On the Relationship between Software Complexity and Maintenance Costs

Abstract

As software becomes more and more complex due to increased number of module size, procedure size, and branching complexity, software maintenance costs are often on the increase. Consider a software such as Windows 2000 operating systems with over 29 million lines of code (LOC), 480,000 pages if printed, a stack of paper 161 feet high, estimate of 63,000 bugs in the software when it was first released [1] and with over 1000 developers, there is no doubt that such a large and complex software will require large amount of money (in US Dollars), social and environmental factors to maintain it. It has been estimated that over 70% of the total costs of software development process is expended on maintenance after the software has been delivered. This paper studies the relationship between software complexity and maintenance cost, the factors responsible for software complexity and why maintenance costs increase with software complexity. Some data collected on Windows, Debian Linux, and Linux Kernel operating systems were used. The results of our findings show that there is a strong correlation between software complexity and maintenance costs. That is, as lines of code increase, the software becomes more complex and more bugs may be introduced, and hence the cost of maintaining software increases.

A Survey of Cascading Style Sheets Complexity Metrics

“…Measurement of software is the process of defining, collecting and analyzing data regarding some software product to understand and control its complexity [16]. Software metrics are used to predict maintainability, reliability, and understandability of software [3,4]. Three steps are required to define and validate software metrics, namely, the definition of new metrics, theoretical validation, and empirical validation [17].…”

“…As is the case with regular software, CSS and its extensions have inherent complexity that keeps on increasing with age. High complexity is undesirable and leads to unreliable, difficult to maintain and understand, and error-prone CSS code [3,4]. Therefore, several complexity metrics have been proposed to measure and consequently, control CSS code complexity [4].…”

Software obsolescence risk assessment approach using multicriteria decision‐making

“…Software sustainment begins as soon as production or delivery occurs. Ogheneovo stated that “60% to 70% of effort is expended on the maintenance phase of the software development cycle.” The COTS software obsolescence drivers (availability of development tools, necessary talent, storage media, etc.) described in Sandborn et al also impact software sustainment support.…”

“…Software sustainment begins as soon as production or delivery occurs. Ogheneovo 29 to build an evaluation model. 32 Inputs for COCOMO include items such as lines of code, annual changes in traffic, and programmer familiarity with program code.…”

A Framework to Assess and Address Human Capacities Needed to Leverage Open Source Software Platforms in Developing Countries

“…The process of changing a soft-ware system or its component to correct faults, improve performance or other attributes, or adapt to a changed envi-ronment after it has been put into operation is called software mainte-nance [23,24]. Software maintenance activities span a system's productive life cycle and 70% of all effort on a software system is estimated to be expended on maintenance alone [25]. A range of human resource capacities are therefore required during both the deployment and operation phases of a software system.…”