NDSA:Future projects

1) Defining goals and principles, presenting new approaches.

One of the immediate challenges to the NDSA is that there is no clear statement of the goals and principles of Data Stewardship. Without such a statement, the formulation of new approaches to meet these goals while adhering to appropriate principles is difficult. Anything that is too different from current approaches may be met with skepticism and challenged as diverging from tradition. More abstract statements of goals and principles that are independent of current practice are necessary in order to allow innovation to be evaluated in a fair and meaningful way.

An example of such a question is this: Does Data Stewardship have a requirement to store the original sequence of bits (format) that defines each digital object preserved, and the sequence of formats into which this original sequence has been transformed? Is it acceptable if the original or some of the intermediate forms are lost or allowed to deteriorate? Is the preservation of bits sufficient, or is the preservation of functionality in interpreting those bits required? What does preservation of functionality mean?

2) Prototype implementations and procedures.

This suggestion is a reaction to the idea that the other working groups will be working on services and infrastructure that are at least "pre-production", meaning that they have some expected level of service. Ideas that are not yet at the stage of stability that these working groups require would go on within the Innovation Working Group, and so would be clearly identified as falling into a different category. Prototype implementations might attract funding from agencies such as the NSF or from industry or from participating organizations. Such activities might be completely contained within the NDSA or they might be a collaboration with other organizations.

3) Experimentation and analysis.

There are a number of experiments that could shed light on the processes of Data Stewardship and measurements and analyses that could be applied to the body of data currently being preserved in a variety of contexts, from the academic and governmental to the industrial and consumer. For example, there is little public information about failure statistics of disk drives. Industry efforts at evaluation tend to be closely held for reasons of commercial sensitivity. A detailed long-term study of disks in controlled environments would be one way to pursue this information. Another way would be to catalog, monitor and tabulate the track record of disks deployed in the field.

4) Digital preservation X-challenges.

There was a lot of interest in challenging the preservation community to innovate and to coordinate and organize such efforts. It would seem that such efforts might be most productive if our working group came up with specific "X-Prize" type challenges that the community could address. If the working group formulated ideas that were then endorsed by the NDSA as a whole, some of them might be appealing enough to attract prizes or recognition for the winners.

An example might be to organize a huge database (an Exabyte? 10 Exabytes) and keep it functional for a minimal period of time (a month? a year?). Maybe that's an uninteresting example, but it gives the idea of a project that would have to involve the community, would attract attention, and would face considerable technical challenges.

5) Working beyond digital preservation

This could mean collaborating with fields that may not currently be considered part of digital preservation (such as Digital Archeology/Forensics) to apply their techniques. Some areas, such as virtualization of programming environments, may be of fundamental importance to digital preservation but are thought of as part of programming languages and operating systems. Other areas, such as material sciences, have a direct impact on what properties of physical objects are measured and recorded during digitization.