Winchester, MA The DuraSpace 2014 Sponsorship Summit took place at the Cosmos Club, founded in 1878, in Washington, DC March 11 and 12. The historic venue provided a good setting to explore how open source governance and funding models could ensure the future of DSpace, Fedora and VIVO projects and communities.
The DuraSpace 2014 Sponsor Summit was focused on providing stakeholders with details about 2013 accomplishments and key strategic organizational priorities for 2014. DuraSpace CEO Michele Kimpton and board president Paul Courant opened the 2014 DuraSpace Sponsor Summit meeting with a review of the full agenda and introductions to members of the DuraSpace board who were in attendance.
Meeting documentation including session notes and slides can be found here.
Courant applauded DuraSpace for their efforts over the past two years to bring in diversified revenue streams to support and sustain the organization beyond grant funding. Courant also noted much effort was spent in the last year to expand the Board to from six to eleven new board members. One significant change was to appoint one member of each steering group to the board. This would provide better representation of the project interests.
Kimpton gave an update on key strategic work for 2014 that includes "building a sustainable business model for the organization". A large part of the meeting was devoted to discussion and feedback on a proposed membership model designed to connect direct project governance with levels of membership as a way to deepen community engagement and commitment.
DuraSpace revenue streams were further diversified in 2013 with a 70% increase in sponsorship support. This increase came as the DSpace and Fedora communities recognized that the code base was more than ten years old and in need of modernization. Redesigning Fedora’s proven architecture to serve the community for the next 5-10 years is already far along with the development of Fedora 4, which is scheduled for a beta release at OR2014. Fedora fundraising took off in 2013 as excitement about the redesign of the product was leveraged by steering group leadership. Kimpton noted that 188 DSpace and 21 Fedora repositories were added in 2013 to illustrate that global use of these platforms continues to increase. Currently there are more than 1800 DSpace and Fedora installations worldwide. A key issue in 2014 will be to engage with the broader community of global users, particularly in developing countries rapidly adopting DSpace.
Establishing project steering groups with direct oversight of the Fedora, DSpace and VIVO projects in 2013 was a key accomplishment. The steering groups are at different stages of development, but the general direction is to tightly engage stakeholders by connecting fundraising and project governance to direct community participation in project leadership.
Kimpton also noted successes in 2013 for DuraSpace services. The recent launch of DSpaceDirect was aimed at smaller institutions in need of simple and cost-effective preservation and archiving of their content. The integration of DuraCloud with Archive-It paves the way for computer services that could be run on arc files in the cloud providing a research opportunity for ongoing analysis. The recent Net Plus Internet2 partnership with general adoption and availability will expand connections with university CIOs. The DuraCloud and Chronopolis integration will enable a Digital Preservation Network (DPN) node at the University of San Diego.
Introduction to the New DuraSpace Membership Model
Jonathan Markow, DuraSpace Chief Strategy Officer, reviewed the evolving membership model for 2014 while reminding attendees that the next day’s sessions would be focused on deep discussion and gathering feedback on this proposal.
"We need to bring more members to our organization," he said, “Membership makes more sense in different cultural contexts than sponsorship does and university library budgets often allow for memberships as opposed to sponsorships.” The membership concept is more consistent with Duraspace’s evolving governance models because it "affords members more of a say in the projects". These are some of the reasons that led DuraSpace to propose a formal governance structure around a membership model.
“Because we don't want to rock the boat with current sponsors we have kept the same levels and have instituted a package of benefits at each of level of membership that are meant to entice folks to come in at higher levels,” Markow explained. The key reason to "join" DuraSpace goes beyond benefits–it is in the mutual interest of both the DuraSpace organization and stakeholders for them to participate in sustaining the projects.
Fedora, DSpace and VIVO updates from project Steering Group members followed.
Fedora Update: Community; Software; Sustainability
Rob Cartolano, Chair of the Fedora steering group and Associate Vice President for Digital Programs and Technology Services for Columbia University Libraries/Information Services, began by clearing up any lingering identity issues around the current work on Fedora 4. In the early days of organizing interest around re-architecting Fedora the effort was named Fedora Futures. Cartolano said, " Fedora Futures = Fedora". They are now one and the same.
“Reimagining Fedora” began because stakeholder institutions came together in 2012 to address the fact that Fedora was not being developed sufficiently to meet current institutional needs. DuraSpace was also not collecting enough Fedora project sponsorship revenue to sustain adequate development.
There are more than 300+ Fedora implementations and only 41 Fedora Sponsors. Project goals include sustaining a process for developing software and building community in the near and long-term.
A beta release of Fedora 4 is due out this summer at the Open Repositories Conference, June 9-13. The team is engaged in a dynamic, iterative development process who are committed to Feodra 4 as a transformational product.
Eddie Shin began work towards a new Fedora with rapid development phases called "sprints" that continue today under the technical leadership of Andrew Woods (12 sprints have been completed towards the release of Fedora 4). The development process has been open and priorities have been selected by the community and is a transparent process. He reminded the audience that by contributing code, use cases and resources the process gets better. Fedora 4 development is great example of successful teamwork.
Cartolano wondered about what the "steady state" for building a well-maintained and vibrant open source software platform would look like over the long term. The Fedora 4 project is initially funded for three years. To keep the vibrant community process alive the team is beginning to discuss options for addressing future sustainability.
Cartolano sees areas for growth and innovations.There is interest in exploring a Fedora and VIVO collaboration. Dean Kraft suggested in comments that the Linked Data for Libraries project will bring VIVO closer to where DSPace and Fedora "live".
DSpace Update: A Year in the Life of DSpace
Stuart Lewis, DSpace Steering Group and Deputy Director of Library and University Collections and Head of Research and Learning Services at University of Edinburgh, gave a presentation focused on the unique DSpace process that interweaves software development with a diverse global community. Currently the DSpace community is focusing on sustainability and governance issues as they relate to an emerging common vision.
After twelve years DSpace continues to evolve with 59 developers from many countries. Each release of DSpace offers the community new features gathered from many contributors. Using GitHub unlocked code contributions from a large number of new committers.
Installations of DSpace continue to grow–currently to almost 1600 users from all over the world. Only 1/5 of the DSpace installations are in North America where most DSpace sponsorship revenue is collected. Developing strategies for increasing DSpace sponsorship contributions outside of North America is a key task for the newly-formed DSpace steering group.
Recently the DSpace community was asked to participate in a DSpace Vision survey to come up with a unified vision for how DSpace might develop in the future. Early ideas include:
1. A focus on modern IR use case
2. Software should be lean and agile
3. Core software functions should be extendable
4. Should be configurable to interoperate and support digital scholarship as it evolves
5. Low-cost hosted solutions and deployments should be encouraged
A discussion of the community vision survey results were scheduled for the following day. Next steps include developing a roadmap based on a common vision, engaging the community, and development/deployment.
VIVO Update: Software–Data–Standards (Ontology)–One Community
Dean Krafft, VIVO steering group and Chief Technology Strategist for Cornell University Library, offered a simple VIVO explanation. VIVO connects scientists and scholars with and through their research and scholarship and is the only standard way to exchange information about research and researchers across diverse institutions. VIVO software is built on VITRO and has had 8 releases in 4 years.
VIVO takes complex inputs from multiple institutions and normalizes them to make sense of authority and connections among resources and people.
Morning sessions concluded with table break-out discussions focused on discussing what have you heard, what concerns you, what inspires you, and what do you have questions about?
Responses were positive with regard to news about increased DuraSpace revenues and initial feedback regarding the change from “sponsorship” to “membership” and growing excitement about deep community engagement in the process of developing Fedora 4. Participants saw opportunities for leveraging the “portfolio” of Fedora products and services with regard to the use of common languages, standards, and knowledge bases. More collaboration opportunities with DuraSpace were desired such as identifying and applying for grants.
Afternoon sessions included a review of national initiatives contributing to the evolving research and scholarship ecosystem.
Introduction to "Research and Scholarship Ecosystem Evolution"
James Hilton, DuraSpace Board and University Librarian and Dean of Libraries at the University of Michigan opened his overview of the Research and Scholarship Ecosystem with a diagram of the ecosystem of scholarly digital content and the associated emerging software stack. He suggested that the diagram is not that exciting because most attendees had seen it previously. He explained that there will be no more 20K payments from stakeholder institutions for anything until how it all fits together is understood. He reviewed the stack and had added VIVO as part of supporting code base.
Integration and interoperability are still on the table as unresolved connectivity issues. As a community Hilton suggests that we should be focused on broad infrastructure development, not on one-off baubles. "Resist the siren song of tight integrations to get to the shiny thing; look at it strategically for the long-haul," he said.
The nature of how we use digital content in almost any context is that we take something and do something else with it. The days of faith-based assertions about the value and impact of this exchange are over. We need to back up assertions with data.
Hilton advocates for loose coupling of technical components to make it possible to establish workflow exit strategies.
Digital Public Library of America (DPLA)
Dan Cohen is the Founding Executive Director of the Digital Public Library of America (DPLA). He introduced the DPLA as a portal for discovery, a platform to build upon, and a strong public option. All DPLA collections are geocoded, content types are collected in content hubs that retain retain digital objects and holdings consist of normalized metadata. DPLA cultivates a peer-to-peer relationship with libraries in trying to gather content that represents "the whole range of human expression".
Service hubs are like DPLAs related to their native geographic areas. He suggested that a pond (collection) > lake (local service hub) > ocean (DPLA content hub) metaphor represents the relationship between service hubs and DPLA content hubs.
The DPLA is one year old and currently holds about 6 million items from 1200 contributing institutions. Content hubs have shown increases in use of the materials over the last year. DPLA provides stats on usage to partners.
The DPLA goal is to partner with everyone.
Coherance at Scale Committee
Chuck Henry, DuraSpace Board member and president of CLIR, discussed the Committee on Coherence at Scale founded by CLIR and Vanderbilt University to foster strategic thinking about how to more rigorously manage the transition from analog to digital in higher education.
There are several large-scale projects being built such as DPN and SHARE. Most facets of higher education interests are covered by one of these projects. Is it useful to look at infrastructure support for a coherent set of these projects.
An Introduction to SHARE
Elliot Shore, Executive Director of ARL, introduced SHARE, (the Shared Access Research Ecosystem) which is an initiative spearheaded by the higher education community. SHARE seeks to ensure that the outputs of scholarly research are preserved, discovered and built upon in a manner which facilitates and accelerates the research process. The Association of Research Libraries (ARL), the Association of American Universities (AAU) and the Association of Public and Land-grant Universities (APLU) have partnered to develop SHARE. Management of SHARE is overseen by a steering committee drawn from the leadership and membership of ARL, AAU and APLU.
One of the first activities of SHARE is the creation of the SHARE Notification Service to make it substantially easier for all parties to manage and oversee the reporting of research results. When complete, the Notification Service will notify any interested stakeholder – at no charge – of the release of research results. It will also be possible to query the Notification Service. “Release of research results” encompasses publication of articles, the dissemination of research data, and the posting of materials in repositories and data archives (SHARE project plan).
Shore asked the audience, “How can we pull this off?” He said that college and university presidents are now hearing about large-scale digital initiatives (DPLA, DPN and others) as parts of a coherent system from many angles. All of these initiatives could be brought together by SHARE.
The SHARE notification system will be looking at research release results initially. This system could live in DuraSpace or several places and it leverages the investments we have already made.
Are there enough pieces out there now that we could begin to build this coherent world now? Digital content/scholarship has been around for about 20 years. SHARE is a system that is being built to work together and could be the glue that holds all of these infrastructure elements together.
Digital Preservation Network (DPN)
James Hilton concluded the overview of "Research and Scholarship Ecosystem Evolution" with a status report on the Digital Preservation Network (DPN). There are ongoing outreach efforts aimed at gathering resources and presenting plans to the Association of American Universities (AAU). The goal is to make sure that the college and university presidents show up to learn more about DPN because they believe in the mission to ensure that the complete scholarly record is preserved for future generations.
DPN has one employee plus technical partners and one strategic partner–DuraSpace–who are working towards developing "a digital preservation backbone for the academy."
Hilton reminded the audience that preserving something "forever" is an untested construct. The technical deliverables reflect a focus on issues related to change over time that include scaling, “brightening” preserved content (making it accessible) is non-trivial, and the fact that the DPN system preserves whatever is deposited, but does not change it.
A limited beta release of DPN will be available at the upcoming CNI member meeting.
Summary of discussions from Day 2 Project Groups
Tom Cramer, Fedora steering group and Chief Technology Strategist and the Associate Director of Digital Library Systems and Services for the Stanford University Libraries, offered a summary of what was discussed during the project’s break-out session. It’s been 18 months since the revolution that began the community push towards re-architecting Fedora into Fedora 4. In one year Andrew Woods has come on board as the project’s technical lead and David Wilcox is now the Fedora product manager.
Notable points in developing a deeply engaging community-driven software development process:
- Andrew Woods has said, "All Fedora 4 development is community-based, and all of the use cases are community-based"
- Collecting ongoing use cases from the community is critical
- Acceptance testing is key to developing a viable product
- We need more members
- The Spring 3.7.2 release which will be the last in the 3.0 series
- Fedora 4.0 beta will be out this summer
- The Fedora 4.0 production release will be out in Fall of 2015 as an MVP
- "Launch and support greenfield Fedora 4 projects" will be the first implementation priority
- Maybe training is going for be required as an opportunity to get people exposed to Fedora
Where does the rubber meet the road? There is agreement on many exciting and gratifying aspects of the Fedora 4 process:
- We love the community process
- The process has helped to “sell it locally”
- There are levels of engagement across the community
- Process is Fast and inclusive
- We love the new features especially the Triplestore and RDF
Main concerns center around migration of Fedora 3 to Fedora 4.
Interesting future directions may include a look into something like FedoraDirect (similar to DSpaceDirect), a VIVO collaboration and a continued focus on establishing governance and a membership model that works for the Fedora community.
Deborah Hanken Kurtz, DSpace Steering Group and Director of the Texas Digital Library (TDL) offered a summary of DSpace discussions. Membership and governance and concern about diversity especially those who can contribute code but not resources were key topics of conversation.
Several good ideas for the future included brainstorming ways to reach out to DSpace users in Japan along with coming up with a lower level of membership that could include in-kind and low fees.
Tim Donohue, DSpace tech lead, reviewed “How DSpace Works”. The DSpace Vision Group, DSpace Ambassadors, DCAT and committers all play a role making sure that the annual release of DSpace is both feature-rich and reflects community priorities.
DSpace Vision Survey results showed that 60% of people who responded had migrated to 3.0 or higher and 20% of the respondent installations are from the US. Casting a wider net to get feedback from international DSpace users was mentioned as being a key factor in moving forward towards a coordinated vision.
The group agreed to continue meeting to work towards presenting a DSpace Roadmap during OR2014.
Mike Conlon, VIVO steering group and Co-director of the University of Florida Clinical and Translational Science Institute, and Director of Biomedical Informatics, UF College of Medicine reviewed the VIVO roadmap that included several related grant activities underway to raise additional development funds.
He reiterated that VIVO is very flexible software but “in the end people are there because the want to share information about scholarly work.” There is an ongoing need for VIVO documentation to get additional people onboard with the development process. The Linked Data for Libraries Project is an opportunity for VIVO to reach out to a group who will understand VIVO’s RDF capabilities.
The core of VIVO is in "sharing" as both a model and a means. Promoting and implementing methods and messages for different types of adopters is part of the way forward for VIVO.