Open Source Development, Communities and Quality

This paper presents the Software Project Governance Framework (SPGF) for characterizing management of software projects, based on mechanisms used for communication and collaboration, the organizational structure of projects, and testing and quality assurance procedures. The framework was developed and validated from interviews and surveys with leaders of more than 70 commercial and community-based software projects, including both closed and open source projects.

Open source has recently become a practical and advocated fashion to develop, integrate, and license software. As a consequence, open source communities that commonly perform the development work are becoming important in the practice of software engineering. A community that is lively can often produce high-quality systems that continuously grow in terms of features, whereas communities that do not gain interest will inevitably perish. Despite their newly established central role, creation, organization, and management of such communities have not yet been widely studied from the viewpoint of software engineering practices. In this paper, we discuss experiences gained in the scope of Laika, an open source project established to develop an integrated software development environment for developing applications that run in Linux based mobile devices.

The Common Criteria standard provides an infrastructure for evaluating security functions of IT products and for certifying that security policies claimed by product suppliers are correctly enforced by the security functions themselves. Certifying Open Source software (OSS) can pave the way to OSS adoption in a number of security-conscious application environments. Recent experiences in certifying Linux distributions has pointed out the problem of finding a mapping between descriptions of OSS security functions and existingtest suites developed independently, such as the Linux Test Project. In this paper, we describe a mechanism, based on matching techniques, which semiautomatically associates security functions to existing test suite such as the ones developed by Open Source communities.

The paper presents a survey of mature Free/Libre Open Source Software communities. The main focus of the survey is the collection of data related to the practices of these communities related to trust elements in their products. The survey is carried out using a structured questionnaire about thoughts and practices followed by Free/Libre Open Source Software communities. The survey focuses on the analysis of the development processes adopted by such communities. The results of the survey confirms basic ideas related to Free/Libre Open Source Software and explains in more detail specific issues related to trust and trustworthiness of the Free/Libre Open Source Software development process.

Process Monitoring represents a big challenge for organizations that aim to manage software projects adopting different development paradigms. In fact, across-process enterprise-level measurement campaigns can be difficult to enact since process attributes to retrieve are semantically diverse and may be difficult to integrate. In this paper, we present PMLite (Process Monitoring Lite) an open source solution to this problem. PMLite is based on an open metamodel and paves the way to the definition of ad-hoc open monitoring frameworks.

Open Source Software Development appears to depart radically from conventional notions of software engineering. In particular, requirements for Open Source projects seem to be asserted rather than elicited.

This paper examines features of the latest major release of the Firefox web browser in attempt to understand how prevalent this phenomenon is. Using archives of mailing lists and issue tracking databases, these features were traced from first mention to release, to determine the process by which requirements are proposed, adopted, and implemented in Firefox. The results confirm the importance of user participation as developers of open source products.

Coordination is one of the keys for the success of open source software (OSS) communities because geographically distributed members need to collaborate on their work using communication tools (e.g., mailing lists, bulletin board systems, bug tracking systems, and so on). In this paper, we investigated the informal social structure among developers and users by analyzing two mailing lists of developers and users in the Apache community based on betweenness centrality, one centrality measure proposed by Freeman. From the analysis results, we found that (1) participants with high betweenness coordinated activities between developers and users and (2) some participants have been functioning as coordinators in the community for a long time.

What happens when an organization form that has emerged in one context is brought into a different context? In this paper, a longitudinal field study approach is used to explore how Hewlett-Packard (HP) molded open source software development (OSSD) into a proprietary software development approach called “Progressive Open Source” (POS). With the help of actornetwork theory, we understand this as a process of translation and find that some central characteristics of OSSD where lost in the translation into POS while other characteristics were gained.

The challenges of commons based peer production are usually associated with the development of complex software projects such as Linux and Apache. But the case of open content production should not be treated as a trivial one. For instance, while the task of maintaining a collection of encyclopedic articles might seem negligible compared to the one of keeping together a software system with its many modules and interdependencies, it still poses quite demanding problems. In this paper, we describe the methods and practices adopted by Simple Wikipedia to keep its articles easy to read. Based on measurements of article readability and similarity, we conclude that while the mechanisms adopted by the community had some effect, in the long run more efforts and new practices might be necessary in order to maintain an acceptable level of readability in the Simple Wikipedia collection.

How do hackers’ practices produce hackers’ identities? In this paper we argue that the association between science fiction and software programs is rooted in hackers’ practices, defining how hackers’ knowledge’ emerge. The mediation is the one of the Heinlein verb “to grok”, part of the Jargon file and of the name of a code browser, OpenGrok, the technology mediating the relationship between OpenSolaris developers and the code base. Starting with a description of the peculiarity of the verb “to grok”, and its connection with a non-Cartesian view of knowledge, we discuss how the history of OpenGrok and its use by developers make this knowledge part of hackers’ practices and identities, as someone involved in a true, deep understanding of software.

This paper extends prior investigation into the social dynamics of free and open source (FLOSS) teams by examining the methodological questions arising from research using social network analysis on open source projects. We evaluate the validity of data sampling by examining dynamics of communication centralization, which vary across multiple communication channels. We also introduce a method for intensity-based smoothing in dynamic social network analysis.

The Free/Open Source Software (F/OSS) research community is growing across and within multiple disciplines. This community faces a new and unusual situation. The traditional difficulties of gathering enough empirical data have been replaced by issues of dealing with enormous amounts of freely available public data from many disparate sources (online discussion forums, source code directories, bug reports, OSS Web portals, etc.). Consequently, these data are being discovered, gathered, analyzed, and used to support multidisciplinary research. However at present, no means exist for assembling these data under common access points and frameworks for comparative, longitudinal, and collaborative research across disciplines. Gathering and maintaining large F/OSS data collections reliably and making them usable present several research challenges. For example, current projects usually rely on direct access to, and mining of raw data from groups that generate it, and both of these methods require unique effort for each new corpus, or even for updating existing corpora. In this paper, we identify several needs and critical factors in F/OSS empirical research across disciplines, and suggest recommendations for design of a global research infrastructure for multi-disciplinary research into F/OSS development.

In today’s business and software arena, more and more companies are adopting open source software. An example of this rising phenomenon is to base software products on highly reusable open source components. In this scenario, the evolution of the software product is coupled with the evolution of the open source component. A common assumption is that component updates are immediately and regularly propagated to the project. This paper investigates this assumption empirically by studying update propagation practices in two popular open source libraries, zlib and FFmpeg. For each library, we analyze various repository sources with information such as bug reports, revision history, and source code. The results of the case studies suggest that update propagation is subject to several technical and non-technical factors including the way the open source library is used, the extent to which updates are documented, and the degree of community involvement. Based on these findings, we propose a set of recommendations that would allow better follow-up of updates and smoother update propagation.

Because of the sheer volume of information available in FLOSS repositories, simple analysis have to face the problems of filtering the relevant information. Hence, it is essential to apply methodologies that highlight that information for a given aspect of the project. In this paper, some techniques from the social sciences have been used on data from version control systems to extract information about the development process of FLOSS projects with the aim of highlighting several processes that occur in FLOSS projects and that are difficult to obtain by other means. In particular, the collaboration between the FLOSS community and a company has been studied by selecting two projects as case studies. The results highlight aspects such as efficiency in the development process, release management and leadership turnover.

Monitoring the performance of processes is often considered critical in classic engineering fields. However, in the area of software engineering (and especially in the Open Source context) it seems that the literature has not yet taken into consideration the problem of identifying the process characteristics and performance of debugging. The aim of this paper is the identification of the performance characteristics of the bug fixing process of Open Source applications, focusing on continuity and efficiency indicators. The importance of such indicators is even more relevant today, since Open Source software is now adopted also in many business contexts. We have analyzed the debugging process of 9 active and popular Open Source projects, collecting a dataset comprising more than 65,000 closed bugs. Results have highlighted four types of bug fixing processes that can be distinguished by considering temporal continuity and efficiency dimensions.

Software development is undergoing a major change away from a fully closed software process towards a process that incorporates open source software in products and services. Just how significant is that change? To answer this question we need to look at the overall growth of open source as well as its growth rate. In this paper, we quantitatively analyze the growth of more than 5000 active and popular open source software projects. We show that the total amount of source code as well as the total number of open source projects is growing at an exponential rate. Previous research showed linear and quadratic growth in lines of source code of individual open source projects. Our work shows that open source is expanding into new domains and applications at an exponential rate.

Is Open Source Software (OSS) undergoing a transformation to a more commercially viable form? We have performed a survey to investigate the adoption of OSS in the Norwegian software industry. The survey was based on an extensive screening of software companies, with more than 700 responses. The survey results support the transformation predicted by Fitzgerald [4]. Close to 50% of the software industry integrate OSS components into vertical solutions serving all major business sectors. In addition, more than 30% of the 95 respondents in our survey have more than 40% of their income from OSS related services or software. The extensive adoption of OSS in the software industry may be a precursor of the OSS adoption in other business sectors.

Most studies of FOSS organizational migration projects focus solely on technological and economical aspects, neglecting the importance of organizational discourse structures for migration decisions as well as success. In looking at the case of the municipality of Munich this paper uses structuration theory in combination with discourse analysis to explain why and how in this case actors were able to overcome strong barriers to migration in the field of desktop software.

Software quality evaluation has always been an important part of software business. The quality evaluation process is usually based on hierarchical quality models that measure various aspects of software quality and deduce a characterization of the product quality being evaluated. The particular nature of open source software has rendered existing models inappropriate for detailed quality evaluations. In this paper, we present a hierarchical quality model that evaluates source code and community processes, based on automatic calculation of metric values and their correlation to a set of predefined quality profiles.

1

Recently, computational agents received significant attention in computer science research community. In fact, intelligent agents is a powerful artificial intelligence technology showing considerable promise as a new paradigm for mainstream software development and able to offer new ways of abstraction, decomposition, and organization that fit well with our natural view of the world. However, despite their promise, intelligent agents are still scarce in the market place. A key reason for this is that developing intelligent agent software requires significant training and skill. Artificial Intelligence methodologies and computer networking tools represent the necessary basic knowledge to design and implement advanced agents oriented systems. This papers introduces an integrated development environment supporting the agents developers to design fuzzy-based agents in a simple and fast way. Proposed framework has been realized by integration of theoretical methodologies as fuzzy logic and labeled tree, together with OSS tools as JaxMe2.

Programmers often search for Open Source code to use in their projects. To understand how and why programmers search for source code, we onducted a web-based survey and collected data from 69 respondents, including 58 specific examples of searches. Analyzing these anecdotes, we found that they could be categorized along two orthogonal dimensions: motivation (reuse vs. reference example) and size of search target. The targets of these searches could range in size from a block (a few lines of code) to a subsystem (e.g. library or API), to an entire system. Within these six combinations of motivations and target sizes, nine repeating motifs, or archetypes, were created to characterize Internet-scale source code searching. Tools used for searching and the criteria for selecting a component are also discussed. We conclude with guidance on how these archetypes can inform better evaluation of Internet-scale code search engines, as well as the design of new features for these tools.

Firefox, a browser targeted at mainstream users, has been one of the big successes of open source development in recent years. That Firefox succeeded where earlier attempts failed is undoubtedly due to the particular choices that were made in the process of development. In this paper, we look at this process in more detail. Mining bug reports and feature requests related to Firefox in Mozilla’s Bugzilla bug tracker system, we find that the attention developers devoted to reports and requests was influenced by several factors. Most importantly, other things being equal, reports and requests from outsiders increasingly tend to be ignored. While such behavior may have helped to shield Firefox from the “alpha-geek power user” in the early stages of development, it also makes it difficult for “mom and dad” to let their voice be heard even after they have adopted Firefox.

Commercial software firms are increasingly using and contributing to open source software. Thus, they need to understand and work with open source software development processes. This paper investigates whether the practice of continuous integration of agile software development methods has had an impact on open source software projects. Using fine-granular data from more than 5000 active open source software projects we analyze the size of code contributions over a project’s life-span. Code contribution size has stayed flat. We interpret this to mean that open source software development has not changed its code integration practices. In particular, within the limits of this study, we claim that the practice of continuous integration has not yet significantly influenced the behavior of open source software developers.

This paper presents an experiment performed on three large open source applications. The applications were instrumented automatically with a total of 10,494 instrumentation points. The purpose of the instrumentation was to collect and store data during the execution of each application that later could be analyzed off-line. Data analysis, on the collected data, allowed for the creation of test cases (test data, test fixtures and test evaluators) in addition to finding object message patterns for object-oriented software.

In this paper, we illustrate our methodology for academic teaching, based on cooperative learning paradigm, which also relies on cutting edge e-learning techniques. We use Web 2.0 resources to fulfill requirements for an interactive-constructivistic “learning space”, where blended-teaching paradigm may be proficiently applied. We extend an existing Free/Open Source Software Learning Management System, to create a cooperative and community-based learning space adherent to our proposal.

Libre (free, open source) software forges (sites hosting the development infrastructure for a collection of projects) have been stable in architecture, services and concept since they become popular during the late 1990s. During this time several problems that cannot be solved without dramatic design changes have become evident. To overcome them, we propose a new concept, the “networked forge”, focused on addressing the core characteristics of libre software development and the needs of developers. The key of this proposal is to re-engineer forges as a net of distributed components which can be composed and configured according to the needs of users, using a combination of web 2.0, semantic web and mashup technologies. This approach is flexible enough to accommodate different development processes, while at the same time interoperates with current facilities.

In this paper we describe the strategy under adoption in Telecom Italia (TI) Technology Department toward open source software. This stems from trying to create synergy among big Telco Player to increase knowledge and influence over strategic communities to the evaluation of the creation of new communities over internally developed applications. In particular here the approach and the expectations in starting the community on WADE (Workflow and Agent Development Environment) is described. This is a platform used to develop mission critical applications and is the main evolution of JADE a popular Open Source framework for the development of interoperable intelligent multi-agent systems. It adds to JADE the support for the execution of tasks defined according to the workflow metaphor as well as a number of mechanisms that help managing the complexity of the distribution both in terms of administration and fault tolerance. The idea is to use WADE as a mean to gather critical information on the opportunity of approaching OS as a strategic mean toward the development of always more important application in Operating Support System for TI, possibly also involving other great Telco Players For this reason great care is being paid in setting up the Community environment and in deciding which metrics are to be extracted from it, since the result will be the input for a strategic decision in TI.

The use of ALT (Advanced Learning Technologies) creates dynamic sharing and exchanging between open source communities that diffuse e-learning systems. In our opinion, the designer practices define new perspectives on e-learning design which are not structured and highlighted enough. This article shows the capabilities of a generic framework to analyse the design practices on a open source project and to explicit these practices. We describe how the semantics for architectural specifications proposed by RM-ODP (Reference Model-Open Distributed Process) framework were applied on an e-learning system project to analyse the principles of invariants, structural and functional.

Innovation in the software sector is a widely debated issue. Which are the most important dimensions to assess innovation in this field? Can we measure innovative processes carried out by software companies and what kind of innovation do they develop? Are FOSS solutions more innovative than proprietary ones? These are the research questions we endeavor to answer in this paper providing some empirical evidence, obtained using an original methodology of collecting experts’ evaluations on the innovation level of 134 solutions provided by a group of Italian Small and Medium Enterprises.

The principles behind the agile development methods and common practise within the Open Source community are vastly different. In recent years there has been a rise of interest in these, in order to detect and inform on areas of compatible shared practises. This paper argues that it is possible to quantify the level of agility displayed by Open Source projects. An indicator of agility, the Mean Developer Engagement (MDE) metric is introduced and tested through the analysis of public project data. Projects sampled from two repositories (KDE and SourceForge) are studied and a hypothesis is formulated: projects from the two samples display a similar level of MDE.

This paper provides two main contributions: first, the MDE metric is shown to vary significantly between the KDE and SourceForge projects. Second, by combining MDE with a project’s lifespan, it is also shown that SourceForge projects have insufficient uptake of new developers resulting in more active, shorter, initial activity, and in a quicker “burning out” of the projects.

Open source projects are typical examples of successful distributed software development projects. Understanding how coordination in these projects takes place can provide important lessons to Software Engineering researchers and practitioners. This understanding has been achieved using different research methods, including, surveys, case studies and social network analysis. However, to conduct these studies each researcher needs to build his own infra-structure from the scratch, a time consuming and error-prone task. This paper aims to alleviate this problem by describing an environment, the OSSNetwork, which allows the automatic data collection of open source repositories. Data collected by the OSSNetwork is aimed to support the construction, visualization, and analysis of social networks. This environment is extensible, therefore facilitating empirical studies of open source projects.

Knowledge sharing between software project participants simplifies a range of decision-making processes and helps improve the way software is being developed, distributed, and supported. However, research in this area has traditionally been very difficult because the source of knowledge, the code, has been a guarded secret and software developers and users inhabit different worlds. F/OSS projects have changed the way we perceive and understand knowledge sharing in distributed software development. This short paper presents our current understanding, and what needs to be done in terms of empirical research in knowledge sharing in F/OSS projects.

In the company open source software (OSS) development context, usability is becoming an important issue due to a growing user population, who is only interested in usable applications, not in their development. Companies try to gain a competitive advantage of OSS by utilizing available components, but the openness is difficult to achieve in the business world with licenses, patents and intellectual property rights. This paper analyses usability and user-interface (UI) development in the company OSS context through an interpretive case study in a software development unit of a large, global corporation. Our initial findings suggest that there is a need for human computer interaction specialists in the OSS context. It is revealed that with the software based on OSS, more time can be spent on usability and UI design. In the company’s viewpoint, there are still many issues involved in deciding what parts of the product will be open source. Especially UI code may remain closed due to competitive advantage and patents.

Open Source Software (OSS) is an increasingly hot topic in the business domain. One of the key benefits mentioned is the unlimited access to the source code, which enables large communities to continuously improve a software application and prevents vendor lock-in. How attractive these benefits may be, the market for OSS however remains limited. In the Netherlands research consisting of 7 interviews and a survey among 206 Open Source Software Service providers (with a 34% response rate) was done to determine whether service providers wanted to cooperate in an Association that will set quality levels and guarantees to its members and their customers.

In this paper, we propose to arrive at an assessment and evaluation of open source projects based on an analysis of their growth rates in several aspects. These include code base, developer number, bug reports and downloads. Based on this analysis and assessment, a well-known portfolio planning method, the BCG matrix, is employed for arriving at a very broad classification of open source projects. While this approach naturally results in a loss of detailed information, a top-level categorization is in some domains necessary and of interest.

Open Source Software development is seen as a panacea by many companies. The promise of community-style development, innovation and cost savings drive the wider adoption of OSS in companies. However, it is still difficult to institutionalize the open and agile culture of sharing innovation especially into larger departmentalized organizations. The aim of this research paper is to investigate the characteristics of one successful OSS development implementation approach limited inside a company (Inner source). Based on our data, we argue that there are possibilities for employing OSS as a new kind of development process within a company and leveraging thus the innovation potential inside the company.

To facilitate the adoption of open-source software (OSS) in industry, it is important to provide potential users (i.e., those who could decide to adopt OSS) with the means for evaluating the trustworthiness of OS products. This paper presents part of the work done in the QualiPSo project for this purpose. A set of factors that are believed to affect the perception of trustworthiness are introduced. In order to test the feasibility of deriving a correct, complete and reliable evaluation of trustworthiness on the basis of these factors, a set of well-known OSS projects have been chosen. Then, the possibility to assess the proposed factors on each project was verified: not all the factors appear to be observable or measurable. The paper reports what information is available to support the evaluation and what is not. This knowledge is considered to be useful to users, who are warned that there are still dark areas in the characterization of OSS products, and to developers, who should provide more data and characteristics on their products in order to support their adoption.

This paper focuses on the common evaluation framework which was developed by the BioSecure Network of Excellence during the European FP6 project BioSecure (Biometrics for Secure authentication). This framework, which is composed of open-source reference systems, publicly available databases, assessment protocols and benchmarking results, introduces a new experimental methodology for conducting, reporting and comparing experiments for biometric systems, participating to standardisation efforts. Its use will permit to make a link between different published works. It provides also the necessary tools to assure the reproducibility of the benchmarking biometric experiments. This framework can be considered as a re-liable and innovative process to evaluate the progress of research in the field of bio-metrics.

This paper introduces eResearch workflow tools as a model for the research community studying free and open source software and its development. The paper first introduces eResearch as increasingly practiced in fields such as astrophysics and biology, then contrasts the practice of research on free and open source software. After outlining suitable research data sets the paper introduces a class of tools known as scientific workflow frameworks, focusing on one—Taverna—and introducing its features. To further explain the tool a complete workflow used for original research on FLOSS is described. Finally the paper considers the trade-offs inherent in these tools.

Open Source Software (OSS) is a multi-faceted phenomenon which has become an issue of strategic importance for many commercial organisations. Stemming from an ideological issue, with emphasis on freedom and community values, we have recently seen a broader interest in the Open Source phenomenon amongst practitioners in many companies. A number of SMEs and large companies are currently exploring the potential of Open Source, and for some it has become core to their business and development activities.

Open Source Software (OSS) for e-Science should make reference to the paradigm of a distributed surrounding over a multi system mix of Web Services and Grid technologies, allowing data exchanging through services, according to standards in the area of the Grid and of Service Oriented Computing (SOC). In fact, biologists, medical doctors, and scientists are often involved in time consuming experiments and are aware of the degree of difficulty in validating or rejecting a given hypothesis by lab experiments. The benefits of OSS for e-Science consider that as many operating nodes as possible can work cooperatively sharing data, resources, and software, thus avoiding the bottleneck of licenses for distributed use of tools needed to perform

cooperative scientific experiments

. In particular, this chapter presents an architecture based on nodes equipped with a Grid and with Web Services in order to access OSS, showing how scientific experiments can be enacted through the use of a cooperation among OSS sites. Such a choice, besides reducing the cost of the experiments, would support distributed introduction of OSS among other actors of the dynamical networks, thus supporting the awareness about OSS and their diffusion. An OSS environment for cooperative scientific experiments (

e-experiments

) can effectively support the distributed execution of different classes of experiments, from visualization to model identification through clustering and rules generation, in various application fields, such as bioinformatics, neuro-informatics, tele-monitoring, or drug discovery. By applying Web Services and Grid computing, an experiment or a simulation can be executed in a cooperative way on various computation nodes of a network equipped with OSS, allowing data exchange among researchers. Our environment formalizes experiments as

cooperative services

on various computational nodes of a grid network. Basic elements are models, languages, and support tools creating a virtual network of organizational responsibility of the global experiments, according to rules under which each node can execute local services to be accessed by other nodes in order to achieve the whole experiment’s results.