BurdMunro98a |
PublicReengineering Wiki |
|
| They describe a 10 step process to identify reusable components of COBOL code. They analyse the code (bottom-up) and interview experts (top-down) to gain
insight into the application.
They reference Tracz and say that 70% of the total life cycle costs are spent on software maintenance.
The mention "software scavenging" as a synonym for finding reuse candidates.
Their method also attempt to estimate the effort required to reengineer the candidates into encapsulated components.
Thy combine 2 previous methods (RE2 and RECAST.
They recognise that their approach does not attain "industrial strength".
The differences that the top-down and bottom-up approaches highlight are generally due to the level of granularity achieved by each approach.
Their case study (I think a BT system) has 21,000 lines of code. most candidates were 200-400 lines lone, their largest was 4100.
They were able to whittle down 20 candidate reuse modules down to four during their ten step process.
Their approach also provides impact analysis (ie what effect does a change have).
Related pages: Bibliography |
| This page last edited on 28 October 1999 |