FORCE2019

Code has become a central methodology for research, and the number of research papers that make use of existing code or develop new code is growing rapidly. Increasing demands on reproducibility and transparency of published papers, as well as recognition of the individual research artifacts that substantiate the results, has heightened the importance of proper documentation, evaluation and accessibility of published code. Nature and other publishers started peer-reviewing custom code over ten years ago for some journals, when that code is central to the main paper. However, wider implementation of this editorial practice has been slow. This talk will discuss best practices for code peer review, drawing on the experience of several stakeholders, as well as giving an overview of recent technological innovations, piloted improvements to platforms and more efficient workflows that help to facilitate peer review and publication of code. We will also discuss how best to incentivize peer review and publication of reproducible and open code, in order to drive uptake of these practices.