For a Glimpse Of The Future, Try Reading A 3.5-Inch Floppy Disk

In his low-tech data recovery operation (old computers), Tom Persky is witnessing the first round of personal digital data decay. Here’s why the next rounds will be worse.

For a Glimpse Of The Future, Try Reading A 3.5-Inch Floppy Disk

Tom Persky’s company,, sells about 250,000 of the 3.5-inch and 5.25-inch square plastic storage cards each year. It has an inventory of about 1 million disks, many of which Persky acquired from competing companies as they went out of business. And he has no plans to restock them as they’re sold.


“We stayed in the business through the good times of the floppy disk business, when it was a very common way to store media, through the trough of what we have now of people not using it very much at all,” Persky says.

That’s not to say floppy disks aren’t still his mainstay. But the business has evolved.

Any remaining modern demand for the once-ubiquitous floppy is driven by machines still made with old data inputs. But for a while, the disks, which first went on sale in the early ’70s, were also the first prevalent medium for storing personal digital data. A generation of college theses, original song lyrics, family photos, and first drafts of novels were saved to them. As the machines that read floppy disks have become obsolete, it’s become difficult for most people to access these artifacts. Instead, they search for “floppy disk” on the Internet. And guess which domain inevitably ends up in their results: Persky’s. Regular requests for data recovery inspired him to expand his services to transferring data from old floppy disks to new CDs.

The first floppy disks, introduced by IBM in 1971, were 8-inches big. Most data recovery firms do not read them.

The problem with digital storage–as the hundreds of people who ship Persky their floppy disks each year are finding out–is that it requires active upkeep. Most computers don’t have the floppy disk drives that were standard 15 years ago. Nor do they run the same operating systems or software programs used to create documents saved to floppies–even if the data is recovered, it may look more today like a garbled mix of symbols than your first novel written in WordPerfect.


As Persky transfers data from old media to new, he is, in other words, witnessing the first round of personal data decay. And in an environment of increasing dependence on digital data, where everything is stored in the nebulous cloud, that decay will only become more widespread and complex as newer media meets the same fate as the floppy.

It is a hassle to open a document saved to a floppy disk that was written in the now-extinct Lotus 1-2-3. But imagine trying to open tweets without Twitter, recover a note from your cloud-based inbox after the service shuts down, or search for family photos in a Facebook account for which you’ve lost the password (and drats, the account is attached to that email account that no longer exists).

Persky uses these computers to transfer data from 3.5-inch and 5.25-inch floppy disks.

Not only are we increasingly saving digital, rather than physical, media, but we’re also abandoning physical storage devices such as CDs, which we store ourselves, in favor of cloud services. We have less control over our increasing amount of personal digital data.

Facebook, Gmail, and other cloud-based services give users some control by providing a data download option. Twitter, for instance, recently added a feature that allows users to download their personal tweets. Several others provide tools for mining your social media past. But in the future, such efforts might read like a Lotus 1-2-3 document opened on a modern-day computer. A tweet is more than its text. It’s dependent on context. Without the Twitter “operating system,” you’ll loose the tweets that surround yours. The interactivity. The other half of your conversations. The links you included that have expired.


Reading Floppies Today, Reading Twitter In 2063

This 9-track drive is the same kind that Kroll Ontrack uses to recover data from old media. According to Jeff Pederson, it sounds like a vacuum cleaner.

Persky’s data recovery methods are low-tech. He owns old computers. When he receives a shipment of disks from a customer, he tries to read them in different drives. Sometimes a disk will mysteriously work in one drive, but not another, and he has no idea why. Whatever data is readable gets transferred to a CD and shipped back to the customer. The service costs $1.99 per 3.5-inch floppy disk or $4.95 per 5.25-inch floppy disk.

“We are doing it by hand,” Persky says. “There is no magic that I am aware of.”

Jeff Pederson, who has worked at a data recovery firm Kroll Ontrack for more than 20 years, has a slightly more sophisticated approach–though it still relies on old machines often purchased on eBay. The 9-track tape drive at the company’s headquarters outside of Minneapolis, for instance, sounds like a loud vacuum cleaner every time he turns it on. It’s about the size of a suitcase. Reels of the tape it reads, once the dominant medium for data storage, haven’t been manufactured since 2002.

Usually it is a government agency or a corporation doing data inventory that asks to have such reels read. Kroll Ontrack can also read other old media such as 3 ½ and 5 1/4-inch floppies. The company recovers damaged media that Pesky’s old computers can’t by using proprietary software that skips over problem areas instead of breaking to issue a read error. Pederson says he rarely encounters data he can’t read. But the firm’s expertise is less about reading data on old media than about translating it.


“Just being able to say you can read something doesn’t mean that it’s necessarily recoverable or that it’s going to mean something to somebody else,” Pederson says. “If we can read something and it’s just a bunch of 0s and 1s and hexadecimal representations, data that nobody can translate, it doesn’t mean as much.”

This screenshot from Knoll Ontrack shows data recovered from old media before it is translated to a readable format.

This translation is complicated enough when you’re moving data from a floppy disk to a CD, but it is only a hint of what we may be dealing with in the future. In the modern era, the problems of content and context meet the problem of volume. A standard floppy disk holds 1.44 megabytes of data. The cloud is virtually limitless.

The Library of Congress, for instance, is saving the Twitter archive. But it has not yet produced a meaningful interface through which to access it.

“We could spend all of our time on digital stewardship and not have any time to do anything new,” says Cathy Marshall, a principal researcher in Microsoft’s Silicon Valley Lab who studies personal digital archiving. “People pick and choose what things they care about. Usually, they just kind of leave it to fate. I think people can care about things, but they just end up leaving it to fate, because it’s so hard to actually do anything about it.“


By the time you decide what is worth saving, it may be too late to recover it. Just as the floppy disk went from domination to obscurity in 30 years, so too may our present-day digital storage systems one day be replaced or forgotten.

Which is why, unlike Persky’s feelings about his floppy disk business, Pederson has no doubt that data recovery is sustainable.

“As long as there are humans running computers, there is going to be data that needs to be recovered and/or converted,” he says. “You’re always going to have to evolve not only the data, but how you’re accessing it.”

[Images: Flickr users Zach Rathore, moparx, Lisovy]

About the author

Sarah Kessler is a senior writer at Fast Company, where she writes about the on-demand/gig/sharing "economies" and the future of work.