Image: de-RSE https://www.de-rse.org/de/index.html

Socializing Infrastructures

de-RSE – Research Science Engineers: A New Association in Germany

An interview by GenR with members of the newly forming German association of Research Software Engineers – de-RSE. The association is being formally constituted in late November 2018 (next week 26th Nov., Berlin) and will provide a new support network for the wider software community. Highlights being a summer 2019 conference in Potsdam, and a fellowship programme. The interview is the first in a series of article in GenR’s theme ‘Socializing Infrastructures‘ and we take the opportunity to ask de-RSE for their comments on Open Science infrastructures and ‘software as infrastructure’.

de-RSE member contributions from:

  • Frank Löffler, frank.loeffler@uni-jena.de
  • Stephan Druskat, stephan.druskat@hu-berlin.de

About de-RSE

Vital stats
URL: https://www.de-rse.org
Twitter: @RSE_de
Conference, de-RSE 2019: https://www.de-rse.org/en/conf2019
Mailing list: https://www.de-rse.org/en/join.html  
Slack: https://de-rse.slack.com/ 
About the association: http://www.de-rse.org/en/association.html 

GenR: What is the organization called, who are the founding members, and why have you formed?

de-RSE (Frank Löffler, Stephan Druskat): The founding of the (official) de-RSE association is about to happen soon (Nov 26th). However, the group has been active essentially beginning with the RSE 2016 conference in Manchester, UK. There, at a conference meant for the British part of the RSE movement, the three founding members Martin Hammitzsch, Frank Löffler, and Stephan Janosch decided to kickstart similar activities for the German community. Since then, the mailing list as the main means of general communication has grown to 222 members to date (12.Nov.2018). Regular audio and video meetings complement this for more specific topics like the founding of the official association (German: Verein), or the upcoming deRSE19 conference. “Historically”, de-RSE has been a small core group that has formed quickly after RSE16 — Stephan Druskat joined as the fourth “member” after being pointed towards the other three at WSSSPE4, which preceded RSE16 — and a wider community, whose activities inform de-RSE and help it gain traction. These include:

GenR: Can you tell us more about the de-RSE conference: its themes, and format?

de-RSE: True to the nature of the whole movement, the themes will be dictated by the community. Submission is open for all topics related to research software and people connected to it. For three days in June next year (4th to 6th June 2019), the international community will gather in Potsdam to network, and collaborate. The website for the conference is at https://www.de-rse.org/en/conf2019, and we plan to publish the call for contributions at the beginning of December 2018. The concept and roles of RSEs are still fairly new themselves, so we are explicitly looking forward to contributions from first-time presenters and people from all walks of science and research, including those that are traditionally deemed rather removed from computing.

GenR: What is the de-RSE mission and programme of activities?

de-RSE: A lot of the initial activities focused on bootstrapping the movement: creating a website, managing the email list and Slack channels, but other activities include providing talks at related conferences and workshops, as well as a platform for general RSE-related questions and offers. The mission itself is multifold, and reflects that of the international RSE community as a whole:

  • boost perception of software as a critical part of research;
  • professionalise software development in research, especially aimed at sustainability and reproducibility;
  • establish basics of software development and engineering as part of the curriculum at universities;
  • establish research software as key element of Open Scholarship*; collaborate with other initiatives that see research software as their object, e.g., eScience and infrastructures;
  • make it easier to write/find job proposals by introducing new, common terminology, and;
  • create career options for software-related jobs in research.

*Note: This denotes what we know as “Open Science”, changed here to reflect a discussion at IEEE eScience that the term excludes all non-capital-S sciences.

GenR: It would be helpful if you could provide a reference for the IEEE discussion on the use of the term ‘Open Science’. At GenR editorial office we have similar concerns as many scholars would not consider themselves ‘scientist’, but instead ‘humanists’ for example. For GenR purposes this is part of the reason that we use the term ‘researcher’ as in ‘Generation Research’ it denotes an even wider umbrella of who could be included.

Stephan Druskat: The issue has been brought to my attention during a discussion session on “Open Science” at the IEEE eScience 2018 conference in Amsterdam, where Daniel S. Katz noted in a comment, that the term “Open Science” should perhaps be dropped in favour of “Open Scholarship”, as Science with a capital S traditionally excludes disciplines such as the humanities, social sciences, etc. But the concept of what has been called “Open Science” does of course impact all academic disciplines, so a more inclusive term might help onboarding people from disciplines outside of STEM.

GenR: What are the ways that people can join, participate, and contribute to de-RSE?

de-RSE: As single point of contact, use http://www.de-rse.org/en/join.html . It describes how to join the group on Slack and the mailing list. Another way to join would be to come to the inaugural meeting of the deRSE association in Berlin on Nov 26th 2018, and be one of the official founding members of the association. (for more details: http://www.de-rse.org/en/association.html)

GenR: What ways are you looking to support the RSE community and what services are you providing?

de-RSE: Our contribution to the developing science landscape and IT-dependent research is mainly, but not exclusively, in the area of research software development. We want to prepare science for the ever-increasing influence of new technologies and new IT concepts, and help it cope with the challenges of the increasing digitalisation of research as well as exploit the resulting opportunities. A more detailed list of the challenges we want to address can be found here: http://www.de-rse.org/en/aims.html .

By teaming up, we give ourselves a voice, and represent our interests together in order to achieve our goals within the German science landscape.

In order to achieve our objectives, the following concrete activities are already being carried out or are planned:

  • providing a forum for the exchange of computer scientists, software developers, RSEs and scientists developing in and for research;
  • running Software Carpentry (SWC) workshops or workshops that follow the SWC concept;
  • implementing a Fellowship Programme in which ambassadors represent and realise the aims of the German RSE community locally, regionally, nationally and internationally;
  • running an annual research software conference not exclusively covering software engineering and not exclusively addressing RSEs;
  • establishing RSE teams in research institutes, universities and other scientific institutions and organizations, and;
  • setting up a national center for research software, covering the German research landscape, and embedding it in the e-Science and e-Infrastructure movement

Infrastructures

GenR: What are de-RSE’s views on ‘software as infrastructure’? How does this work?

de-RSE: Software is part of the computing infrastructure, and always has been. The awareness that most science done today relies on software infrastructure is slowly growing. One of our aims is to make sure this infrastructure does not fail, which includes supporting those that develop, manage, support and fund it. As for the term “infrastructure”: as an RSE-centred organisation, we acknowledge that both large, service-based infrastructures as well as smaller-scale, and offline software are part of the backbone of research, and we address the whole set of software used in and developed for research as infrastructure.

GenR: How do you look to get non-programmers involved in software processes: development, maintenance, learning, etc?

de-RSE: Everybody can be a programmer. Most self-proclaimed non-programmers only need to lose their fear to “break something” or to “embarrass themselves”. Three lessons are to be learned here. First, programming is not that different from what most of us do in their daily life. Second, you don’t need to be a master programmer to have a big, positive impact in your own life. And third, and hardest to accept, mistakes are part of the process. Everyone makes them, and making them is OK. More generally, we need to enable people to code by teaching them. Basic foundations of software development and software engineering — as, e.g., covered by the The Carpentries‘ curriculum — should be part of domain curricula in all places that educate researchers. More informal means of teaching may also be effective, such as hack events, workshops, mentorship, etc. Because not only can everyone be a programmer, everyone can also be an RSE.

GenR: About infrastructure ownership and governance: new options for ownership, governance and role the of private sector (how should institutions be formed, kept, changed, or renewed)

de-RSE: We believe that there is no one-size-fits-all. Different project have different needs, and will need to find their own way of organizing themselves.

GenR: Can procurement policies mandate free and open source software (FOSS), if so how and what are good examples? And what needs to go in place to accompany FOSS policies?

de-RSE: Funders like the DFG (Deutsche Forschungsgemeinschaft) can, and do, have a big impact when they enforce open licenses on software developed within projects they fund. This not only brings the issue of licenses to scientists that otherwise would not even think about a license at all (and thus, leave software non-free and unusable), it also helps scientists argue that the software they develop should be free if their institute policy would otherwise default to a non-free license.

GenR: What are the missing bits of Open Science infrastructure that you think should be there ‘the known unknowns’?

de-RSE: At the moment, there is no complete (or almost complete) database of scientific publications. There are several projects under way to change this, but mostly due to resistance by conservative publishers, their data is incomplete. This in particular includes the creation of a free (as in freedom), (almost) complete citation graph.

Another topic of interest especially for the scientific community, and related to software, is reproducibility. This does not only include developing software technically in a way that allows the reproduction of results, or to publish software in an way that allows free distribution, use, and modification. It also includes efforts revolving around proper handling, storage, annotation, and publication of scientific data.

Additionally, there is a profound lack of tools that help researchers find and evaluate research software. Some data repositories also list software, as do indexers, but the metadata models are often too generic to cater for usable and helpful search. An actual evaluation of research software based, e.g., on such search results would presuppose among other things the openness of the software, otherwise proper evaluation is impossible.

GenR: Open Science and infrastructure: hardware, methodologies, scientific knowledge, human capital, training and education, scientific inventions. What is on your list of what constitutes Open Science infrastructure?

de-RSE: All of the mentioned points are part of the science infrastructure already. Not all can completely be open realistically, though. This includes specifically hardware that cannot easily be shared. Other aspects can be made fully, or at least almost fully open, including scientific knowledge, and training and education. One key methodology is surely the creation and establishment of best practices for Open Science, and particularly open research software. To achieve this, training and education, again, are important factors.

DOI: 10.25815/

Z21J-SY37

Citation format: The Chicago Manual of Style, 17th Edition

Generation R, De-RSE, Frank Löffler, and Stephan Druskat. “An Interview with De-RSE: Supporting Researchers Using Software,” 2018. https://doi.org/10.25815/Z21J-SY37.

References

‘RSE 2016 | Manchester Conference’, 2016. http://ukrse.github.io/conf2016.

‘De-RSE.Org – Email List, Join’. Accessed 23 November 2018. https://www.de-rse.org/en/join.html.

‘4th Workshop on Sustainable Software for Science: Practice and Experiences (WSSSPE4) – WSSSPE’. Accessed 23 November 2018. http://wssspe.researchcomputing.org.uk/wssspe4/.

‘DH-RSE, Working Group of the German Association for Digital Humanities’. Accessed 23 November 2018. https://dh-rse.github.io/.

‘Digital Tools — Software and Services | Schwerpunktinitiative “Digitale Information”’. Accessed 23 November 2018. https://www.allianzinitiative.de/fields-of-action-projects/digital-tools-software-and-services/?lang=en.

‘Helmholtz Open Science: Helmholtz Open Science Workshop “Access to and Reuse of Scientific Software”’. Accessed 23 November 2018. https://os.helmholtz.de/bewusstsein-schaerfen/workshops/helmholtz-open-science-workshop-zugang-zu-und-nachnutzung-von-wissenschaftlicher-software/.

‘De-RSE.Org – DeRSE19 Conference’. Accessed 23 November 2018. https://www.de-rse.org/en/conf2019/.

Druskat, Stephan. ‘@danielskatz at #eScience2018 #OpenScience Discussion: It Should Be Called #OpenScholarship to Make It More Inclusive for Non-(Capital S)-Scientists.’ Tweet. @stdruskat (blog), 30 October 2018. https://twitter.com/stdruskat/status/1057303813051346944.

‘The Carpentries’. The Carpentries. Accessed 21 November 2018. https://carpentries.org/index.html.