Patrick Schmitz of the UnivCollectionSpace logoersity of California, Berkeley, has been a driving force behind the CollectionSpace application, which was built with the Nuxeo open source content management platform. He even gave a talk on the project at the Open Source Business Conf erence in San Francisco, May 2011.  

CollectionSpace is an Patrick Schmitzopen source, web-based software application for the description, management, and dissemination of museum collections information. The CollectionSpace project team is made up of museum professionals, software engineers, and interaction designers.

It’s a compelling project, partly because it fulfills a truly unmet need for affordable software in small- to mid-size museum settings. In addition, it’s (yet another) interesting use case for a content management platform. 

The Interview

Tell me about your project.

We want to create a cost-effective software solution to support collections from art history to zoology. Museums traditionally have had to cobble together a software solution that is relevant to their domain. Fine arts are one of few that have a lot of money. Other domains have specialized needs, but struggle with the budget to obtain a solution to track and manage their collections.

CollectionSpace is not a single application for single set of users. We are building a framework and a series of applications on top of it. Each application is directed at the same kinds of stuff, but each individual domain has specialized requirements – the metadata are different, the UI is different.

For example, the Phoebe A. Hearst MuSchema_extension_model-1seum of Anthropology at the UC Berkeley campus is an early adopter. They have a certain tracking requirements that museums in other domains do not have. In anthropology, adherence to NAGPRA (Native American Graves Protection and Repatriation Act) requires a very specific kind of documentation in order to comply with this Federal regulation. In CollectionSpace, this is represented by specific metadata around workflows. Once the model for this is effectively established, it can be shared with other anthropology museums and organizations.

How do museums use CollectionSpace?

This is not just about cataloguing collections. The descriptive metadata is only 1 or 2 services we provide, out of several dozen. Other services are used to track metadata around activities, such as loan/borrow, intakes, acquisitions, movement records, transport records, insurance, research visits, cited in publications. There are authority services to represent people, organizations, storage, taxonomic authorities for life sciences, stratographic data for earth sciences, etc.

CollectionSpace is like a records management system - it tracks activity around objects.

Each museum has 4-10 internal users. There is also an interest in exposing this data to the public, because outreach is important in this domain.

Why did you choose Nuxeo Enterprise Platform?

We looked at xml databases, existing museum applications, Alfresco, Fedora, and Nuxeo. We also considered a solution based on ESB / SOA toolkits such as JBoss, Mule, etc.

Nuxeo had a presence, was participating in public spheres, and the messaging around API as a model was important. Nuxeo presented itself first and foremost as platform, and secondarily as an application, and that made a difference for us.

Any comments about the development work you’ve done with the Nuxeo Enterprise Platform?

Nuxeo has nice flexibility around repository and domain configuration. This made it straightforward for us to define multi-tenancy for CollectionSpace, which was an important part of our hosting structure. We have one domain per tenant, in some cases, a repository per tenant.

The current support for database templates has made it easy for us to support multiple database  backends - we recently transitioned from MySQL to Postgresql as the preferred backend, for performance reasons.

Tell me more about Community Source.

This is technical and community-oriented work. We are trying to establish a social foundation for contributing to these models, so that members of different domains can use this platform to build applications that service their particular requirements. We would like to see this happen for many different domains, so that accepted models can be established and re-used.

We call it community source because we have vested partners in it for long run, and sub-communities of interest within this. The project is currently working to establish a sustaining foundation to provide long-term support for the community.

What have been your biggest challenges for this project?

The development teams are spread across different time zones and different organizations, so we have to put some effort into collaborating and coordinating the project. We have people here at UC Berkeley, at the University of Cambridge in the UK, at OCAD University in Ontario, and some people at different museums.

We follow a traditional agile development approach, where we crank out new functionalities sprint by sprint.  A sprint is typically 3-6 weeks long. Defining the scope of a sprint and collaborating with different stakeholders takes a significant amount of time.Yoda

What do you like most about this project? 

I love hanging out with these people from a truly diverse array of museums. They're teaching us about all kinds of amazing things. I got to see some of the amazing collections of Native American artifacts at the Hearst Museum, I get to learn about dinosaurs, insects, and mammals with other campus museum experts, and saw the original Yoda at MMI (Museum of the Moving Image) in New York.


-- @JaneZupan, who is now planning a trip to NYC to visit to the MMI