Catálogo de publicaciones - libros

The Capability Maturity Model Integration (CMMI) is a structured representation of software development processes that can support an organi-sation’s software process improvement (SPI) strategies. However, CMMI and SPI initiatives generally exhibit low levels of adoption and limited success. One of the major reasons for these shortcomings is that many organisations undertake SPI initiatives without knowing whether or not they are ready to undertake them. Our previous research has enabled us to develop a software process improvement readiness model/framework to address this problem.

This paper reports on the implementation of the SPI readiness model in three large-scale case studies. We have found that organisations with higher CMMI levels are more ready for SPI initiatives than organisations with low CMMI levels. We suggest that organisations at higher CMMI levels have developed capabilities that enable them to further leverage SPI than organisations at lower CMMI levels.

In modern organizations teamwork is considered a key factor for succeeding in business. A growing emphasis on team culture culminates with a great number of articles analyzing different aspects to improve teamwork practises. Since software development projects are normally team efforts, teamwork improvement in software organizations should also be considered essential. In these companies, software process improvement programs based on international maturity standards are current issues in software engineering investigation. In this article, we firstly establish the teamwork key factors for succeeding in software development projects. Secondly, these key factors are analysed taking the ISO/IEC 15504 as a reference improvement framework.

We have conducted face-to-face questionnaire based survey sessions with twenty-three Vietnamese software practitioners in order to determine software process improvement (SPI) de-motivators. The main objective of this study is to provide SPI practitioners with some insight into designing appropriate SPI implementation strategies and to maximize practitioners support for SPI.

We asked practitioners to choose and rank various SPI de-motivator against the five types of assessments (high, medium, low, zero or do not know). From this, we propose the notion of ’perceived value’ associated with each SPI de-motivator. We have identified ’high’ and ’medium’ perceived values demotivators that can undermine SPI initiatives. We have identified what demotivates developers and managers to be actively involved in SPI initiatives. We have also identified SPI de-motivators of small-medium and large sized organisations.

We present the application of the process workshop method to define revised work processes in software development companies. Through two empirical action research studies, we study the impact of company premises and goals on the execution and subsequently on the results of the method. We conclude that both premises and goals will influence the workshops, and suggest how the focus of the workshops should be altered to achieve better results depending on the context. We also strengthen previous claims that the process workshops are a good arena that fosters discussion and organizational learning, and that involvement in the workshops leads to higher acceptance and usage of the resulting process.

Reference processes are supposed to be the basis for collaboration and mature cooperation in software development. Large business organizations need large and diverse reference processes. However, process conformance is a constant concern. There are many explanations why a project may deviate from its reference process. This is especially true in larger software companies with a lot of different projects and variants modeled in a single reference process. During an industrial cooperation we have identified a phenomenon that adds to the problem: Unclear and incorrect information flows. Process modeling notations and practices in many large organizations nurture information flow anomalies. We improved the information flows in the reference software process by means of information flow analysis and flow patterns. A comprehensible reference process with reasonable information flows is easier to understand and therefore gains acceptance in the project team.

In dynamic and constantly changing business environments, the need to rapidly modify and extend the software process arises as an important issue. Reasons include redistribution of tasks, technology changes, or required adherence to new standards. Changing processes ad-hoc without considering the underlying rationale of the process design can lead to various risks. Therefore, software organizations need suitable techniques and tools for storing and visualizing the rationale behind process model design decisions in order to optimally introduce future changes into their processes. We have developed a technique that support us in systematically identifying the differences between versions of a process model, and in connecting the rationale that motivated such differences. This results in a comprehensive process evolution repository that can be used, for instance, to support process compliance management, to learn from process evolution, or to identify and understand process variations in different development environments. In this article, we explain the underlying concepts of the technique, describe a supporting tool, and discuss our initial validation in the context of the German V-Modell XT process standard. We close the paper with related work and directions for future research.

Because of a shortage of IT specialists, many companies which are not involved in software development business are forced to use employees who have minimal or no any knowledge about software development and IT as testers (let’s call them non-IT testers). The author of this paper has used years of experience in working with such testers to provide a description of them, looking also at their most typical testing styles and the problems which occur for testers, their colleagues and bosses, and the overall software development processes. Non-IT testers often feel like second-class employees, because they are forced to work in an environment in which they do not have sufficient skills. This paper reviews issues which should be taken into account when training these testers, including the question of what exactly they should be taught. Examples and conclusions are used to provide advice on the more effective use of non-IT testers to achieve positive results.

System products need to be developed faster in a global development environment. A more efficient user requirements collection and product feature analysis become more important to meet strict time-to-market and quality constraints. The goal of this research is to study and find the best practices to support distributed business requirements management during the early phases of product development. The paper describes the process of mining require-ments management organizational patterns. The experiences and improvement ideas of requirements management have been collected from a large company operating in the sector of the process automation industry. The results present issues that were found important when managing requirements in a distributed environment. The results are further generalized in the form of an organiza-tional pattern which makes it easier for other companies to reflect on and to apply the results to their own cases.

It has been demonstrated by several case studies that SPI programs generate a substantial benefit, especially for organizations with immature processes. However, experience shows, that there is little buy in to SPI programs from these organizations. The reason is the step by step approach in ISO 15504 and as well in CMMi (Staged) especially the lack of ability to deliver sufficient data for planning and confirming improvements. In these cases a measurement oriented approach might work better because it increases the probability of senior management commitment by focusing aspects which are in the range of management perception.

Code coverage and mutation score measure how thoroughly tests exercise programs and how effective they are, respectively. The objective is to provide empirical evidence on the impact of pair programming on both, thoroughness and effectiveness of test suites, as pair programming is considered one of the practices that can make testing more rigorous, thorough and effective. A large experiment with MSc students working solo and in pairs was conducted. The subjects were asked to write unit tests using JUnit, and to follow test-driven development approach, as suggested by eXtreme Programming methodology. It appeared that branch coverage, as well as mutation score indicator (the lower bound on mutation score), was not significantly affected by using pair programming, instead of solo programming. However, slight but insignificant positive impact of pair programming on mutations score indicator was noticeable. The results do not support the positive impact of pair programming on testing to make it more effective and thorough. The generalization of the results is limited due to the fact that MSc students participated in the study. It is possible that the benefits of pair programming will exceed the results obtained in this experiment for larger, more complex and longer projects.