FAIRIES Hub@UiO is about FAIR research software, Infrastructure, Expertise and Support at the University of Oslo

Jean Iaquinta

jeani-at-uio.no

Max 400 ord

The Software FAIRIES Hub@UiO is about FAIR research software, Infrastructure, Expertise and Support at the University of Oslo. It is based on the fact that “SOFTWARE IS NOT DATA”, as stated by Lamprecht, Anna-Lena et al. in ‘Towards FAIR Principles for Research Software’ (DOI: 10.3233/DS-190026).

Software is commonly referred to as “a collection of instructions and data that tell a computer how to work” (https://en.wikipedia.org/wiki/Software). Data are facts or observations that provide evidence. In contrast, software is the result of a creative process that provides a tool for doing something, for example with data. Software is developed and ought to be properly maintained and published, which involves versioning, documentation, licensing, testing and validation, etc.

The term FAIR (Findable, Accessible, Interoperable and Reusable) has been circulating for some time now but mostly refers to data (static). However, while open science and reproducible research are pillars at UiO, neither the University Library nor USIT clearly provide clear guidance about practicalities (what has to be done), responsibilities (who has to do it), and even less the necessary means, supporting staff and infrastructure, training/mentoring or accompanying implementation measures.

CodeRefinery (https://coderefinery.org) fulfills the training needs but does not address the specificity of each UiO faculty as it is done for research data management, in particular when it comes to long-term archiving of the research software along with all the necessary environment to reuse it. We believe that what fundamentally hinders a wider adoption of “FAIR software” at UiO are:

• A general lack of understanding regarding the difference between software and mere data;
• No clear guidelines to researchers on what to do with software;
• No targeted communication towards each faculty (meet people where they are and take into account their current knowledge and practices);
• No examples to showcase for each faculty;
• Little information/guidance and training on available infrastructure and how to use them (mybinder, NREC and other cloud services) and tools (package managers such as conda, containers, workflow management systems);
• Unlike publication of research articles or requirements of Research Data Management by funding agencies, there is no real incentive or recognition for those who would be tempted to “do things correctly” as opposed to “do the bare minimum”;
• UiO does not have a way to provide advice on software licensing.

The purpose of the present hub is to address each of these issues.

Dette inkluderer også en kort beskrivelse av hva man ser for seg etter endt sentral finansiering. Det legges vekt på en sunn "exit-strategi".

The project will deliver:

• A website with online documentation explaining the differences between software and data, and why it is paramount to adapt the FAIR principles to software in the context of open science and reproducible research;
• Presentations (face-to-face or online) of FAIR software principles and practical recommendations such as use public repository with version control, add a license, register your code in community registry, enable citation, etc. (see for instance https://fair-software.nl/) to University Librarians, USIT research support and faculty administrators as well as academic staff that are current and future PIs of research projects;
• A “cookbook” intended to researchers and support staff (research software engineers, USIT research support staff, etc.) from various faculties on what to do with software in practice;
• Workshops for training researchers on best software practices (with CodeRefinery);
• Open call(s) for FAIR software champions at UiO (basis for showcases);
• Showcase examples of FAIR software targeting each faculty (domain specific) with proper usage of the available infrastructure (mybinder, NREC and other cloud services) and tools (package managers such as conda, containers, workflow management systems);
• Publish FAIR Software champion success stories for broad dissemination at UiO;
• Provide counseling on software licensing.

Once a network is formed, and the community as a whole has recognized the importance of FAIR software, this hub will nicely complement the strategy devised by USIT and the University Library, taking full advantage of the services for management and curation offered for FAIR research software (and not only “static” data).

Software FAIRIES hub @ UiO is not meant to remain as a hub-node. The aim is to transfer all the FAIR software and other best software practices activities to more permanent services served by USIT. Nowadays, apart from training activities (under CodeRefinery framework) there is no service unit to support high-quality research software developments and publication. We are aiming at setting up a new service similar to what is provided by the UK Software Sustainability Institute (https://www.software.ac.uk/) and foresee several steps:

1. Transfer hub/node activities to USIT that would naturally provide day-to-day services to researchers for their research software;
2. Identify synergies at national and Nordic level: Norway is a tiny country with few inhabitants, and qualified staff in the domain of software engineering in academia is scarce. We think that having a center of excellence at national level would make more sense, and we believe UiO-USIT can lead these activities e.g. extend what will be done within UiO to other universities in Norway.

Ikke bindende

Department of Geosciences

Ikke bindende

Anne Fouilloux

Research Software Engineers in the Local IT units will be involved and benefit from the training as well as knowledge sharing and expertise acquired within the hub.

University Library will be involved and contribute to all aspects related to the management and curation of FAIR software (as they are doing for FAIR data), but also by arranging and possibly hosting (if physical) training events.

USIT will be involved in the hub as provider of the infrastructure including storage, cloud computing resources and virtual machines on demand, and support staff. USIT would also host a help desk with a dedicated contact address in the user support system. Oppgi noen aktuelle miljø som kan være noder i nettverket

• Naoe Tatara, University Library and Department for Data Capture and Collections Management
• Sabry Razik, Section for Research Infrastructure (FI)
• All UiO departments where software is developed
• Local IT units and Research Software Engineers
• University Library - Department for digital services and collections
• USIT - Department for Data Capture and Collections Management

Max 300 ord

With the Software-FAIRIES-Hub@UiO we primarily aim at:

1. expanding knowledge about “best software practices” through

   • end-user workshops and training (there is a gap between simplistic “Hello world” examples and the real-life questions): these trainings will be provided with CodeRefinery https://coderefinery.org/ (some of us are CodeRefinery members), and lessons/material will be developed within CodeRefinery too;
   • hackathons targeting new communities (to introduce FAIR software in practice, with examples from their application areas); each node being a new community, we will start with department/institutes where Carpentries instructors were identified (through the Nordic Carpentries & CodeRefinery communities; see https://tinyurl.com/carpentries-nordics);
   • training of Research Software Engineers (or equivalent) in each UiO Unit so that expertise does not vanish when students, post-docs and temporary staff move on to new pastures.

2. providing an infrastructure to facilitate FAIR software development and archiving

   • through access to suited resources internal to UiO (due to restrictions and/or confidentiality, all software may not be made public) for computing on the cloud; for storing and curating FAIR software (with metadata, inputs/outputs, license, DOIs, etc.), in collaboration with FAIR@UiO (https://www.usit.uio.no/prosjekter/fair-uio/);
   • helping with using of package managers (like conda) and containerization (for reproducible and portable programming environments);
   • encouraging the usage of workflows in which one or more tasks involve computations performed in a container;
   • by promoting automated testing of software (to ensure that it delivers what it is supposed to).

3. offering expert support

   • to address issues which are often blocking for novices (and overwhelming for those getting started);
   • putting in place a mentoring program for closer interactions with users who would have answered a call (possibly once or twice a year);
   • gathering examples for demonstration of what can be achieved and advertising “success stories”;
   • counseling on software licences.