In my master project, two colleagues and I got the chance to develop a powerful web-application, using Asp.Net, Azure and TFS, to help the developer answer his daily questions using the information from multiple software repositories.
In a typical workday, developers have to answer several questions, such as “Who is working on what?” or “Which is the most popular class?”. Today’s tool support is limited, as we only found tools, where the usage was tedious and time consuming, where the user has to learn a new query language or where the license costs are very high. Besides this issue, there are enormous amounts of information a developer has to manage. A solution to increase the efficiency of answering everyday questions is needed to support the developer in keeping track with the growing complexity of the information. Fritz and Murphy developed a concept, information fragments, which compares and merges different data sets from different repositories using an id and text matching algorithm between the connections of these repositories. These nodes and edges are aggregated to a graph, the composed fragment, and presented to the user. We base our work on this approach and developed an extensible web application prototype that lets the user easily manipulate and filter the composed data by using an easy to understand abstraction of the model. Additionally, the data is represented using five different visualizations, each meaningful for different situations. The usefulness of the approach and its implementation was evaluated, using four usage scenarios. Finally, interesting directions for future work have been presented and discussed.
You may read our report here.