The Accidental Archive: On Finding My Grandfather in a Forgotten FTP Server
I wasn’t looking for him. I was looking for a specific strain of heirloom tomato, the kind my grandfather used to grow in his backyard. The seeds, like him, were gone. In a last-ditch effort, I fell into a rabbit hole of old gardening forums, the kind that predate slick social media platforms, their text-heavy interfaces a relic of an earlier web. A post from 2001 mentioned a 'seed savers' FTP server, a digital repository for gardeners to trade rare varieties. The address, a string of numbers and a tilde, was still listed. On a whim, I plugged it into a client.
It connected. The server, running on some university’s forgotten machine, was still alive. Its contents were a time capsule from the late 90s: text files with planting guides, scanned diagrams for building cold frames, and directories upon directories named for vegetables and flowers. It was a quiet, unadorned archive, built not by an institution but by a community. And then I saw it: a folder simply titled ‘Tomatoes_M’. I clicked.
Inside was a list of text files, each named for a variety. And there it was: ‘Mortensen’s_Yellow_Stripey.txt’. My grandfather’s name. My hands went cold. I downloaded the file. It was just a few lines of plain text. A description of the plant’s height, its fruit’s sweet, low-acid flavor, and ideal harvesting time. At the bottom, a note: ‘Shared by Ed Mortensen, Cleveland, OH. A generous grower.’
I sat back, staring at those words on my screen. This was not a scanned photograph or a digitized letter. This was a data point, a single entry in a collective, open dataset. It was the most ‘him’ thing imaginable—practical, helpful, and shared without any desire for credit. The university IT department, by neglecting to decommission this one server, had inadvertently preserved a piece of him. It was digital preservation at its most accidental and most human.
This experience changed how I view the sprawling, messy landscape of open data. We often think of it in grand terms: government datasets, scientific research, public records. We don’t always see the intimate, personal archives that exist in the forgotten corners of the web. These are not the meticulously curated collections of the Wayback Machine, but the scattered, fragile remnants of communities that built their own repositories. They are vulnerable, living on borrowed time on a server someone might decide to clean up next Tuesday.
Finding that file was like finding a note in a bottle, a digital message that had traveled twenty years through the volatile currents of the internet to reach me. It was a powerful reminder that data, in its plainest, most unadorned form, is never just data. It is a footprint. It is a voice, quiet but persistent, speaking from a past we assume has been overwritten. It is the reason we fight to keep these old links open, not for the sake of preservation itself, but for the singular, personal connections waiting to be remade.
Notes & further reading
A few pages I came back to while writing this: