OMA Archive is my own digital research library, for the kind of material most libraries don’t keep: fringe scholarship, underground publishing, samizdat, zines and grey literature. Instead of one app trying to do everything, it’s a small stack of proper open-source library software, each part doing one job.
InvenioRDM, the repository software CERN built, holds the master copy of every item and its metadata. ArchivesSpace keeps the archival structure: collections, series and finding aids. Omeka S is for curated exhibits and notes. VuFind is the search catalogue, and Kavita is where I actually read, with EPUB, PDF and comics in the browser. Calibre-Web provides a catalogue feed too.
The glue is a set of scripts I wrote. They move books from Calibre into the repository in bulk, upload the files, make clean reading copies, and export everything as CSV, JSON or BibTeX. Preservation is taken seriously: every file gets a SHA-256 checksum that can be re-verified, and items can be bundled into proper archival preservation packages (OAIS). There’s a full backup to the NAS and a health check across every service. Each item also carries my own fields, from provenance and OCR status to subculture, research priority and, my favourite, a “weirdness” score.
Specs
| Status | Running; login-only at oma-archive.org |
| Built | May 2026 |
| Repo | Private oma-archive |
| Services | InvenioRDM (repository), ArchivesSpace (archives), Omeka S (exhibits), VuFind (discovery), Kavita (reading), Calibre-Web (catalogue, OPDS) |
| Infra | Docker Compose per layer; Postgres, MySQL, Solr; NAS backup over rsync |
| Scripts | Calibre → Invenio export and ingest, file upload, metadata export (CSV/JSON/BibTeX), OAIS preservation packages, reading-copy generation, checksums + verification, ID assignment, backup, smoke test |
| Metadata | 19 custom fields, incl. provenance, OCR status, subculture, research priority, archival sensitivity, weirdness |
| Not yet | A public view of the collection |