In this talk 1 will present my experiences in transitioning and validating an automated software architecture measurement system in two large multinational corporations. I will describe the measures that we employed and the tool chains that we constructed to automatically calculate these measures. I will also describe how we got the development teams to accept and apply these measures through pilot studies, surveys, and constantly adjusting the measures based on feedback and correlations with productivity measures. This experience shows that it is critical to guide the development teams to focus on the underlying problems behind each measure, rather than on the scores themselves. It is also critical to both quantify architecture debt and prove to development teams and management alike that these measures matter, and that we can calculate the return on investment of paying down the debt.
展开▼