Door and window are placed via preset and moved with a bespoke wall-bound
`move-tool` (affordanceTools.move), not the draw tool — so the previous inline
triangle never showed for the paths actually used. Migrate both:
- move-tool (move + preset, the community path): publish the on-wall ghost pose
to `useFacingPose` in the same building-local frame the ghost renders in,
dropped to the floor under the wall (the ghost Y is the opening centre); clear
on every off-wall / hide / reveal / unmount path
- draw tool (standalone from-scratch path): publish the on-host pose, clear on
fallback/hide; frame depth read via a ref to keep the setup effect deps clean
Removes the now-dead `FacingIndicator` public export (the editor-side overlay is
its only consumer, via relative import). The unified overlay now covers every
placement/move path: items, column/shelf, stair, and door/window.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Every placement and move path now publishes its ghost pose to a single
`useFacingPose` store, drawn by one editor-side `<FacingPoseIndicator>` overlay,
instead of each path drawing its own triangle (which left the nodes-package and
PlacementBox paths invisible):
- column/shelf presets + all moves (PlacementBox via move-registry, and
DragBoundingBox) now publish the pose, so the triangle finally shows
- stair create + move use a declarative `facingIndicator: { reversed: true }`
(new registry resolver) so the triangle sits before the entry pointing out —
resolved in one place, so create and move match automatically
- stair placement defaults to single and respects the shared `point`
continuation (C) toggle, like the other placement tools
Checkpoint on the placement-interaction epic: also carries the in-flight
continuation-profile extraction (lib/continuation), grid surface (item #8), and
HUD work. Door/window still render their own legacy inline triangle and are
migrated to the overlay next.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
interaction-scope.md + review-architecture skill: the sanctioned Alt-as-toggle
(wall/fence chain mode, the one place Alt-as-force is meaningless), and a second
known-legacy pattern — `event.altKey` alignment-bypass in the roof/polygon/slab
previews + ceiling/slab snap paths (migrate-on-touch; wall+fence already done).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the legacy held-Alt mechanism on wall and fence drafting with a
mode toggle, mirroring the snapping-mode chip:
- Wall: `wallChainMode` room (auto-close on loop) / single. Room mode
finishes automatically when the new endpoint lands within the join-snap
radius of the chain's first vertex; single commits one wall per click.
- Fence: `fenceChainMode` continuous (chain until double-click/Esc) /
single. Fences are linear barriers, so continuous has no auto-close.
- Both: Alt-tap cycles the active drafting tool's chain mode (clean-tap,
scoped to wall/fence drafting); a clickable HUD chip shows the mode.
Persisted + migrated in `useEditor`.
Migrate wall and fence off held-Alt-bypass-alignment to the unified
convention: alignment now follows the magnetic snap mode, which frees Alt
for the toggle. 2D floorplan parity kept in sync with the 3D tools.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The floor grab-offset comment claimed floorStrategy.move reads localPosition
"under Alt (free place)"; it reads event.position with mode-governed snapToGrid
and has no Alt branch (Alt is force-place-only). Describe the real reason both
frames carry the offset: it's computed local-space but the strategy consumes the
world point.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The snapping model (Shift = cycle mode, Alt = force/free, mode-driven reads via
isGridSnapActive/isMagneticSnapActive/isAngleSnapActive, snapProfile-declared
context) lived only in code and the plan; tools.md still preached the legacy
"Shift = bypass snapping". Close the drift so the architecture review refuses
tool changes that revert to the old pattern:
- tools.md: replace the held-Shift-bypass manipulation policy with the unified
mode-driven model + the single snap read path.
- interaction-scope.md: new "Snapping mode & modifiers" section (contexts, read
path, modifiers, the chip-needs-a-scope rule) + a Rules bullet + the
known-legacy MEP movers (migrate-on-touch) incl. the dual-path constraint
(a bespoke mover must not open a `moving` scope — it re-mounts the generic
mover via useMovingNode).
- review-architecture skill: add interaction-scope.md to the reads and a new
"F. Interaction scope, snapping & modifiers" checklist — new shiftKey-bypass,
ungated grid step, missing snapProfile, a new useEditor interaction flag, or a
bespoke mover opening a moving scope are blockers; touching the legacy MEP
movers forces migration.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The MEP run/fitting/terminal tools anchored their cursor readout pill at
~+0.35m above the placement point, so with the tall CursorSphere line
(badge at +2.7m) the pill sat right on the cursor and overlapped it,
especially when zoomed in. Editor main (#438) already raised the duct
pill to +1.45m; this long-lived branch predates that merge. Bring every
MEP tool that uses the tall cursor onto the same +1.45m anchor:
duct/pipe/liquid/lineset runs, duct/pipe fittings, and duct-terminal.
hvac-equipment (height-aware anchor) and pipe-trap (no cursor line) are
left as-is — their HUDs already clear the ghost.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The last two hot per-pointer-move 2D states still lived in FloorplanPanel's
useState, re-rendering the ~10k-line panel on every move:
- marquee (box-select): the whole drag struct moves to a dedicated
use-floorplan-marquee store; down/move/up/cancel read+write it via
getState() (panel holds nothing), and a FloorplanMarqueeOverlay leaf
subscribes to the moving corner and renders the rect alone. Drops the 3
bounds memos + the useState.
- reference-scale: the rubber-band's moving end was always equal to the
shared cursorPoint (written every move anyway), so drop the `cursor` field
from the draft and read it from useFloorplanDraftPreview in a new
FloorplanReferenceScaleDraftLine leaf. The draft now carries only the
per-click guide + start anchor, so it no longer re-renders the panel.
Closes out the 2D edition perf pass — every build/edit/select hot path now
writes a store, not panel state.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Roof-segment edit/move ignored the active snap mode and showed no chip:
- add snapProfile:'structural' so a body-move resolves the no-angle polygon
context (grid/lines/off) like every other structural move;
- resize uses getSegmentGridStep() (0 outside grid mode = the "smooth" resize
that used to need a held Shift), dropping the captured gridSnapStep + Shift;
- move drops its Shift bypass;
- the affordance dispatcher opens a boundary reshape scope for the resize so
the snapping chip shows and the context resolves.
2D rotation handles now match the 3D gizmo across all six rotate affordances
(column / elevator / roof-segment / shelf / spawn / stair): a shared
rotateAffordanceDelta snaps to the 15° step unless Shift (free), the
dispatcher opens the same ROTATE_HANDLE_DRAG_LABEL handle-drag scope the 3D
gizmo uses so the contextual HUD shows the "Shift = rotate freely" hint, and
the live degree readout snaps to match the committed rotation.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wall and fence endpoint-move affordances hard-snapped to the grid via a
hardcoded WALL_GRID_STEP and always ran Figma line-alignment, ignoring the
active snapping mode. Now:
- grid step follows getSegmentGridStep() (0 outside grid mode), so lines /
angles / off no longer force a grid snap the mode chip says is inactive;
- Figma alignment is gated on isMagneticSnapActive() (the lines mode);
- angles mode angle-locks the endpoint off the fixed corner (free length),
mirroring the draft tool;
- fence drops its legacy Shift-bypass to match the wall's unified model
(Alt stays as linked-segment detach).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The 2D affordance dispatcher (`startAffordanceDrag`) now begins the matching
reshaping interaction scope (boundary / hole / curve / endpoint) on pointer-down
and tears it down on release/cancel, matched by node id. This makes the
contextual snapping HUD show the right chip during polygon vertex/edge and wall
endpoint/curve edits, and lets `getActiveSnapContext()` resolve the correct
per-context snapping mode the affordance snap math already reads.
Wall creation is owned by the 3D `WallTool`, which is dead in 2D-only view
(canvas `display:none`). Mirror the slab/ceiling 2D-only committers: commit
locally via `createWallOnCurrentLevel`, gated on `viewMode === '2d'`, chaining
the next segment from the committed wall's resolved end. Split/3D keep their
single-owner tool commit.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Same fix as slab: ceiling is committed by its 3D registry tool, dead in 2D-only
view. Commit it from the panel on both close paths, gated to viewMode==='2d'.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Slab is committed by its 3D registry tool, which accumulates the grid:click
vertices the 2D panel emits and commits on close. That path is dead in 2D-only
view — the 3D canvas is display:none, so the tool never commits and the slab is
never created (split/3D work because the 3D side is live). Mirror the zone
pattern: the panel commits the slab itself on both close paths (double-click +
click-first-vertex), gated to viewMode==='2d' so split/3D keep their single-owner
tool commit (no double-create).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The shared polygon-vertex affordance snapped via snapPointToGrid(rawPoint), whose
default step is the hardcoded WALL_GRID_STEP (0.5m) — so slab/zone/ceiling vertex,
edge, and add-vertex edits always quantized to half-meters regardless of the
active mode OR the user's grid-step setting (plan open bugs #1-2). Use the
mode-aware getSegmentGridStep() (0 in non-grid modes) so grid quantizes to the
live step, lines/off pass through to the wall-snap/alignment resolver. Drop the
legacy shiftKey bypass from the slab/ceiling magnetic resolvers (they already
gate on isMagneticSnapActive).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Perf: move the per-move wall/fence/roof draft END points into
useFloorplanDraftPreview; a new FloorplanLinearDraftLayer leaf owns the live
draft polygon + fence segment + wall measurement, subscribing to the store. The
shared FloorplanDraftLayer keeps only the per-click anchors. Wall/fence/roof
drafts now have zero per-move panel setState — buttery smooth like slab/zone.
Parity: migrate the remaining legacy Shift=bypass paths to the unified
mode-driven model. roof (move + click) honored only always-grid + bypassSnap —
now grid/lines/off (footprint → no angle). wall + fence click-commit still used
the legacy bypass while their move-preview didn't — now consistent. Wall Alt
stays 'commit single wall' (open product decision, untouched).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The polygon-draft snap path used the legacy model: bypassSnap = shiftPressed and
angleSnap = pointCount > 0 && !bypassSnap, so the 15deg angle lock engaged after
the first vertex regardless of mode — hijacking grid/lines/off into angle-snap
even though the HUD chip showed the right mode.
Migrate all three placement paths (move preview, single-click vertex, double-
click close) to the unified model: angleSnap = isAngleSnapActive(); grid flows
through snapToHalf (step 0 in non-grid modes); wall-snap/alignment already gates
on isMagneticSnapActive(). Behavior now matches the chip — grid quantizes,
angles locks 15deg rays, lines snaps to walls/alignment, off is free. Drop the
now-dead bypassSnap param from snapPolygonDraftPoint.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The 2D build/edit tools republish the snapped cursor point AND the screen-space
coordinate-badge position on every pointer move. Both lived in FloorplanPanel
useState, so each move re-rendered the whole ~310ms panel (the badge fires on
every pointermove while any build tool is active — the dominant culprit).
Move both into a useFloorplanDraftPreview store written via getState() (no panel
re-render); render the crosshair + live polygon-draft edge from a
FloorplanDraftCursorLayer leaf and the coordinate badge from a
FloorplanCursorIndicator leaf, each subscribing to the store. Same store+leaf
pattern as the stair build preview. Slab/zone/ceiling drafts now have zero
per-move panel setState. Wall/fence/roof draftEnd are follow-up slices.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The stair tool held its 2D build preview in FloorplanPanel useState, so every
grid:move re-rendered the whole ~120-220ms panel. Move the preview into a
dedicated useStairBuildPreview store written via getState() (no panel re-render)
and render it from a FloorplanStairBuildPreviewLayer leaf that subscribes to the
store directly — the same pattern that keeps column/elevator placement smooth.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Migrate roof/stair/elevator draft tools + the 2D floorplan move overlay off the
legacy Shift/Alt=bypass model onto mode-driven snapping (isGridSnapActive /
isMagneticSnapActive); Alt dropped (no validity gate). stair/elevator now use the
live grid step.
These three are placed as footprints, not directional draws, so the angle-lock
mode was meaningless: add NodeDefinition.snapDraftDirectional (default true; false
for roof/stair/elevator) so their draft resolves to the no-angle 'polygon' context
(grid / lines / off). snapContextOf takes an injected draftDirectionalOf, like
profileOf. Add toolHints to stair/elevator so they route through the contextual
HUD and show the snapping chip. Fix one stale Alt-bypass comment in the item
placement coordinator (#9: already force-only). +snapping-mode test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The 5 bespoke MEP movers (duct/pipe-segment, liquid-line, lineset, duct-fitting)
now read the active snapping mode (isGridSnapActive / isMagneticSnapActive)
instead of shiftKey=bypass. The moving scope already carries the node
(setMovingNode → begin('moving')), so the per-kind context resolves with no
extra wiring. Grid and alignment are now independent reads.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Migrate all 9 MEP kinds' placement tools onto the unified snapping model:
declare snapProfile ('item' for point-placed hvac-equipment / duct-terminal /
duct-fitting / pipe-fitting / pipe-trap; 'structural' for directional runs
duct-segment / pipe-segment / liquid-line / lineset), and replace the legacy
shiftKey-bypass reads with mode-driven isGridSnapActive / isMagneticSnapActive /
isAngleSnapActive. For runs the 45° lock becomes the cyclable 'angles' mode;
Alt stays the vertical-riser modifier (run drafting has no validity gate to
force). Port mating gated on "mode != off". Dropped stale "⇧ smooth/free" hints.
The bespoke MEP move-tool/selection (endpoint) tools stay on the legacy model —
they use setMovingNode(null) so no moving-scope context resolves yet; migrating
them needs scope-wiring first (follow-up).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The 3D MoveDoor/MoveWindow tools wrote useScene every frame during a move
(freeFollowAt + applyPreview alternating): the wall:move (R3F) / grid:move
(DOM) de-dup compared event.timeStamp across two event systems with different
clocks, so it never matched and the floor free-follow ran during on-wall
slides too, ping-ponging the host and churning the nodes ref → framerate
collapse in both 2D and 3D. Replace it with a single-clock wall-ownership
window (performance.now, ~4 frames): the floor follow stands down while a
wall/roof hit is fresh. On-wall slides now write no scene per frame (mesh +
useLiveTransforms only). Lower the live wall-cutout throttle 120→60ms now that
the per-frame churn is gone.
Also completes the door/window modifier-model migration (#10): Shift=cycle /
Alt=force-place, fully mode-driven snap, snapProfile:'item'; exclude
ground-line candidates from along-wall opening alignment; emit the move SFX
once per snapped step.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Migrate the remaining floor-placed kinds onto the unified snapping/modifier
model and generalize floor collision so any solid floor kind blocks any other.
- shelf/column/spawn declare `snapProfile: 'item'` → contextual snapping chip,
Shift=cycle, Ctrl=grid step during placement; their tools read the active
mode (grid/lines/off) instead of legacy Shift/Alt bypass; spawn fresh
placement now respects alignment ("lines") like its move.
- Resize/radial handles claim the handle-drag scope (new RESIZE_HANDLE_DRAG_LABEL)
so the HUD shows no select-mode shortcuts mid-resize.
- Column move migrated to the generic MoveRegistryNodeTool (declare `movable`,
drop the bespoke move-tool) — gains mode-driven snapping, alignment, R/T,
slab lift, grid SFX, and the collision box for free. 2D move still routes
through `floorplanMoveTarget`.
- Cross-kind floor collision: new declarative `FloorPlacedConfig.collides`
(item/shelf/column opt in; spawn/MEP/stair stay off). `canPlaceOnFloor` now
treats every colliding floor kind as an obstacle (was item-only), reading the
declarative footprint; the generic move tool's red/green placement box gates
on `collides`. Column footprint uses the visible `columnFootprintHalf` extent
so the box/slab-lift/collision track the real (round/square) column size.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Generify the snapping/modifier HUD off the FSM scope and node declarations
instead of wall-creation-shaped, leaking pills.
- Per-context snapping (`snappingModeByContext`, persisted): wall / item /
polygon mode-sets with exclusive modes (grid | lines | angles | off), each
doing exactly what its chip says. Context is node-declared via the new
`NodeDefinition.snapProfile` ('item' | 'structural'); the resolver maps
(profile × action) → context with no per-kind switch.
- Scope-driven HUD: helper-manager reads the interaction scope; reshaping
(endpoint/curve/boundary) and item move get their own chip, no select-hint
leak. Rotate R/T rounds to 45°; Alt = force-place only (hidden for
structural kinds); Shift = cycle everywhere.
- Slab/ceiling drafting: Shift=cycle, mode-aware grid/angle, Enter finishes
(minDraftVertices); polygon boundary vertex/edge drag begins a reshaping
scope. Fix grid/angle being ignored on boundary edit + slab creation:
make resolveSurfacePlanPointSnap exclusive (alignment gated on magnetic) so
grid/angles keep the snapped fallback instead of the raw cursor.
- Painter application scope: node-derived (single/object/matching/room) from
the hovered node, cyclable via Shift, single-source HUD chip.
- Remove the redundant GridSnapControl from view-toggles (grid step lives in
the contextual HUD now).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The 7th and last of the legacy interaction flags. The node being placed/moved
now lives inside the interaction scope's `placing`/`moving` variant (carried
inline, since fresh-placement/duplicate drafts aren't in the scene yet), read via
`useMovingNode()` / `getMovingNode()` / `movingNodeOf(scope)`.
- scope.ts: `placing`/`moving` carry `node: AnyNode`; add `movingNodeOf`.
- use-interaction-scope.ts: `useMovingNode` (hook) + `getMovingNode` (imperative);
no useRef snapshot needed — the node is set once at `begin`, stable for the gesture.
- use-editor.tsx: drop the `movingNode` field + the `set({ movingNode })` writes.
`setMovingNode` still drives the scope and still sets `movingNodeOrigin` /
`placementDragMode`, so cross-store subscribers keep firing. Param + ~90 call
sites unchanged.
- migrate ~17 reader sites to `useMovingNode()` / `getMovingNode()`; drop
`movingNode` from lib/scene.ts; export `movingNodeOf`.
Every interaction flag is now derived from the single authoritative scope; only
`movingNodeOrigin` + `placementDragMode` intentionally remain as useEditor flags
(they outlive the scope / gate companion behavior).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two fixes for the split-view FPS cliff when moving a wall or opening:
- wall/move-tool: publish the live preview to useLiveNodeOverrides instead of
writing useScene.updateNodes every frame. The store's `nodes` ref no longer
churns per frame (which had re-rendered every useScene(s => s.nodes) subscriber
app-wide). Matches the existing 2D wall drag + 3D wall-system override pattern;
the final plan still commits atomically as one undoable change.
- floorplan-registry-layer: replace the single global siblingEpoch with a
per-node epoch bumped only for the nodes affected by the live drag (dragged
wall -> walls at its old + new junctions + child openings; door/window -> host
wall; gutter -> roof-peer gutters), unioned with the previous frame's live set
so a cancelled drag reverts. Dragging one wall/opening now rebuilds a handful
of geometries instead of all the level's walls + openings.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Migrate `curvingWall`, `curvingFence`, `movingWallEndpoint`, `movingFenceEndpoint`
(and their setters) off `useEditor` onto the authoritative interaction scope.
The `reshaping` scope variant gains an `endpoint` discriminator; existence
checks read `useIsCurveReshape()` / `useEndpointReshape()`, and the few sites
that need the node (affordance-tool mounts, wall-vs-fence type checks) read it
from `useReshapingNode()` — a frozen drag-start snapshot, mirroring the old
flags so the tools' own per-frame writes don't feed back. `MovingWallEndpoint`
/ `MovingFenceEndpoint` move to the kind-owned tools that consume them.
`editor-api` is simpler: endpoint engagement is kind-agnostic, and the
`engageMove` reshape clears are gone (the scope is single-owner).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Migrate the two pure-mirror interaction flags off `useEditor` onto the
authoritative `useInteractionScope`. Both carried payloads byte-identical to
the scope union, so this is a zero-behaviour-change refactor: readers use the
reference-stable `useEditingHole()` / `useActiveHandleDrag()` hooks (or the
`getEditingHole()` imperative read), and producers drive the scope directly
with guarded `endIf` so clearing one interaction never stomps an unrelated
scope. Adds the `holeEditScope` builder + a no-leaked-flag invariant test.
Closes the first slice of the legacy-flag deletion; the rich-payload flags
(movingNode, curving*, *Endpoint, placementDragMode) remain.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Implements plans/editor-placement-interaction-overhaul.md: an authoritative
interaction-scope state machine plus the catalogued placement/interaction
fixes, and split-view floor-plan performance.
- Interaction-scope spine (lib/interaction/* + store/use-interaction-scope),
driven from central useEditor setters; overlay scoping (zone labels,
context badges, floating action menu) reads resolveOverlayPolicy.
- Bug tracks A/B/D/E/F/G/H: handle/cutout raycast, footprint validity,
auto-slab loop, ceiling hosting, B-key tool desync, 2D drop offset,
per-frame jank.
- Snapping modes (grid/lines/angles/off) + contextual HUD chips; modifier
model (Shift=cycle, Alt=free place, Ctrl=grid step).
- Item move now tracks the cursor 1:1 (was a laggy per-frame lerp); handle
rig hides during a whole-node move; rotate gizmo advertises Shift=free
rotation in the HUD and hides the move cross while rotating.
- Floor-plan perf: pause live reactivity while in 3D-only view; per-node
geometry cache so only changed nodes rebuild on a drag; hoist wall miters
to a once-per-pass ctx.levelData (O(N^2) -> O(N) on wall/opening drags).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Selecting a textured wall/slab/item replaced its surface with a flat
purple wash. Root cause: the selection highlight clones the material and
tints it, but `NodeMaterial.clone()` on the WebGPU backend drops the
texture-map node assignments, so the clone rendered flat — and a strong
albedo blend + emissive washed whatever was left.
Fix: re-attach the maps from the source material after cloning (shared by
reference) and drop the albedo tint, keeping only a gentle indigo emissive
so the real material/texture stays readable with a soft "selected" glow.
Applied to both highlight paths:
- generic editor highlight (slabs/items) in selection-manager
- wall path (walls are excluded from the generic one), built lazily +
cached/self-healing so it survives the wall finish's async texture load
Removes the now-dead eager wall `highlightedVisible`/`selection` profile.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds a Section B rule to the review-architecture skill: a new node kind or
geometry change that exposes paintable parts must carry overrides in a
schema `slots` record resolved via def.capabilities.paint (and thread it
through clone/duplicate/preset paths), and texturable def.geometry must
emit UVs in metres (1 unit = 1 m) so catalog finishes tile. Surfaced in the
skill description too.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds a `horizontal` fence style that builds composite-cladding panels —
horizontal boards stacked between square intermediate posts — instead of
the vertical pickets that slat/rail/privacy all produce today.
- New `style: 'horizontal'` (core schema + inspector Style control).
- Posts march along the whole span at `postSpacing`, each with a cap.
- New `postCap` control (none/flat/pyramid, default pyramid); pyramid
caps render as a 4-sided cone fence part.
- New `slatGap` control for the board reveal — 0 collapses to one flush
panel so the stacked-board seams don't read as lines.
- `postCap` + `slatGap` are horizontal-only (visibleIf); postCap uses a
labelled dropdown so its options aren't a context-free switch.
- Lower the `postSpacing` inspector minimum from 0.2 to 0.05 m.
- 2D floor plan reuses the solid-panel marker for horizontal.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Duplicating an item drops to the catalog placement flow, which rebuilds the
draft from the asset + transform and never carried node.slots — so the copy
lost every painted slot override. Thread slots through the draft create path:
useDraftNode.create seeds it onto the draft and commit() forwards it to the
final node, the placement coordinator passes it to its lazy wall/ceiling
draft creates, and the item move tool supplies node.slots for both the floor
(direct create) and wall/ceiling (coordinator) duplicate paths.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Registers a paintApply SFX (with rate/volume jitter + a 60ms gap so rapid
multi-face painting doesn't machine-gun), wires the sfx:paint-apply bus
event, and emits it from the material-paint click chokepoint in the
selection manager (fires on apply, not hover/preview). Adds the audio
asset to the editor app's public resources.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Documents the content-author contract for catalog item GLBs: slot_
material naming + deriveSlotId rules, authored defaults and
pascal_material extras, the cutout reserved mesh, the ~1 UV unit/m world
scale, vertex colours for fixed detail, and the validated Blender + glTF
export recipe (Custom Properties export gotcha, gltfpack -ke extras
preservation). Indexes the page in the architecture README and links it
from materials-and-themes.md (previously a dangling "phase-6 Blender
recipe" reference).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace 89 bundled item model.glb files with the authored/optimized
versions currently served from prod storage (items/system/<slug>/model.glb),
so the committed source of truth matches what live projects load and a
catalog re-seed no longer clobbers authored geometry. Items with no prod
object (404) are left as-is.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Layout is now: fixed eraser/reset + category tabs (no longer sticky-inside-
scroll), a single scrolling catalog grid, and an always-visible scene-material
footer. The custom-material "+" moved out of the colors-only grid cell onto the
scene-material section header, so a custom material can be added from any
category (creates a blank scene material, selects it as the brush, opens its
inline editor).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- MaterialPaintPanel owns its scroll: the eraser/reset row stays pinned and
the category tabs stick to the top, so only the material list scrolls.
- Selected catalog swatch + active scene-material card use the same
`ring-1 ring-primary ring-inset` outline as item/preset tiles.
- Choosing a material category auto-selects its first material.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Slabs now expose two paintable faces instead of one: `surface` (top, keeps
the wood floor default and the prior slot id so painted slabs are
unaffected) and `side` (vertical walls + underside, default light grey).
The merged slab buffer is split by per-triangle face normal into a top mesh
and a side mesh, each tagged with its slot id for paint resolve/preview;
legacy whole-slab material maps onto the top only. Textures-off still
collapses to the floor role.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Custom-create now pre-creates a scene material and opens its inline editor
in the build pane (no separate right-side PaintPanel, which is removed);
the brush + "Paint with" use a scene: ref so painting stores the ref and
edits propagate everywhere. Slot preview (shared + item) resolves scene
refs so hover shows the real material.
Material properties editor uses the shared SliderControl for roughness/
metalness/opacity; row action buttons use Tooltip instead of title; the
color input renders as a clean filled swatch.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Retire the inline material fields on every slot-model kind, moving them
onto the unified node.slots model on load so painting, edit-propagation,
and the picker behave uniformly.
Walls: slots {interior,exterior}; slot-first viewer resolution threading
sceneMaterials (content folded into the wall material hash); WallRenderer
subscribes to the scene-material palette so a scene-material edit
re-renders live; wallPaint rebuilt on createSlotPaintCapability.
Load migration generalizes legacy -> slots across slab/ceiling (surface),
fence (posts/infill/base/rail), column (shaft/base/capital/frame), shelf
(shelves/frame/back), and stair (per-role tread/side/railing). Library/
scene refs pass through; inline customs mint a deduped scene material;
legacy fields cleared. No visual change (renderers already fell back to
the legacy fields). Roof/chimney/dormer/vents intentionally stay on their
role system.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The compass needle rotation was `-floorplanUserRotationDeg`, but the
2D-floorplan SVG and 3D camera projections both place true north at
`+floorplanUserRotationDeg` (= cameraAzimuth − 90°). The negated sign
made the needle rotate opposite to north as the camera orbits, so it
appeared to track the camera instead of pointing north.
The formula was written for the 2D floorplan and reused verbatim when
the compass was portaled into the 3D viewer; both instances had the
flipped sign. "Align to north" is unaffected (userRotation 0 → up).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The earlier gable fix world-referenced only V, leaving U in segment-local
space — so gable and wall lined up vertically but slid horizontally. Replace the
V-only scalar offset with the segment's full world matrix (roof group composed
with segment transform) and project vertical faces exactly like the wall kind:
U = ±worldX/Z, V = 1 - worldY. Gable now tiles continuously into the walls in
both axes. Sloped shingle UVs are unchanged.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Walls used THREE's ExtrudeGeometry UVs, which restart at each wall's local
start/end — so finishes seamed at wall joins, showed a mid-face axis-switch
stripe, and didn't line up with the roof gable. Re-project the render mesh's
UVs in WORLD space (1 unit = 1 m), matching roof-system's pushRoofUv exactly:
vertical faces U = ±worldX/Z, V = 1 - worldY. De-indexes so each triangle uses
its own face normal (no edge seams). Applied only to the render mesh; collision
and floorplan geometry keep the original UVs. Now walls tile continuously
across segments and meet the gable seamlessly.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The gable wall band tiled its V in segment-local space, so its texture phase
didn't line up with the wall below (whose ExtrudeGeometry UVs map V = 1 -
height-from-base). Make vertical roof faces tile V in world space
(V = 1 - worldY) via the segment's resolved world Y (parent roof + segment
position), so the band matches a ground-floor wall's vertical tiling at the
eave seam. U is unchanged (stays in the face's local run).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The posts/infill and the accent rail shared the same panelDepth*0.35 depth, so
their faces were coplanar where they cross -> z-fighting. Make the verticals
0.001m shy so the rail wins the depth test cleanly.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
mergeGeometries throws on an empty array, so an absent slot group (infill with
showInfill off, base on a floating fence) crashed the geometry build. Return an
empty BufferGeometry for an empty group; the renderer already skips meshes with
no position attribute.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Add 'concrete-drywall' (Prepared Drywall) catalog finish + 512 KTX2/webp
runtime assets; default walls (WALL_SLOT_DEFAULT interior+exterior) and the
roof wall/trim band to it so roofs read continuous with walls.
- fence: replace the 2-slot panel/rail split with 4 slots that match the panel's
build options — posts / infill (showInfill) / base (grounded only) / rail —
each its own mesh with userData.slotId; conditional slots track the build
state. Defaults: posts/infill/base charcoal, rail wood-finewood27.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- fence: split into two paint slots — panel (posts/base/infill, default charcoal)
and rail (cap, default wood-finewood27) — as separate meshes with userData.slotId.
Fix applyFenceUVs to continuous world-space 1 UV unit = 1 m (drop the per-part
min origin that broke tiling across parts). New generateFenceSlotGeometries.
- roof: real catalog defaults for the segment surfaces via getRoofMaterialArray
(the actual default path): wall/trim concrete-plate (matches walls), deck +
soffit soft-white, shingle terracotta; textures-off role escape hatch kept.
Align nodes getRoofMaterials no-parent fallback to match.
- stair: body slot default -> preset-lightgrey.
- (biome formatting normalization of the round-1 merged renderer files rides along.)
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Elevator exposes four finish slots — cab / doors / shaft / glass (glass only
when shaft or door style is glass) — on node.slots. The finish materials now
resolve through a per-node material set (context provider) keyed on node.slots
+ scene materials: textures-off joinery role -> node.slots ref -> declared
default (cab preset-softwhite, doors metal-steel, shaft preset-lightgrey,
glass preset-glass), with glass transparency flags re-applied. Cab/door/shaft/
glass meshes tag userData.slotId; buttons, indicators, control panels, and
queue strips stay untagged (functional UI, not paintable). Interaction system
untouched. Monochrome unchanged.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Column exposes shaft / base / capital / frame slots on node.slots (base when
baseStyle!=none, capital when capitalStyle!=none, frame only for non-vertical
support styles). The single material context is widened to a per-slot map plus
a ColumnSlotContext; each part subtree is wrapped in <ColumnSlot> so every
mesh primitive resolves its material AND stamps userData.slotId from the same
context. Decorative shaft parts inherit shaft, base/capital carvings inherit
their parent slot. Resolution: textures-off wall role -> node.slots ref ->
legacy material/preset -> declared default (shaft/base/capital concrete-plaster,
frame metal-steel). Monochrome unchanged.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Stair exposes three paintable slots — treads / body / railing (railing only
when railingMode != none) — on node.slots. The 2-material-index body mesh
maps materialIndex 0->treads, 1->body via userData.slotIds; railing meshes
tag userData.slotId='railing'. Per-slot resolution layers over the viewer's
base body/railing materials only in textures-on mode: node.slots ref ->
legacy per-part field (preserved) -> declared default (treads wood-woodplank48,
body wood-woodfine2, railing metal-steel). A custom preview swaps the targeted
body material-array index and whole-mesh railing materials. Monochrome
unchanged. Stops using DEFAULT_STAIR_MATERIAL.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Fence joins the node.slots paint model with one paintable 'surface' slot
(merged post/base/rail geometry is a single mesh). Resolution order matches
slab: textures-off joinery role -> node.slots.surface ref -> legacy inline
material/preset -> declared default (library:wood-finewood27). Tags the mesh
userData.slotId='surface' and wires capabilities.slots + paint. Stops
borrowing DEFAULT_STAIR_MATERIAL.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace the flat #e5e5e5 placeholder on the roof shingle surface (material
index 3) with the catalog roof-weatheredshingles finish via
resolveSlotDefaultMaterial. Wall/trim, deck, and interior surfaces and the
textures-off role array are unchanged. Roof keeps its existing per-segment
role material system (no node.slots migration).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Pre-adds the unified slot-ref field (Record<slotId, MaterialRef>) to the
four kinds being migrated onto the paint-slot model, so per-kind work can
proceed in parallel worktrees without cross-package core edits.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Remove height/displacement/mask/mortar/joint texture files the material
schema never samples (it reads only albedo/normal/roughness/ao/metalness).
Computed as present-on-disk minus catalog-referenced in material-library.ts.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace the editor-only EditorEnvironment wrapper with a SceneEnvironment
component exported from @pascal-app/viewer, mounted as an opt-in <Viewer>
child (still not baked into the Viewer component). One source of truth the
editor and the community public viewer both inject; embed/thumbnail
surfaces simply don't mount it. Sunset preset at environmentIntensity 0.6.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Inject an EditorEnvironment wrapper (drei prefiltered sunset HDRI at
environmentIntensity 0.6) as a child of the editor Viewer, not baked into
the Viewer component, so read-only/embed viewers stay lightweight. This
gives PBR metals their reflections and lifts lighting on vertical walls
that flat directional + hemisphere lights cannot.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Builds on the explicit per-mesh slot tagging (currentDoorSlot/currentWindowSlot):
- Per-part painting: door = panel/frame/glass/hardware, window = frame/glass,
each independently paintable. The recessed door/window body sits behind the
wall, so the proud invisible cutout wins the scene raycast over the wall and
the shared resolveSlotByReRaycast() re-raycasts the kind's own subtree to pick
the exact part under the cursor (panel↔frame↔glass↔hardware). Hover tracks the
cursor via a re-eval (idempotent, no flicker).
- Door frame is its own slot (separate frameMaterial); hardware = new flat
'metal-chrome'.
- Library defaults (generic): panel/frame -> library:preset-softwhite, glass ->
library:preset-glass (flipped preset-glass to FrontSide — DoubleSide poisons
the WebGPU MRT pass; it's the only glass we use).
- Catalog: add flat (non-PBR) 'metal-chrome' + 'metal-brass'; drop metal
metalness 1 -> 0.6 so metals are lit by existing lights (no env needed).
- World-scale UVs (1 unit = 1m) on door/window box meshes via shared box-uv.ts,
so finishes tile at real-world scale instead of stretching.
- PaintResolveArgs gains an optional for subtree re-raycasting.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The editor's /icons assets were ~10MB of oversized PNGs (a single
toolbar icon up to 1.4MB). Convert every non-PWA icon to WebP (quality
92) and repoint all /icons/*.png references to .webp across
packages/editor, packages/nodes, and apps/editor. PWA/platform icons
(apple-touch-icon, icon-192, icon-512) stay PNG.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Snapshot/preset capture is for framing a thumbnail, not editing, so it
should only move the camera:
- gate the selection manager, transform handles, floating menus, and the
tool manager (which renders the site boundary flags) behind
isCaptureMode, so clicking the scene no longer selects items or shows
editing controls
- hide zone meshes and the HTML zone tags while capturing
- force 3D on entry (the 2D/split floorplan panes render nothing useful
for a thumbnail) and restore the prior view mode on exit
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The real interceptor was the 'cutout' mesh, not the hitbox: it's a 1m-deep CSG
helper whose front face sits 0.5m proud of the glass, and current three.js
raycasts invisible meshes — so every paint/hover ray resolved to 'cutout'
(no slotId → role null). Disable its raycast; combined with the hitbox noop,
paint/hover rays now land on the tagged frame/glass (panel/glass) children.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- post-processing: hoverHighlightMode was a dependency of the pipeline-build
effect, so every hover rebuilt the entire pipeline. The hover style is already
pushed to uniforms in a separate effect, so the rebuild was pure waste —
removed it from the deps (and the build log).
- selection-manager: temporary [paint-debug] logs for window/door hover to trace
why their paint dispatch drops (to be removed once diagnosed).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The window/door root is an invisible hitbox the system gives full-depth
BoxGeometry; its front face intercepted every paint/hover ray, so the hit
resolved to the hitbox (no slotId) → role null → paint silently disabled.
Disable the hitbox's own raycast in the visual path so R3F's recursive
intersect returns the tagged frame/glass (panel/glass) children instead;
selection still works because those child hits bubble to the root's event
handlers. Restored to the default raycast each build, and kept for 'opening'
windows/doors (no visuals to paint, still need a selectable hitbox). The
'cutout' child is visible=false so the raycaster already skips it.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Windows and doors build all visuals in their viewer systems from module-global
materials, so this threads per-node slot materials + userData.slotId tags
through those builders without restructuring them:
- window: 'frame' + 'glass' slots. door: 'panel' (body = casing + leaf) +
'glass'; the opening reveal keeps its own material.
- Each system captures per-frame viewer state, then updateWindow/DoorMesh points
the builder-facing base/glass materials at the node's resolved slot override
(recomputed per node, so the next node resets without a restore). Meshes are
auto-tagged in the shared addBox/addShape helpers by which material they got.
- Textures-off still collapses to the role material (escape hatch); a slot
override only applies in colored mode.
- Editing a referenced scene material re-dirties the window/door (these systems
aren't covered by GeometrySystem's scene-material re-dirty).
- New paint capabilities (resolve role from userData.slotId, preview by
userData.slotId) + capabilities.slots; window/door dropped from the paint
disabled list. Shared previewSlotByUserData helper.
Defaults unchanged: unpainted windows/doors render exactly as before (the slot
fallback is the existing frame/glass material), so no visual regression.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Brings slab, ceiling, and wall onto the unified slot contract the shelf
established, so each declares its paintable slots with a declarative default and
(slab/ceiling) is painted through the registry capabilities.paint dispatch.
- Shared helper packages/nodes/src/shared/slot-paint.ts: a node.slots-based
PaintCapability factory (commit/resolve/effective generic; preview injected).
Distinct from surface-paint.ts, which writes the legacy inline node.material.
- slab: schema slots; def.geometry resolves node.slots.surface -> legacy
material -> declared default, tags the mesh userData.slotId; slabPaint +
capabilities.slots. Retires DEFAULT_SLAB_MATERIAL in the slab path.
- ceiling: schema slots; material builders extracted to ceiling/materials.ts
(shared by renderer + paint preview, built BackSide so the hover preview is
visible from below); renderer resolves the slot; ceilingPaint + slots.
- wall: WALL_SLOT_DEFAULT in core; the viewer's getMaterialsForWall renders an
unpainted face with its declared default instead of the themed wall role;
capabilities.slots (interior/exterior). wallPaint's inline interior/exterior
fields are unchanged (node.slots migration is a later step).
- selection-manager + material-paint: drop slab/ceiling from the legacy
single-surface arms (now registry-driven).
Behavior change (intended, matches the shelf precedent + the phase-5 plan):
colored-mode UNPAINTED slab/ceiling/wall surfaces now render their fixed slot
default (#e5e5e5 / #f5f5dc / #ffffff) instead of the theme role colour. The
textures-off (monochrome) role collapse is unchanged — the escape hatch.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Wire KTX2 into the catalog finish path and add 15 textured finishes across four
new families, all as GPU-compressed KTX2 (512px) + a webp picker thumbnail.
- Shared `ktx2-loader.ts`: one KTX2Loader (Basis transcoder) used by both the
GLB loader and catalog textures; `ensureKtx2Support(renderer)` runs once at
viewer init (GPUDeviceWatcher) so `.ktx2` finishes load even with no GLB in
the scene. `use-gltf-ktx2` now reuses the shared instance.
- `materials.ts`: texture loaders pick the KTX2 loader for `.ktx2` urls, the
image loader otherwise (all three load paths).
- `material-library.ts`: 15 entries (fabric ×6, leather ×2, concrete ×4,
metal ×3). Neutral albedo for tinting, `flipY: false` (compressed textures
can't flip), `repeat` per real-world tile size, metals `metalness: 1`.
Normals encoded UASTC, data maps ETC1S/linear.
Assets generated from raw sources via the new community
`scripts/build-material-textures.ts`. 1024/256 tiers + raws kept out of git.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Cubby without a bottom board: the lowest cell opens onto the floor, but its
column divider still started at the (missing) bottom board's top, leaving it
floating ~thickness above the floor. Anchor it at y=0 when there's no bottom.
- Bookshelf full-height divider: it crosses the continuous shelves, so it can't
share their depth plane (proud at the front, coplanar with the back panel —
z-fighting down the centre back). Recess its depth to sit fully INSIDE the
boards' depth: embedded at each shelf crossing (board occludes it) and tucked
inside the back panel.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The cubby column divider was full-depth while the boards are depth-recessed,
so it (a) sat proud of the shelf fronts — the earlier Y-tuck then made it poke
past the top/bottom boards — and (b) shared the back panel's rear plane,
z-fighting down the centre of the back.
Give the divider the same depth recess as the boards (via boardGeometry) and
drop the Y-tuck: it now sits flush with the shelf fronts and its back tucks
inside the back panel. Flush top/bottom is fine — those board faces are
back-to-back, not co-facing, so they don't fight.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Cubby per-cell dividers landed exactly flush on the board below and under the
board above — coplanar faces that shimmer and read as merged into the shelf.
Extend each divider 1mm into both boards (centre unchanged) so it tucks under
the top board and onto the bottom board with a solid, seam-free join.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two follow-ups to the plate-recess fix:
- Top surface: full-height frame members that pass UNDER the top board
(back panel, bookshelf column dividers, corner posts) had their top face at
unitHeight — coplanar with the top board's top — so they z-fought through it
(the seam on the top). Drop their top 1mm (cappedFrameY); bottoms stay on the
floor so nothing floats.
- Plate width: the previous pass recessed every board's width, which opened a
1mm gap where boards ABUT the side panels (cubby / bookshelf-with-sides).
Width is now recessed only where boards span OVER posts (open-rack /
no-sides bookshelf); abutting boards keep full width and meet the sides
flush. Depth stays recessed everywhere (gap-free, fixes the back-panel fight).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Open-rack / no-sides shelves span their boards across the full footprint, so a
board's end / front / back faces land exactly on the corner posts' outer faces
(and, with a back panel, on its rear face). Coplanar surfaces the depth buffer
can't order flicker as z-fighting (visible as seams where plates meet posts).
Recess every board 1mm on width + depth via a `boardGeometry` helper so each
plate sits just inside the frame — the meshes still overlap (no visible gap),
but no two faces are coplanar. Frame parts (posts/sides/back/dividers/brackets)
keep their full size as the silhouette.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Proves the unified (nodeId, slotId) slot contract on a procedural generator,
beyond items and walls. A shelf now exposes three paintable slots — shelves /
frame / back — painted through the same PaintCapability dispatch and the same
node.slots: Record<slotId, MaterialRef> shape items use.
Foundation (shared, reusable by future procedural kinds):
- core: SlotDeclaration type + capabilities.slots(node) registry declaration;
GeometryContext gains `materials` so a pure builder can resolve scene:<id>
slot refs without importing useScene.
- viewer GeometrySystem: threads the scene material library into every builder
ctx, and re-dirties (bypassing the geometryKey skip) any geometry node that
references a scene material when that material changes — so editing a custom
colour propagates to every shelf using it, matching items.
Shelf:
- schema: slots: Record<string, MaterialRef> (mirrors ItemNode).
- geometry: per-slot material resolution (slot override -> legacy whole-shelf
-> declared default colour); every mesh stamped with userData.slotId;
DEFAULT_SHELF_MATERIAL retired (declared default gives identical off-white).
- paint.ts: PaintCapability (resolveRole from userData.slotId, scene-material
commit for one-off colours, preview restricted to __fromGeometry meshes so
hosted items aren't ghosted, getEffectiveMaterial incl. legacy fallback).
- definition: paint + slots capabilities; slots folded into geometryKey.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Pool slabs (generatePoolGeometry) were the one procedural surface emitting
normalized [0..1] floor UVs, so a textured finish stretched to fit the pool
instead of tiling at real-world scale like every other surface. Switch the
floor to shape-space metres (x, -z) — the same mapping generatePositiveSlab-
Geometry already uses for its caps. Pool walls were already in metres.
Also document the contract in wiki/architecture/materials-and-themes.md: every
procedural surface generates UVs in metres (1 UV unit = 1 m), GLB slots follow
the same ~1 unit/m authoring convention, and a catalog material's `repeat` is
therefore a per-material world-scale setting (tiles per metre), identical for
every surface that uses it — never per-item.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The textures axis had no UI control. Add Colored (textures on — show item
materials, textures, vertex colours) / Monochrome (textures off — flat clay
by surface role) options to the Render dropdown, wired to useViewer.textures.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
