Most specifications (describing requirements, tests, constraints/hazards, and design approaches) for complex systems are hosted either in databases, in documents, or in wiki's. All approaches have their strengths and weaknesses. Database-based system tend to be expensive and proprietary, and conversion systems are needed to generate document "views" onto the data that can be read by stakeholders and persisted in a document control system. Database approaches provide great flexibility and enable linking between requirements and related artifacts like tests, design details, constraints, parent requirements, and hazards. Document-based systems are inexpensive and easy to set up, but lack a mechanism to link requirements to related artifacts. It can also be hard to maintain consistency in document formatting and content with multiple spec authors at work. Wikis provide easy access to project stakeholders, but don't lend themselves to structured and consistent information formatting. Linking wiki pages and table records can be manually intensive, and creating print-quality documents can be difficult.
LSpeX combines the power of a database-driven RMS system with the simplicity and low cost of document and wiki-driven systems. All requirements and related artifacts are persisted in XML, with one XML file created per requirement, test, hazard, or design specification. Each specification type has an associated schema to ensure compliance to organization-specific specification norms. XSL-FO is used to transform the XML-based specifications into a set of PDF documents. The key feature of the resulting PDF documents is a rich set of internal and external links. The links help management complexity by making it easy to navigate between related specifications. For example, links connect individual requirements to the test protocols that demonstrate their complete implementation, and to the constraints or hazards that motivate their existence. PDF redlines can be automatically generated for change management activities.
The entire system is customizable through a set of XSLT files, and is powered by Xalan and FOP (both open source). Plugins can be created to connect the system to wiki's and document control systems to make specification reading and controlled editing easy.
Ohloh computes statistics on FOSS projects by examining source code and commit history in source code management systems. This project has code locations but that location contains no recognizable source code for Ohloh to analyze.