Discussing The Research Software Ecosystem

In line with the international definition [Barker et al., 2022, Gruenpeter et al., 2021], we define research software as software created during the research process or for a specific research purpose. This definition encompasses source code, scripts, computational workflows, executables, and related artefacts, as well as a variety of research software types and categories, ranging from a few lines of code to large research platforms. Research software is used to collect, analyse, simulate, process, visualise, or manage research data, including text, images, audio, sensor data, and other digital resources, as well as to generate scientific models, control research instruments, and optimise research procedures. Additionally, aspects that play a role in its development are often included in research software, such as technical documentation, user manuals, metadata, parametrisation, management plans, and digital notebooks [Deutsche Forschungsgemeinschaft, 2024], although these are not specifically included in the international definition.

When discussing the research software ecosystem (Footnote: In a strict sense, a more nuanced term would be "research software anthroposystem", but this term is rarely used.), we refer to the dynamic, complex, interconnected, and interdependent network of people, communities, institutions, software, infrastructure, practices, activities, policies, and resources that collectively support the sustainable development, deployment, use, sharing, and maintenance of research software. In this ecosystem, research software has three interconnected functions: i) it provides use functions by acting as a means for research tasks such as data analysis, simulation and modelling, instrument control, and workflow management, ii) it delivers archive functions by preserving the intellectual work behind research and ensuring that it is findable, accessible, interoperable and reusable (FAIR), while supporting reproducibility and iii) it enables community functions by fostering collaboration, learning or training, advancing research through iterative development, and building reputation.