Stephen Hawking’s Floppy Disks & the Digital Legacy of Science: An Interview with Leontien Talboom

As part of the University of Cambridge’s Future Nostalgia initiative, Leontien Talboom is working with collaborators to preserve knowledge trapped on legacy media before it’s lost. Through this project, she is building practical guidance and raising the profile of the work required to recover and safeguard data stored on floppy disks. After reading recent BBC coverage of Leontien’s work to preserve data from Stephen Hawking’s floppy disks , I reached out to her for this interview to explore the broader implications of her efforts for the future of history of science collections and research.

In what follows, we discuss the goals of Future Nostalgia, lessons learned from working with Stephen Hawking’s floppy disks, why this work matters for the future of the history of science, and what guidance she has for archivists and scientists stewarding today’s born-digital record.

Trevor Owens: What motivated the Future Nostalgia initiative at this moment, and what specific gaps in practice or documentation are you aiming to address?

Leontien Talboom: The Future Nostalgia project emerged when I started working on legacy digital material at Cambridge University Library. I’ve been running our Transfer Service for over three years, and when I began, we had acquired a 5.25-inch floppy disk drive with an FC5025 controller , and a USB 3.5-inch drive. I assumed we had the right equipment for most of the floppies we would encounter, but I was so wrong.

Floppy disks are incredibly nuanced. It’s not just about the format; there are variations in density, drive type, controllers, operating systems, and so on. When trying to research these issues online, the information is scattered across forums, interest groups, and YouTube videos. Even when you find something, it’s often opinion-based and rarely documented with results.

This led to a workshop I ran at iPRES 2024 with colleagues Elizabeth Kata, Tyler Thorsted, and Chris Knowles. Future Nostalgia is an extension of that work; pushing further toward shared knowledge and practical guidance for dealing with floppy disk within memory institutions.

TO: From your work with Stephen Hawking’s floppy disks what most surprised you, whether technically or archivally?

LT: What surprised me most was the sheer diversity of material stored on the disks. Intellectually, I knew floppies were everyday working media, but seeing it firsthand really brought that into perspective; they contained everything from drafts of writing and bits of early code to spreadsheets, calendars, and personal notes. It showed how digital storage reflected the full range of someone’s working life rather than a single purpose.

I was also struck by the variety of systems, formats, and file extensions represented. Even within a small set, you encounter different operating environments and software dependencies, which really highlighted how much contextual knowledge is needed; not just to image a disk, but to actually access what’s on it. Sometimes the files may be technically preserved yet still inaccessible without the original system or software.

Overall, the experience reinforced that floppy disks aren’t interchangeable carriers; they are unique artifacts with their own histories and dependencies. It underscored how much valuable scientific and cultural material still sits on legacy media, and how urgent it is to recover it before that opportunity disappears.

TO: This article from the Science Museum Group Journal on the background on the Hawking Archive highlights that many digital processes are represented by paper materials that may reference or connect with digital files that are no longer accessable. How does the prospect of “absent digital files” in that work shape your preservation and access strategy?

LT: I think these absent files raise an important point more generally: archives often inherit collections where the physical traces of digital work survive; printouts, annotations, disk labels, even when the files themselves don’t. In those cases, rather than trying to fill in those gaps, our priority is to record them clearly so that researchers understand what is present, what is missing, and what might never have been transferred or retained.

More broadly, the possibility of absent files reinforces the need to document collections as thoroughly as possible at the point of transfer. That includes capturing information about the systems, software, and storage media involved, even if not all of the digital content can be recovered. It also shapes how we communicate access: being transparent about limitations is just as important as making material available. So while the focus of my current work is on recovering what is still on the disks, acknowledging and documenting what isn’t is a key part of this work too.

TO: What do you see as the implications of your work for historians of science using these archives now and in the future?

LT: I hope it broadens the conversation about preserving computer systems beyond the nostalgic home-computing machines beloved by retro enthusiasts. Many business and research systems, where scientific and historical material resides, are poorly documented. Even if you can create a disk image , you may not be able to access the files. They may rely on obscure disk structures, and furthermore no emulators exist for those systems.

We need to recognize that imaging alone isn’t enough. Preservation must include the systems, because access depends on them, especially as the people who built and used these machines retire or pass away.

TO: From your experience, what pragmatic steps would you suggest that labs and individual scientists take today to ensure their digital records remain preservable?

LT: The most important thing is to be mindful of the file formats you choose. Wherever possible, create and store material in open, well-supported formats rather than proprietary ones, as these are far more likely to remain accessible over time. My colleague Kim Clugston and I explored this in a blog post for Cambridge’s research repository, Apollo, which highlights simple format choices that can make preservation much easier.

It also helps to organize and document digital material while you are actively working with it. Meaningful filenames, keeping related files together, and avoiding multiple uncontrolled copies can save significant time later on. Even a brief README or note about software versions, instruments, or workflows provides context that future researchers, and even you in a few years, won’t be able to reconstruct once memories fade.

TO: For teams at libraries and archives beginning to tackle legacy floppies, what first steps and partnerships do you recommend?

LT: Connect with your local computer museum or retro computing enthusiast group; they are incredibly knowledgeable and generous communities. It may take time to establish relationships and explain why archival approaches differ from hobbyist ones, but it’s absolutely worth it. I’ve learned so much from the Centre for Computing History here in Cambridge.

As a first step, and this is a bit of self-promotion, I’d recommend the Floppy Disk Guide .

Also, put out calls for hardware. Either in local list servs or institutions wide groups. You’d be surprised how much equipment people have in attics and garages. I’m now known as ‘the floppy disk person’, and people regularly send me legacy hardware; which is both lovely and incredibly helpful!

TO: What’s next for the Future Nostalgia initiative? Are there other born-digital science collections you’re currently working with or hope to engage in the future?

LT: There is still a lot more research to be done around floppy disks. It really feels like we’ve only scratched the surface. Even within this single format, there are questions about different encodings, drive behaviour, and failure patterns that we haven’t fully explored.

I’m also interested in looking beyond floppy disks to other storage interfaces that are now becoming urgent, such as SCSI-based data cartridges in our collections. There is very little consolidated information about these in our community, so there may be a related strand of work in the future.

Another area I would like to focus on is the wealth of knowledge that isn’t online. Computer magazines, manuals, and technical documentation from the 1980s and 1990s hold a huge amount of practical information that could help us understand systems we’re trying to re-access. This is still largely untapped, except for the digitised copies that you can find online such as in the Internet Archive.

Finally, I hope to run more public and community-focused events like the Copy That Floppy Café , which was incredibly engaging and received great feedback. There is definitely more to uncover, and I see Future Nostalgia continuing through both ongoing research and wider collaboration across the digital preservation community.