Crispy Gamer

Saved Games: Preserving the New TV

As more and more people do their research and get their information online, companies have worked to digitally preserve everything they can. From Google's controversial efforts to digitize and archive every book ever written to Hollywood's continuing project to convert its films to bits and bytes, a great percentage of humanity's cultural achievements will soon be saved for future generations.

Saved Games: Preserving the New TV
The Library of Congress is committed to preserving digital media, even games.

Ironically, it is the digital achievements that have had the least institutional structure for preservation. Though the Internet Archive has been trying to save Web sites and online discussions for years, only now is there a concerted effort to save the history of computer gaming -- preserving the software, the hardware and the virtual worlds. The International Game Developers Association has released a white paper on how game developers can help preserve their own history. But with so much of gaming history already on the edge of oblivion, four American universities and the Library of Congress have joined forces to archive what they can. And the Digital Preservation Project's game-saving efforts are not as easy as you would think.

Jerome McDonough of the University of Illinois at Urbana-Champaign says that gaming needs to be protected and preserved. "I'd argue that computers are the first really new medium to emerge for artistic endeavor since the creation of television," he says. "What you're seeing today in computer games is the birth of a whole new genre of artistic work. Preserving these early efforts is as important today as it would be to try to preserve Edison's first moving image materials.

"They are a tremendous social phenomenon, as well as a new artistic medium. If we can't preserve computer games and interactive fiction, we can't preserve a very significant part of our modern culture."

Four universities are involved in the Digital Preservation Project, with a couple of corporate partners including Linden Labs, the people behind Second Life. The scholarly approach, with its institutional constraints and measured pace, and the limited budget, mean that they can't save everything. But each school in the project brings a special skill to the table.

"Stanford University Libraries has one of the better collections of computer game material in the country," says McDonough. "The folks at Maryland have longstanding connections with the electronic literature community, which is also very concerned with how this literature will be kept accessible. Andy Phelps and his gang at Rochester Institute of Technology are very interested in problems of virtualization of platforms, and how that might be applied to games in the digital preservation realm. I've been involved in metadata problems (and some other problems) with digital preservation going back five or six years now, from before I was working as a professor here."

It is the collaboration that makes the Digital Preservation Project so potentially powerful. With one of the partner schools right on my doorstep, it only made sense to get a ground-level view of the process and problems of archiving virtual worlds and other digital media.

The Maryland project

Matt Kirschenbaum, a board gamer and a scholar, is sensitive to the role of institutions in preserving our cultural heritage. Kirschenbaum is an English professor at the University of Maryland and Associate Director of the Maryland Institute for Technology in the Humanities (MITH). "One reason we have cuneiform tablets is that we have big buildings to keep them in, and they aren't lying out in the street. Not everyone has access to them. While there are technological challenges to preserving games, there are also social and institutional challenges to preserving this type of digital media.

"If you think preserving games and virtual worlds is a sexy, adventurous thing for the Library of Congress to be involved in, you're right," he jokes. "Most of the grant program that is funding us is for preserving state records and things like that."

Setting the processes and priorities has proven to be challenging, even "arbitrary and opportunistic," says Kirschenbaum. With only two years of funding from the Library of Congress, the schools could not do everything -- so they adopted a case study approach, choosing significant historical milestones such as the original Spacewar! (1962), often considered the first videogame. When Stanford University discovered the original tapes for the William Crowther game Adventure (1975), which is in the public domain, it became a natural starting point for Maryland's emphasis on interactive fiction and virtual worlds.

Saved Games: Preserving the New TV
If a game was played on teletype, do you need the teletype to get the full experience?

Games to be archived were also chosen for the technical problems they presented, argues Douglas Reside, one of Kirschenbaum's colleagues. Not all of Second Life can be preserved, so only a couple of its islands will be preserved as examples. "The challenges there are very different from the challenges presented by Adventure, which developed in a multi-author way over the course of a decade," he says. "The game was originally played on something like a teletype, and then ported to a number of different platforms. Then you have Mindwheel, which was in shrink-wrapped boxes for five different platforms, but development stopped with the commercial release."

Kirschenbaum also sees the project as research into the very nature of archival work. With so many different types of games to archive, it quickly became apparent that there would be no "silver bullet" for preservation. Formal similarity between games did not mean that they could be saved in the same way. And emulation, the most familiar form of preserving digital entertainment, is not the same as preserving the original game experience. Saving multi-authored experiences like Adventure and Second Life requires ways of reflecting both the process of change and the core experience itself.

"For Second Life," Kirschenbaum says, "we are using videography of the world. That's where the Internet Archive comes in. We're also experimenting with collecting the data that gets passed from the server to the client, and using XML to recreate a 3-D environment. Theoretically, that would allow someone in 50 or 100 years to create an avatar and walk around the world. It wouldn't, however, allow social interactions. But we have the video for that."

For the Maryland team, this is not about preserving games as much as it is about preserving the gamer experience. They are expanding to the preservation of what Kirschenbaum calls "paratextual materials" -- walkthroughs, FAQs and other user-created content that is unavoidably part of the game space. Rachel Donohue, a graduate student attached to the program, argues that this sort of approach is crucial to a real archive.

Saved Games: Preserving the New TV
The most famous game in the world is on Stanford's list.

"This goes toward the archival concept of context," she says. "You have a game and it can stand in isolation, but to understand if it had an impact on society you really need the external information." She points to DOOM, a game that Stanford is working on preserving, as an example. It was, after all, endlessly modded and adapted, and this very customization was what put it at the center of the Columbine Massacre argument.

Context will be preserved largely through descriptions of the game and the environment in which it appeared. Donohue cites "contemporary reviews, articles, playthroughs [and] speedruns" as examples. The rapidly expanding academic and retail literature on gaming, Kirschenbaum notes, helps situate what they research. "What we are doing is not happening in a vacuum. It's not necessarily the case that we need to explain what Second Life is from the ground up. This is a two-year exploratory project, after all."

The art and hassle of archiving

The Preservation Project has run into a number of problems unrelated to technology or finding material to save. Donohue points to its experiences with Mindwheel, an adventure game that was co-written by Poet Laureate Robert Pinsky, as a sign of how murky the intellectual property rights could be.

"Synapse Software has changed hands so many times that it is relatively unclear who precisely owns Mindwheel at this point. It was bought by one company that sold some of its assets to another company who sold some of those assets. Most of the people we've gotten in touch with only say that they don't support the product anymore, so we have to go by the active license -- [it's] really a form letter."

UIUC's Jeremy McDonough does not see the IP rights as a serious long-term problem, however. "Libraries have managed to preserve works without running afoul of intellectual property law in the analog realm for some time. Second Life does present a slightly more complicated case than many, in that users retain IP rights over their creations (rather than it all ending up the property of Linden Lab). That means that archiving any significant chunk of Second Life means a large number of individual negotiations for permission to copy. We're currently working on software to help speed the process of identifying rights holders within a particular island of Second Life and engaging in discussions to enable preservation. So, games can be complicated in terms of their IP, but actually not much more so than films and television."

Saved Games: Preserving the New TV
The only game from a poet laureate and one of Maryland's core preservation efforts.

For McDonough, the big impediment to preservation is Congress's 1998 Digital Millennium Copyright Act, which limits who can copy a work, and how. Libraries and archives must apply for exemptions from the DMCA.

"If we can't copy a digital work legally, there's no way we can preserve it; all media die eventually," McDonough notes. "While the DMCA provides for exemptions, forcing the library world to continually reapply for exemptions -- to engage in what should be routine preservation activity -- places a significant burden on libraries to no particularly good end.

"Technological protection measures [e.g., DRM], combined with the DMCA's prohibitions against circumventing such measures -- even to engage in preservation copying -- are a particularly problematic combination when trying to preserve computer games. As a result, it's actually proving easier for us to contemplate preserving games from computing's early history (Spacewar! and Adventure) than some of the more recent games."

Maryland's Doug Reside takes a more pragmatic approach to the problem. "As of yet, there aren't any standards for digital preservation the way there are for paper preservation -- even beyond games, for archiving things like email or Word documents. Add to that the complexity of preserving something interactive and graphical. With DRM." Because the problems with the "boring records" on the Internet haven't been settled yet, the archiving community hasn't even come up with a general approach to how game or virtual world preservations should be done. There are only a few "best practices" in place that everyone can agree on.

This gets especially complicated for those early games that shipped with tchotchkes that were often used as either copy protection or as central parts of the gaming experience. Even if this experience can be preserved at the physical archive, it cannot be perfectly replicated at remote locations that may have online access to the archive materials. What can be preserved, and who is the prospective user of these archives? Future game designers may be interested in the games just for tracking influences, where future scholars may be very interested in the material history of the hobby, making preservation of working hardware integral to the preservation of the software.

Looking ahead

So much depends on the project's next step. The existing grant has a two-year term, but McDonough is confident that the Library of Congress is committed to this project. "If, at the end of those two years, we've identified practical methods for trying to preserve computer games and interactive fiction," he says, "I think we can build on what is essentially a research initiative, and moving that knowledge forward into practical use. Both Stanford and UIUC have gaming collections, and I think gearing up our libraries' preservation programs to handle these materials would be a good idea."

At Maryland, Kirschenbaum is already thinking of the next step in his preservation efforts. "MOOs and MUDs," he says. "They're the natural bridge between interactive fiction on the one side and virtual worlds on the other. It's not that we don't think they are important; we just don't have the time or resources right now. Those would present some interesting and unique problems.

"We also want to look into establishing more formal relationships with the amateur preservation community on the Web (MAME developers, the abandonware community) and bring them in, in a more organized way."

Rachel Donohue wants to take the project in the other direction, engaging the publishers and developers with preservation in a more direct way. "I've talked to a few people in the industry, and, you know, they just don't do it. They don't worry about records management on the design side, only what they are legally required to keep on the business side. I think that's really tragic."

Left out of this discussion, of course, is the average gamer community. This is an archival project, and not an effort to make out-of-print games or materials as easily available as out-of-print books or public-domain written work. But for game researchers, developers and serious students of the industry, this joint effort by the American government and various academic and business partners is an important first step in saving what can be saved and how it can best be kept for future analysis. Once best practices are in place, it should be easier for archivists all across the country to contribute in their own small way.