The Sheridan Libraries at Johns Hopkins University (JHU) have launched the first US university-based open source programs office (OSPO). The JHU OSPO represents a foundational effort to develop a model for universities to engage more effectively with the open source community and to foster greater social impact through partnerships with individuals, community centers, companies, and government. Furthermore, it represents the framework from which universities can build upon the university pillars of research, education, and translation, especially as it relates to social impact. The JHU OSPO is working with other key players in the university context such as the Provost’s office, research administration, academic departments and institutes, and technology transfer. This piece outlines the important, unique role the OSPO can play in the broader university landscape.
While the OSPO is a widely used construct in the tech sector as evidenced by the TODO Group, the use of OSPOs in other contexts such as universities and governments is nascent. From an operational perspective, there are many similarities for a university OSPO. JHU’s OPSO manages the university’s enterprise GitHub account and has begun the process of migrating existing GitHub repositories into this enterprise account. Doing so raises the possibility of sharing common tools, processes, etc., developing new ones (e.g., GitHub actions), and lowering costs throughout the university. JHU has also signed an agreement with Bitergia to develop analytics and dashboards. Through a grant from the Alfred P. Sloan Foundation, the JHU OSPO is working with Duane O’Brien from Indeed.com and Alyssa Wright from Open Collective to implement a FOSS Contributor Fund during the 2021 calendar year, which will result in a playbook and set of tools for other universities. This set of activities will result in an inventory of open source software, an initial set of analytics and metrics, and a means to understand and support the software supply chain dependencies within a university.
JHU’s OSPO has been fortunate to receive advice and support from other OSPOs and open source leaders but a university-based OSPO requires a different strategy and implementation for several reasons.
JHU’s OSPO was formed as part of a Provost Fellowship for the author of this piece with a focus on open scholarship. In the university sector, a great deal of attention has been paid toward open access to articles. As part of his Fellowship, Choudhury explicitly addressed the role of open data and open source software as primary research objects in their own right. This emphasis is especially important in the context of reproducibility. There is growing recognition that even with open access to articles and data, for certain disciplines, the software is essential to properly interpret and reproduce results.
The ecosystem and infrastructure (including metrics and rewards) for articles has been established over decades, if not centuries. There has been progress toward rebalancing the relationship between authors, universities, libraries, and publishers though much work remains to be done. This established set of players and associated infrastructure experience a certain inertia that can result in friction when changes are proposed or implemented. While open data is different in important ways, the explicit connection between articles and cited data leverage existing connections within current ecosystems and infrastructure (e.g., publishers hosting data as supplemental materials).
Open source software offers a unique opportunity to build upon an ecosystem and infrastructure without prior encumbrances or constraints. Many university researchers who are familiar with choices for managing and curating articles and data will readily note that they do not have such knowledge when it comes to managing and curating software. While existing efforts such as the CHAOSS framework represent promising pathways to new metrics for open source software, only recently have there been efforts to extend them to the academic community. This current knowledge gap creates an opportunity to establish new policies, processes, workflows, etc. without being held to prior constraints. This opportunity is especially important when one realizes that the relationship with the private sector for open source software is also fundamentally different. While the relationship between authors, universities, libraries and publishers is perhaps neutral at best or even contentious at worse, there is no such tension when it comes to open source software.
Technology companies such as Microsoft and Google produce open source software and foundations as the Linux Foundation and Eclipse Foundation manage open source software that is used extensively by universities. In a fundamental sense, the private sector has led the effort to embed open source software into the fabric of our society in a manner that is useful to universities. If anything, universities have been largely insulated from this movement even though there are almost certainly supply chain dependencies and untapped opportunities for contributions from universities. There are noteworthy efforts such as the US Research Software Sustainability Institute but it is worth noting that all of the principals, senior personnel and advisory board members have academic backgrounds or affiliations. Even if universities and corporations wish to work together, the often opaque, complex nature of universities can be challenging. University based OSPOs can become the organizational API on multiple dimensions between universities, the private sector and government. The OSPO++ working group has become a valuable network and community focusing on building such partnerships.
The JHU OSPO is the operational entity for the Institute of Applied Open Source, which represents a model for building educational programs and workforce development related to open source software. In partnership with JHU’s Department of Computer Science, Microsoft, and MossLabs, the Institute will launch a series of “semesters of code” starting in the Fall 2021 semester. Building upon hackathons and Google’s summer of code, the semesters of code represent longer, immersive experiences for students to work on open source software using industry best practices that address university priorities, especially relating to social impact. The initial set of projects at JHU focus on priorities such as public access to research, citizen engagement and empowerment through digital municipal services, precision medicine, and understanding challenges to democracy.
As an example, the JHU OSPO has been partnering with the City of Paris for further development of Lutèce, an open source municipal platform offering hundreds of digital services to the citizens of Paris. The St. Francis Neighborhood Center, a local community center in West Baltimore, wishes to use Lutèce for digital services as part of its smart center strategy. By including Lutèce as one of the platforms for semesters of code, the JHU OSPO will be able to energize students who wish to make a social impact, learn about open source software development practices, and better prepare and position themselves for post-college careers. By working directly with open source leaders in the tech sector, these students will also learn the non-technological aspects of open source including mentorship, teamwork and community engagement, which resonates with the idea of open source as a “verb rather than a noun” – a powerful means of cooperation and way of working together. As the Institute evolves, the JHU OSPO will play a key role in identifying open source projects from within and beyond JHU and engaging additional partners including students at other universities, community organizations and government offices.
The OSPO embraces the opportunity for open source software as an additional pathway for translation and impact. Often in a university context, translation is directly related to technology transfer or commercialization, particularly since the passage of the Bayh-Dole Act in 1980. This method of translation remains important but it is not the only method for impact. University researchers have identified a desire to explore additional pathways specifically related to open source software. As mentioned earlier, these researchers are not as familiar with the related choices available for such pathways and impact.
One related point in this context is the institutional home for a university OSPO. At JHU, the OSPO has been launched within the Library, which is considered a neutral academic division serving all other divisions. At other universities, it may be appropriate to house the OSPO in a the research administration or technology transfer office. The key consideration relates to the scope and type of impact being sought. The Library OSPO emphasizes the curation of open source software as a research object in its own right and the support of educational programs such as semesters of code. Even in this case, it is critical to work closely with other units that touch upon open source software such as the technology transfer office.
Regardless of the arrangement within a university, the OSPO represents a powerful new pathway for translation. The City of Paris and the JHU OSPO have partnered under the auspices of the open source license for Lutèce. If both groups had to secure a grant or sign a memorandum of understanding, it is certain that the administrative handshake would take considerable effort and time. Another case of new translation involves an open source software at JHU (Public Access Submission System or PASS) and the Eclipse Foundation. The JHU OSPO led the effort to become Eclipse Foundation members to explore the possibility of offering PASS as an official Eclipse Foundation product. In addition, multiple researchers at JHU have expressed their interest in joining Eclipse Foundation working groups to advance their research interests.
Perhaps most importantly, 2020 has been a challenging year on so many fronts ranging from the pandemic to racial injustice to economic duress. Universities are rightly being called upon to make direct, substantive impact on addressing these and other challenges. While the OSPO features an important operational role, its greatest potential in a university setting relates to its central role in a growing set of partnerships related to the open source way as a means to address the biggest challenges of our time.
Sayeed Choudhury is the Associate Dean for Research Data Management and Hodson Director of the Digital Research and Curation Center at the Sheridan Libraries of Johns Hopkins University. He leads the University’s Open Source Programs Office. His full bio is available here.
Acknowledgements: Many people have offered invaluable advice and support to the JHU OSPO but it is important to note the role of MossLabs in particular. The JHU OSPO journey began with a seemingly innocent conversation with Jacob Green of MossLabs and has bloomed in amazing ways. It was Jacob’s inspiration to launch an OSPO at JHU.