Most of us probably aren't used to thinking about this, but all publications have an expected shelf life. A properly cared-for hardback book will probably outlast its author, perhaps even its author's grandchildren, but at some point it will molder into dust. 3D digital environments and artifacts have lifespans better measured not in generations, but in short spans of years, even months. This means that without great care on the part of digital content creators, their creations are at high risk of disappearing from view entirely. This obviously raises prudential worries, who wants to pour months or years of time and energy into a project that will disappear from everyone's view within months or years? But in cases where projects are funded by grants, or supported by release time by a home institution, or even just potentially significant, the potentially short shelf life of digital projects raises moral worries as well. If one has a moral obligation (of whatever sort) to make one's digital project accessible, then it bears thinking about strategies for preserving 3D digital environments and artifacts.
Here it is worth noting that many of the 3D projects we have mentioned above have become relatively or entirely inaccessible. We referenced Arkansas State University's 3D interaction reconstruction of the Lakeport Plantation in Second Life. It is now entirely offline (and it is only one of a number of ASTATE heritage reconstructions that are now entirely offline). The Virtual Rosewood Museum in Second Life is similarly offline. The website hosting Drama in the Delta, mentioned above, is still live, but you cannot readily enter the 3D environment, because it was constructed to run on a version of the Torque game engine that is now unsupported on most current platforms. The Digital Hadrian's Villa Project, also mentioned above, is still navigable. But since it was built in a dated version of the Unity game engine, there is not much you can do or see there. I could go on, for the number of inaccessible 3D models created in even the last decade is large, indeed.
But reflection on the differing natures of digital projects suggests that this might not be as great a tragedy as it might initially appear. For probably not all DH projects should be preserved into the indefinite future. So for any given DH project, perhaps the first question to ask when it comes to sustainability is "for how long should this project continue to be accessible?"
In general, there is a tradeoff between being on the cutting edge of technology and lastingness. If you published your project 15 years ago using .txt files and .jpg images, it would still be fully available today. But it obviously wouldn't be very flashy or interactive. So once you determine how long a DH project should last (ideally), the next step is to think about how much you are willing to trade sustainability off against goals for your project that can only be accomplished using technology more advanced that .txt or .jpg files, and how far you are willing to push the technology envelope.
For any given project, there is unlikely to be an obvious optimal tradeoff between sustainability and technological innovation. But once you make a choice here, and commit yourself, it is worth thinking about deploying a backup option or options. Most of the DH projects mentioned above (for instance, the Virtual Rosewood Museum, Drama in the Delta, the Digital Hadrian's Villa Project) are still being discussed today only because they employed multiple deployment options.