Fading away: the problem of digital sustainability

The recent past is littered with unsuccessful attempts to store information for the longer term. The problem is technology changes incredibly quickly. Some readers will remember microfiche - small sheets of film viewed through a large light box - at libraries. How many computers have a slot for 3.5 inch floppy disks any more? And forget about 5.25 inch disks. Regardless of the format you are currently storing your precious digital baby photos on, they are likely to be unreadable in 10 years time.

But while your digital baby album may be a personal priority, the issue of digital sustainability is a massive problem on a worldwide scale. Consider the necessity of maintaining crime scene evidence (increasingly stored digitally as images or instrument readings) for a statutory period of time, or the far more serious problem of long-term records of nuclear waste.

On a more mundane level, government records, articles in academic online-only journals and the data that was gathered in the name of research all need to be looked after for an extended period of time.

There is a new emphasis on the reuse of data, with the perception that the initial government investment in research should not be “single use”. Potentially the same data can be used by other researchers for different research later down the track or to check the accuracy of analysis. The changing focus of research means that while information today is usually collected in digital format, previous data collected in analogue form needs to be digitised for preservation or analysis.

One group, based at ANU, is taking the issue of digital sustainability very seriously. The Australian Partnership for Sustainable Repositories (APSR) is working towards a future where all major public institutions have a digital repository - not only as their back-up plan, but as a way of enabling research.

APSR was established in 2002 with a brief to share software tools, expertise and planning strategies; participate in international standards; maintain a technology watching brief; and support national teaching and research with technical advisory services.

A partnership between the University of Sydney, University of Queensland, ANU and the National Library, APSR has developed several practical open source software tools to assist the use and running of repositories.

The data problem

A UK report released last year showed that less than 20 per cent of UK organisations surveyed have a strategy in place to deal with the loss of or degradation to their digital resources.

Of the large amount of data collected in academic contexts, much of it is unavailable after it is used for analysis. This is partly because it may be sitting in a box under a researcher’s desk or because it is in a format that only an old computer can read. Indeed, as Margaret Henty and Kevin Bradley discovered when interviewing academics for a survey on data habits, some researchers keep old computers in their offices so they can read old disks that are storing data from previous research.

But this does not have to be the case. There are data archives which look after data in a sustainable way.

The Australian Social Sciences Data Archive (ASSDA), based at ANU, was set up in 1981 with a brief to collect and preserve computer-readable data relating to social, political and economic affairs and to make the data available for further analysis.

Great care was put into the Archive when it was established, explains Margaret Henty, the National Services Program Coordinator at APSR. “The use of internationally recognised standards has meant that data can be exchanged and sustained with the minimum of intervention, even though the hardware and operating systems have changed over time,” she says. “This emphasis on standards has been maintained, and the ASSDA has been an important contributor to international work in this field.”