software offers great potential in terms of software reuse [Mey88]. Much as object-oriented programming allows for the creation of more reusable components, it is the reuse of the design of an application that is most promising for attaining the goals of reusability. For this reason, our work concentrates on application frameworks, an import...

Software reuse is advocated primarily as a technique to improve programmer productivity. Reuse permits various artifacts of software development to be used on more than one project in order to amortize their development costs. Productivity is not the only advantage of reuse although it is the most widely publicized. By incorporating reusable parts into a new product, the parts bring with them whatever qualities they possess, and these can contribute to the quality of the new product. This suggests that reuse might be exploited for improving dependability as an entirely separate goal from improving productivity. If useful properties pertaining to dependability could be shown to be present in products as a direct result of software development based on reuse, this might be a costeffective way of achieving those qualities irrespective of the productivity advantages. In this paper, we address the issue of certifying reusable parts and exploiting certification to establish properties of sys...

We survey the past and presence of a recently emerging computer science research subdiscipline: software engineering's research on reuse. The aim of this paper is to learn from software engineers' experiences with reusing software, try to avoid their errors, and to adapt their know-how according to the special needs of knowledge engineering. So the emphasis lies both in concrete approaches and techniques as well as in the abstraction of their methodologies respectively the ideas behind them. The justification for our investigation is that, since the software engineering discipline is older than the knowledge engineering discipline, it seems promising to check its results on their applicability in knowledge engineering. Of course, not all findings and outcomes can be simply transferred due to some inherent differences between the respective areas of discourse. On the other hand there is much overlap between the fields. We therefore identify some more general experiences as well as some ...

FOOM (Formal Object-Oriented Methodology) is a method which assists in understanding and modelling organisations and thus information systems. The method promotes: ! highly appropriate and precise requirements specifications for information systems ! models of organisational and inter-organisational processes, which may form the basis for organisational and network reengineering.