From the archives: Preserving virtual worlds

(I originally published this article on on February 26, 2013)

Preserving Virtual Worlds
Professor Henry Lowood (Liebenwalde Ware in Second Life) giving talk on SJSU’s SLIS island in Second Life

A hundred years from now, when historians want to understand the ways we use Second Life®, how will they know? Computers won’t be anything we even recognize as a computer and won’t run current software or even be able to read installation disks.

Questions like this, but about software created decades ago, were the subject of a talk last night by Stanford University’s Henry Lowood (avatar name Liebenwalde Ware) at San Jose State University’s SLIS (School of Library & Information Science) Island in Second Life. The talk was titled, “Look Back on the Preserving Virtual Worlds Projects – and (Nearly) 15 Years of game Preservation.”

There are several parts to the problem, and they are true whether we are talking about a virtual world created twenty years ago or if we are talking about Second Life as historians will see it a hundred years from now.


The first part of the problem is media. Decades ago, media used for software installation and data storage are obsolete today, such as data cassettes (like the old audio cassettes) and floppy disks. These media are notoriously unreliable and don’t age well. Even if they do hold up over time, it becomes more difficult over time to find the hardware to read them.


The second part of the problem is hardware, not just media readers, but the computer itself. Professor Lowood’s presentation included a slide that showed part of this problem: a computer, looking like it measured about three feet wide, three feet deep, and about six feet tall, that was the equivalent of a video card today. Hardware like this from decades ago is scarce to nonexistent, along with the computers that used it, making playing the games that used it impossible. It’s the same problem that historians will have a century from now with Second Life: even if they manage to find a NVIDIA or other video card that will run Second Life well today, they won’t have the computers it goes in.


The third part of the problem is software. Even if you have the installation media, the computer, and all the peripherals needed to run the software, how do you run it? What metadata is required for it to run? Moreover, it’s not enough to have the software for the game or virtual world. You also need the operating system and any components needed for the game.


A fourth part of the problem that may not be obvious at first. It’s cultural: just how did people use the software or game? This is best viewed from the Second Life perspective. A century from now, if historians have the software to run Second Life and they manage to put together a computer for it to run on, what will they see and what will they do?

Will the installation software contain all the user-created content that we see today? Will these future historians see a Second Life from 2004 or 2014? Will they see a 2004 Second Life that even to us looks crude – this writer was in Second Life in 2004 and remembers what it was like – or will it have features that we haven’t yet seen?

Even more of an issue will be the fact that future historians will be the only avatars in that world. They won’t know the very many, very different ways we use it, and the very important role it plays in many people’s social lives. For some people, it’s a creative tool, for others it’s a dance hall, for others it’s a workplace, and for others, a vital part of their spiritual lives. There’s no way future historians will be able to experience that for themselves. They will be in a virtual world that’s like a physical world that’s been stripped bare of human life by a neutron bomb, leaving only buildings and prim, sculptie, or mesh vegetation.


A fifth problem is one that will only affect researchers, writers, and students: if you’re writing a paper or book and need to include software in a foot note, how do you create a citation for software? For example, if you’re citing a book, you give the page number of the reference, but how do you do the equivalent with software so that others can go into the software and see exactly what you’re referring to?

The scope of the overall preserving virtual worlds problem can be seen in a collection housed at Stanford, the Cabrinety Collection, which the university acquired in 1998. It’s only one collection for long-ago games and worlds, but it consists of 15,000 pieces of software that take up over 800 linear feet of physical storage space. Far more software has been created than is in this collection and is of a far greater complexity. Archiving it all will be a gargantuan task.

Four U.S. universities are taking part in the preserving virtual worlds and software effort: Stanford University, University of Illinois, University of Maryland, and Rochester Institute of Technology.

For more information

If you’d like to learn more, you can read Professor Lowood’s blog, You can also follow him on Twitter at @Liebenwalde (his avatar name is Liebenwalde Ware).

Leave a Reply

Your email address will not be published. Required fields are marked *