This grant supports the development of a publishing workflow for the computational notebooks used by researchers including the American Geophysical Union (AGU)’s community of earth and planetary scientists. Computational notebooks improve the interactivity, reproducibility, and reuse of journal articles by combining text, visualizations, and other elements into narrative documents, but until now they have primarily existed as supplementary materials to traditional journal articles. Grant funds will allow the AGU to to explore ways of integrating computational notebooks with publishing platforms by bringing together representatives from scientific societies, publishers, repositories, and federal agencies for a series of events, ultimately developing a workflow to preserve the interactivity of computational notebooks into their final publication.

As part of a set of connected activities under the banner of the US Research Software Sustainability Institute (URSSI), this grant will support the hiring of a full-time URSSI community manager who will focus on cultivating and growing URSSI’s community while identifying new needs and potential collaborations, as well as providing support with engagement and dissemination activities. This new position will draw on expertise and support from Code for Science and Society and the Center for Scientific Collaboration and Community Engagement, as well as PI Ram’s deep experience growing the rOpenSci community.  

As part of a set of connected activities under the banner of the US Research Software Sustainability Institute, this grant funds an effort by Kyle Niemeyer, mechanical engineering professor at Oregon State University, to develop and run four, weeklong “beyond introductory” winter/summer schools for researchers who want to deepen their software engineering skills. Beyond directly training an anticipated 130 scientific and engineering researchers in sustainable software development, Niemeyer will develop, hone, and release a scalable mid-level curriculum for the training of non-expert scientists in best practices in software engineering and his project will provide useful data on the size of demand for such skills training events.

Software has been critical to research for decades, but only in the past handful of years has it become a topic of broad national conversations about the scientific enterprise. Much of this discussion has centered on the difficulties of maintaining research software beyond initial grant funding, especially when funder support has historically focused on new projects rather than maintenance of existing research infrastructure. Unlike instruments, research software is rarely commercialized, and “sustainability” usually relies on persistent, ongoing engagement by developers cross-subsidized by institutions and other grants.   As part of a set of connected activities under the banner of the US Research Software Sustainability Institute, this grant funds a project led by Daniel S. Katz at the University of Illinois, Urbana-Champaign to collect and analyze research software sustainability policies across a host of key institutions. Katz’s targets are guidelines, rules, and practices related to research software created by institutions such as universities, laboratories, and industry as well as the funders, professional societies, and publishers that exert strong influence on the norms of scientific practice. Katz will then use the collected policies to answer several important big picture questions about research software sustainability, including characterizing the variation of strategies and institutional structures for the support of research software, the alignment or misalignment of such strategies, and funding sources and opportunities for the maintenance of research software.

In most contemporary software development, investments in user experience (UX) design are viewed as highly prudent expenditures that result in products that are intuitive, meaningful, and relevant for users. When it comes to scientific software projects, however, it’s rare to find a UX designer amongst any core set of contributors. Recent research on the topic suggests that greater investment in the UX of a piece of software (especially early in the lifecycle of a project) could have large positive downstream impacts: easier adoption of software, lowered barriers to its use by a broader set of scientists with different backgrounds and levels of training, diminishing maintenance costs, and healthy cycles of user feedback informing how to build more useful software.   This grant funds a project lead by Lavanya Ramakrishnan at the Lawrence Berkeley National Lab (LBL), to help codify how to effectively implement UX design in scientific software projects through the study of LBL’s extensive corpus of UX research on myriad scientific collaborations. Ramakrishnan’s team will use the research to generalize best practices for implementing UX into the development of a diverse typology of software types (i.e., data pipelines vs. instrument control systems), and then prototype a UX design system to be used for common scientific tasks and workflows.