Software Reengineering Patterns

Funding Body: EPSRC

• Grant value 138,534 pounds

The Edinburgh Team

GRANT HOLDERS:

Dr Rob Pooley, Dr Perdita Stevens

OTHER PARTICIPANTS:

Dr Rick Dewar

Objectives

• To establish that the pattern paradigm is useful for describing and transferring expertise in systems reengineering by identifying candidate patterns, by observation of projects and by interviewing experts in industrial collaborators
• To validate these patterns and modifying them in the light of further interaction with experts and thus to produce a catalogue of important, validated reengineering patterns

Summary

CONTEXT:

Evolutionary component-based reengineering (CBRE) of legacy systems to support business process change.

PROBLEM:

Engineers have difficulty becoming expert at CBRE, particularly where the structure of a system must be incrementally improved, not abandoned. Effective learning materials do not exist, and although reengineering experts exist, they are rarer than design experts. So people learn largely by their mistakes, which is exceedingly expensive.

SOLUTION:

A similar problem of transferring expertise exists in design. Here PATTERNS have recently been enthusiastically adopted by software practitioners to help solve the problem. A pattern is a description of an expert solution to a common problem in context. Learning the pattern includes understanding the context, the problem, the solution, and its merits and demerits relative to other solutions. A pattern is an effective learning unit: the expertise embodied in a pattern is transferable. We will apply the same paradigm to reengineering of legacy systems, producing a catalogue of software reengineering patterns in order to help engineers learn to be more effective. We will interact closely with experts (in BT and elsewhere) and with reengineering projects to find and validate patterns.