From preservation to re-execution: Bridging the gap between repositories and computational replay

Abstract

It is worthwhile to make institutional repositories interoperable with Web-based computational reproducibility tools to foster the reuse of scholarship. While repositories excel at long-term preservation, they were not made with the interactive capabilities required to execute complex code or analyze dynamic data directly. There are several open-source platforms, including JupyterHub, BinderHub, Whole Tale, and Emulation-as-a-Service Infrastructure, which provide ‘sandbox’ environments for replaying research that can be integrated with repositories to add this functionality. By integrating these scholarly tools, institutions can move beyond static file storage to offer seamless, in-browser execution of scholarly materials. This unified approach aims to reduce technical barriers for secondary researchers, ensuring that computational work remains both FAIR and re-executable. This approach advocates for a research lifecycle where active preservation and active discovery and reuse are deeply interconnected through sustainable infrastructure.