How to Identify Risk in a Large Code Base

Identifying risk in software and code is a relatively young science. In 25 years, we’ve moved from using naïve metrics (like bugs-per-line-of-code) to using nuanced models like those applied by automated testing practitioners at Etsy, Google, Facebook and other successful continuous delivery organizations. In these organizations, many people are responsible for understanding and identifying risk in the code base, which is often large and ever-changing in these large applications.

In this Masterclass, Noah demonstrates different techniques and tools he regularly uses to identify and examine risks in large codebases. 

To view the example data and code for the tools mentioned in this talk, you can go to the GitHub links provided and read through the code.

To use the code for the tools mentioned in the talk, you may need to either download the tool or fork the GitHub repo. This will require your own GitHub account (free for personal use) and using the GitHub tutorial on forking repos.

Jupityr/iPython notebook and example data can be found at this GitHub link: https://github.com/textarcana/code-risk/blob/master/code-risk-notebook.ipynb

This notebook allows you to see the results of using each tool described in the talk, so you know what you can expect the outcome to look like when used on your own code base.

Cloc code-repository visualisation tool: http://cloc.sourceforge.net

Scripts Created by Noah for wordcounts, developer contribution, popular files: https://github.com/textarcana/code-risk/tree/master/bin