Contribute Media
A thank you to everyone who makes this possible: Read More

Bug forecasting by visualizing code evolution


PyCon Finland 2015 Tommi Penttinen - Bug forecasting by visualizing code evolution

I would like to introduce and expand the ideas presented by Adam Tornhil in his work (Your Code as a Crime Scene – applying forensics to mining software repositories - where do the bugs propbably lurch). Tornhill's main ideas are combining cyclomatic code complexity metric with repository history -calculated "code/knowledge ownership". Adding test coverage data, replicated production-system web traffic and their call graphs adds whole new dimensions to the data available from the repository.

About the author: I'm a(n almost-university-dropout) mathemathics major, software engineer. I've worked professionally with Python (and Django) since 2009, having also had the privilege to work only with DVCS. I've worked in different organizations ranging from freelancing and small companies to multinational corporations. I enjoy automation, debuggers and meta-programming. Most of my non-programming spare time goes into a cappella singing, improvisational theatre and finding ways to express my acrobatic tendencies.


Improve this page