Skip to Main Content
To understand the evolution of software, researchers have developed a plethora of tools to parse, model, and analyze the history of systems. Despite their usefulness, a common downside of such tools is that their use comes with many strings attached, such as installation, data formats, usability, etc. The result is that many tools are only used by their creators, which is detrimental to cross-fertilization of research ideas and collaborative analysis. In this paper we present the Churrasco framework, which supports software evolution modeling, visualization and analysis through a web interface. The user provides only the URL of the Subversion repository to be analyzed and, if available, of the corresponding bug tracking system. Churrasco processes the given data and automatically creates and stores an evolutionary model in a centralized database. This database, called Meta-base is connected to Churrasco through object-relational persistence. The persistency mechanism is meta-described in terms of the EMOF meta-meta- model and automatically generated based on any given evolutionary meta-model. In case the meta-model changes, the persistency mechanism is automatically updated. After providing a detailed description of Churrasco, we provide evidence, by means of an example scenario, that it allows for collaborative software evolution analysis, based on visualizations available on our analysis web portal.
Date of Conference: 1-4 April 2008