Conclusion: A fragmented World

are combining forces. As such, F/OSSprocess improvement is directly bound to the ability of projects to interoperate. Thus this issue also need to be considered in theF/OSScontext.

Interoperability is considered to be achieved if the interaction can, at least, take place at three levels:

data, application and business enterprise. First, technical Interoperability is needed to exchange data between applications or projects and can be achieved through the use of common communication means and languages such as XML or RDF and open standards, semantic Interoperability is needed to make applications understand exchanged data, this can be achieved through specific ontologies, finally orga-nizational Interoperability deals with the issue of connecting and making interoperable organizations having made their own business, technical and ontological choices. While the technical interoperability issue is currently mostly well known, interoperability research work is mostly related to the seek for semantic interoperability, but organizational interoperability is simply left apart.

The Flink group [53] has defined a Linux ontology and aims at demonstrating the benefits that result from formalizing knowledge about the Linux operating system. In [73], the authors discuss the application of ontology-based knowledge engineering to Linux. Various possible applications such as package management or information search are discussed which would all benefit from a comprehensive ontology of the domain. The use of an ontology to describe Linux is similar to our approach as it provides a common ground of understanding.

The Friend of a Friend (FOAF) [58] project is creating a Web of machine-readable pages describing people, the links between them and the things they create and do. It is a community driven effort to define an RDF vocabulary for expressing metadata about people, and their interests, relationships and activities.

FOAF is tackling head-on the wider Semantic Web goal of creating a machine processable web of data.

It facilitates the creation of the Semantic Web equivalent of the archetypal personal homepage: My name is Leigh, this is a picture of me, I’m interested in XML, and here are some links to my friends.

And just like the HTML version, FOAF documents can be linked together to form a web of data, with well-defined semantics.

DOAP [31] is a project to create an XML/RDF vocabulary to describe open source projects. This project aims at providing an internationalizable description of a software project and its associated re-sources, including participants and Web resources as RDF schemas. The project provides basic tools to enable the easy creation and consumption of such descriptions in all the popular programming languages and ensures interoperability with other popular Web metadata projects (RSS, FOAF, Dublin Core). Fi-nally, DOAP vocabulary is extensible for specialist purposes. Note that DOAP does not aims at handling software releases, nor at planning data internal to the project such as task assignments or milestones.

QualiPSo [153] is a recently started European IP project working on open source process improve-ment. It’s goal is to investigate and implement development processes through an open forge, including business models, methods and tools to foster the wide adoption of Open Source Software by European organizations from ITC players to end users. It focuses on aspects such as license management, doc-umentation and information management, and interoperability. Unlike other projects focusing on the technical or semantic aspect of interoperability, QualiPSo aims at handling all the aspects of interoper-ability mentioned above. It will provide a model for information contained inF/OSSincluding elements such as project, participants, tasks and planning, requirements, bugs, documents, new functionality pro-posals, source code, project versions, mail, forums, meetings, source code management.

3.7 Conclusion: A fragmented World

As seen in this chapter, organization, process management, interoperability and information display are hot topics. As they all contribute to process improvement, they have thus to be considered in theF/OSS

context. Nevertheless, none of existing approaches and tools provides consistent support for projects to enable F/OSSprocess streamlining. None considers the F/OSSProcess as a whole highlighting how activities are interwoven. Further, the community management issue is globally tackled poorly.

44 CHAPTER 3. EXISTING APPROACHES AND SOLUTIONS Fragments of solutions exist to tackle specific issues. However, these attempts to handle F/OSS

subprocesses are not integrated in a global view, and some, such as project metrics management, still need to be tackled. In fact, most issues involving interaction between F/OSS activities, or involving information exchange are unaddressed.

From the tool perspective, the same issue can be highlighted. A large number solutions aim at tackling issues related to specific activities of theF/OSSprocess but locally without being integrated in a large picture. Let cite, for instance, defect management tools (Bugzilla [15]), testing tools (Bugzilla test runner [16], Fitnesse [52], Salome [163]) and other tools (SourceForge) [170], GForge [72], Alioth [2], Libresource [105]) providing an integrated solutions for managing F/OSS projects. The dashboards provided by some projects [35,68] are a good symptom of this issue as they focus on specific information hardly usable.

We have seen thatF/OSStools for process and workflow management exist, however these tools are often designed for the enterprise environment, and are not considering the specific stacks and require-ments of theF/OSSenvironment we explored in Section2.4and Section2.3. Further these tools imply a machinery which while being mandatory in the enterprise context and very useful in both enterprise and

F/OSScontext, might be difficult to implement in an open environment.

In such a context we can argue that the issue ofF/OSSprocess improvement is left open. A transver-sal approach to the process management issue is required. It has to offer means to handle project as a whole considering involved activities, their content, community, and allowing reasoning aboutF/OSS

processes, tasks, metrics, etc. in a distributed context in order to enableF/OSSprocess streamlining.

Part II

Model

45

Chapter 4

Model Requirements

This chapter defines the requirements of a model for describing and using Free and Open Source re-sources and managing the underlying process. The purpose of this model is to provide the means nec-essary to design Information Systems able to streamline theF/OSSprocess. These requirements can be split in two groups. Information management requirements are introduced in Section 4.1and Process Management requirements are presented in Section4.2. The last part of this Chapter, Section4.3, lists selected elements of solution.