Redefining legacy : a technical debt perspective

Organisations that manage legacy systems at scale, such as those found within large government agencies and commercial enterprises, face a set of unique challenges. They manage complex software landscapes that have evolved over decades. Current conceptual definitions of legacy systems give practitioners limited insights that can inform their daily work. In this research, we compare conceptual definitions of large-scale legacy and technical debt. We hypothesise that large-scale legacy reflects an accumulation of technical debt that has never been through a remediation phase. To pursue this hypothesis, we identified the following question: How do practitioners describe their experience of managing large-scale legacy landscapes? We conducted 16 semi-structured open-ended, recorded and transcribed interviews with industry practitioners from 4 government organisations and 9 large enterprises involved with the maintenance and migration of large-scale legacy systems. A snowball sampling technique was used to identify participants. We adopted an approach informed by grounded theory. There was consensus among the practitioners in our study that the landscape is fragmented and inflexible, consisting of many dispersed and fragile applications. Practitioners report challenges with shifting paradigms from batch processing to near real-time customer-focused information systems. Our findings show there is overlap between challenges experienced by participants and symptoms typified by technical debt. We identify a novel type of technical debt, ``Ecosystem Debt'' which arises from the scale, and age, of many large-scale legacy applications. By positioning Legacy within the context of Technical Debt, practitioners have a more concrete understanding of the state of the systems they maintain