Systems Development: Basic Flaws in the Current Culture - Ideas for Rectifying Some of the Problems

Systems Development:-Basic flaws in the current culture-Ideas for rectifying some of the problems It is nov 21 years since the term 'Software Engineering' vas coined. In this time relatively little progress seems to have been made in getting the development of computing systems under control. This paper identifies three basic flavs in computing systems development cUlture, it tries to identify the reasons for the flaws and suggests some ideas for rectifying them. The flays identified are: (i) A lack of a quality culture in computing systems development. (ii) Application domains not clearly defined/understood/ researched. (iii) Too often computing research and development seems to be fragmented and heading in all directions simultaneously. An example of the lack of an understanding of quality and its consequence is as follows: "The perception and assessment of the quality of things vith which ve are familiar is an accepted and natural skill, Also the people responsible for developing these products normally know hov to achieve the required quality level. Abstract (Contd) With respect to computing systems, in general the doers and supervisors do not appear to have developed the skills of assessing the quality of a software product and/or its components. Hence hov vill then deficiency affect the selection of alternative sources of reusable code? ["It is not enough for everyone to do his best. Everyone is already doing his best." Dr U E Deming.]" Much, probably most, research and development in computing systems is computing or softvare technology led (eg The Alvey Program). The paper suggests that more research and development should be 'Application domain' led. Initially to identify the domains and the (particular tools/skills/design needed, both technically and managerially. If the 'Application domains' are understood then it is· possible that softvare engineers vill be able to move from being mechanics or technicians (equivalent to 1880 engineers) to engineers who understand the application area in which they work (equivalent to 1990 engineers). Also if research and development is application domain led then it is probable that the development of quality cultures in specific domain areas may accelerate.