Investing in People: Anticipating the Future of Research Software
Authors: Kim Hartley and Michelle Barker
We are proud to report that the second International Research Software Funders Workshop, co-hosted by the Digital Research Alliance of Canada (the Alliance) and the Research Software Alliance (ReSA) from 18-20 September was a great success. People are at the heart of research software sustainability and funders can lead the way by investing in the people who develop and maintain research software. That is why the workshop theme was “Investing in People: Anticipating the Future of Research Software”. The event aimed to set the future agenda for government, philanthropic, and industry funders to support sustainable research software and communities.
In total, 50 representatives from more than 35 organisations – including funders and organisations that support research software from around the globe – attended the hybrid workshop held in Montreal, Canada, and virtually. During the three-day workshop, experts explored how research software funders can facilitate capacity building through funder practices; research software platforms, infrastructure, and communities; and new initiatives. Moreover, Version 1.0 of the Amsterdam Declaration on Funding Research Software Sustainability (ADORE.software) was released and is now available for signing.
This second iteration of the funders workshop built on the inaugural International Funders Workshop: The Future of Research Software, co-hosted by the Netherlands eScience Center and ReSA in November 2022 in Amsterdam. During this first workshop, more than 40 funding organisations gathered to set the agenda for supporting sustainable research software and drafting of the ADORE.software. Following public consultation in 2023 and 19 organisational expressions of interest in becoming signatories, Version 1.0 of the Declaration was released in advance of the Montreal workshop and is now available for signing. ADORE.software is the first step towards formalising, on a global level, the basic principles and recommendations related to funding the sustainability of research software, including the people needed to achieve this goal.
The first workshop marked a further step in the development of the research software community. Stakeholders have continued to build on this momentum through involvement in the Research Software Funders Forum, convened by ReSA, and its funder-led working groups. The Alliance and ReSA, together with the workshop Steering Committee, organised the 2nd International Research Software Funders Workshop to continue this important work and facilitate global collaboration among funders and other key decision makers and influencers across the research software ecosystem.
2nd International Research Software Funders Workshop
Pre-workshop - September 18
The event commenced with a pre-workshop, including a session on How to Explicitly Support Research Software as Part of Your Funding Program, delivered by ReSA, and a session on CiteSoftware, led by a group of stakeholders aiming to drive the adoption of a common research software preservation and citation guidance resource for the research community.
Workshop Day 1 - September 19
The first day of the workshop included presentations that set the scene for participants to explore the crucial role funders play in supporting sustainable research software and communities.
Carole Goble from the Software Sustainability Institute (SSI)/ELIXIR-UK/University of Manchester provided an overview of what research software is and why it is critical to the research endeavour. Carole highlighted that the people who make, use, and resource software are fundamental to research software sustainability. Moreover, research software needs resources and both direct (e.g., money) and indirect (e.g., community, reward, recognition, open-source support, etc.) forms of support for sustainability.
Josh Greenberg from the Alfred P. Sloan Foundation delivered a keynote that provided a funder’s perspective. He asked attendees several pertinent questions: How much (and when) should we worry about technical debt? How much attention should we pay to user interface design in research software? Who should do which kinds of software? How central should “open source” be? How can we help institutions build capacity for research software development (and maintenance)?
The recommendations in ADORE.software informed the structure of the workshop’s breakout sessions, with a focus on research software practice; research software ecosystem; research software personnel; and research software ethics. During the first breakout, attendees discussed capacity building through funder practices. Key takeaways include:
- Funders benefit from investing both before (community/need development) and after (ongoing maintenance and re-use) research software development.
- International collaboration and collective action are critical and valuable (e.g., building a multilateral funding initiative due to the global nature of research software).
- Funders should consider how to support and sustain Research Software Engineers (RSEs) but also incentivise researcher-developers to develop better software engineering practices.
- The research software community needs a set of high-level classes to measure the impact of research software, such as citations and mentions of software maturity, or combining mention analysis and dependency analysis.
Malvika Sharan from The Alan Turing Institute and Open Life Science provided a research software community perspective in her talk on “Exploring ‘Do No Harm’ Principles in Open Research Communities.” Malvika’s compelling keynote encouraged funders to become leaders in the research software ecosystem by incentivising and recognising all types of research roles; investing in capacity-building goals that focus on bridging scientific and economic divides; and applying community-oriented frameworks to extend the benefits of research to the broader community that ‘do no harm’, thereby ensuring open science practices enable equitable research and research outcomes.
During the second breakout session, participants investigated capacity building through research software platforms, infrastructure, and communities. One group explored the role of Open Source Program Offices (OSPOs), noting that OSPOs as centres of software could ultimately help research institutions understand their software investments. Another group considered why and how to support research software communities, highlighting the need for more evidence-led recommendations and exploring the possibility of reversing the funding mechanism to fund projects after the fact, as a function of how much they were used. In discussing research software in research assessment reform, participants noted that very few research software outputs are being assessed and indicators are fragmented and field-specific. Some solutions include sharing guidance on how software is being evaluated in a single place, sharing job profiles for RSEs, and sharing existing examples of research indicators that include software. Finally, another group highlighted that the environmental impact of research software is an emerging field, and persuading researchers to think about making their code efficient is a challenge.
Workshop Day 2 - September 20
The final breakout session focused on building capacity through new initiatives. Participants explored how funders can implement FAIR for Research Software (FAIR4RS) in funder policy. As some funders are developing guidance, questions arose about whether funders should create guidance individually or collectively and how the community should help funders do this. Through a discussion on how funders use persistent identifiers (PIDs), the need to engage publishers around software citation has become clear. And building on the theme of software citation, participants explored how the Research Software Directory, a free open-source platform, supports software citation.
For the final keynote, Jean-Baptiste Poline from McGill University provided a researcher’s perspective on the impact of funders' investments in research software. Jean-Baptiste presented the role of research software from multiple viewpoints, offering key takeaways such as the value of building a community of researchers through collaborative projects; the need for community-endorsed standards and practices that support open and FAIR (Findable, Accessible, Interoperable, and Reusable) scientific research (e.g., INCF Network) to maximise research value; the importance of addressing incentives first; and the idea of funding new technologies and large labs but giving precedence to well-governed communities more than centralised projects. He recommended creating the tools needed to change culture, improving training and open practices, developing community standards, and changing incentives.
The workshop culminated with a Do-a-Thon, “a short collaboration where participants from different perspectives and skill sets work together on challenges, projects, or to learn something new” (OpenCon Cascadia). Prior to the workshop, participants were invited to submit do-able ideas. During the Do-a-Thon, attendees worked in small groups on the following topics: the Map of Open Source Science; linking software to community usage; and ten things applicants should do to make it easier to assess the impact of funding programs, to build on Ten simple rules for funding scientific open source software by Carly Strasser et al. This dynamic collaboration resulted in participants considering next steps, crowdsourcing tips, and determining some practical and generalisable approaches to address the challenges explored.
Overall outcomes and how you can get involved
The workshop solidified the importance of international collaboration to support research software. There is a widening group of funders with awareness of the need to support research software and an understanding of how to do this. The Research Software Funders Forum has been building momentum since its inception in early 2022; its dedicated funder-led working groups are expanding their reach by engaging with other stakeholders and organisations, such as the Global Research Council. Moreover, the workshop highlighted areas where coordination by funders would be beneficial, for example, in supporting the Amsterdam Declaration on Funding Research Software Sustainability. Participants also expressed an interest in exploring opportunities for forums, for example, to engage publishers around software citations (along with the existing ReSA Task Force on Code Availability).
We thank all the participants for their engagement during the workshop sessions. We’re grateful to the workshop Steering Committee for their efforts in organising a successful hybrid workshop. A full report, written and edited collectively by workshop participants, is forthcoming.
To stay up to date on ADORE.software, visit www.adore.software. If you want to show your support for the Declaration, you can become a signatory (for funders) or supporter (for non-funders) and help share the news. To receive updates on news related to the Declaration and future activities, subscribe to the ReSA newsletter. Funders are invited to join the (free) Research Software Funders Forum by contacting ReSA at firstname.lastname@example.org.
If you’re interested in learning more about ReSA and its activities, join our next Research Software Community Leaders Forum on 15 November at 20:00 UTC. ReSA’s Community Leaders Forum is open to all, and you can sign up for calendar invites.
We would like to express our sincere gratitude to our community, Founding Members, and Organisational Members – without whom none of this work would have been possible. This project has also been made possible in part by grant 2021-000000 from the Chan Zuckerberg Initiative DAF, an advised fund of the Silicon Valley Community Foundation; and the Alfred P. Sloan Foundation.