Agile Analytics: Growing Your Product by Absorbing Your History

The old cliche says that you have to understand where you've come from in order to know where you're going. In an agile coding environment the best indicator of your past is the code that you've generated. Therefore the best path to the future of your product is analysing that code.

So what can your code tell you about the health of your project? What other information do you have at your disposal that you might not be taking advantage of yet? How can analysing this information help your project in the long run? Armed with the right tools, and enough data, you can learn an enormous amount about the health of your project ... and how to improve it.

This talk will introduce existing tools that are available for analyzing your source code, as well as some new tools that will help analyze the softer areas of your project. Effective use of these tools provides invaluable information about the quality of your source code, your team, and your product. The tools will shine a light on what has and hasn't worked, the areas that require the focus of your efforts, and pitfalls you should avoid.

Once you learn to analyze the data you already have about your project's history, you'll be able to make the most informed decisions about the direction your project should take.

--

This talk focuses on my work analysing source code in order to become more effective at spotting pain points, making suggestions about how to improve things, and understanding exactly where effort is being spent. There are a variety of tools in Ruby already that make this work possible: flog, flay, reek, roodi, toweilie, dust, castigate, and churn are among some of the more popular ones.

My work has revolved around using the output from these tools, combined with output from custom tools to identify pain points. The point of this talk is to share what I've learned.

I have also developed tools that map project history, and the inevitable changes, to measuring quality, efficiency, and predictability.

I will be demonstrating the new tools that I haven't shown off yet. They are impressive. There will be lots of pretty pictures and excellent explanations. The audience will be wowed. Some may cry a little.