-This paper provides motivation and an initial plan for creating a professional society (or extending the scope of an existing organization) in research software. The aim of this paper is to initiate a dialogue on this topic.
This paper is concerned with ”research software”, by which we mean the software which is needed to conduct research, whether it is simulation codes, workflow scripts, visualization and data analysis tools or low level system software for new computing environments.1
Research software is behind many modern advances in science and engineering, facilitates new understanding and research methods in the humanities, social sciences, education and the arts, impacts the research and development agenda of our major industries, and contributes to national security. In addition, as we investigate new mechanisms of teaching and learning in our schools and colleges we expect research software to play an increasing role across the spectrum of educational activities.
The development, use and sustaining of research software
takes place around the world primarily at university campuses
and national laboratories by a distributed, unconnected and
often unrepresented set of students, postdoctoral researchers,
programming staff and early-stage faculty. It is a matter of
debate as to whether the current credit mechanism can properly
assess or reward the broad workforce, and the broad activities,
involved in providing and supporting research software. The
current accepted mechanism is via research or experience
publications such as to the SIAM Conference on Computational
Science and Engineering, the IEEE Journal of Computing in
Science & Engineering [
Many researchers and practitioners in traditional disciplines
are members of professional societies — for example the
American Physical Society [
This paper proposes the development of an independent entity which could act as a “professional society” for those involved in research software. The broad aim of such an entity would be to bring together those developers, educators, researchers, and professionals with an interest in research software to promote dialogue, share resources and experiences, address challenges and support professional development and furthering the profession of research software development and support. In addition, such a society would represent the public interest in promoting the role of research software.
This paper lays out initial thoughts on the role this society might play, and how it might be constructed, funded and sustained, in the hope that this could be the start of an ongoing conversation in the WSSSPE community. The end of the short paper lists potentially similar activities, and if there is interest in moving ahead with the ideas here, a first step should be to investigate if the activities would fit within an existing organization, or an existing organization might be modified to include these activities.
This section collects together potential roles that a professional society for research software might fulfil. 1) Be the leading voice for research software and an authoritative source of information for the advancement of research software and its role in benefiting society. 2) Support professional growth of members through opportunities for life-long learning, career development, and professional networking. 3) Advocacy and lobbying for research software and research software providers to federal agencies and congress including funding and career paths. 4) Coordinating body to collect and analyze information about research software. 5) Provide effective programs in support of the research software community and the conduct of the development and sustaining of research software. 6) Collaborate with national scientific and research societies for the advancement of research software, research software education and through this the science community. 7) Advertising of jobs in research software development, potentially leading to a mechanism to facilitate contracting across sites for software expertise. 8) Coordination of awards to reward aspects of software development. 9) Investigating research software alumni and contribution to society. 10) Hold and sponsor conferences and workshops in research software 11) Support journal or other communication medium. 12) Promote an active, engaged and diverse membership, and support diversity and inclusion across the research software community. 13) Provide public outreach around research software. 14) Establish and reward outstanding mentors who can help develop the community. 15) Provide material to help influence the promotion of research software developers, for example providing information useful for improving reference letters, resumes, etc. 16) Develop and monitor education programs in research software development. 17) Through membership establish professional qualifications, e.g. fellows.
Funding mechanisms discussions need to include both short term plans which are focused around start up activities, and longer term plans focused on appropriate and realistic sustainable funding models. Realistic understanding particularly of longer term funding is essential.
1) Short term initial funding from foundations such Gordon and Betty Moore or Sloan to provide necessary flexible funding in preliminary stages. 2) Short term funding for people networking from NSF
RCN-like solicitation. 3) Initial donations from large sites with a vested interest in promoting research software (e.g. current HPC centers) 4) Membership fees for individuals, sites, and software projects. 5) Gifts/membership fees from vendors and industry.
Going ahead with establishing a society would be dependent on there being individuals interested, motivated, and with the right skills to organize and lead the activity. This motivates the need to understand the essential elements needed to bring together into a comprehensive plan. Generic elements towards forming a non-profit organization in the US include 1) Find funding to start planning and some operations at a supportive host university or other organization, then establish a non-profit organization and apply for taxexempt status. 2) Establish a group to develop and implement below plans, think about leadership, mission statement, advisory team/board. 3) Articulate well the scope of issues for the organization and the demographic and stakeholders it will serve. 4) Establish a business plan and staff plan, e.g. Executive Director, communications coordinator, administrator, come up with a budget plan. 5) Make sure there really is a need and interest in the organization, and potential funders. 6) Choose a name (with free domain name), and state in which non-profit will be registered. 7) Assemble a board of directors (3-6 people) with appropriate experience and diversity across the target demographic. 8) Follow steps to incorporate the organization, which will be state dependent. This will include finalizing the name, filing articles of incorporation and creating bylaws. 9) File for nonprofit status (involves significant fees, paperwork associated with federal and state tax exemptions). 10) Develop a sustainable fundraising plan and apply for foundation grants after being awarded 501(c)(3) status. 11) Create a marketing plan to recruit donors and volunteers. 12) Provide good communications and online presence including social media.
As mentioned in the introduction, if the idea of a professional society for research software is of interest, then the very first task should be to consider where there is overlap with existing societies, if an additional society is needed, and whether existing societies might be extended instead of forming a new society. Also, there are mature domain-specific societies in some disciplines which should also be looked at in regard to their potential role. A close second task should be to consider whether there might be additional new communities that could be included under the same discussions — for example in reproducible science or cybersecurity, as well as thinking about how the research data community overlaps with the interests of the research software community — and whether it could make sense to pursue a society focused on both research data and software.
a) Coalition for Academic Supercomputing Centers
(CASC): CASC [
b) Association for Computing Machinery (ACM): The
ACM [
c) Research Software Engineers in the UK: In the United Kingdom, the Research Software Engineers (RSEs) in the UK group citerse is based around staff engineers rather than the field of research software as a whole. It’s objectives are to create a community to represent UK RSEs, raise awareness of RSEs; campaign for the recognition and reward of RSEs; campaign for RSE to be a formal role in academic; organize networking events for RSEs.
d) Software Sustainability Institute (SSI): Based at the
Universities of Edinburgh, Manchester, Oxford and
Southampton, the mission of the Software Sustainability Institute [
e) NumFOCUS: NumFOCUS [
f) Association for Learning Technology (ALT): In the
United Kingdom, the registered charity ALT [
g) Advanced Computing Systems Association (USENIX):
The USENIX association [
This short paper gathers very preliminary ideas around the need for a professional society of research software, or for incorporating research software more fundamentally under the wings of an existing professional society. The aim of this paper is to stimulate discussion among potential members and stakeholders of such a society.
The author would like to thank the anonymous reviewers whose suggestions and comments helped improved the paper.