The Amsterdam Declaration on Sustainable Research Software
Press enter or click to view image in full size
Software is now a key part of every research project. It’s either a direct instrument in the data ecosystem, such as the vast geospatial data, or material simulation, or provides data science tools for making sense of the results. Yet in many ways, research software lags behind industry standards — in quality, documentation, tests, performance, — overall, what we call software sustainability. And so a lot of it becomes abandonware, subject to /bitrot/ — when a PhD candidate graduates or a professor goes to another university, the project they developed to support their work becomes abandoned. Without documentation or support it quickly becomes unusable.
There are forces in play working against this trend. Eg, in EU, publicly funded software should generally be made available as OSS. However, that, by itself, does not prevent abandonware.
In November 2022, the National eScience Center and Research Software Alliance convened a meeting of research funders devoted to the Future of Research Software. I’m proud to be the IBM representative in this process, drafting the Amsterdam Declaration on Sustainable Research Software. The declaration will consist of several parts, including principles. recommendations, and implementation guidelines.
I’ve worked with the drafters to ensure that a vibrant and viable community is recognized as a prerequisite for any sustainable software. Furthermore, we agreed that clear delineation of the roles required to sustain a community, such as a community manager, is key. Not only should it be OK for one in a grant application, it should be encouraged by the funders.
You can see the current draft here and comment on it. There is now a website hosting the declaration, called ADORE.software.
The meeting included multiple national innovation agencies, nonprofit foundations and international consortia such as OECD. The funders are actively seeking guidance on the best practices ensuring their investment persists and maximizes the public good ensuing from the public funding of research and innovation. The Amsterdam Declaration process allows us to have a seat at the table among those whose contributions will define the future of research software, solving the hardest challenges facing humanity.
