Experiences reverse engineering manually

Abstract

Better understanding of manual reverse engineering can make it and any associated systems reengineering more effective. We reverse engineered a version of a system (referred to as BOS/X) in support of a broader reengineering effort. System reengineering goals and other circumstances dictated a focused, limited duration, manual reverse engineering exercise. This presented an opportunity to study the BOS/X reverse engineering separately from other reengineering activities. We studied the BOS/X reverse engineering, the results achieved, and some limited reverse engineering metrics. The paper describes the: systems reengineering context; circumstances preventing application of automated techniques and motivating manual reverse engineering; reverse engineering results; resources required to produce the results; and an evaluation of the reverse engineering effectiveness. Combined, these results may be used as measures (standards of comparison) that can be studied further, for example to determine potential areas for future automation application.