Through a mirror darkly: How programmers understand legacy code

Abstract

Y2K – the Year 2000 software update challenge – highlighted for the general population what has been a recurrent issue in the software development community. That problem is the requirement to maintain and update legacy software long after the original programmers have gone, taking with them perhaps the only concrete knowledge of how the software actually works. Legacy code is older code that is still in use [28]. This research investigated how programmers understand legacy code in order to maintain and improve it. The findings were used to develop a theoretical model on how programmers understand legacy code, integrating difficulties and methods, including the need to understand why software was written in a particular style, format, or functional structure. The primary conclusions from this research are that knowledge sharing and collaborative structures in the enterprise could be significant factors for programmers and that even poor documentation is useful.