Systemförvaltningsmodeller

The term Software Maintenance was coined in the early 1970s. The subject has significance from several points of view. The costs associated with software maintenance are as a rule considerably higher than what the system cost to develop, which is borne out by many studies conducted in different countries over a period of twenty years. Scientific interest in software maintenance, on the other hand, has been sparse in contrast to interest in system development. The area of problems dealt with in the present work is software maintenance with an emphasis on software maintenance models. The general problems related to software maintenance have been detailed in several com-prehensive studies. On the other hand, knowledge of the effects of using software maintenance models is lim-ited, despite substantial investment in various organisations in order to develop and implement software main-tenance models. A software maintenance model is an idealised picture of recommendations as to how system maintenance should be carried on. The problems upon which this research study focuses have been concre-tised and condensed into three main issues: 1) what are the main characteristics of software maintenance mod-els; 2) in what way do software maintenance models contribute to the stated purposes; and 3) in what ways do software maintenance models contribute to solving the perceived problems in connection with software main-tenance? The approach chosen was to study ten different software maintenance models from ten different large organisations in Sweden and two international standards. The material was analysed on the basis of a framework that I created, containing eleven different groups of characteristics. Firstly, the results show the main characteristics of the software maintenance models studied, which primarily indicate a higher degree of specialisation and a greater level of detail in the software maintenance models. Secondly, it appears that the software maintenance models at best contribute to approximately half of the defined purposes. Thirdly, it is clear that the software maintenanance models only contribute to solving a third of the known problems associated with software maintenance. There is thus a relatively broad scope for improvement in this area