Trustworthy 100-year digital objects

Abstract

How can an author store digital information so that it will be reliably intelligible, even years later when he or she is no longer available to answer questions? Methods that might work are not good enough; what is preserved today should be reliably intelligible whenever someone wants it. Prior proposals fail because they generally confound saved data with irrelevant details of today's information technology---details that are difficult to define, extract, and save completely and accurately.We use a virtual machine to represent and eventually to render any data whatsoever. We focus on a case of intermediate difficulty---an executable procedure---and identify a variant for every other data type.This solution might be more elaborate than needed to render some text, image, audio, or video data. Simple data can be preserved as representations using well-known standards. We sketch practical methods for files ranging from simple structures to those containing computer programs, treating simple cases here and deferring complex cases for future work. Enough of the complete solution is known to enable practical aggressive preservation programs today.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 29 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.