Building on the success of other publishers using Software Heritage, like eLife and IPOL, Dagstuhl is now implementing this approach for archiving research software.

Collaborating for Open Science

Dagstuhl has been the publishing arm of the Leibniz Center for Informatics since 1990 and ventured into open-access publishing in 2008. They’re best known for their LIPIcs series, which has proceedings from top computer science conferences. But they also publish reports, manifestos, the LITES journal (embedded systems), TGDK (graphs for data management), and DARTS (evaluated artifacts).

Dagstuhl manages its own storage for these software artifacts, publishing them separately with individual metadata and DOIs. This approach ensures the code is accessible and discoverable, aligning with the broader goal of preserving research software.

“Linking software to publications ensures availability and enables citation and reuse by enforcing proper documentation and archiving. For this long-term archiving, Dagstuhl relies on Software Heritage, a non-profit organization supported by UNESCO, which is the world’s leading software archiving service.” —Saadet Bozaci, Dagstuhl

Dagstuhl is participating in the European Union-funded FAIRCORE4EOSC project. It highlights Software Heritage’s growing role in the European Open Science Cloud (EOSC).

Dagstuhl’s work focuses on connecting articles with their related code.
This involves:

• Archival: Automation of software code archival related to articles.
• Referencing: Implementing permanent links (SWHIDs) to code on publication pages. The capability to add an iframe on the publisher’s platform, adding an image.
• Description: Enriching software discoverability through metadata addition.
• Citation: Streamlining of software credit attribution through citation information provision.

Publishing systems are connected to the Software Heritage archive using APIs and connectors.

Where it started

This trend in publishing was initially described in the Scholarly Infrastructures for Research Software report (EU Commission, 2020). The Dagstuhl team—Saadet Bozaci, Michael Wagner, and Michael Didas—played a key role in making it happen. The European Commission’s FAIRCORE4EOSC project (GA 101057264) is funding this effort.

How it works 

Dagstuhl Publishing has expanded both its submission system DSUB and its publication server DROPS to support supplementary material (especially research software) associated with scholarly articles (either contributions to journals or conference proceedings). This submission system provides support to authors, editors, and publishers throughout the scholarly article preparation process.

In the extended workflow for submitting peer-reviewed, camera-ready articles, authors must include related supplementary materials, such as research software. To ensure proper citation, authors should provide the software’s URL and descriptive metadata (CodeMeta standard), including the SWHID.

If the software is not yet archived on Software Heritage, authors are encouraged to request archiving. The system also provides support with pre-filled metadata forms. Metadata, if already available, is automatically read from the hosting platform, such as GitHub, or other sources, such as citation.cff files, and presented to the author in the form for revision and approval. 

Authors have the option to synchronize contributor details with the published article. The publisher integrates this metadata into the scholarly article during publication preparation for accurate referencing and citation. The publishing team verifies the metadata for completeness and accuracy. This curation process ensures critical metadata, like affiliations and funding details, are accurate and current. The publishing team resolves any metadata inconsistencies or fills in missing information. Authors can review and make any final adjustments to the updated metadata and reference before publication. This curation process ensures the article and its supplements have high-quality metadata, properly integrated into the publication. After publication, the reference/citation data is also part of the published metadata of the scholarly article. The metadata will be used in the scholarly article. If it hasn’t been published elsewhere, it will also be published on the DROPS publication server. This published metadata will also be transferred to Software Heritage for long-term availability.

Get involved

If you’re a publisher or organization interested in joining this effort, you can become a Deposit Partner or a Deposit Interest Group Member. Just reach out to dig@softwareheritage.org

The post Inside Dagstuhl Publishing’s drive to preserve code appeared first on Software Heritage.

Abramatic has played a crucial role in EOSC, as part of the Executive Board and as the architecture working group chair. At the end of 2020, the Executive Board produced a set of recommendations for the implementation of the Horizon Europe program. FAIRCORE4EOSC is one of the awarded EU projects throught the Horizon Europe program.

The RSAC (Research Software APIs and Connectors) components developed through the FARCORE4EOSC project intended to improve interoperability between various infrastructures catering to research software. This component has a significant impact on the four pillars of the SIRS report, namely Archive, Reference, Describe, and Cite. The goals are listed below:

Archive: Creating interoperability between the infrastructures to preserve research software artifacts and their metadata in the universal source code archive, Software Heritage.

Reference: Adoption of the SWHID (Software Heritage Identifier) to identify software artifacts accurately and reference specific versions of the software. 

Describe: Exchange metadata about research software using the CodeMeta vocabulary, thereby enabling an interoperable ecosystem.

Cite: Align the citation export formats for research software following the biblatex-software specialised entry types for software 

We have traveled a thoughtful and progressive journey together from the initial SIRS discussions to the practical implementations showcased in the workshop through FAIRCORE4EOSC. This journey reflects our ongoing commitment to improve scholarly infrastructures for research software.

Looking back to our 2023 deposit sprint onsite at the Inria Research center:

Demo-mania: showcasing results from the FAIRCORE4EOSC EU project

The workshop was structured around a series of “Demo-mania” sessions, moderated by Morane Gruenpeter, where service managers demonstrated their integrations and discussed the challenges and opportunities for Research Software preservation.

The workshop was structured into three segments by infrastructure type:

• Scholarly repositories showcased how they facilitate access to and citation of archived software, emphasizing the importance of these resources in academic research.
  • Zenodo, presented by Manuel Alejandro de Oliveira da Costa (CERN)
  • HAL, presented by Yannick Barborini (CCSD)
  • DANS Dataverse instance, presented by Wilko Steinhoff (DANS)
• Aggregators highlighted their role in connecting software artifacts with scholarly literature, enhancing metadata, and improving visibility
  • SwMath, presented by Maxence Azzouz-Thuderoz (FIZ)
  • OpenAire, presented by Serafeim Chatzopoulos (Athena)
• Publishers shared their strategies for embedding Software Heritage archives directly into their publication workflows, underscoring the critical role publishers play in supporting software preservation
  • Episciences, presented by Raphaël Tournoy (CCSD)
  • Dagstuhl, presented by Saadet Bozaci (LZI)

Source code is fragile; it can disappear. It’s important to note that software source code plays an essential role in research and should be archived properly, alongside data and publications. The generic workflow below shows how software built for research as part of the open science ecosystem should be archived, referenced, described, and cited.

Live feedback and collaborative dialogue

Each demonstration was followed by a feedback session, where participants discussed potential improvements and strategic approaches to overcoming existing challenges.

Participants were asked “Which colleague in/outside my institution would I like to show this demo with? Why this person, what are my expectations regarding this person?”

• Researchers and librarians who could focus on software deposit and improving metadata quality, respectively.
• IT developers who could enhance the integration between publications and software, leveraging Zenodo’s capabilities to link DOIs and SWHIDs.
• Various stakeholders, including the SciCodes consortium, CRIS developers, and InvenioRDM managers, are appropriate colleagues who could explore existing workflows and develop further integrations.

These stakeholders could help push for a culture change within their organizations, suggesting roles such as training coordinators or open science charter updates to embed software archiving practices further.

Reflecting on the SIRS Report

We concluded the workshop with an activity to reflect on the progress made since the publication of the SIRS report. Participants used a metaphorical “mood boat” to describe their journey through the recommendations and visions outlined in the report. This session provided a poignant illustration of the community’s journey, highlighting both achievements and areas where more concerted efforts are needed.

This boat represents our journey through the SIRS recommendations/vision. Using the different situations  Inspired by the scenarios illustrated by Mathilde Riou and Vincent Chaillou, participants discussed their experiences and future plans.

The workshop concluded with a consensus on the need for continued collaboration and advocacy to promote the recognition of software as a critical research output. While adoption is still a challenge, the participants committed to taking specific steps toward engagement:

• Assisting researchers in depositing their software, enhancing the connection between publications and software.
• Providing guidelines, documentation, and training for publishing and archiving software.
• Enhancing the visibility and usability of SWHIDs.
• Sharing workflows for other infrastructures to join the deposit partners community.
• Organizing more events like this to help bring stakeholders together.

Looking ahead

As we build on the momentum from this workshop, we invite all stakeholders in the research and software communities to join us in advancing these critical infrastructures. By working together, we can ensure that software, a cornerstone of modern research, is preserved and accessible for future generations.

Acknowledgments

A big thanks to all the presenters, participants, and the organizing team, whose dedication and enthusiasm made this workshop successful.

Special thanks to Jean-François Abramatic for his introduction and to all our partners who continue to support and advance the mission of Software Heritage.

The post Deposit Partners Workshop 2024: Key takeaways appeared first on Software Heritage.