The qualities of digital media are well-known: their agility, portability and scalability, among others. Their one down-side seems to be the result of their usefulness: we can’t keep up the ratio of preservation to advances in technology. A couple of years ago, we attended a week-long workshop on digital preservation at Cornell University, and one of the slogans of this really excellent workshop was “there is no silver bullet” in terms of preservation of digital assets. We came away from that workshop with strong ideas on how to get started and created a campus-wide survey on the types of digitalia being employed and methods used to preserve them.
On November 14 & 15, 2007, David Conners and Diana Peterson attended a workshop held at the new PALINET headquarters in Philadelphia entitled “Stewardship of Digital Assets,” which the organizers subtitled “sustaining digital collections.” About 40 people attended the PALINET workshop from a variety of institutions: museums, public libraries, community colleges, universities, government, and businesses. Haverford was one of the smaller institutions represented; most attendees came from government institutions like NARA, the Pennsylvania and Delaware State libraries, even the Supreme Court of Canada. The trainers included Liz Bishoff, former Executive Director of the Colorado Digitization Program; Tom Clareson, Program Director for New Initiatives at PALINET; Robin Dale, former Program Manager for Preservation at RLG; and Katherine Skinner, Co-Director of the MetaArchive Cooperative.
While the expression “there is no silver bullet” was still uttered, it was not a slogan. In the final analysis, we wrote up a strategies statement based on our assessment of risk of loss in which we listed: increased institutional buy- in for digital preservation; an expansion of the previously formulated survey; an improvement of our backup procedures for digital objects; transfer of digital objects on superseded formats and CD and DVD to spinning disks; and consideration of the future of digital objects that are produced in proprietary formats. We listed the personnel and stakeholders we considered vital in this undertaking and broke down our preservation plan to the level of implementation. Finally, we projected financial needs and potential resources.
Following are some of the points presented in the workshop that merit notice in this report and some further detail to the strategies statement outlined above
1) A short definition of digital preservation as prepared by ALCTS suggests that it is the combination of policies, strategies and actions that ensure access to digital content over time. Integral to these propositions are such factors as assigning responsibility for digital preservation to staff, technical specifications, reliable master files, sufficient descriptive, administrative and structural metadata to ensure future access and a quality control procedure. In addition are such factors as disaster preparedness, programs to refresh, migrate and emulate digital data, and, of course, sources of funding. One way to get started may be to hire a consultant.
2) We had always thought of back-up of our various digital formats (think: websites, email, data sets, GIS, images, audio, video, among others) as synonymous with preservation, but, while this continues to be of significance, it lacks the necessary aspect of long-term preservation or structural integrity. Best practices also dictate the employment of standards, oversight to curry to new media and a system of LOCKSS (lots of copies keep stuff safe).
3) While we might aspire to being a Trustworthy Repository (www.crl.edu/content.asp?l1=13&l2=58&l3=162&l4=91) that ingests all digital material produced by the college and provides it safe, updated and authenticated harbor until its egress at any point, the cost in terms of personnel and money may be greater than what Haverford can afford. There is a good deal of literature on life cycle preservation for digital media, and a number of institutions that are gearing up to step into this role.
4) Katherine Skinner gave one of the more interesting presentations when she talked about the founding and organization of the MetaArchive Cooperative (MetaArchive.org). The MetaArchive is a cooperative made up currently of Emory, Georgia Tech, Virginia Tech, Florida State, and Auburn. They use a LOCKSS (www.lockss.org/lockss/Home) -based system of servers to deposit their digital objects. If any particular copy of a digital object is compromised on an individual node, the other copies can overwrite it. Two aspects of the MetaArchive were interesting to us. First, the cost to deposit our digital data in the MetaArchive is only $200/year. Second, the MetaArchive Cooperative is interested in exporting its model to other groups, which has already been implemented for a state-wide digital preservation program in Alabama. Is this something PACSCL or PALCI would be interested in?
Next steps: we hope to build on the survey that was circulated across Haverford’s campus, stressing the need not only to back-up critical digital data, but also to engage in best practices for long-term preservation, including providing counseling on practices (safety, security, longevity, breadth, authenticity, reporting) and prioritization analysis. This is all possible as long as there is institutional interest and commitment.
Diana Franzusoff Peterson and David Conners