IF IT AIN'T BROKE ...
... RE-ENGINEER IT! That's what many businesses are saying, both large and small. According to Infosys, a leading Indian IT consultancy company, "Up to 80 per cent of a western bank's annual IT budget goes on adapting or fixing outdated systems" (FT 25/02/04). Frank Muehleman, from Dell, further adds "Small businesses are finding that they need to migrate their legacy systems. They are finding that running their business with Excel spreadsheets does not scale." (FT 31/03/04)
What is re-engineering, and why is it necessary? Existing "legacy" software systems, outdated though they may be, can contain valuable information about how the business works. Moreover, the software is tried and tested, and it usually works (which is more than can be said for much of new software!). Thus, the software isn't "broke"---it may just need to be web-enabled, or moved to a more robust platform. Re-developing it from scratch may require detailed human knowledge of how the software works, which may not be easy to obtain (for example, if the program is written in an obsolete language like COBOL, or the team that developed it no longer exists). Even if we had this knowledge, re-developing from scratch will inevitably introduce new bugs! Researchers in re-engineeering study ways of finding automated tools that will examine a legacy code-base and discover important facts about how it works, and ways of automating the transformation of the code. Re-engineering tools can drastically reduce the cost of updating legacy systems, and minimise errors as well.
This is why Leicester is involved in a partnership with ATX Software called "From Legacy Systems to Services on the Net" (Leg2Net). Leg2Net is sponsored by the Marie-Curie programme - a European framework for the transfer of knowledge. Under this partnership, ATX makes available their state-of-the-art technologies on System Re-engineering and Generative Development for supporting teaching and research at Leicester. Furthermore, ATX seconds two of their staff to Leicester for a total period of 4 years during which a second generation of these technologies will be developed as a joint research project.