Strategic Developments in Our Dynamic Open Source Communities

Thu, 2008-09-11 12:47 -- Anonymous (not verified)

By Sandy Payette

It’s been an exciting summer with the release of Fedora 3.0, Mulgara 2.0, and the kickoff of the new collaboration of Fedora Commons and the DSpace Foundation.

In July, Michele Kimpton, Executive Director of the DSpace Foundation, and I stood on the steps of the Library of Congress to shake hands as a symbol of our commitment to collaboration. This is a very exciting development for our organizations and our communities for several reasons. First, our communities will benefit from our efforts to bring the DSpace and Fedora repository systems closer together and make them more interoperable. We have heard from many users that the ideal solution for them would be to have the “best of both” and we are committed to determining the best strategies to meet this requirement. Second, our non-profit organizations will benefit from the strategic alliance we have formed. Both the DSpace Foundation and Fedora Commons are situated in an exciting and dynamic context that is influenced by both technical and social forces. With many new players in the terrain that might loosely be described as “repositories” we believe that together DSpace and Fedora can continue to provide a unique value proposition over the long haul. We both have strong communities, successful open source products, deep understanding of the challenges of digital preservation and scholarly communication, and close relationships with universities, libraries, scholars, and scientists. Working together, we will continue to serve our mission by leveraging new technologies and by developing new strategies for integrating repositories in Web-based knowledge spaces and emerging cyberinfrastructure. We look forward to lots of community participation as refine our ideas and translate them into software development plans. Stay tuned for more information over the next few months.

Fedora 3.0 Milestone

Both the core Fedora development team and committers and contributors from the wider community have done a great job in delivering Fedora 3.0. I consider this an important milestone for several reasons: (1) it streamlines the core Fedora repository service, (2) it introduces the first in a series of new features that will make the Fedora repository integrate better with new and emerging Web technologies and standards, and (3) it provides a simple starting point for creating and storing “content models” that describe different genre of digital objects.

I’ll mention a few things about better integration with the Web. With feedback from our community, we are focused on providing new lightweight interfaces for the Fedora repository that make it much easier to integrate with Web applications. Also, we would like to expose the repository in a manner that promotes interoperability using common Web protocols and formats. In Fedora 3.0, we took our first steps by introducing a new REST interface for managing the repository and supporting ingest/export using the Atom syndication format. In the next six months we will kick off a number of new innovative projects to improve the repository fit within the Web, both from the perspective of front-end access to the repository, and from the perspective of integrating Web storage providers at the back-end of the repository.

Here’s my simple explanation of the new “Content Model Architecture (CMA) unveiled in Fedora 3.0…. Given that you can register a description of what a particular type of digital object should look like (for example a “book” digital object), it’s a lot easier to take the next step, which is to assert which digital objects in the repository conform to the model. With Fedora’s new CMA features, the relationships between digital objects and their content models can be easily recorded via RDF. At a minimum, this enables you to “round up all the suspects” during times when you would like to find all the “book” objects for maintenance or migration. Better yet, you can use content models as the basis for which transformative services can be dynamic associated with digital objects.

Mulgara Semantic Triplestore

Finally, I would like to draw your attention to the new release of the Mulgara semantic triplestore, whose development is supported by Fedora Commons. This week Mulgara 2.0.5 was released. The new Mulgara 2.0 line is a major milestone that continues to advance the state of this very powerful RDF triplestore in open source. With the release of Mulgara 2.0 we now have SPARQL query support, plus an array of new features that greatly improve reliability and scalability. Check out the details at the Mulgara web site.