Should the OAIS standard be a strait jacket or a diving board?
Years ago I heard direct criticism of PREMIS from those close to the OAIS process because it didn’t directly employ the OAIS information model. But now it’s referenced in the OAIS Roadmap. I still feel there are a variety of interpretations of what a ‘reference model’ should do and how it should be ‘implemented’. But more importantly I don’t know of any proposed methods for validating the reference model against systems that claim to implement it in the real world.
The Applicability section notes that the model:
“does not specify a design or an implementation. Actual implementations may group or break out functionality differently”
But does not provide us with guidance on how such alternate perspectives can be mapped from the OAIS, or how existing implementation might be benchmarked against the standard; despite this the OAIS remains the:
“basis for further standardization in this area”
Without a methodology for linking reference model to implementation (and, ideally, from implementation through to ISO16363 audit) the range of functional entities and the data/metadata and messaging transfers between them can’t be analysed in terms of real world practice. Will there be a time when the authors of OAIS can demonstrate that their work has been validated as meaningful, appropriately scoped and the right tool for reference when creating an OAIS implementation?