Selected research challenges





Understanding Software Systems Using Reverse
Engineering Technologies
Research and Practice


Hausi A. Müller, Department of Computer Science, University of Victoria



Selected research challenges


    
A research agenda
    
Reverse engineering research recommendations
    
The Year 2000 Problem



A research agenda


    

[HKM+93,
Fei93]

    
More case studies

    
Reverse engineering benchmarks

    
Migration paths

    
Cost-benefit analysis of reverse and reengineering

    
Integration and application of domain knowledge

    
Reverse engineering education

    
Issues of scale, extensibility, and applicability


Reverse engineering research recommendations


    
[SWC93]

    
Use real-world programs as examples

    
Use systems as examples

    
Do empirical studies

    
Use multiple information sources

    
Focus on economic impact

    
Strive for utility above all else

    
Develop semi-automatic toolsMake goals explicit


The Year 2000 Problem


    
Problem

    
Software that uses just two digits instead of four to represent
the year
    
Length of time between two dates is often computed by
subtracting two-digit years from each othe
    
Phone call across time zones billed for 100 years
    
Problem is buried deep inside software and hardware
    
Invoices, payrolls, credit card transactions, bill payments,
inventory systems, loans, databases that sort by year
    
Banking, financial services, insurance industry, phone
companies


    
Fixing this problem will cost between $300 billion and $600
billion according to industry analysts [CNN]

    
A challenge for the entire IT industry and the world economy


Year 2000 Conversion Strategy


    
Assess application portfolio and identify impact

    
Implement a project plan based on application research

    
Define scope of conversion project

    
Identify problem areas with impact analysis tools

    
Identify source code to be modified through code reverse
engineering tools

    
Develop conversion approaches

    
Develop testing procedures

    
Validate conversion plan using a pilot project

    
Automate as much of the process as possible

    
Minimize risk


Year 2000 Vendors


    
IBM

    
Formal Systems, Ottawa

    
Viasoft

    
Intersolv

    
Millenium Dynamics Inc. (MDI)

    
IBS/Solution 2000

    
Peritus

    
Many others

    
http://www.year2000.com



Next Chapter:
Summary and conclusions






Copyright © 1996 H. A. Müller, University of Victoria
[e-mail]