How privacy labels become redactions

Open Memory Archive does not hide private material with a visual overlay. It reads privacy labels before export, writes a public archive with placeholders, and leaves private source text or private media out of that public archive.

Context: The privacy annotations shown below are typically set by the upstream platform, not by the end user. The archive builder reads them and applies redaction automatically. End users see only the final result.

This demo is not a rule-editing screen. The current public toolkit shows the archive format, builder, redaction layer, and offline viewer. Archive organisers mark privacy in source data such as CSV or JSON. A future application can expose the same fields as checkboxes or dropdowns.

The process

1. Mark the source material

Each person, event, chapter, place, or media item carries simple privacy fields.

2. Build a public archive

The builder uses the public-demo profile to decide what can be shown.

3. Publish only the safe output

The public archive contains placeholders, public media, checksums, and a redaction log.

See the difference: source vs. exported text

Below is actual text from the family archive example. The left column shows the source material; the right column shows what appears in the public export.

Source (family-only) Public export
Anna Korhonen. Lives in Turku. Manages the family's archive and decides what is shared publicly. [REDACTED] - living person, consent not confirmed
Family argument at the summer cottage, July 1991. Consent for this entry was withdrawn by Anna in 2024. [REDACTED] - consent was withdrawn
Anna's mother was estranged from the family between 1988 and 1995. Matti's notebook mentions it three times but Anna asked to keep those pages private. [REDACTED] - marked as private or family-only
Harbour Market Day, summer 1967. Matti's notebook records 52 boats and a queue around the block for smoked vendace. Harbour Market Day, summer 1967. Matti's notebook records 52 boats and a queue around the block for smoked vendace.

How someone marks privacy before export

Before export, each item is marked with simple privacy fields:

In the file-based demo, these fields appear as CSV columns such as privacyScope, consentStatus, livingPersonStatus, containsPersonalData, and containsSpecialCategoryData.

Community memory demo: 3 redactions

The main demo archive is built from ordinary Markdown, CSV, and local media files. It produces the current public demo with 3 items redacted.

Source item Privacy labels Public archive result
person-anna / Anna Korhonen family, unknown consent, living, containsPersonalData: true Redacted person placeholder
event-private-meeting family, unknown consent, living, containsPersonalData: true Redacted item placeholder
media-private / Private participant note family, unknown consent, living, containsPersonalData: true Redacted media; source file not copied
person-matti / Matti Korhonen public, not_required, deceased, containsPersonalData: true Displayed in the public archive
media-note / Harbour market photograph public, not_required, not_applicable, containsPersonalData: false Copied into the archive media folder

Redacted family demo: consent and private media

The second demo uses a more privacy-focused fictional family archive. It shows withdrawn consent, family-only scope, private text, private media, and signed URL removal.

Source item Privacy labels Public archive result
section-private / Private detail family, unknown consent, containsPersonalData: true Redacted section; private text removed
person-living / Anna Korhonen family, unknown consent, living, containsPersonalData: true Redacted person placeholder
event-withdrawn / Withdrawn private event private, withdrawn consent, living, containsPersonalData: true Redacted item; reason is withdrawn_consent
media-private / Private family recording family, unknown consent, living, containsPersonalData: true, signed URL present Redacted media; private file and signed URL not included

What is actually published

The public archive includes the viewer, structured archive data, a privacy summary, checksums, and public media files. It does not include private source text, private media files, private storage paths, or signed URLs that were removed by the redaction policy.

Open community memory archive Open redacted family archive