Software Archiving

After raising this topic in PLANETS and us evangelizing at the OR2010 it was great to see quite a bit of practical approach to software archiving at Archives New Zealand.

When working in the area of long-term access to digital objects (and when your objects are anything more complex than your basic image file) a software library containing the original creating and rendering applications becomes inevitable. Many digital objects are not properly accessible with any modern software packages such as WordPerfect, WordStar or different types of database files. The first step when accesing older digital objects is to get an idea of their intellectual value or relevance by opening, for example, text documents in some editor or text processor. This may be of some use but the rendering results are (far) from the original: The number of pages can change significantly, garbage is added to the page content, footnotes are moved to the end of the file, formatting is completely lost. If this was what memory institutions served up as authentic versions of older files, it would leave the reader with a number of serious doubts regarding the authenticity as it is not easily provable that the original document wasn’t altered in a malicious way, that all original content is there and that significant meaning wasn’t lost.

This is much more serious for database files. It is sometimes possible to load them to a text or hexadecimal editor. But the displayed results depend heavily on the content of the database cells and often are not easy to interpret. And, in most cases all the relational information of the data contained is lost completely. Those are only two examples tested at the National Archive of New Zealand on born digital content from their holdings and holdings of contributing agencies.

Thus, for a number of reasons the Archive is investigating the feasibility of setting up an Software Library/Archive of relevant applications and has begun to acquire copies of relevant software while the wider implications of such a library (such as legal provisions) are being explored. As many of those applications could not be reliably executed on todays operating systems the relevant operating systems are archived with them too. Beside this certain additional software components are needed like the hardware drivers to install the operating systems to original reference hardware or to emulators like QEMU, VMware or VirtualBox. More items might be required in the library if the relevant digital objects of interest are compressed or combined into a package. E.g. the Archive holdings contained 39 Apple HFS formatted disks which itself contained the relevant material compressed with StuffIt, a popular but now nearly extinct archival format.

Nevertheless, there are a number of practical and legal issues are still to be solved.


Leave a Reply

Join the conversation