Experiments on the impact of deviations to process execution

Autor: Reda Bendraou, Xavier Blanc, Marcos Aurélio Almeida da Silva, Marie-Pierre Gervais
Rok vydání: 2013
Předmět:
Zdroj: Ingénierie des systèmes d'information. 18:95-119
ISSN: 1633-1311
DOI: 10.3166/isi.18.3.95-118
Popis: Software Process Models formalize the way developers in Software Development companies cooperate. Once formalized, these models allow companies to analyze, improve and optimize the way people work. It turns out that developers take these models simply as guides during execution, deviating from them at will. In this work, we investigate whether these deviations have a significant impact on the outcomes of the process. Because of the complexity of such investigation, mainly because of the number of variables involved (the process model, the developers, the development projects, the companies in which they are executed... ), in this work we limited ourselves to a single kind of process model: the issue resolving processes. These processes offer the advantage of being fairly standard in software development and of having their execution to be commonly mediated and logged by tools. We analyzed the logs produced by three open source projects in the aim of comparing what happens in situations in which developers deviated from the process and in situations in which they did not. Ourfindings confirm our initial intuition that when people deviate more from the process model the overall execution takes longer and needs more rework.
Databáze: OpenAIRE