Repository navigation
LIBRARY SPEC
Written 2026-07-28. Implements North Star §5.2 (named journeys, editorial voice) at the surface where RISE holds its texts. Draws its organizing principles from an early media-cluster note (since retired) and the Neorenaissance corpus.
Status: SUPERSEDED as policy by docs/specs/ARCHIVE-CANON-SPEC.md. This is
the critique that produced the canon — kept because the faults it names are why
the shelf is now fifteen works rather than a hundred, not because its remedies
are the current plan.
The Library has four faults, and only the last is about quantity.
- It is organized by FORMAT, not by idea. "Literary", "Research", "Sacred Texts" describe what a file is, not what a reader wants. A person does not arrive thinking "I would like some literary content."
- It leads with the wrong thing. Sacred Texts lands first, which presents the whole system as a wisdom-quote app — the "woo" register the software has outgrown.
-
It carries material that contradicts its own seriousness.
Declassified · Released intelligence documents 👁️holds four short excerpts (~5K chars total): Gateway Process, Project Stargate, Soviet Psychotronics, Holographic Matrices. This is the single clearest instance of the register problem, and retiring it costs nothing. - It is thin. 8 literary excerpts (~16.5K chars) and 6 wisdom texts. A "world-class housing of public-domain texts" is not 25K characters.
From the Archive: organize by resonance, not genre.
A work earns its place by what it does to a reader, not by which shelf a bookshop would file it on. The Archive names four functions:
- induce state — shift consciousness through encounter
- install pattern — leave structural residue in the mind
- generate connection — rhyme with other works in the corpus
- serve recursion — reward repeated engagement
This is already how the rest of R.I.S.E. thinks. The Chapel does not offer "religious content"; it offers the Passion beside the paintings that depict it. The Library should be no different.
The Library is the reading face of a Renaissance philosophy of education — the conviction that intelligence and meaning are not in tension. In the Neorenaissance corpus this is the claim that symbolic depth is load-bearing infrastructure, not ornament, and that a civilization which lets it erode suffers the "deathless death": systems that keep running while the reason for running quietly departs.
That conviction shapes the shelving; it is never printed on it. (Decision, 2026-07-28.) The tract's own argument is that meaning must be structural rather than decorative — a library that announced its philosophy in a banner would be doing the opposite of what it claims. A visitor should meet a corpus with evident conviction behind it and be able to infer the conviction from the curation alone.
No manifesto page. No named framework. No author byline.
Four collections, replacing the five format categories.
Names are plain; the editorial line beneath carries the voice. (Decision, 2026-07-28.) A shelf a visitor cannot predict is a shelf they will not open — obliquity at the door costs more than it earns. The atmosphere belongs in the orienting line, where it does work, not in the label, where it only obscures.
Structure, proportion, the grammar of things. Klee's Pedagogical Sketchbook, Kandinsky's Point and Line to Plane, Euclid, Vitruvius, the harmonic traditions. Induces: the sense that structure is legible.
Marcus Aurelius, Epictetus, Seneca, Montaigne, Pascal, Thoreau, Emerson, Rilke. The inner fortress and the practices that maintain it. Induces: self-possession. Most of the current Literary corpus is already here.
The Upanishads, Tao Te Ching, the Heart Sutra, Meister Eckhart, the Hermetica, Boethius, Plotinus, Blake, Dickinson. Where language ends. Induces: the recognition that the map stops.
Plato's Timaeus and Critias, Hesiod, the mythographers, Bruno, Kepler's Harmonices Mundi, Vico. Induces: the sense of something transmitted across the silence.
Open question for scale: at four shelves, Euclid and Kepler are ambiguous between FORM and THE RECURRENCE. These are a first cut, not the whole taxonomy; finer structure within a shelf may be needed once the corpus is large. Do not treat the four as final.
Retired: Declassified (register), Research (an ArXiv abstract feed is a different product; live abstracts are not a corpus and cannot be curated).
Not a shelf: the Chapel keeps its own door. Scripture is not filed beside Nietzsche. That separation is already load-bearing and stays.
Not a category grid. A short editorial statement — three or four sentences with a point of view — and then the shelves, each with its own line of orientation. The Archive's own voice is the model: dense, declarative, unhedged.
The current registry stores title, author, description, tags. It
should also carry:
- why it is here — one or two sentences of editorial judgement, in the Archive's voice. Not a blurb. Not a Wikipedia summary.
- its function — which of the four resonance functions it serves.
- its rhymes — other works in the corpus it speaks to. This is what turns a list into an archive; it is also what makes "generate connection" real rather than asserted.
- its provenance — translator, edition, year, and the public-domain basis. A world-class housing is honest about what edition you are reading and why we may show it.
An archive that does not remember is a catalogue. The Library should know what a reader has read and let them return to it. Local-only, no account. This is the Archive's "recursion" made literal.
Every text is public domain, and the basis is recorded per work. This is the textual analogue of curation-only for imagery, and it is non-negotiable for the same reason: the system would rather show nothing than show something it has no right to.
The Archive's Books tiers name many works still in copyright — Calvino, Borges, Flusser, Michaux. They inform the taste and cannot be ingested. Where a work is unavailable, the shelf is honestly shorter. A reader is never shown a text we cannot justify holding.
Per work, record: source (Gutenberg id, Archive.org id, or an ingest script), translator and edition, year of publication, and the public-domain basis (pre-1930 US publication, author death + 70, US government work, or an explicit CC0/PD dedication). Translations have their own copyright — a 2003 translation of Marcus Aurelius is not public domain even though Marcus Aurelius is. The existing corpus already gets this right (George Long, 1862) and the pattern should be enforced rather than assumed.
Architecture first. Ingest is its own stint.
In scope:
- The four shelves, replacing the five format categories.
- Retire Declassified and Research.
- Editorial voice: the door statement, per-work "why it is here".
- The work entry: function, rhymes, provenance fields.
- The reading record.
- Re-file the existing corpus onto the new shelves.
All six items landed. Notes on what the work turned up:
A fifth shelf was necessary. The 19 starter sequences are R.I.S.E.'s own compositions, not found works — several already carry the framework explicitly ("Neohumanist core", "Neohumanist identity"). Filing them among inherited texts would break the Archive's central promise, so they sit under COMPOSED: written for this system, in its own voice. A reader should always know whether they are meeting a received text or one written here.
FORM is empty, and stays empty. We hold no Euclid, Vitruvius, or Kandinsky. The shelf is honest about that rather than padded with near-misses; the ingest stint fills it.
Four texts were retired for copyright. Starter sequences reproduced verbatim in-copyright translations: Calvino's Cosmicomics (Weaver, Harcourt 1968), Borges' The Aleph (di Giovanni 1970 / Hurley 1999), Flusser's Technical Images (Roth, Minnesota 2011), and Klee's Pedagogical Sketchbook.
The Klee case is worth recording because the answer was not guessable. Klee died in 1940, the German original is 1925, and 93% of 1953 US books lapsed for non-renewal — the prior strongly favoured public domain. But the Stanford renewal database holds RE090004, filed 26 Jan 1981 against registration A78572 (published 12 Jan 1953), claimant Hattula Moholy-Nagy Hug, whose new matter is explicitly "introduction & translation & concluding notes." Renewed in the final eligible year. The German source may well be public domain; this English text is not.
The lesson for the ingest stint: publication date never settles a pre-1964 US work. The renewal record does, and it must be checked per edition.
Out of scope, next stint: bringing public-domain texts in at volume (Gutenberg / Archive.org ingest, rights verification, chunk formatting). The shelving must exist before it is worth filling, and the fields defined here are what the ingest will populate.
- Chapel separation holds. The Library never absorbs scripture.
- The three-layer law. A work's shelf and rhymes are CONTENT; the runtime follows. Curation never teaches the cortex about pericopes.
- Reverent degradation. A text that will not load is absent, not a broken entry — and per §4 absence is now observable to the maintainer.
- Public domain only, basis recorded per work (§4).
-
Authored boundaries are sacred. Ingested texts must not acquire
|phrase markers automatically; see PHRASE-CHUNKING-STUDY.md §4 — the runtime cannot yet tell an authored boundary from a derived one, and a bulk ingest that guessed would poison that distinction before it is built.
Sixteen works ingested from SOL's acquisitions dossier; 53 texts across five shelves. Form went from 0 to 6.
| shelf | texts | ingested |
|---|---|---|
| Form | 6 | 6 |
| The Interior | 10 | 3 |
| The Limit | 11 | 3 |
| The Recurrence | 8 | 4 |
| Composed | 19 | — |
Machinery: scripts/archive-ingest.mjs. A work declares its edition,
rights, source, and what its headings look like; shared code does the
rest. Every generated module carries the SHA-256 of the artifact as
fetched alongside the payload checksum — the first says this is the
file the world served us, the second says this is what we made of it.
Check the END of a text, not only the beginning. This caught six defects that a first-lines check would have missed entirely: Vitruvius ending on his editor's essay, Boethius running into Gutenberg's licence and then into the translator's citation list, Dow closing on "THE END", Ross on a printer's colophon and a paragraph index, Rasmussen on a list of tellers and page numbers.
Losing the first section is the quietest failure. Parker and Beckwith both lost tale 1 to a contents-skip neither edition needed. A collection that begins at its second story reads perfectly well and is simply wrong.
Structure is per-edition, not per-work. Five Gutenberg files state the same three things five different ways. Counting headings to find the body failed on Dow (six in contents, eight in the body); a contents block is contiguous and body headings are far apart, so DISTANCE is the signal. Boethius defeated the shared helper entirely and got an explicit parser rather than another option on the general one.
The reading unit is an editorial decision. Montaigne parsed into three books of ~940,000 characters each — sixty hours a section, which is a file rather than a reading. He is read one essay at a time; 102 of them.
Lazy loading is not optional at this scale. Six works statically imported took the content bundle to 1.63 MB. Metadata is now eager and text is lazy: the eager bundle is ~116 kB with sixteen works, and Montaigne alone is a 2.8 MB chunk nobody downloads unless they open him.
- Westervelt's Hawaiian legends — the plain-text edition has no machine-detectable chapter structure. Guessing would produce sections that are not the work's own divisions.
- The Upanishads and Hermetica remain withheld. Both need rebuilding from Müller and Mead alone, and no clean Gutenberg edition of either translation was found.
- The Heart Sutra and Gospel of Thomas cannot be cleared at all: Conze and Lambdin are modern, and the Nag Hammadi codices were unearthed in 1945.
- Marcus Aurelius (Selected) is withheld as a duplicate, awaiting a reading-route feature the Archive does not yet have.
- Twenty-seven of SOL's 43 cleared editions remain un-ingested, including the four requiring Internet Archive scans (Euclid, Alberti, Helmholtz, Owen Jones) and the image-dependent works SOL placed last.
Form · The Western Canon · The Eastern Canon · Indigenous Traditions · Composed. The functional shelves (Interior, Limit, Recurrence) are retired.
Organising by what a work does is a real property and it flattened a more important one. "The Limit" filed Lao Tzu beside Nietzsche and the Dhammapada beside Emily Dickinson — as though the tradition a mystic comes from were incidental to them. For a mystic it is most of what they are.
"The Recurrence" was worse. Four collected oral traditions sat with the Hermetica and the I Ching under a heading meaning the same intuition appearing where there was no contact — which turns living traditions into evidence for a thesis. That is precisely what SOL's dossier warned against, and the caveat had been written into the curation while the shelf that caused it was left standing. A warning in a comment does not fix a heading.
Eastern and Western name literate canonical traditions; Indigenous names a relationship to colonisation. Yuwaalaraay, Kalaallit, Cherokee, and Jamaican Maroon traditions share nothing with each other except that outsiders collected them.
The shelf is kept and named plainly, with the shared property stated in its own orientation rather than hidden: held in memory and speech, written down late — and each work's line says whose framing stands between the reader and the tradition.
Vitruvius, Dow, Ross, Crane, Kandinsky, and Dresser are Western by geography, but they are read as craft rather than as canon. Filing Principles of Decorative Design beside Thus Spoke Zarathustra would be the same error in the other direction.
sacred-wheel-of-time, filed as "Toltec Tradition, translator
Traditional," is Carlos Castaneda — the verses are verbatim don Juan
quotations from Tales of Power (1974). In copyright, and the deeper
problem is the attribution: Castaneda's don Juan is a twentieth-century
literary creation whose ethnographic basis is disputed by scholars of
the people it claims to describe. It could not simply move to the
Indigenous shelf; it had to leave.
Asking where a work belongs geographically is what surfaced it. The functional shelves never had to ask.
Each canon divides into Classical · Literary · Esoteric, in reading order: the founding texts, then what was written in their light, then where the tradition goes quiet and strange.
Deliberately the SAME three on every shelf. A reader learns the vocabulary once, and the parallel is itself a claim worth making — that every tradition has a classical core, a literature, and an esoteric edge. The Archive makes that claim by structure rather than by saying so.
| shelf | classical | literary | esoteric |
|---|---|---|---|
| Form | Vitruvius | Dow, Ross, Crane, Dresser | Kandinsky |
| Western | Marcus Aurelius, Epictetus, Boethius | Montaigne, Thoreau, Emerson, Rilke, Whitman, Dickinson, Nietzsche | Hermetica, Emerald Tablet, Blake, Julian |
| Eastern | Tao Te Ching, Gita, Dhammapada, Yoga Sutras, I Ching | Rumi, Kabir, Okakura | Zen koans |
| Indigenous | all four collections | — | — |
Some fits are looser than others, and the curation says which and why. The Tao Te Ching is CLASSICAL because it founds a tradition, not because it resembles Marcus Aurelius. Blake is ESOTERIC rather than literary because he claimed transcription rather than composition. Julian sits with the esoteric because a visionary is not a systematist.
Indigenous Traditions is all-classical and that is correct, not an omission: four collected oral corpora, with no literature written in their light and no esoteric commentary the Archive holds. A shelf with one division renders as a flat list rather than labelling the obvious.
"All" stays flat. Divisions across four canons at once would interleave traditions and say nothing; the grouping only means something when a reader is standing at one shelf.
Form & Design is a subject, not a fourth region. It sat in one filter
row beside Western, Eastern, and Indigenous, implying they were
alternatives. They are not — Vitruvius is Western AND a book about form.
Each shelf now declares an axis, and the panel shows two labelled
rows, mirroring the Collections panel for imagery.
The subject row filters on axis !== 'tradition', so a shelf that
forgets to declare one appears somewhere rather than vanishing from
both. A test asserts every shelf appears exactly once across the rows.
The subject axis is deliberately open. What belongs on it — Cosmos, Number, Craft, Attention — is a philosophical question rather than a filing one, and worth deciding rather than accreting.
The Western shelf holds thirteen essayists, poets, and philosophers, and no imaginative literature at all: no Homer, Dante, Chaucer, Shakespeare, Cervantes, Milton, Goethe, or Melville. Verified — all twelve names return absent.
This is a defect in the acquisitions brief, not in SOL's work. The brief asked for "aphorism, dialogue, verse, and the essay in its older sense" and warned that "prose that is merely informative reads poorly here." Each clause is true. Together they made short an unstated requirement, and SOL delivered exactly what was asked.
Paradise Lost is compressed. The Inferno is compressed. They are simply also long — and length was never the constraint, since payloads load lazily.
A second curation brief corrected it, and carried one instruction the first brief needed: do not let a stated preference become an unstated exclusion — if something in this brief would narrow the result in a way we have not noticed, say so before beginning.
It also asks SOL to challenge the divisions. If forty works of imaginative literature arrive, "Literary" holds both Montaigne's essays and Melville's novel, which are not the same kind of thing. The parallel across shelves is worth defending, but not against evidence.
Published from docs/vision/LIBRARY-SPEC.md. Edits made here are overwritten on the next push to main.
Start here
- COMPOSER-FIRST-ROADMAP
- VISION
- PROJECT-KNOWLEDGE
- ARCHITECTURE
- SYSTEM-DESIGN-REVIEW-2026-08-22
- ACCOUNT-SAVES
- RELEASING
- PLUS-CUSTOMER-LIFECYCLE
- AGENTS
Release
Pilot
- LIVE-CURRENT
- LIVE-HANDOFF
- LIVE-MCP
- DIRECTORY-SCREENSHOTS
- INSPECTOR-RUN-2026-10-10
- CLAUDE-RUN-2026-10-10
- CREATIVE-CONTROL-EVALS
- PERCEPTION-EVAL
- SOUND-LEVELS
- WORDING
- RUN-2026-10-08
- COMPOSER-INSTRUMENTS
- CHATGPT-DEMO
- EMBED-WITNESS
- RISE-GATE0
- CHATGPT-HOST-2026-10-04
- LIVE-002-HOST-2026-10-05
- EMBED-STAGE-HOST-2026-10-07
- RELEASE-GATES-2026-10-07
- EMBED-SAFETY-2026-10-05
- LIVE-GEMINI
- LIVE-EVALUATION
- LIVE-RED-TEAM
- CURRENT-CONSOLIDATION
- RISE-CHANGES
The Archive - texts and their editing
- ARCHIVE-CANON-SPEC
- ARCHIVE-CLEANSING-SPEC
- CORPUS-REVIEWER-PROMPT
- TYPESETTING-CANON
- PHRASE-CHUNKING-STUDY
- CHUNKER-AWARENESS-STUDY
- LIBRARY-SPEC
- SOL-PD-ACQUISITIONS-DOSSIER-LITERATURE-2026-07-28
Authoring - scores, rooms, and rendering
- EXPERIENCE-PROGRAM-SPEC
- LIVE-CURRENT-EVENTS-V1
- RISE-CURRENT-V1-SLICE
- RISE-SDK
- SCRIPTORIUM-SPEC
- WORKSHOP-COMPOSITION-STUDIO-SPEC
- AGENT-COMPOSITION-AND-RENDER-SPEC
- NARRATION-LANE-SPEC
- RECITATION-SPEC
- SCRIPTORIUM-STRENGTHENING-SPEC
Reading surfaces
- PAGE-MODE-SPEC
- RHYTHMIC-VISUAL-PRESENCE-SPEC
- LATERAL-TRAVERSAL-SPEC
- CONTINUOUS-FIELD-SPEC
- JEV-VARIANCE
- INVOCATION-SKINS
- PHASE-2-SAFETY-SPEC
- SPATIAL-CHAMBER-SPEC
- Premium_Mobile_Chamber
Imagery - what may be shown, and on whose authority
- MUSEUM-ATLAS
- SOURCE-CURATION-SPEC
- SOURCE-EXPANSION-SPEC
- ATRIUM-IMAGERY-SPEC
- PERICOPE-IMAGERY-SPEC
- icon-museum-request-draft
- TEXT-ATTUNED-IMAGERY-SPEC
The Chapel
Direction and unbuilt work
Affect and the Emotions room
Reading decision service and retired gate
Product direction and team tracking
- README
- READER-DIRECTION
- CONSOLIDATED-READER
- COMPOSER-FIRST-ROADMAP
- TRACKER-DESIGN
- 2026-10-03-direction-decisions
- 2026-10-04-composer-decision
- 2026-10-05-consolidated-reader-decisions
- 2026-10-05-embed-stage-decision
- 2026-10-06-score-speech-service-roadmap
- 2026-10-07-current-with-a-look
- 2026-10-07-reader-control-and-replay
- 2026-10-05-one-colour-vocabulary-design
- 2026-10-07-chrome-follows-theme
- 2026-10-07-gallery-follows-with-works
- 2026-10-05-canonical-home-design
- 2026-10-04-reader-evidence
- 2026-10-original-performance-roadmap
- 2026-10-08-voice-hardening-2-findings
- 2026-10-09-feelings-sounds-parked
- 2026-10-10-connector-directory-readiness
- 2026-10-10-connector-submission-packet
- 2026-10-16-mystery-school-complete-library
- 2026-10-13-mystery-school-plan