GSoC/GCI Archive
Google Summer of Code 2009


Web Page:

Mailing List:

The mission of the Hackystat project is to provide a framework for collection, analysis, visualization, interpretation, annotation, and dissemination of software development process and product data. Work on Hackystat began in 2001 as a research activity in the Collaborative Software Developmnt Laboratory ( in the Department of Information and Computer Sciences ( at the University of Hawaii ( From 2001 to 2006, the project grew a substantial code base (approximately 350,000 LOC), user community (over 800 users of the "public" Version 7 Hackystat server, with an undetermined additional number of users on private servers), and developers (dozens of contributors to the code base from approximately 20 academic and industry sites). A significant number of publications occurred during this time (a partial list of which is available at In addition, there was one commercial, non-open source spin-off based upon a subset of this project called Sixth Sense Analytics ( A summary of Project Hackystat's release history is available at


  • Issue analysis with software ICU Software ICU is already a great tool for monitoring software project's health. However, one significant vital sign is missing, that is the project's "issue". Issue tracking system is now one of the most important and popular tools for task management. And it usually consists of formatted information that it is easy to extract information automatically for sensor data. The use of issue tracking system reflects the task management habit of a software project, which is an essential part to its success. The project will be introducing issue tracking as a vital sign in software ICU. The components need to be done includes DPD analysis of "issue", several telemetry analysis of issue that showing different aspect of it, and the correct coloring mechanism to demonstrate issue in a meaningful and concise way.
  • Master Dissertation- A middelware realising autonomous services for Hackystat (1) Service management. We now have about six or so services that run as a standard part of a Hackystat system, with more on the way. Unfortunately, all of these services need to be configured and run individually. Also, if one goes down, we have no way to knowing. A huge benefit would be to create a framework that allows easy downloading, deployment, and monitoring of these services.
  • Social Network Impact on Programming Practices Hackystat provides numerous tools for allowing users to analyze patterns in their programming behaviors; I will build upon this functionality by adding features that will allow users to identify trends in the software development community at large. A relational graph constructed using data from social networking sites in addition to Hackystat data will be mined to develop predictive models that users can query to gain a better understanding of how social networks impact coding practices.
  • Towards linked sensor data Currently Hackystat provides an API to access its collected sensor data from external resources, however publishing sensor data as RDF Linked Data on the Web can significantly increase benefits for both Hackystat end users and third-party applications. I plan to transform current sensor data in new “linked sensor data” represented as RDF descriptions and to start working from the 20th of April.