Matias Vierimaa

Empirical Studies in ESERNET

The goal of the ESERNET project has been to combine empirical studies research and their results into a common repository. The main objective of the empirical studies in ESERNET has been to further expand the available empirical knowledge about selected software engineering practices that are relevant for the software industry. The approach selected has been to perform co-ordinated and repeatable empirical studies across varying software development environments. By producing reliable and successful empirical studies, ESERNET has aimed to continue the interest of the companies and academia to further support ESERNET and Experimental Software Engineering (ESE) repository. The ESERNET project created a specific plan for experimental studies that has been used as a basis for empirical studies. The studies followed a common process and produced specified work products. At the end of studies the results were stored to the ESERNET Knowledge Repository for other project participants to evaluate, replicate, utilise and learn. This chapter describes the selected approach.

Enabling Comprehensive Use of Metrics

Although software measurements have been used in industry for several decades, it is still difficult for companies to make the most of them. An efficient use of measurements is a fundamental issue - without it, measurement is waste of time and effort. In this paper we present an approach to supporting and boosting the utilisation of measurements at both project and organisational levels. The underlying idea behind the approach is to create a practical measurement utilisation plan. We present how the utilisation plan can be developed and how this approach can motivate software developers and managers to collect and utilise metrics. We also describe the experiences and feedback gained from a case study. We consider our approach as a vital enhancement to existing measurement methods, one to be adopted when aiming at a comprehensive utilisation of measurements. Moreover, our approach prepares the ground for packaging the measurement results in an organisation.

Automatic Measurement at Nokia Mobile Phones: A Case of SDL Based Software Development

Software measurement forms a sound basis for monitoring software development process and software product quality. Implementing software measurement is, however, difficult and creates several challenges and possible pitfalls. These challenges include the establishment of measurement practices, effective metrics collection and the use of collected metrics to improve development practices or product. Automated software measurement can have a significant role when adopting software metrics for systematic use. This paper is based on the results of a measurement automation project at Nokia Mobile Phones during the years of 1999-2000. This paper describes the experiences from defining metrics for controlling new SDL based development practices. The process through which the automation was implemented is described in detail and implications of the findings are discussed.