Do Software Processes have value in and of themselves? I believe that as process researchers and practitioners, we would like to think so. But fundamentally, I believe that we engage in processes to achieve certain business, organizational and technical outcomes. Moreover, the achievement of these desired outcomes are dependent on the individual and collective efforts of highly skilled and experienced human agents. We can improve the chances of achieving our desired outcomes through intelligent design of the process, through increased skill sets and experience, and through the use of better tools. In this issue of Software Process: Improvement and Practice, we have four articles on process and method design. We look at this issue from several perspectives. In the first article titled, Requirements for the Design of Processes, Kesav Nori discusses the requirements for the design of processes. His premise is that on an industrial scale, processes need to be repeatable and designed to address customer needs and to deliver what it is that customers value. Moreover, process design needs to assert these intents and purposes of the process when the human agents of the process conduct their intellectual tasks. Mr Nori emphasizes the important role of human agents and presents a responsibility framework that defines a scope for both individual and collective action within the organization. In the second article titled, Method Engineering: Towards Methods as Services, Professor Collette Rolland from the University of Paris discusses process design from the perspective of method engineering. Her premise is that methods need to be engineered or adapted to meet the requirements of the task at hand. These requirements are derived from customer needs as well as the requirements of all stakeholders involved. The aim of Method Engineering is to provide techniques for composing (i.e., retrieving, adapting, and assembling) new methods from reusable components. Professor Rolland's article provides a survey of important results pertaining to the defining and assembling of components. These results indicate that the full set of benefits of method components may be achieved by moving to the notion of Method as a Service (MaaS). The notion of MaaS is defined and an approach to MaaS is outlined in this article. In the third article, Scoping Software Process Lines, the authors look at the challenge of adapting complex, heterogeneous processes consistently within an organization is discussed. Often times in real-world settings, it is desired to have one standard organizational process that is generic enough to cover most if not all development projects. At the same time, there is the need to provide enough process detail and granularity to address the specific needs of the project and to provide useful guidance. To address this challenge, the authors introduce the concept of Software Process Lines which are analogous to software product lines with the goal of reducing the complexity and effort required to tailor organizational processes for each project. This article proposes the major steps to be taken to scope a software process line including: scoping, modeling, and architecting. The authors describe the scoping process in detail and discuss experience from determining the scope of space process standards for satellite software development. In the fourth article titled, A Customizable Pattern-based Software Process Simulation Model: Design, Calibration and Application, the authors address the issue of process change and assessing the impact of these changes quantitatively. The authors present a process simulation tool, GENSIM 2.0, that uses generic process lifecycle model templates (macro patterns) coupled with models that are tailored to specific project instantiations (micro patterns). The complete tailored model is then coupled with empirical data from the project and used to quantitatively assess important questions about the process such as preferred process improvement options and optimal V&V strategy. Real-world applications of this model are presented in the article. The four articles presented in this issue were based on papers that were selected as being among the best papers of the 2008 International Conference on Software Process (ICSP) held in Leipzig, Germany in May of 2008. All of the papers went through an additional reviewing process and were significantly expanded from their original conference versions. The acceptance rate for these papers was under 8%.
Search from 250 M+ research papersSearch from 250 M+ research papers
Open Access
Aug 13, 2010
Stefan Hoehme + 1
