An Environment for Sustainable Research Software in Germany and Beyond: Current State, Open Challenges, and Call for Action

Hartwig Anzt, Felix Bach, Stephan Druskat, Frank Löffler, Axel Loewe, Bernhard Y. Renard, Gunnar Seemann, Alexander Struck, Elke Achhammer, Piush Aggarwal, Franziska Appel, Michael Bader, Lutz Brusch, Christian Busse, Gerasimos Chourdakis, Piotr Wojtek Dabrowski, Peter Ebert, Bernd Flemisch, Sven Friedl, Bernadette Fritzsch, Maximilian D. Funk, Volker Gast, Florian Goth, Jean-Noël Grad, Sibylle Hermann, Florian Hohmann, Stephan Janosch, Dominik Kutra, Jan Linxweiler, Thilo Muth, Wolfgang Peters-Kottig, Fabian Rack, Fabian H. C. Raters, Stephan Rave, Guido Reina, Malte Reißig, Timo Ropinski, Jörg Schaarschmidt, Heidi Seibold, Jan P. Thiele, Benjamin Uekermann, Stefan Unger, Rudolf Weeber

The deRSE Society for Research Software, a key stakeholder in the German research software landscape.

Introduction: The Academic Software Crisis

Modern scientific breakthroughs, from imaging black holes to modeling global climate patterns, are increasingly dependent on complex software. However, research software is frequently treated as a disposable byproduct of the research process rather than a core academic asset. This systemic neglect has led to a "software crisis" in academia, where code developed by PhD students or postdocs on temporary contracts often becomes undocumented, non-functional, or lost once their funding ends. This lack of sustainability not only wastes public resources but also exacerbates the "replication crisis," as findings cannot be verified without accessible, working software.

Methodology and Community Engagement

This position paper is the culmination of a broad community effort, synthesizing insights from major workshops held in 2019, including the deRSE19 conference. The authors employ a persona-based approach, following a fictional researcher named Kim to contrast the current state—where software development is often a career burden—with a proposed future where code is a recognized, career-enhancing asset. Furthermore, the paper introduces a Selection Criteria Framework to assist funding agencies in identifying software projects that warrant long-term, sustained investment based on their maturity, quality, and community impact.

Key Pillars for Sustainability

The roadmap for sustainable research software rests on several critical concepts:

The Research Software Engineer (RSE)

The paper champions the "Research Software Engineer" (RSE) as a distinct and vital professional role. RSEs bridge the gap between domain-specific scientific knowledge and professional software engineering. To prevent the drain of technical expertise, the authors argue for the creation of permanent staff scientist positions ("Mittelbau") dedicated to software development within research institutions.

Beyond FAIR Data: Executable Software

While the FAIR principles (Findable, Accessible, Interoperable, Reusable) are well-established for research data, software presents unique challenges because it is executable. Survival requires continuous maintenance to adapt to evolving hardware and library dependencies. The authors advocate for a "Public Money, Public Code" paradigm, suggesting that software developed with public funds should default to Free/Libre Open Source Software (FLOSS) licenses.

The Sustainability Triage

Recognizing that not every script requires long-term preservation, the paper proposes a three-tier maturity model:

  1. Class 1: Scripts for personal or single-group use.
  2. Class 2: Tools used beyond a single research group.
  3. Class 3: Standard infrastructure for an entire field.

Only Class 2 and 3 software should be prioritized for dedicated, long-term sustainability funding.

A Strategic Call for Action

The paper concludes with specific recommendations for German and international policymakers to modernize the research ecosystem:

  • Institutional Support: Establish a National Software Sustainability Institute to provide centralized consulting, training, and code review.
  • Funding Reform: Shift from short-term "innovation" grants (typically 2 years) toward longer 5-year cycles focused on maintenance and stability.
  • Curriculum Integration: Embed software engineering fundamentals into the Bachelor and Master level curricula for all scientific disciplines.
  • Academic Recognition: Reform the "Hirsch-index" culture to ensure that high-impact software contributions carry significant weight in tenure and promotion decisions.

Conclusion

While the focus of this roadmap is the German research landscape, the challenges identified are global. Addressing the tension between the academic drive for "new" innovation and the practical necessity of "boring" maintenance is essential. By professionalizing the RSE role and integrating software into the formal recognition structure of science, the research community can ensure that the digital foundations of modern knowledge remain robust and reproducible.

Summary updated

BibTeX
@inproceedings{anzt2020environment,
  author = {Hartwig Anzt and Felix Bach and Stephan Druskat and Frank Löffler and Axel Loewe and Bernhard Y. Renard and Gunnar Seemann and Alexander Struck and Elke Achhammer and Piush Aggarwal and Franziska Appel and Michael Bader and Lutz Brusch and Christian Busse and Gerasimos Chourdakis and Piotr Wojtek Dabrowski and Peter Ebert and Bernd Flemisch and Sven Friedl and Bernadette Fritzsch and Maximilian D. Funk and Volker Gast and Florian Goth and Jean-Noël Grad and Sibylle Hermann and Florian Hohmann and Stephan Janosch and Dominik Kutra and Jan Linxweiler and Thilo Muth and Wolfgang Peters-Kottig and Fabian Rack and Fabian H. C. Raters and Stephan Rave and Guido Reina and Malte Reißig and Timo Ropinski and Jörg Schaarschmidt and Heidi Seibold and Jan P. Thiele and Benjamin Uekermann and Stefan Unger and Rudolf Weeber},
  title = {{An Environment for Sustainable Research Software in Germany and Beyond: Current State, Open Challenges, and Call for Action}},
  booktitle = {F1000Research},
  year = {2020},
  url = {https://openreview.net/forum?id=NGvYqc7OVk},
}
Figure