In part 1 of this story, I described the difficulty of accessing a commercial CD-ROM published in 1989. Eventually, out of frustration, I questioned its value and wondered who actually cared about outdated software. So I consulted some colleagues.
It turns out that some gamers care, especially those who are fanatical about the original look and feel of favorite old games. Many are dedicated enough to build their own emulators in order to continue playing the games.
And some cultural historians care. Geoffrey Brown, professor of computer science at Indiana University, says that cultural historians with technological interests may want to access old CD-ROMs to study the design decisions.
For example, the Library of Congress’s American Memory was initially created for Hypercard. Carl Fleischhauer, digital initiatives project manager at the Library of Congress, worked on American Memory from its inception. Fleischhauer said, “American Memory was put onto CD-ROM between 1990 to 1994. Then the Mosaic browser came out and it became possible – and better – to put it all onto the web. We migrated all the content; nothing was lost. But even though the original disks still reside here in the Library, we can’t play them to see examples of our own original work.”
Brown has worked extensively with vintage CD-ROMs and he said, “Technology isn’t the problem, just lack of demand.” Indeed, just as a cylinder recording circa 1900 will still play on a cylinder player, the Hypercard application will play on the appropriate hardware and software. But lack of demand makes it difficult for the average person to find free or low-cost resources.
Some institutions have created labs comprised of vintage technology, which they use for research. The Maryland Institute for Technology in the Humanities at the University of Maryland, for example, has collected hardware and software that date back to the early 1980s.
MITH’s preservation services are not open to the general public — it mostly serves the university and funded projects — but some outside researchers can find help there. Matthew Kirschenbaum, associate director of MITH, said he recently helped an out-of-state artist who wanted to restore some of his old digital artwork because he — the artist — valued its original appearance.
The Archeological Media Lab at the University of Colorado at Boulder serves a similar function as MITH. Such labs are becoming increasingly necessary as institutions acquire collections that contain digital media, diskettes, CD-ROMs and even whole computers. Emory University is the recipient of such a collection, the Salman Rushdie archive, which includes the author’s personal computer resources.
In this video, Emory staff discuss the uniqueness of the collection for recreating Rushdie’s writing environment, such as Rushdie’s penchant for desktop sticky notes and the way his pages fit his computer screen (suggesting that the screen size may have influenced some of his composition decisions).
The Rushdie collection demonstrates the value of preserving vintage technology to support versions of artist’s digital work. It’s similar to preserving paper drafts of an author’s work to compare with the finished version, such as the drafts of Langston Hughes’s poem “Ballad of Booker T.” or Beethoven’s manuscript of Opus 109.
Geoffrey Brown notes that preservation of electronic games still has not improved and in time our modern games will, too, become inaccessible because of hardware and software dependence. “Look at today’s game platforms,” Brown said. “It’s virtually impossible to emulate them.”
Leslie Johnston, manager of technical architecture initiatives in at the Library of Congress, said that even our precious smart phone apps are destined for the same fate. “An app isn’t any better than a Hypercard stack,” said Johnston. “Because an app, twenty years from now, isn’t going to run easily in the same way that a Hypercard stack isn’t. That app requires a certain operating system and certain hardware to operate.”
So the app, the game, the interactive CD-ROM…all eventually become museum pieces. And that’s exactly where the salvation for curating and preserving vintage technology is likely to come from: museums, libraries and archives such as MITH, the Archeological Media Lab and Emory’s Rushdie archives.
Matthew Kirschenbaum defends the preservation of vintage technology in select cases, if only for the primary experience itself (as opposed to simulating it with modern emulation methods). He said, “What you risk missing out on, even with emulation, is the entirety of experiencing the work in its original environment. Sometimes the complete physical platform, the hardware, the original technology is important and it is integral to what that thing is….Even if the graphics don’t look great by today’s standards, that particular look that we associate with a certain moment in computing – the pixels and the jagged appearance, the 8-bit graphics and tinny sound effects – that’s important for that work and for a lot of people it has a lot of emotional power behind it.” Or an appreciation by younger audiences for creations from the past.
Those of us considering what to do with our old software or hardware may face difficult appraisal decisions.
How about the stuff you created? Artwork? Documents? Kirschenbaum said, “If your personal data is important to you then you should become a preservation activist and know about the steps you can take to get it back.”
A lot depends on formats, of course. A twenty-year-old GIF file should be easy to open in a modern graphics program but a text document might be more difficult depending on the word processing program you initially used. And, of course, once you restore your files, if they have personal value you should back them up on modern media. If they don’t have personal value — if it’s old homework, for example — delete them.
Commercial software and hardware are different. They have broad appeal, since many people enjoyed them at one time and, by dint of their widespread usage, they are true cultural artifacts.
One archiving solution is to donate them. If you can upload your software, the Internet Archive has a software archive that accepts donations. The Computer History Museum in Mountain View, California, accepts software and hardware donations. And they have an online checklist of what they want and what they don’t. The Computer History Museum is open to the public, so generations of visitors may be able to appreciate your donation.
And that’s what I decided to do with my CD-ROM of the Electronic Whole Earth Catalog: donate it. Better to have it curated by professionals and enjoyed by the public then neglected in my basement.
Thanks for the mention of our old “pre-Web” American Memory CDs, Mike. Your blog brings to three themes to my mind–these are the same things you said, expressed in different words. First, digital content will very often be changed by our preservation actions. We don’t have a lot of experience with this but it seems likely that some items will be changed in striking ways while for other items–the majority, we hope–the changes will be subtle and mostly in the realm of “look and feel.” I’m not troubled about such changes: for decades, brittle books were profoundly transformed when microfilmed for preservation; they became sets of pictures of book pages. Meanwhile, in the digital realm, there is a belief that changes in content will result from preservation-via-format-migration and not from preservation-via-system-emulation. But as early as 2001, Margaret Hedstrom and Clifford Lampe reported on some end-user testing that suggested that _both_ migration and emulation change the content (http://webdoc.gwdg.de/edoc/aw/rlgdn/preserv/diginews/diginews5-6.html#feature1). Which leads to the second theme: is there any way to identify the _essential_ features (sometimes called significant features) for different categories of digital content? Which aspects of a given category do we _not_ want to lose when we carry out preservation actions? Clearly, for some items look and feel is essential (games, for sure) while for other items (emails), not so much. The third theme is about sorting out when we seek to preserve, um, the content-qua-content from instances of seeking to preserve the “experiencing” of the content. In the analog realm, this latter question has received wonderful discussion among moving image archivists. Some film archivists highlight the importance of theatrical projection (preferably with a carbon arc lamp, in a real theater) as the final chapter in the story of successful film preservation. It was fun to see some of these same themes elucidated in Jefferson Bailey’s recent Signal blogs on the “artifactual” side of digital content (//blogs.loc.gov/digitalpreservation/2011/11/the-artifactual-elements-of-born-digital-records-part-1/). Jefferson asked about the implications for preservation if we value how digital content was experienced by its first generation of users.
Thanks for your comments Carl. What you say is fair and even-handed. There’s a lot to consider regarding migration and emulation, and no one correct answer. And thank you for mentioning Jefferson Bailey’s well-thought-out blog on the subject, which I neglected to do; our stories relate to each other.
“Geoffrey Brown notes that preservation of electronic games still has not improved and in time our modern games will, too, become inaccessible because of hardware and software dependence. ‘Look at today’s game platforms,’ Brown said. ‘It’s virtually impossible to emulate them.'”
I found this comment somewhat strange. Game preservation is one of the best-off fields of digital media preservation, with 15 years of open development and discussion behind it. The game emulation community has accomplished some truly marvelous things, both on the accuracy and aesthetic fronts. I’d love to see a deeper partnership between the “real” digital preservation community and the hobbyists, who are writing cycle-perfect emulators and designing CRT-like filters.
The remark about current hardware being hard to emulate is also odd, given that modern game systems are now made using many off-the-shelf parts. Dolphin has good-enough emulation of still-current systems, due to the Wii’s nearly-generic PowerPC hardware (for comparison, a quad-core processor is needed to accurately emulate the Super Nintendo’s custom RISC processors/coprocessors). Special hardware is and always will be a problem, but game system emulation is becoming easier instead of harder.
In my experience, a lot of material from the HyperCard era is of just the right vintage to be very useful as patent prior art right now, too.
It is frustrating to think about how much stuff we could really use is buried and collecting dust, totally offline, in someone’s garage, somewhere. I sincerely hope that in another couple decades, more of our current work is at least gathering dust on the internet, instead!
Amen, Kevin. Your comment might even spark a side discussion on “internet dust.”
Thanks for the mention and I agree that this is a rich and complex topic. As Brown notes, emulation is less a technical problem and more a financial one. Because of that, enthusiastic user communities like the one Alex speaks of will be important collaborators and contributors to emulation projects. At the same time, some formats (or collections) will not garner the same enthusiasm or excitement. Those will require a decision about how to preserve them and what parts of them to retain. And it is those decisions, and Carl’s three themes, that prompt the most interesting questions. Determining those “essential” features and the “experience” of the item were not problems of the “paper” age.
We will have to decide what to preserve not just content-wise, but context-wise. That seems the most difficult (and maybe even frightening) challenge: that we not only have to choose what to save, but we have to choose what parts to save of what we choose to save. It is archival appraisal, redux. But it’s encouraging that communities — gamers and practitioners alike — are testing different ways to make those decisions.