http://www.computer.org/software IEEE Software's mission is to build the community of leading and future software practitioners. The magazine delivers reliable, useful, leading-edge software development information to keep engineers and managers abreast of rapid technology change. The authority on translating software theory into practice, the magazine positions itself between pure research and pure practice, transferring ideas, methods, and experiences among researchers and engineers. Peer-reviewed articles, topical interviews, and columns by seasoned practitioners illuminate all aspects of the industry, including process improvement, project management, development tools, software maintenance, Web applications and opportunities, testing, usability, and much more. en-us Wed, 4 Jan 2012 11:00:01 GMT http://csdl.computer.org/common/images/logos/software.gif List of recently published journal articles http://www.computer.org/software http://doi.ieeecomputersociety.org/10.1109/MS.2011.132 The growth of cloud computing is becoming significant and attracting the attention of the IT industry. Cloud-based applications and services are continually being developed. As applications and services migrate to the cloud, testing can be expected to follow the same trend. Therefore it is necessary to understand the dynamics of testing in the cloud. For this purpose, our study is based on views obtained from interviews with eight organizations that are using cloud computing. The results suggest that cloud computing can make testing faster and enhance the delivery of testing services. Cloud computing also highlights important aspects to testing that require attention, such as integration and interoperability. Organizations need to develop a better understanding of cloud-based testing before they can adopt it into use. http://doi.ieeecomputersociety.org/10.1109/MS.2011.132 http://doi.ieeecomputersociety.org/10.1109/MS.2011.91 Quantitative management requires an understanding of the nature of variation and utilizing this understanding to improve process performance. Popular process reference models like CMMI embrace quantitative management at high maturity levels. However, even at high maturity levels, there is limited number of available studies on the application of quantitative techniques. In this article, we describe a systematic approach with a well-defined, detailed guideline intended for use by software organizations in assessing their processes and applying quantitative techniques to understand their potential for improvement. More specifically, we introduce an Assessment Approach for Quantitative Process Management (A2QPM) to evaluate the suitability of a software process and measures for quantitative analysis and we discuss the application of this approach in 12 processes in 6 different organizations. The results show that as systematic approaches and supporting tools become available, software organizations can readily apply quantitative techniques to improve their processes. http://doi.ieeecomputersociety.org/10.1109/MS.2011.91 http://opac.ieeecomputersociety.org/opac?year=2012&volume=29&issue=01&acronym=software IEEE Software http://www.computer.org/portal/site/software/ http://doi.ieeecomputersociety.org/10.1109/MS.2011.159 Lack of precise specification is a well-known syndrome in the software industry. This article covers some peculiar aspects of the problem and its causes in the automotive software industry. We describe how the situation motivates engineers to grasp reverse engineering methodologies to comprehend third-party components. We tell our story of a reverse engineering practice that we have developed and adopted in a recent project on testing embedded systems of a modern vehicle. http://doi.ieeecomputersociety.org/10.1109/MS.2011.159 http://doi.ieeecomputersociety.org/10.1109/MS.2011.158 Recently, cloud computing platforms, such as Microsoft Azure, are available to provide convenient infrastructures such that cloud applications could conduct cloud and data-intensive computing. To ensure high quality of cloud applications under development, developer testing (also referred to as unit testing) could be used. The behavior of a unit in a cloud application is dependent on the test inputs as well as the state of the cloud environment. Generally, manually providing various test inputs and cloud states for conducting developer testing is time-consuming and labor-intensive. To reduce the manual effort, developers could employ automated test generation tools. However, applying an automated test generation tool faces the challenge of generating various cloud states for achieving effective testing, such as achieving high structural coverage of the cloud application since these tools cannot control the cloud environment. To address this challenge, we propose an approach to (1) model the cloud environment for simulating the behavior of the real environment and, (2) apply Dynamic Symbolic Execution (DSE) to both generate test inputs and cloud states to achieve high structural coverage. We apply our approach on some open source Azure cloud applications. The result shows that our approach automatically generates test inputs and cloud states to achieve high structural coverage of the cloud applications. http://doi.ieeecomputersociety.org/10.1109/MS.2011.158 http://doi.ieeecomputersociety.org/10.1109/MS.2011.157 In a recent IEEE Software Article, Sallyann Freudenberg and Helen Sharp unveiled a burning question: “Sociological studies---what were the personalities in successful/failed agile teams?” [Fre10]. This points to a widely held belief that if we can find the right mix of individual personalities, we should end up with a successful team. Intuitively, this makes sense: If group members are all introverted, they might be too shy to communicate well; if they are all type-A personalities, they might all clamor over one another to lead and no one would follow any instructions. The aim would be to pick and mix personalities to get the right team. http://doi.ieeecomputersociety.org/10.1109/MS.2011.157 http://doi.ieeecomputersociety.org/10.1109/MS.2011.156 Service-oriented architecture (SOA) has gained significant attention as a means of developing flexible and modular systems. Academic studies of SOA as a systems development philosophy abound, and recent industry surveys indicate that most firms are also actively pursuing SOA initiatives. This article uses a rigorous case-study methodology to examine five main benefits of SOA – business flow transparency, plug-and-play capability, leveraging legacy systems, rapid product development time, and reduced costs – as perceived by the organizations that have implemented SOA. Participants in this study report that not all stated benefits are realised due to, among other things, a failure of service-oriented thinking at an organisational level, problems allocating financial responsibility for services within and between organisations, and a lack of mature tool chains. These issues are significant because they are, according to the participants in the study, critical to leveraging investments in SOA. http://doi.ieeecomputersociety.org/10.1109/MS.2011.156 http://doi.ieeecomputersociety.org/10.1109/MS.2011.153 The large scale, dynamic, and heterogeneous nature of clouds, poses numerous security challenges. The main challenge is providing a robust authorization mechanism for distributed clouds that incorporates multi-tenancy and virtualization aspects of resources. In this paper, we present a distributed architecture that incorporates principles from security management and software engineering and propose key requirements and a design model for the architecture. http://doi.ieeecomputersociety.org/10.1109/MS.2011.153 http://doi.ieeecomputersociety.org/10.1109/MS.2011.131 We present an experimental approach to evaluate the potential of Google App Engine for high-performance parallel computing. We designed a generic master-slave framework that enables fast prototyping and integration of parallel algorithms that are transparently scheduled and executed on the Google Cloud infrastructure. Compared to Amazon EC2, App Engine offers one order of magnitude lower resource acquisition latency and a cheaper pricing model for shorter jobs below one hour. We show experimental results that demonstrate good scalability of a Monte Carlo simulation algorithm. Although important speedup can be gained, it is diminished by two main obstacles: middleware overheads and resource quotas. http://doi.ieeecomputersociety.org/10.1109/MS.2011.131 http://doi.ieeecomputersociety.org/10.1109/MS.2011.129 Although COTS integration presents clear advantages in a wide variety of engineering fields, several problems and limitations may arise partly due to their heterogeneous nature. Few works have studied the integration of particular COTS to specific domains. This approach has been used on a project performed at the University of Vigo for reusable software development in the engineering domain. The underlying hypothesis was that it could be worth making easier COTS integration in order to facilitate the development of applications in a specific domain. The solution consists on a two-layers wrapping approach: COTS domain model is captured in the first layer, facilitating general-purpose functionality integration, as well as the development of a second layer of wrappers that provide better integration for more specific-domain functionality. Several real software projects using the proposed solution have been developed, and results yielded notable effort savings, showing that this approach could be interesting to reduce COTS integration efforts http://doi.ieeecomputersociety.org/10.1109/MS.2011.129 http://doi.ieeecomputersociety.org/10.1109/MS.2011.126 Estimates of effort and duration for a new software project often have to be adjusted to deal with an imposed target delivery date or a constraint on staffing. Estimating methods assume an effort/duration trade-off relationship based mostly on theory or expert judgment. This paper describes a process for analyzing actual project effort and duration data which is designed to explore the trade-off relationship. I assume a reference relationship of a simple power-curve with variable power ‘N’ and use this (a) as a means of comparing the trade-off relationships assumed by four well-known estimating methods, and (b) as the basis for a process to analyze actual project data. Results are presented of applying the process to 16 sub-sets of project data. These suggest, for example, that the value of ‘N’ differs between new development projects and enhancement projects. http://doi.ieeecomputersociety.org/10.1109/MS.2011.126 http://doi.ieeecomputersociety.org/10.1109/MS.2011.120 Codification and testing of business and rules in application programs has historically been a challenge in software engineering. The business rules approach has recently been adopted by many organizations to formalize and compartmentalize business rules as a separate component from application code. This paper investigates and presents the impacts of the business rules approach on testing activities in the software development life cycle (SDLC) at a large Fortune 500 corporation. Our findings suggest that the business rules approach has the potential to engage testing personnel early on in the process and to improve the efficiency and effectiveness of testing activities. We also discuss the process changes, challenges, and overall impact on SDLC associated with the business rules approach. http://doi.ieeecomputersociety.org/10.1109/MS.2011.120 http://doi.ieeecomputersociety.org/10.1109/MS.2011.119 The dilemma of migrating business software applications to the cloud is a dominant IT topic among consultants, software managers and executives. It's obvious that the fairly broad general interest that cloud computing concentrates is motivated by the quick, painless deployment and maintenance of applications that are now a burden of the provider. Yet one should be able to make a thorough analysis of his business IT problem to decide whether the features and cost of a cloud solution are appropriate and the risks are reasonable and remain under control. http://doi.ieeecomputersociety.org/10.1109/MS.2011.119 http://doi.ieeecomputersociety.org/10.1109/MS.2011.118 Agile software development involves self-managing teams that are empowered and responsible for meeting project goals in whatever way they deem suitable. Management are required to place more trust in such teams than under a more traditional development methodology. This paper highlights how the use of agile practices can enhance trust amongst agile team members. It also presents challenges that agile teams may now face as a result of using agile practices, which are based on the findings from three case studies of agile software development teams. http://doi.ieeecomputersociety.org/10.1109/MS.2011.118 http://doi.ieeecomputersociety.org/10.1109/MS.2011.116 Safety-critical software-dependent systems such as those found in the avionics, automotive, maritime, and energy domains often need to be certified by a licensing or regulatory body based on one or more safety standards. Safety standards do not specify the details of the evidence that needs to be collected for the certification of a particular system because these standards need to be generalizable and applicable to a wide variety of systems. Without an upfront agreement between the system supplier and the certifier about the details of the evidence that needs to be collected, there will invariably be important omissions in the evidence information provided by the supplier, which will need to be remedied after the fact and at significant costs. The contributions of this article are twofold: we present both a flexible approach and a publicly available supporting tool for assisting suppliers and certifiers in developing an agreement about the evidence necessary to demonstrate compliance to a safety standard. The approach is model-based; specifically, the safety standard of interest is expressed via an information model. The supporting tool, which is available online, takes this information model as input and assists system suppliers and the certifiers in reaching a documented and consistent agreement about the safety evidence that needs to be collected. http://doi.ieeecomputersociety.org/10.1109/MS.2011.116 http://doi.ieeecomputersociety.org/10.1109/MS.2011.92 Collaboration between industry and academia supports improvement and innovation in industry and helps to ensure industrial relevance in academic research. This paper presents an exploratory study of factors for successful collaboration between industry and academia. A survey was designed for data collection and was firstly conducted in Sweden then replicated in Australia. The context for the two studies is different thus forming a starting point for potential generalizations in the future. From the two studies we conclude that the industrial side of collaboration is the key element for successful collaboration, with key factors being “Buy in and support from company management” and “Champion at company”. Context-specific factors were also identified based on differences in the context between the two studies. These findings may help industry and academia to set up successful collaborative ventures. http://doi.ieeecomputersociety.org/10.1109/MS.2011.92 http://doi.ieeecomputersociety.org/10.1109/MS.2011.69 Employing a software platform is an approach to achieve a higher degree of software reuse since multiple software products can share the platform-provided services. However, the platform development usually involves stakeholders from different application domains. Their application situations vary widely and thus Non-Functional Requirements (NFRs) for the software platforms must address a wider range of needs than those for a single product. This paper describes lessons learned in developing NFRs for a large software platform, the challenging issues we faced, and the techniques we used to address those issues. Our techniques are pragmatic and helped with NFR reconciliation and management, and improved the quality of the NFR specifications. The improved quality of the NFR specifications has permitted automation of platform performance testing for the past two years. http://doi.ieeecomputersociety.org/10.1109/MS.2011.69 http://doi.ieeecomputersociety.org/10.1109/MS.2011.24 For projects that rely on empirical process control and deliver frequently working versions of software, developers and project managers regularly need to examine the status of their software quality. Our studies illustrate that simple goal-oriented changes or extensions to the existing data of their respecting defect tracking systems could provide valuable and prompt information to improve their software quality assessment and assurance. http://doi.ieeecomputersociety.org/10.1109/MS.2011.24