| Summary: | Computer Aided Software Engineering environments have shown some promise in meeting the challenge of the software crisis. CASE tools have however generally supported only one fixed method each, whereas organisations modify methods as they use them. This lack of flexibility in CASE tools is being addressed through configurable metaCASE tools, which allow modification of the methods supported by the CASE tool. MetaCASE tools are still comparatively in their infancy, and have yet to be widely taken into use. To help them mature and become more usable, the process, concepts, and tools for configuring them to support methods should be improved, and they should be able to support a wider range of interlinked, evolving methods. In addition, the support offered to the normal CASE user must be improved to meet that now expected of normal CASE tools. The thesis extrapolates from existing metaCASE tools, trends in methods and investigations of CASE tool use to present a set of three extensions needed for the next generation of metaCASE: multiple tools to represent and manipulate the same data to multiple simultaneous users, working as part of projects which employ multiple methods. Improvements are also needed in the basic underlying data models used in CASE, in particular concerning relationships and complex
objects. The research follows an evolutionary constructive research paradigm. A new data model, GOPRR, is designed to answer the needs of metaCASE. The proposed extensions of metaCASE are implemented and tested in the building of the GOPRR-based MetaEdit+ metaCASE environment and its repository.
|
|---|