Preservation Metadata for Software

Software preservation is concerned with the long-term storage of and access to computer programs in order to keep software alive. As archival objects, software presents us with many challenges. The notion of software covers a huge variety of applications that have complex relationships with often specific hard- and software environments that are maybe considered quite standard at the time of release, but over time become obsolete. In order for archived software to remain accessible we need a lot of information about the original environments in which the software worked at the time of release so that it can be reconstructed, either by using original hard- and software or, more likely, by setting up an emulation of the original environment. This report addresses the challenges of how to capture information about the original environment in (preservation) metadata. It will first provide an overview of the types of metadata that one needs to consider when describing software, but also look at standards that will facilitate standardization and interoperability with systems for accessing software collections. It will then zoom in on system requirements: lists of properties that inform the user of software about some basic requirements that the environment must adhere to. In the following two chapters we first evaluate PREMIS as a model for describing system requirements. We then explore Wikidata as a potential semantic database that can be used as a vocabulary. Both of these chapters are exploratory in nature and will critically reflect on the degree to which PREMIS and Wikidata can play a role in describing system requirements.