Published July 7, 2021 | Version v2
Dataset Open

The Temporality of Technical Debt Introduction on New Code and Confounding Factors

  • 1. Institute of Mathematics and Computer Science, University of Groningen, Netherlands
  • 2. Department of Applied Informatics, University of Macedonia, Greece

Description

Code Technical Debt (TD) is intentionally or unintentionally created when developers introduce inefficiencies in the codebase. This can be attributed to various reasons such as heavy workload, tight delivery schedule, or developers' lack of experience. Since a software system grows mostly through the addition of new code, it is interesting to study how TD fluctuates along this process. Specifically, in this paper we  investigate: (a) the temporality of code TD introduction in new code, i.e., whether the introduction of TD is stable across the lifespan of the project, or if its evolution presents spikes; and (b) the relation of TD introduction to the development team’s workload in a given period, as well as to the experience of the development team. To answer these questions, we have performed a case study on 47 open source projects from two well-known ecosystems (Apache and Eclipse) as well as additional isolated projects from GitHub (not selected from a specific ecosystem) and inspected the number of TD issues introduced in 6-month sliding temporal windows. The results of the study suggested that: (a) overall, the number of TD issues introduced through new code is a stable measure, although it presents spikes; and (b) the number of commits performed, as well as developers' experience are not strongly correlated to the number of introduced TD issues.

Files

Files (368.9 kB)

Name Size Download all
md5:b1b8ef3f6794bdcde1e02f91a19e3167
368.9 kB Download