Technical Debt Is Not Free: an #InflectraCON2013 Live Blog

 

Chad Green

Director of Architecture, Glennis Solutions

When you hear the term "Technical Debt", what does that mean to you? Often the term "a quick and dirty fix" is used with the idea that it will be taken care of later. In short, anything you have to revisit later because of the limitations of today is specifically technical debt. 

The fact is, many of us have probably participated in the process of developing technical debt, whether we intended it or not. Even mature teams find themselves in technical debt, sometimes by active means, and sometimes by inaction. I remember working with an organization that had a great automation framework and it was very robust, with a lot of code libraries to support it. It was a great system, as long as the developers that created it were there to maintain and support it. However, we came to a point where our developers that did support this were no longer there and the team members that were there did not have the level of expertise necessary to maintain it. What was once a vital linchpin of our efforts became outdated and in some ways dangerous to do any work on. We went from having a reliable system to a need to modernize. In short, we woke up with a technical debt through no intentional effort on our part but we had to address it. We ultimately did but it took time and effort and a lot of iteration. We learned from that experience that we needed to make sure that what we developed had a broad base of support so that any of us could work on and maintain it.

Going into debt is not a crime or even a bad situation by itself. It can certainly become a bad situation if it's not addressed or worse, ignored. In many ways, we have to be more careful and make sure we are doing things in ways that are maintainable and understood. I went through this recently with some changes I proposed to a system that had a different way of handling API data (this came from suggestions of one of our devs). When I submitted it, the comments back were, "This is interesting and a way that we hadn't considered. We're not saying "no" but we may want to make sure we understand why we'd want to do it this way. Can we revisit this in the next sprint?" That's a perfectly reasonable request. Let's understand what making this change might be and how it might modify our testing methodology.

Technical debt sounds scary but there are ways to limit it or mitigate it. Take the time to do actual code reviews and make sure that the changes being made are understood. Talk about these things in Stand Up if necessary but surface these issues, as well as discuss these in your sprint reviews. Much of the time technical debt will be hiding in plain sight. The best way to deal with technical debt is to look for it and identify it as early as possible.