pipelines/SLICE_SCOPE_RECONCILE_2026-07-27.md
Read-only analysis, 2026-07-27. Axis: THE BUILD-ORDER RECONCILIATION. Not a ruling; a cost-and-collision
brief for the director. Every load-bearing claim carries its repo evidence. Where a number is an
extrapolation rather than a count, it is marked as one.
---
Josh's instinct is structurally right and his ruling is also right, and they are not actually in conflict —
because they are about two different things that the current plan conflates into one word, "slice."
currently has no relationship with, and the amulet chain he just authored has no opening without the
Prologue. That hole is real, it is measurable, and it exists today.
has no GIS path (the terrain chain that made six maps affordable has no input for a dimensional region),
and its power level is comparative — it can only be calibrated against a build space that does not exist
yet.
one of the four that separates them, and it is already 80 percent written into the ratified roadmap.
creation selections are unbuilt (P2.11, BUILD-NOW, unranked) and every shape owes an answer. Ch 13 — the
chapter Josh named — is one of the two chapters those selections BIND (Ntala's gender and starting build).
Shipping to Ch 13 without resolving P2.11 ships the binding without the bound.
Ch 13 is confirmed as the healer's-child join. docs/spine/CH_13.md — Ntala, CHAR_0002, joins as active
companion, the journal opens in her hand, hard line HL_0070. Josh's memory of the chapter number is exact.
---
MapForChapter in HumanityQuestSliceDriver.cpp routesCH_02/CH_03 to FloresIsland, CH_04 to Bali, CH_05 to SamosirToba; BorobudurMerapi, PalembangMusi and
AngkorKhmer exist as staged Ch-5 sites outside the router. Two cross-map seams (IsCrossMapSeam:
CH_03→CH_04, CH_04→CH_05).
SliceChapterId = "CH_02" is a static constant and NextChapterFor is a chain. Adding chapters — at thefront or the back — is a data-and-content operation, not an architectural one. The engine does not resist
a prepended Prologue.
page_status Research Phase and env_profile_status stub.Source/ returns onlylandscape-weightmap false positives).
| Span | Chapters | Regions | Care tiers | Hard-line nodes |
|---|---|---|---|---|
| Ch 2-5 (today) | 4 | 4 (all have DEMs) | standard ×4 | 0 |
| Ch 6-13 (the add) | 8 | 8 (all pending, env blank) | standard ×3, elevated ×4, high ×1 | 1 (Ch 8) |
| Prologue + Ch 1 | 2 | 1 (no DEM — dimensional) | none (invented) | 2 |
Sources: docs/spine/CH_NN.md cultural_depiction_tier; T0_Chapter_Index canonical_hard_line_flag;
T0_Region_Index coordinates / page_status / env_profile_status.
This table is the analysis in miniature. The current slice is the cheapest four chapters in the game on the
care axis — every one standard, zero hard-line nodes. Ch 6-13 is not three times the work; it is three times
the regions carrying five elevated-or-high chapters and a hard-line node. The roadmap's own P1 honest-risk
clause already says the high-care load "does not parallelize as cheaply and will pace their own regions"
(ROADMAP_POST_5090_TO_SHIP.md P5 risks). Ch 10 at HIGH is the /Gwi San — a living Indigenous community —
and the DEEP_BOOK contract's recommendation is per-record author review for the high-care three.
From docs/spine/CH_PROLOGUE.md and T0_Region_Index @ fairy_realm:
Familiar Communion — that exist nowhere else in the built or planned ability model.
Ch 1's BE_01 is verb ENDURE, "no win and no fail-wall." The shipped combat model is HP-lethal plus
composure-break plus withdrawal (D-DEFEAT-AXIS, ruled). Neither SEAL nor ENDURE is expressible in it today.
fairy_realm.coordinates = "N/A (dimensional)". The whole deterministic chain — derive_region_bbox → Copernicus GLO-30 → Gaea erosion → WorldCover weightmaps (PRE_5090_BUILD_PLAN.md
W4.2-W4.5) — takes a bounding box of a real place as input. The fairy realm has none. Every other region
in the game, including all nine mythological realms' earthly approaches, is a real place with a DEM. The
thing that made six maps affordable does not apply here at all.
env_profile_status stub, and per the realm charter no vril-site row existsfor the fairy realm at the Prologue.
These are the honest case for a longer slice, and they are stronger than the emotional framing alone.
Hole 1 — the Ch-2 opening beat is currently hollow. docs/spine/CH_02.md CH02_B01: "Wake on cold stone
in the dark of Liang Bua after the rift-fall; the Child Familiar bites once, bolts, and is watched dying from
a crack in the rock — arrival as loss." The Child Familiar is the player's own Ch-1 selection from twelve
(T3_Core §2.2). In shape (a) the game opens by asking the player to grieve a creature they never chose,
never named, and never met — and then holds that grief for fifty-six chapters to the Ch-58 reveal. The beat
is designed as a payoff of a bond the shipped product does not create.
Hole 2 — the four selections are unbuilt and Ch 13 binds them. P2.11 "Character creation — four
selections only" sits in PRE_5090_BUILD_PLAN.md as BUILD-NOW and is absent from the ranked 30. Its four
outputs all bind forward: the Child Familiar to the Ch-58 slot, the best-friend gender to **Ntala's gender
and starting build at Ch 13**, appearance to the voice preset, and the name. Two of the four bind inside
Ch 2-13. So shape (b) — earthly-only to Ch 13 — ships the chapter whose companion is defined by a selection
the player was never offered.
There is also a live tension inside P2.11 itself: it is specified as a selection screen ("Writes base
BiographyState"), while Ch-1 canon rules the same four selections "woven as childhood, not a menu"
(docs/spine/CH_01.md header, T3_Core §2.2). Those are two different implementations of one canon. Shape
(d) is the only shape that needs both, and — see §4.3 — that turns out to be a feature.
---
The ruling, verbatim from the charter's restatement of Josh's directive: "the fairy realm built last after
all mythological realms, with the Prologue and Ch 1 game-build last of all"
(REALM_DESIGN_PROGRAM.md §0.1). Its reasoning is §0.2 bullet 3, and it is not scheduling:
The Prologue and Ch 1 are the only scenes that show the fairy realm's physics as ORDINARY, felt from
inside by a being who has never known anything else. That scene cannot be authored until the nine partial
views exist to be the thing it is ordinary relative to.
Two distinct claims live inside that. Separating them is the whole reconciliation.
claim about the text.
the whole arc: demonstrated and FELT so the seventy-five-chapter climb back to it lands." This is a
calibration claim about the systems.
The refraction claim is partially discharged already: the Prologue and Ch-1 spine entries are authored
(249 and 261 lines, spine_status drafted), and Josh authored the emotional spine on 2026-07-27. The ceiling
claim is not discharged and cannot be discharged early — see §3.2.
Collision: none. Fully compliant with the ruling.
protagonist. That is survivable — Early Access buys narrative latitude that a full release does not — but
it should be a chosen cost, not an unnoticed one.
Collision: none with the fairy-realm ruling. This shape never touches the origin pair. Its collision is
with the density curve, the care doctrine, and the P0 book intake — not with the build order.
Cost, itemized against the current build:
page_status pending with blank environment profiles.genuinely superlinear, because per-record review does not fan out.
Sumatra/Java → Ch 6 Sri Lanka → Ch 7 South India → Ch 8 Swahili Coast is an Indian Ocean crossing. Fork C
ruled tiered traversal with playable compressed crossings and "important crossings become staged/open
dungeons"; an ocean crossing is the hardest instance of that ruling and the slice has never built one.
ledger holds 264 UNROUTED seeds from four regions (QUEST_SEED_HOUSING_LEDGER.md — 400 rows, 264 seeds);
at the same per-region density, eight more regions imply roughly 500 additional seeds to type and author.
*(Extrapolation, explicitly marked — the real number is a P1 output, not a desk number.)*
What shape (b) buys: Hole 1 stays open, Hole 2 gets WORSE (Ch 13's companion is bound to an unoffered
selection). It buys the healer's-child join and the amulet gift beat's natural window — but the gift beat's
emotional charge is the guilt engine closing, and the guilt engine OPENS in the Prologue. Handing Ntala the
amulet in a game with no Prologue is handing her a nice item.
That is the honest verdict on (b): it triples the most expensive axis in the plan to reach a beat whose
payoff is gated on the content it deliberately excludes.
Collision: direct, total, and on the ruling's strongest reasoning.
This requires the fairy realm EARLY. Working it honestly rather than dismissing it:
contributes nothing and 100 percent of the geography is authored or generated.
seven non-pilot boards.
And the ceiling problem, which is the part that cannot be bought with effort. See §3.
Shape (c) is not "the slice plus ten chapters." It is the slice, plus the eight most care-expensive earthly
chapters, plus the single most expensive art surface in the game, plus two novel systems — landed before
the pipeline has run its own practice laps. The charter's own §9 objection already concedes the honest form
of this: "the four later lenses and the volume of board work genuinely should not compete with the Ch 2-5
slice."
Collision: none. It is the ruling's build order expressed as a release schedule.
And it is largely already ratified. ROADMAP_POST_5090_TO_SHIP.md P4 rules Early Access; P5 states "each
finished chapter (or chapter batch) ships as a Steam content update through the P4 chain." Shape (d) adds
exactly one thing to the ratified plan: the assertion that the Prologue and Ch 1 are the LAST drop rather
than an unscheduled one — which is the build-order ruling restated in release terms.
Details in §4.
---
Josh's question implies one: the Prologue's INTERIOR or castle scope only, with the realm splendor deferred.
Worked seriously, because it is the right question to ask.
The Prologue's ceiling is expressed AS the realm's atmosphere, not as anything containable in a room. The
spine's own prose is explicit and it is the design, not decoration:
saturating everything, so that to breathe here is to be full."
own saturation and anything that should not be there simply thins and fails."
Sovereign Vril Projection, Temporal Sight and Rift Sealing are all expressed through open realm space.
Familiar Communion is "twelve kinds of great luminous animals that are not his pets but his peers." An
interior-scoped Prologue delivers the birth and cuts every faculty that makes the ceiling legible. It would
ship the emotional content and drop the structural function — and the structural function is the reason the
Prologue exists.
Verdict: interior-only is not a reduced Prologue. It is a different scene that happens to contain a birth.
Two independent failures, and either one is disqualifying.
makes the player's first thirty minutes a grey approximation of the place the game claims is the most
beautiful in it — and the AAAAA posture is an OUTPUT-defined tier. Every other surface can be greybox in
private because the QA fleet is the only player (THE JOSH GATE). A shipped Prologue cannot.
higher than everything in seventy-five chapters, felt. Calibrating that against four chapters of ability
content (the foundry's eleven abilities, Ch-5's honest zero) sets the ceiling against roughly five percent
of the build space. Then the AAAAA balance doctrine lands hundreds-to-thousands of viable builds under it
and the Phase-5M/5H build-space sweeps measure breadth against it. A ceiling set early is a ceiling
re-authored later — and for a set-piece, re-authoring is not re-tuning a number, it is rebuilding the
scene.
**The ORIGIN PRELUDE: deliver the emotional spine early as a non-realtime surface; build the playable
origin pair last, replacing the prelude in place.**
What it preserves:
Familiar, meets the friend, and falls — so Ch 2's opening beat lands as designed.
calibrated, because a cinematic ceiling is a promise, not a measurement — which is precisely the
distinction the ruling's reasoning protects.
chain and the eleven scene-type templates; only generated participants and video generation are 5090-gated.
Its real wall, stated honestly: a prelude at AAAAA bar still needs the fairy realm's VISUAL GRAMMAR — the
art bible the charter schedules after all nine road realms.
But there is a non-obvious asymmetry that makes this cheaper than it looks. Per the charter's own
generation_class axis (§ costume rule), the fairy realm is invented — "no real source tradition exists
to be misrepresented. Lane G legal on Panel A and Panel B. Atlantis and the fairy realm, and nothing else on
the road." Every other realm needs reference assembly, a rights register, and L-R3's living-tradition
determination before a single plate is legal on the accuracy panel. The fairy realm needs none of it.
The fairy realm is research-CHEAP and render-EXPENSIVE. The build-last ruling is grounded in render and
in refraction. It is not grounded in research. So the fairy realm's art bible can ride early — as a third
pilot alongside Atlantis and the Duat, and arguably the cheapest of the three to research — without
violating the ruling's own logic. What genuinely cannot ride early is the realm's PHYSICS-AS-ORDINARY
framing, which is the refraction claim and which the charter is right about.
This is the seam. Art bible early, playable realm last, prelude in between.
---
exactly the ratified D-STEAM-FORMAT shape. Plus, per §1.4, an answer to P2.11.
says this. Ch 6-13 is a natural first batch — it completes the African arc, lands the healer's-child join,
and gives the amulet gift beat its window.
build-order ruling and the release schedule become the same statement.
P5 already carries the per-chapter factory, the density-curve scorecard, and content-update shipping. What
shape (d) adds is not machinery but two pieces of sequencing discipline:
to the end of the game. This one prepends to the front and re-frames everything before it. It is the only
drop that changes the meaning of shipped content rather than adding to it. The P5 scorecard's care/reveal
row is the natural home for the check that the prepend does not violate a reveal window that shipped
chapters established.
(d)'s honest framing is stronger than shape (a)'s, not weaker: "you are beginning in the middle; the
beginning arrives last" is a real EA promise with a real delivery date attached to a real build order.
This is shape (d)'s genuine upside and it is not available to any other shape.
The doctrine "built last and read first" is, in a staged release, a live event. On finale-drop day the game
reveals its own beginning, and every existing player has a reason to start over — not a marketing
contrivance but the literal structure of the work. The Prologue reframes the whole arc in hindsight
(CH_PROLOGUE.md: "reads as innocent childhood/kingship on first play, and only in HINDSIGHT reframes the
whole arc"), and by finale-drop day the EA audience has the hindsight. They are the only audience that will
ever experience the Prologue the way the doctrine intends and also already know what it means.
Mechanically it is cheap, because the architecture already supports it:
UQuestSaveGame.ChapterId), so an existing save simply never enters a chapterahead of its pointer. A prepended Prologue is new-game-only by default — which is the correct default.
SliceChapterId is a constant and NextChapterFor is a chain. Prepending is a data edit.without discarding progress. Small, and best specified before players have saves rather than after.
Shape (d) needs both implementations of the four selections, and this is where that becomes an asset:
BiographyState). Necessary regardless, because Ch 2 needs a chosen familiar and Ch 13 needs a bound
best-friend gender.
returning-player skip path — the standard shape for a game that adds a prologue.
Neither is wasted, and the canon ruling ("not a menu") is honored in the version that is canon, at the time
that version exists.
Stated plainly, because this brief is not an advocacy document.
early-game chapters deliberately positioned as "lighter, seed-rich" on the density curve is a thin first
impression at any price, including twenty dollars.
factory run to be kept.
---
The plan currently uses one word for two things, and every shape reads differently once they are separated.
does it land, is it fun, is the care right. Governed by the Josh Gate ruling.
does it stand alone, does it justify its price, does it read as the honest start of something.
Today they are welded: the Josh Gate sits at P3-exit → P4-entry and "his verdict gates the P4 Steam
listing." Nothing requires them to be the same content, and the arguments for their sizes point in
different directions.
polish per chapter, not chapter count. Ch 2-5 is well sized. Making it twelve chapters delays his first
contact by the whole Ch 6-13 program and tests nothing extra.
changed, and coherence is about where it starts and stops, not how many chapters it has.
What each shape means for EACH, stated separately:
| Shape | Josh-gate slice | Ship slice (EA base) |
|---|---|---|
| (a) | Ch 2-5. Unchanged, earliest possible date. | Ch 2-5. Thin; opens on a hollow beat; owes P2.11. |
| (b) | Ch 2-13 — delayed by the full 8-chapter program, five of them elevated-or-high care. Tests no more ceiling than Ch 2-5. | Ch 2-13. Substantial. Still no origin; Hole 2 sharpens at Ch 13. |
| (c) | Prologue-13 — delayed furthest, and gates on a ceiling that cannot be calibrated yet. | Complete-feeling, at the cost of shipping the ceiling early and badly. |
| (d) | Ch 2-5, unchanged — the gate keeps its earliest date and its correct size. | Ch 2-5 at launch, growing to Prologue-77. The ship slice becomes a schedule, not a size. |
The clarifying observation: shape (d) is the only one that lets the two definitions take their own natural
sizes instead of forcing one compromise. It is also the only one under which Josh's first playthrough date
does not move at all — and under the Josh Gate, that date is the date the human backstop finally arrives
after a program that has run entirely without one.
A recommended variant worth naming: decouple them explicitly and let the Josh gate fire TWICE. Once at
Ch 2-5 P3-exit (the ceiling judgement, gating the EA listing) and again at the finale drop (the origin
judgement, on content whose care register — the King, the mother, the fall — is his own authored material
and the thing he most deserves to rule on from play).
---
Against PRE_5090_BUILD_PLAN.md ▶ THE PRE-5090 WINDOW PLAN (ranked 30, 29 of 30 pre-5090-workable).
Ranks 1-30 stand as ordered. The one addition below is owed regardless.
The ranks stay in dependency order; several triple in size and cannot complete inside the two-week window.
two, including the Indian Ocean crossing, against Fork C's staged/open-dungeon ruling for important
crossings.
page_status pending.their seeds do not exist until P0/P1. This is the honest reason shape (b) cannot be front-loaded into the
handshake window even if Josh rules it today.
the charter's own §9 answer argues against.
freeze-scheduled.
This is the finding with the highest ratio of value to cost, and it holds under every shape:
currently unranked, and it is owed by (a), (b), (c) and (d) alike. Two of its four outputs bind inside
Ch 2-13 (the Ch-58 familiar slot; Ntala's gender and starting build at Ch 13 — the exact chapter Josh
named). The Ch-2 opening beat is unshippable without the familiar selection in any honest reading.
future front-prepended Prologue is reachable by an existing save. Cheap now (SliceChapterId is a
constant, NextChapterFor is a chain, saves are chapter-keyed); expensive after players have saves. This
is the one piece of shape (d) that genuinely wants to land before the EA base ships, and it is small.
Everything else in shape (d) is a P4/P5 sequencing decision with no pre-5090 cost. The window plan does not
re-prioritize.
---
Shape (d), with the origin prelude, and the two slice definitions decoupled.
window — and by then the Prologue's emotional content has reached the player via the prelude, so the gift
beat's guilt engine actually has its opening.
generation_class = invented — theone realm needing no reference assembly, no rights register, and no L-R3 determination.
non-realtime surface over the P2.7 Sequencer chain, closing Hole 1 without calibrating the ceiling.
prelude in place, and triggering the replay-from-the-beginning event that only a staged release can stage.
This honors both of Josh's positions rather than trading one against the other. His instinct — that the
slice needs its emotional entry — is satisfied at the prelude. His ruling — that the origin pair builds last,
at the grandest-display bar, after the nine realms — is satisfied exactly, with its own reasoning intact.
A cinematic prelude may be the worst of both worlds. It spends real budget on a surface that is
explicitly temporary; it delivers the game's most important emotional content in the one register the whole
design philosophy distrusts — the non-playable one, in a project whose Pillar-3 hard line is that every
chapter is playable and whose §17.9 discipline is show-don't-tell; and it risks the finale drop landing as a
remake of something players already saw rather than a revelation. If the prelude is good, it steals the
finale's thunder. If it is bad, it poisons the opening. Josh's ruling that these scenes must be the
grandest display is, read strictly, an argument that a lesser version of them should not exist at all — and
"ship nothing until it is the grandest display" is a completely coherent position that shape (a) already
implements.
The honest answer to the objection, and where it lands. It lands on the prelude's SCOPE, not on its
existence. A prelude that dramatizes the King's ceiling is the objection's target and should not be built. A
prelude that stays with the child — the amulet passed at the birth, the mother, the run-back, the friend,
the fall — shows none of the ceiling, spends little, and closes Hole 1 completely. The finale drop then adds
the thing the prelude never showed: the realm, the King, the sovereign faculties, the seal. That is not a
remake. It is the same event from the other side of the loss, which is what the doctrine says the origin is.
If Josh rejects the prelude entirely, the fallback is shape (a) plus P2.11 plus the prepend-safe entry
architecture — the smallest honest thing, with the origin arriving whole and last. That fallback is also
defensible, and it is what the ruling implements as written.
---
here and the factory cadence is explicitly a range the first post-slice chapters must MEASURE.
roughly 500 more for eight regions is an extrapolation and is marked as one at every appearance.
twelve chapters is a ratio applied forward.
are all Josh's, and this brief reads them rather than amending them. Shape (d) is presented because it is
the shape under which none of them has to move.
AMULET_PROVENANCE_BRIEF.md's own terms, and this analysis consumes it as ruled direction, never as a
landed row.