The maintenance problem of application software: An empirical analysis
Ghazi I. Alkhatib · Journal of Software Maintenance Research and Practice · 1992
Abstract This paper presents findings from the analysis of data contained in reports and documents maintained by an MIS department of a major pharmaceutical and nutritional manufacturer. To guide data collection, the paper develops a conceptual framework consisting of a descriptive scheme and an explanation model. The descriptive scheme classifies maintenance activities into four categories: adaptive, corrective, perfective, and ongoing support. The explanation model identifies meta‐factors that affect maintenance activity. The meta‐factors are age, size, programming language, processing environment, structured programming, modularization, analysis and design methodologies, end‐user involvement, documentation generation, and maintenance management. The paper analyses three business functions with high, middle, and low average time per repair maintenance project (ATM) measured in person‐hours over a period of one year. Each function has several systems, and each system consists of modules. The analysis of available data for all functions maintained shows that maintenance (both repair and update) consumed 49% of data processing (DP) resources. For the three functions, the majority of these resources are devoted to perfective maintenance and corrective maintenance. Guided by the explanation model, several variables belonging to the meta‐factors are identified. The following variables have the most significant influence over maintenance: real‐time processing, database processing, end‐user ongoing support, module size, number of runs and runtime per module, and number of reports and number of copies per function.