The Cathedral and the Starlight: Two Visions for Preserving the Digital Commons

When we talk about safeguarding our digital heritage, we often picture a singular, monumental effort—a great library of Alexandria reborn in silicon. This instinct leads to what I’ve come to think of as the "Cathedral" approach. It’s embodied by projects like the Internet Archive’s Wayback Machine: centralized, institutional, and built with the grand ambition of capturing the entire web. The scale is breathtaking, the engineering heroic. It is, in essence, a fortress against time, built to protect the public record within its formidable walls.

Yet, there exists a fundamentally different philosophy, one that feels less like architecture and more like astronomy. I call this the "Starlight" approach. It’s the model championed by the decentralized archiving protocol known as IPFS, the InterPlanetary File System. Instead of one central repository, IPFS fragments data across a peer-to-peer network. Each piece of preserved content is given a unique, content-based identifier—a cryptographic hash. To access it, you don’t ask a single server for a file at a specific, fragile location (a URL); you ask the network, "Who has this hash?" The data is served from wherever it exists, from anyone who has chosen to "pin" it. The archive exists as a constellation of shared points of light.

Contrasts in Philosophy and Vulnerability

The contrast is profound. The Cathedral offers a single, authoritative (if incomplete) version of history. It provides continuity, a familiar interface, and a clear chain of custody. But its centralization is its greatest vulnerability. It faces legal challenges, financial pressures, and the existential risk of being a single target. If the Cathedral’s doors are shuttered, what then?

The Starlight model, by its nature, is resilient and diffuses responsibility. No single entity can take it down. It thrives on distributed participation. Its weakness, however, is its potential ephemerality. A piece of data only persists as long as at least one node in the network cares to host it. The archive becomes a living, collective agreement, not a permanent edifice. It can feel chaotic, harder to navigate, and lacks the curated authority of an institution.

This isn’t a simple debate of right versus wrong. It’s a necessary tension. The Cathedral gives us a common reference point, a baseline of truth we can all point to. The Starlight gives us a hedge against catastrophe, a system that can’t be erased by a single stroke of a pen or a failed funding round. One creates a pillar; the other, a web.

The most promising future for our readable public records may lie in a conscious symbiosis of these models. Imagine the Internet Archive, a canonical Cathedral, routinely backing up its collections into the Starlight of a decentralized network. Imagine community archives using the resilience of IPFS to underpin their local collections, while also contributing copies to larger institutional efforts. The Cathedral provides focus and accessibility; the Starlight provides redundancy and radical distribution. Together, they create a preservation strategy that is both a monument and an ecosystem—a structure that can withstand storms, and a night sky where our data, once shared, can continue to glow long after its original source has faded.

Notes & further reading

A few pages I came back to while writing this: