feat(Sprite3d): anchor support via settings.anchorPoint + named presets engine-wide#1538
Merged
Merged
Conversation
Adds a settings.originPoint option (preset string or {x,y}) so the
quad's pivot can be placed anywhere relative to pos, e.g. bottom-center
for ground-anchored 2.5D characters. Baked as a constant local offset
applied after flips, so it composes with billboarding and
trimmed/rotated atlas frames. Default stays centered — no behavior
change for existing code.
Updated the billboard example to anchor the character at its feet
instead of using a manual GY offset fudge factor.
Closes melonjs#1514
Member
|
Hi, Thanks for the PR, before merging I'm gonna make some further changes however that I think makes it a better fit with the legacy/existing API :
|
Contributor
Author
|
@obiot thanks for the detailed feedback! Happy to implement this myself, just let me know if you'd rather I push these changes, or if you're planning to handle it on your end. Either way works for me. |
Member
no worries, I will do it, already on it :) |
…ine-wide
Rework of the originPoint proposal into the hybrid agreed in review:
Sprite3d anchors through the SAME settings.anchorPoint key, convention
and centered default as the 2D Sprite — no second anchor API — while
keeping the (verified-correct) vertex-bake mechanism from the original
commit.
- new shared resolver (renderable/anchorPoint.ts): the named presets
("center", "top", "bottom", ..., "bottom-right") now work everywhere
settings.anchorPoint does — Sprite, Entity, Collectable, ImageLayer,
Text, BitmapText (and subclasses), plus Sprite3d. On the 2D classes
invalid values keep their historical silent (0, 0) outcome for full
backward compatibility, but log a console warning; Sprite3d (a new
surface) throws. Values are never clamped; Vector2d instances and
bare {x, y} objects pass through unchanged.
- Sprite resolves the anchor once, up front, and forwards the RESOLVED
pair into the texture-cache descriptor (spritesheet atlases store it
as the cached per-frame pivot), so presets and {x, y} behave
identically on both paths and a raw string or aliased Vector2d never
lands in the shared cache.
- ImageLayer normalizes its bare-number shorthand pre-super (its Sprite
base consumes the same setting and the strict resolver rejects bare
numbers).
- Sprite3d wires the inherited ObservablePoint as the LIVE anchor:
setMuted + setCallback post-super; runtime anchorPoint.set() re-bakes
the quad (region path via _applyFrame, plain-quad path via
bakeQuadVertices) and re-derives the cull bounds (direct point
mutation — the Rect width setter's recalc() walks Mesh's repurposed
normals array and would throw).
- Mesh gains the _anchorBaked opt-out consulted by preDraw's per-frame
applyAnchorTransform recompute, so the anchor never leaks into the
renderer transform on either camera path (plain Mesh behavior
unchanged, webgl_mesh_anchor.spec.js untouched and green).
- cull bounds fixed: side = max(max(w,h), sqrt2*hypot(hw+|ox|, hh+|oy|))
— the sphere exactly encloses the anchored quad in any billboard
orientation (a bottom-anchored 130x225 character was ~47% under-sized
and popped out on screen); centered default byte-identical.
- billboard example: anchorPoint: "bottom", camera-target framing
restored (GY * 0.6 became dead code aiming at the floor), stale
comment fixed.
- tests: new tests/anchorpoint-presets.spec.js (cross-renderable
contract, 72 tests, failing-first) + 9 geometry/runtime/cull tests in
sprite3d.spec.js (failing-first). Full suite 4698 passed / 0 failed.
Master merged in (CHANGELOG + the melonjs#1539-melonjs#1541 fixes).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019t8kP8vp58ZrvaD2FqJU6A
obiot
changed the title
Member
|
Thanks again @asadullahbro — the vertex-bake mechanism you built was verified correct end-to-end and is kept as-is. I've pushed the rework we discussed on top of your commit ( API change: Your presets went engine-wide: the names now work everywhere Also fixed along the way:
Tests: new 🤖 Generated with Claude Code |
…mpat pins
Sprite3d (9 new adversarial cases in sprite3d.spec.js):
- anchor survives animation frame changes
- combined torture: anchor + flipX + frame change + runtime re-anchor,
then unflip — exact vertex assertions at every step
- anchor on a BOTH rotated and trimmed atlas region
- anchor round-trip (bottom → arbitrary → out-of-range → center)
restores geometry AND cull bounds byte-exact vs a fresh sprite
- single-axis anchorPoint.y = ... property writes re-bake
- out-of-range anchors: cull sphere encloses every projected corner
across billboard yaw/pitch (property invariant, 3 anchors × 3 poses)
- spherical billboard: corner distances from pos are rotation-invariant
- frame-derived sizing (no explicit width) computes offsets from the
frame grid
Sprite 2D backward compatibility (7 pins in anchorpoint-presets.spec.js):
- spritesheet preset ≡ {x, y} equivalence
- PIN of the pre-existing shared-atlas pivot inheritance (first
sprite's anchor becomes the cached pivot for later same-image
sprites) — documents the wart, guards against accidental change
- caller Vector2d mutation no longer retro-poisons the sprite or the
cache (the one deliberate safety divergence from master)
- garbage on a shared sheet keeps legacy (0, 0) for both sprites
- anchor survives 2D animation frames and flips (untrimmed pivot rule)
- Entity preset reaches entity AND child sprite; NineSliceSprite
inherits presets
Master-parity proof: the whole spec was run against master's engine
src — all 31 pure back-compat pins pass there; the 48 failures are
exactly the new-feature surface (presets, warn-on-garbage, Sprite3d's
new API) plus the documented aliasing fix. Full suite 4713 / 0.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019t8kP8vp58ZrvaD2FqJU6A
An honest audit of the anchor coverage found five holes; all closed:
- FULL preset-table validation: only "bottom"/"top-left" were
value-verified — a typo'd pair in any of the other 7 presets would
have passed every test (same failure class as the CSS color table).
All 9 names now assert their documented values.
- Tiled property chain: applyTMXProperties keeps a string anchorPoint
("bottom") intact through coercion and it resolves correctly in a
renderable constructor; the legacy numeric expansion ("0.5" →
{x, y}) is pinned too — the CHANGELOG's Tiled claim is now tested.
- pool recycling: a pooled Text resolves presets on pull AND
re-resolves new settings on recycle (onResetEvent path).
- Sprite3d flipY + anchor (only flipX was covered): bottom anchor
stays pinned, unflip restores byte-exact.
- REGRESSION GUARD for the cull fix: pins that the legacy max(w, h)
bounds would NOT enclose a bottom-anchored 130×225 quad (farthest
corner 234.2 vs old radius 159.1) while the shipped bounds do —
the pop-in bug can't silently return.
anchorpoint-presets.spec.js: 93 tests; sprite3d.spec.js: 49. Full
suite green. All sprite/anchor-heavy examples verified running clean
in-browser (platformer, sprite, aseprite, text, isometric-rpg, ui,
tiled-map-loader, texture-packer — zero console/page errors).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019t8kP8vp58ZrvaD2FqJU6A
Member
|
Follow-up hardening pushed (
Full suite: 4727 passed / 0 failed. 🤖 Generated with Claude Code |
- Mesh class doc + preDraw JSDoc: note the _anchorBaked subclass opt-out (Sprite3d) — "the 2D path keeps the anchor" was no longer the whole truth - Renderable#anchorPoint property doc: document the named presets accepted by settings.anchorPoint across all consumers, and that a plain string preset now works as a Tiled property (the old note implied the json: expression was the only way) Audited: zero originPoint references remain anywhere; all 8 settings.anchorPoint @param lines carry the widened type + presets. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_019t8kP8vp58ZrvaD2FqJU6A
Member
|
all done, thanks again for the PR ! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds a settings.originPoint option (preset string or {x,y}) so the quad's pivot can be placed anywhere relative to pos, e.g. bottom-center for ground-anchored 2.5D characters. Baked as a constant local offset applied after flips, so it composes with billboarding and trimmed/rotated atlas frames. Default stays centered — no behavior change for existing code.
Updated the billboard example to anchor the character at its feet instead of using a manual GY offset fudge factor.
Closes #1514
Description
Type of change
Checklist
pnpm lintpasses)pnpm testpasses)pnpm build)Related issues