By SSI Fellow Matthew Bluteau.
Code review plays a crucial role in ensuring quality in the software development lifecycle, and it is a great practice for knowledge transfer within teams. However, like with most standard software engineering practices, it is much less prevalent in the world of research software development. It was this observation that led to the creation of the Research Code Review Community (RCRC) by Hollydawn Murray (Health Data Research UK). I would like to take some time in this post to reflect on the activities of the RCRC and some of the results for me personally that came out of my involvement.
The Research Code Review Community
Hollydawn led a rallying call at the beginning of 2021 to everyone in the research software community to build consensus and awareness around good practice in code review. I came across it via a lightning talk at SORSE. Acknowledging that fostering this type of culture practice would require change at many levels, the RCRC (then CRC) set out five working groups:
- Diversity, equity, and inclusion
- Code review during development
- Code review at the time of publication
- Recommendations for stakeholders
- Training and education
I was involved with the “Code Review During Development” group, or “dev-review” for short. We made steady progress towards defining guidelines to implement code review in the research software development workflow. The final outcome was a website with flowcharts to guide anyone developing research software towards the practice of code review.
Sadly, like many volunteer-driven projects, the entire RCRC started to fizzle out after about a year. It is an interesting topic why this happened in our particular case as with lots of similar projects more broadly, but that is outside the scope of this current post. Personally, it was a hugely valuable experience to collaborate with people from diverse domains, roles, and countries, and it was one of my first true introductions to the power of the “open source” model. However, I was still left with the feeling that our resource needed some further advertisement.
Our dev-review group gave one last kick at the can at CW22, running a workshop to get feedback on the website and raise awareness about the project. But alas, we weren’t able to garner the interest we needed.
RSECon22: Making Connections
The story of the RCRC doesn’t quite end there. We all know how conferences can be great places to connect with peers, and RSECon22 is no different. I was lucky enough to bump into another SSI Fellow, Hannah Williams. In the course of our conversation, Hannah mentioned she had heard about my work with the RCRC and said she would be interested in having the guidelines about code review presented at her institution, the UK Health Security Agency (UKHSA).
Thankfully, Hannah followed up on this, and I was invited to give a talk at the UKHSA Code Review Workshop in December 2022.
UKHSA Code Review Workshop
I delivered these slides at the workshop, and they seemed to be generally well received. It was interesting to see some of the other code review practices happening at UKHSA, which contributed to my own awareness of the state of code review in UK public sector bodies. Being an employee of a public sector body myself, it is useful to know about the practices other organisations in similar positions are using.
My journey in the RCRC showcases the transformative power of collaboration but also some of its limitations when exclusively run by volunteers. I like to think that code review in research has improved because of the RCRC efforts, but admittedly it is only a small step forward. More positively, a chance encounter at RSECon22 facilitated by the SSI Fellowship shows how conferences and fellowships can bridge gaps and extend the impact of community efforts.