Digital art dies twice: once when its hardware fails, and once when its intent is forgotten. The central problem of preserving digital installations is that the work exists simultaneously as fragile technology and as an idea, and the two decay at different rates. The field's answer is a structured decision process: first define what makes the artwork itself, then choose among three established strategies—migration, emulation, and documentation—based on what must survive. This article explains that process and the reasoning behind each choice.

The Ephemeral Nature of Digital Art: Core Challenges

Digital artworks are hostage to their dependencies. As the Digital Preservation Coalition notes, software-based art is often behaviourally and functionally complex—interactive or simulation-driven—and carries extensive dependencies on specific hardware, system software, and external interfaces. Compiled code is frequently a "black box," making its internal workings difficult to ascertain. Because these works are typically bespoke projects coded without long-term maintenance as a goal, institutions inherit them as what the Coalition describes as legacy systems: deployed, finished, and poorly understood.

The conservator Alice Watkins, writing for the Institute of Conservation, observes that contemporary works may combine synthetic materials, electronic media, food, digital components, and found objects in a single piece, each degrading at different rates and demanding different treatments. Traditional methods often fail here, and collaboration with specialists becomes routine. The stakes are concrete: without intervention, the work becomes unplayable or unviewable.

Defining the Artwork's Identity: Artist Intent and Significant Properties

Before any technical decision, the question is philosophical: what is the work? Project TRACKS, a Belgian preservation initiative, puts it plainly: artists need to define the essential characteristics of their artworks, because if others can understand the original intention, they can take better-informed preservation decisions. These "significant properties" are not fixed; they are, in the project's account, the result of negotiations with artists and curators and of the work's exhibition history, and they can extend into the present and future.

The artist's own voice matters here. A contributor to Studies in Digital Heritage describes a career-long effort to resolve "this dichotomy between materiality versus intent" in ephemeral, digitally-born work, choosing to vary approaches over time and using both migration and emulation "to preserve my intention." The same essay invokes Jill Sterrett's argument that conservators must remain flexible, making choices that are "open-ended" in keeping with what the artist attempted to convey. Reconstruction, the author cautions, borders on producing wholly new objects.