For many years, Australia and New Zealand have been at the forefront of developments in digital preservation. As digital preservation is core business for all archives, ICA 2012 (International Council on Archives, Brisbane) presented an excellent opportunity to compare notes on different approaches. In what follows (and the next post) I shall be mixing and mashing notes from three plenary and breakout sessions with those from a full-day’s dedicated digital preservation workshop on Friday during which the attendees did a SWOT analysis of different Australasian approaches to digital preservation. – by Inge Angevaare
Setting the scene: the Australian National Archives
Michael Carden of the Australian National Archives (NAA) expertly set the scene for the discussion of Australasian digital preservation approaches, taking his cue from the 2002 Green Paper which defined the concept of “performance” as the “thing” to be preserved. NAA have chosen to “normalize” all incoming file formats to standards-based open file formats and developed the Xena software to do this. The so-called Digital Preservation Recorder (DPR) manages this process and leaves an audit trail – all details to be found in Carden’s informative full paper. Carden asserted that digital preservation is “core business” for archives, and therefore not something to be outsourced. All of the technology is developed in-house and made available to the community in open source, “which gives transparency to the process; others can check us.”
Carden was very confident in the NAA’s ability to deliver authentic and reliable records. He mentioned scaling as the major challenge ahead. During the workshop, another possible weakness was mentioned: the NAA is a preservation-only system. Very few people have access. This makes the system quite secure, of course. I also learned that it was a matter of funding: funding was only provided for a preservation system. So far that is not a problem, because the records in the system (from government agencies that have been decommissioned and some private records) are not yet eligible for access (due to the legal transfer period 20 years, which equals the embargo period on access). But, eventually, more funding will have to be secured to build an access module. Also, the original NAA design deals only with records that have been legally transferred to the archive. As we shall see below, thinking has meanwhile evolved into a more proactive direction.
Victoria Public Records Office: “The challenge is to get the records”
Andrew Waugh of the Victoria Public Records Office (PROV) readily admitted that his institution’s experience “stands on the shoulders of giants”, referring, a.o., to the National Archives. But the PROV system was designed to provide access right away, in compliance with the legislation of the State of Victoria. PROV, like the NAA, normalizes the records it receives, while keeping a copy of the original. However, access is preserved only to the preservation format copies for the indefinite future – “if rendering engines are not available, we will either build our own or migrate the long-term preservation formats.” Adding that “We do not consider that this will occur for many years,” echoing what David Rosenthal from LOCKSS has been arguing for some time in the library community.
PROV is built on the basis of commercially available, but fully compliant, components (Centera). It seeks to integrate physical and digital records into one system, as “a record is a record”. According to Waugh, one of the major flaws of the OAIS Reference Model is that it looks at digital records in isolation and does not make room for existing systems. The PROV system keeps digital objects and metadata together for fear of linkages being broken.
Waugh drew special attention to the challenge of “getting the records”. This involves making sure that digital records are created and that they are transferred. “Currently,” Waugh asserted, “it is not economic for agencies to deal with archiving.” “We must work with agencies to make it less expensive to transfer records, and to transfer early. This includes dealing with non-mature (temporary) records.” This is a change from the traditional position of archives to deal only with mature records, and it is a change in attitude that was echoed by other speakers (including Hans Hofman). (Sorry, no full paper available for Waugh.)
UK National Archives: “Every decision should be based on usage”
Oliver Morley of the UK National Archives (no full paper available) disagreed with his Australian colleagues about file format normalization. He argued that “digital formats have standardized” by themselves: 99% of the records received have top-20 file formats, and most of those are regular Office formats. Besides, Morley argued, only 25% of what archives keep will ever be accessed, so normalizing everything can be a waste of effort.
Another non-problem for Morley are web archives. “The challenge is not quite as big as I thought,” he said. “Public sector agencies typically produce small websites, and what is published is captured automatically.” Morley does acknowledge that e-mail is a problem. “We have no solution for e-mail yet. There are huge selection problems.”
Here are the principal features of the UK NA’s preservation policy:
So much for now; more (especially on the workshop) in the next post, scheduled for a few days from now. I cannot resist the temptation to go out and (try to) see some of the whales that are migrating past Australia’s east coast during the Winter months.
Categorised in: Uncategorized