学术文献库

Software engineering often no longer involves building systems from scratch, but rather integrating functionality from existing software and components or implementing packaged software. Conventional software engineering comprises a set of influential approaches that are often considered good practice, including structured programming, and collecting a complete set of test cases. However, these approaches do not apply well for packaged software (PS) implementation; hence this phenomenon requires independent consideration. To explore PS implementation, we conducted ethnographic studies in packaged software development companies, in particular, to understand aspects of the feasibility study approach for PS implementation. From an analysis of these cases, we conclude that firstly; the analyst has more of a hybrid analyst-sales-marketing role than the analyst in traditional RE feasibility study. Secondly; the use of a live scenario software demonstration in order to convince the client to buy into the PS may lead to increased perceived feasibility and reduced resistance to PS implementation. Thirdly; the assessment criteria that are used to estimate the effort and time needed for PS implementation are new features, level of customization, software 'output', and technical needs. Fourthly; the feasibility study for PS implementation differs strongly from traditional RE as the analyst mainly considers how to deal with requests for modifications to existing functions.