What this harness is, what ingesting it cost, and where the ingested copy is meant to end up. These pages are about the work; the catalogue itself is served by the library handler, and the replica rendered from it is this instance’s own front page.
This is not a copy of IRIS. who-iris models the WHO
Institutional Repository for Information Sharing by reference:
235 nodes of a 273,559-item 361.55 GiB repository, of which 12 items are held here. Everything on these pages was transcribed from captures the
owner supplied — iris.who.int is not reachable from the
environment that renders them, so nothing here was fetched. The figures come
from the catalogue rather than from this sentence, which is why they are the
same ones the replica shows.
The requirements the ingestion produced, rendered from
skills/iris-dspace.md rather than restated — so the page
and the skill cannot drift. Includes what the captured DSpace records gave
that inference could not.
The general pattern this instance is a worked example of: select → serialize → package → sign → distribute → verify, with the trust anchors read off the WHO SMART Trust IG and the transport left undetermined rather than guessed.
System requirements, pipeline placement, multi-graph partition model, and browser WASM SPARQL query engine architecture for WHO-IRIS (with scale plan for 300,000 items and zero binary blobs in RDF).