A Software Maintenance Project Size Estimation Tool Based On Cosmic Full Function Point

Abstract

Software size estimation is a good reference to assess the project's manpower, schedule and budget in the initial project planning. Nowadays, many organizations are concerned about the cost of software maintenance, because the cost of software maintenance is constantly growing. It takes almost eighty-percent budget of software companies on maintenance, which means that the costs of software development are out-weighted by maintenance costs. Function point methods are widely adopted in estimating the size of a software under development. However, the function point calculation of a maintenance project is different from a development project. In COSMIC-FFP, when analyzing a system developed from scratch, all messages in a sequence diagram will be accumulated as function points. However, when analyzing maintenance project, if the messages passed to objects which are the same as the previous version rather than the newly added or changed, such messages will not be counted as function points. It is necessary to distinguish between the unchanged and new or enhanced functions in order to distinguish between the maintenance costs in the future and the development costs in the past. This study proposes a method to distinguish between existing and new or changed function points for maintenance projects and develops a supporting tool for COSMIC-FFP. By comparing the differences between two versions of a system, we obtained the number of lines of code per Cfsu from a regression analysis, so that we can assess more accurately the size involved in a maintenance project.