An Empirical Study Of Component Based Software Development Process For Evolving A Legacy System (T-0748) (MFN 6913)

Abstract

Systems that were developed way back with the techniques that were useful then but are hardly considered now and also those modern software systems which are based on modern languages but are not able to fulfill the business requirements are said to be legacy systems. Despite of many definitions of legacy system there is one thing common in all definitions which is its importance and significance to the particular organizations. These systems play the role of backbone in any organization as they possess precious data and are very robust. There is no doubt in their importance for any organization but the growing demands in the business value, process and software agility legacy systems has to be evolved to newer standards while staying with the same business value and assets. Though evolution is one of the best approach for reliving the dying systems but it is not easy by any means as at times it gets very complex because many of the aspects have to be focused before approaching and while implementing the evolution process. In our work we emphasized on these factors and found key factors that that needs to be focused and which force organizations to evolve their legacy systems. Besides factors evolution approaches has to be examined thoroughly as currently many evolution approaches are used while evolving the legacy system. One of the mostly used approaches is component based software development which makes the organization with quickness and trust as this approach is open to accept already used components which cut out the development time and risk. Also, CBSD develops system in such a way that it can be used again and again for creating another system, which makes CBSD one of the preferred approaches for evolving legacy system. The following research is exploratory research which is primarily based on comprehensive literature review and industrial surveys. In the literature review the basic purpose was to get deeper understanding of legacy system and the approaches used for the evolution of the legacy systems. While to find out the main factors which influence evolution we conducted a survey in the local industry of Islamabad, Pakistan. We distributed questionnaire in 11 different software development companies and we obtained 196 meaningful feedbacks in total. Results of this study figured out the main factors behind the evolution of the legacy system. After thorough analysis of the feedbacks from the survey in SPSS we identified 7 major factors which are analyzed and are discussed in detail in the later part. As per our work the performance degradation of the legacy system is the most important factor as more than 70% of the respondents considered it. While the inconsistency in the functionality is the second most important with 61%, Code quality is third with 31%, missing documentation is fourth important with 29%. Staff expertise is fifth with 18% and maintenance cost with 12% is the last most important factor for evolution as per the respondent’s feedbacks.