EI-3.5 — Archive Representation
Status: COMPLETE
Date: 2026-08-08
Objective
Represent historically retained material without converting the 72 EI-1.5 ARCHIVE records into first-class current Writing.
Authority
The final EI-1.5 classification workbook/CSV is the item-level authority. Its final quality gate is:
- PUBLISH: 16
- REWRITE: 15
- ARCHIVE: 72
- DELETE: 125
- TOTAL: 228
- Unclassified: 0
The final EI-1.5 version incorporates the privacy correction: records referencing Ben and/or Sophie are DELETE and are not archive records even when an earlier working workbook placed some of them in an Archive Queue.
Archive model
ARCHIVE means preserve, but de-emphasize. It does not mean republish.
An archive record may exist in one of two states:
- Existing public historical post. A legacy item already present in Jekyll may remain reachable through the chronological Archive, but it is not automatically Featured or part of the primary Writing path.
- Metadata-only historical record. A WordPress item absent from production remains represented by the EI-1.5 ledger and archival source material. Its body and media are not migrated merely to make the archive exhaustive.
This distinction prevents preservation work from recreating the accumulation problem that EI-1 was designed to solve.
Public archive behavior
The /archive/ page is the chronological index of material actually published on the current site. It is not a mirror of the old WordPress database.
The page explicitly distinguishes:
- Published archive — current production pages/posts, including retained historical material.
- Legacy record archive — the 72 EI-1.5 ARCHIVE decisions retained in the editorial ledger and archival sources without automatic republication.
A legacy record can be promoted later only through a new explicit editorial decision. Promotion requires privacy review, source verification, and an intentional PUBLISH/REWRITE override; archive status alone is never sufficient.
Privacy boundary
DELETE material is not an archive tier. It is excluded.
The 90 Ben/Sophie privacy records must never be migrated, republished, restored, surfaced in a public metadata index, or used as sources for production media. Associated media remains outside the production repository.
Because this rule is stronger than historical working files, any pre-final EI-1.5 Archive Queue entry that conflicts with the final privacy decision is ignored.
Media boundary
ARCHIVE status does not authorize media restoration. Google Drive remains the archival/master repository. GitHub receives media only when required by an approved production item under docs/media-manifest.yml and docs/MEDIA_ARCHITECTURE.md.
Discovery rules
- Featured: never implied by ARCHIVE.
- Writing: should prioritize current/retained Writing rather than archive-only material.
- History: may contextualize historical collections and selected computing history, but does not expose every archive record.
- Archive: chronological preservation/discovery surface for material actually in production plus an explanation of the legacy-record boundary.
- Search/RSS: archive-only material absent from production is not regenerated merely for indexing.
Source preservation
The authoritative historical bodies remain recoverable from the scoped paco.org WordPress SQL/archive material in Google Drive. EI-3.5 does not copy those bodies into GitHub. The final EI-1.5 classification artifacts remain the disposition ledger for all 72 archive records.
Closeout
EI-3.5 establishes a non-destructive archive model while avoiding wholesale republication:
- 72 ARCHIVE decisions remain preserved by the EI-1.5 ledger.
- No ARCHIVE classification is automatically migrated.
- Existing historical production posts remain chronologically discoverable.
- Missing archive-only bodies/media remain in archival storage rather than GitHub production.
- DELETE/privacy material remains excluded from both public archive metadata and production media.
Next sprint: EI-3.6 — Navigation & Discoverability.