An update from Oxford

Michael Popham is Head of Digital Collections & Preservation at Bodleian Libraries, University of Oxford

It has now been six months since the Polonsky funded “Digital Preservation at Oxford and Cambridge” project (www.dpoc.ac.uk) officially came to a close, but the impact of this work is still causing ripples across both organizations.

Within the Bodleian Libraries at the University of Oxford, we have been seeking funding to support a number of business cases created as a direct result of recommendations arising from the work of the Polonsky Fellows. The digital assets in our care have been acquired over an extended period of time (three decades or more) and are extremely varied: consisting of digital images and textual transcriptions of items in our physical collections, research data and outputs, born-digital archival deposits, databases used to catalogue discrete collections of specialist material, and assorted A/V files (created for even more assorted reasons), employing almost every technology and file format that has been popular over the past 30 years. As the Bodleian Libraries seek to collect and create ever-increasing amounts of digital data, the scale of the challenge we face is growing exponentially.

We have no immediate plans to acquire a commercial off-the-shelf digital preservation system, and in the meantime prefer to rely upon open standards and solutions to hold and manage our digital treasures. To this end, we’re working with our peers in the preservation community to develop approaches like the Oxford Common File Layout (OCFL) as an application-independent way of describing the storage and layout of versioned digital objects we want to preserve; colleagues have produced a useful introductory article on this topic and readers are warmly invited to help shape the development of this specification.

To actively monitor the health of our digital assets, we have begun to build and implement a collection of micro services which combine the functionality of a number of existing applications which we already employ, to take regular snapshots of the files in our digital collections. The results of such monitoring are combined into a suite of reporting tools and dashboards which give collection managers easy-to-understand summaries of the state of their data, and can be configured to alert them to any changes which might be unintended or undesirable. Combined with a robust and rigorous approach to backup and archiving, these micro services can be used to identify issues as they emerge, and so give us the time to take any necessary mitigating actions.

Whilst growing quietly confident about our ability to face the future (and very much hoping that pride doesn’t come before a fall), we are conscious that we cannot neglect the wealth of digital assets which we have already acquired. Many of these take the form of aging websites and associated applications, which – sometimes to our surprise – have been quietly ticking over from one year to the next with a minimal amount of care or intervention. Some of these sites present few technical preservation challenges (e.g. simple html webpages with embedded images) but even though we now have the infrastructure and applications in place to redeliver the image content via a IIIF server, it’s much harder to decide what to do with, say, the accompanying academic commentary – too big (and specialized?) to be usefully captured as metadata, potentially too valuable to be simply taken down, and so taking an archival snapshot of the website is an obvious way forward. This should at least enable us to preserve most of this data whilst we assess each site and enact a suitable preservation plan. We will undertake this work in collaboration with colleagues from BEAM (Bodleian Electronic Archives and Manuscripts), who already provide a web archiving function on behalf of the library.

Perhaps the biggest preservation challenge that remains to be confronted is the one which we share with so many other members of the DPC: achieving genuine and lasting cultural change within our institution, and recognition that the most challenging aspects of digital preservation are often about changing people’s minds and attitudes, rather than the application of new technologies and infrastructure. It is relatively(!) easy to persuade colleagues of the need to invest in resilient preservation infrastructure, or the urgency of developing and implementing tools to support digital preservation functions – but securing the resources to develop and maintain training materials and documentation, to collect and disseminate information about good preservation practices, to ensure that our colleagues across our extensive institution all know what they need to know about digital preservation – well, that’s proving a little more difficult. Sustainably supporting soft skills is a hard problem but if we cannot find a resolution it might put at risk all our investment in other areas of digital preservation.

p.s. Any readers who are keen to know more about the wrap-up of DPOC are very welcome to drop by our poster presentation at the forthcoming iPRES Conference.