06 / Project
Archive 01
A digital archival system for preserving visual collections.
2024 · Software Engineer · Archival System
Overview
Archive 01 is a preservation system for visual collections — photographs, prints, and documents that need to outlast a hard drive. It treats metadata, derivatives, and access as part of the same conservation practice.
The Problem
Private collections and small institutions were storing scans in folders with inconsistent names and no checksums. Colour profiles were stripped. Captions lived in separate documents. When a drive failed, the collection’s meaning failed with it.
The Approach
Ingest is slow on purpose. Each object receives a stable identifier, checksums, preservation and access derivatives, and a cataloguing record that can be incomplete without being chaotic. Viewing is calm: one object, its relations, its history. Public access is a separate, narrower surface.
Implementation
A Node.js pipeline writes originals to object storage with integrity checks, then produces access images. PostgreSQL holds the catalogue, rights, and provenance. The TypeScript interface is built for cataloguers first. Exports follow simple, documented packages so the archive is not trapped in the application.
Outcome
Collections can be held with a record of what arrived, what was derived, and who may see it. Recovery is possible from storage plus the catalogue, not from folklore about how the folders were arranged.
Technology
TypeScript · Node.js · PostgreSQL · S3
