Mining RSE repository timelines on GitHub: How long will it live, and who will notice?

Do your research software repositories have licenses? Do you use issues? Have you added contributing guidelines? And more importantly, does it make a difference? We mined ePrints publications and GitHub repositories to find out. We first scanned around 200,000 publications from UK universities for links to GitHub. Then, we manually identified links that pointed to software created for the publication to make sure that we were in fact looking at research software. Finally, we pulled a broad range of data about those repositories using the GitHub API and transformed those into a time-lined format. This allows us to analyse the lifecycle of a repository – when do potential users and contributors discover it, and when do they start engaging with the code, if at all? What aspects make or break the community's engagement with your source code? Among other findings, our exploratory analysis suggests that leading by example encourages externals to submit issues and commits as well, and that repositories that manage to grow their contributor pool continue to increase their stars for longer. The resulting code base and data are publicly available on GitHub (https://github.com/softwaresaved/rse-repo-analysis). A recording of this session is available on YouTube: https://youtu.be/MjhhCI0GOU8