The Code Handoff: Pitfalls and Good Practices for Handing Code off to a New Developer

Geoffrey Boushey
-
Proctor

A surprising amount of code essential to academic research is written by postdocs, graduate students, and programming staff on limited grant appointments in research labs.  The process of transitioning ownership of code to a new research programmer is challenging under the best circumstances. Unfortunately, a lack of consistent code versioning practices, build procedures, supporting data, use documentation, integration testing, and outside collaboration often makes this transition considerably more difficult.  By addressing these elements continuously during development, research labs can greatly minimize the expense and delay involved in bringing a new developer up to speed with a code base.  Ideas will be explored by incrementally improving a sample application with the following issues: 1) missing code, 2) missing build processes, 3) missing data, 4) missing documentation, 5) missing tests, and 6) missing collaborators.

Session Files