Skip to content

Commit a8a376c

Browse files
obiotclaude
andauthored
CHANGELOG: house order and length for 20.7 (#1680)
Two drifts, both previously corrected for 20.6 and repeated here. `### Fixed` sat above `### Changed`. The house order puts Fixed last — Added, Changed, Performance, Fixed — as in 20.3.0 and 20.4.0. And the entries had grown back into paragraphs. Trimmed to the one-or-two sentence form: the pivot entry loses a clause about where the covering spec built its renderable, which is commit-history material rather than something a reader acts on. The `onBodyUpdate` entry is dropped entirely. The fix stays in the code and in its commit, but only the deprecated `Entity` passes that callback — the adapter builds bodies without one — so nothing on the modern path consumes it and a reader has nothing to do about it. Claude-Session: https://claude.ai/code/session_01NGvtaUNATVCVxD2qcbiY4t Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
1 parent 05b9987 commit a8a376c

1 file changed

Lines changed: 5 additions & 6 deletions

File tree

packages/melonjs/CHANGELOG.md

Lines changed: 5 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -2,13 +2,12 @@
22

33
## [20.7.0] (melonJS 2) - _unreleased_
44

5-
### Fixed
6-
- Body: rotation pivoted about the wrong point for any renderable not at the world origin. `body.bounds` is already renderable-local, and the visual pivot subtracted `renderable.pos` from it a second time — so a 40x40 body on a renderable at `(100, 50)` turned about `(-80, -30)` rather than its own centre. Only correct at the origin, which is where the spec covering it built its renderable
7-
- `Body.rotate()` no longer throws on a shape that cannot rotate. It called `shape.rotate()` unguarded, and neither `Box3d` nor `Point` implements one — so the documented way to rotate collision shapes crashed on exactly the bodies most likely to carry a 3D sensor. Such a shape keeps its orientation and still contributes its bounds
8-
- `Body.rotate()` reports the change to its owner again. The `onBodyUpdate` notification was commented out, so an owner that folds the body's bounds into its own kept the pre-rotation extent — and since the broadphase indexes the renderable's bounds, the rotated body was queried at the wrong size
9-
105
### Changed
11-
- Loader: `load()` no longer names asset types while resolving a `src`. Every parser is registered with the same shape, and a type whose `src` is not a bare path declares `normalizeSrc` and `needsBaseURL` — which is how `fontface` unwraps a `url(...)` descriptor before the base URL goes on, and leaves an installed `local()` family alone ([#1648](https://github.com/melonjs/melonJS/issues/1648), thanks @ICOM725)
6+
- Loader: `load()` no longer names asset types while resolving a `src`. A type whose `src` is not a bare path declares `normalizeSrc` and `needsBaseURL` instead — which is how `fontface` unwraps a `url(...)` descriptor before the base URL goes on, and leaves an installed `local()` family alone ([#1648](https://github.com/melonjs/melonJS/issues/1648), thanks @ICOM725)
7+
8+
### Fixed
9+
- Body: rotation pivoted about the wrong point for any renderable away from the world origin. `body.bounds` is already renderable-local and the pivot subtracted `renderable.pos` from it a second time, so a 40x40 body on a renderable at `(100, 50)` turned about `(-80, -30)` rather than its own centre
10+
- `Body.rotate()` no longer throws on a `Box3d` or `Point` shape, neither of which can rotate. Such a shape keeps its orientation and still contributes its bounds
1211

1312
## [20.6.0] (melonJS 2) - _2026-09-16_
1413

0 commit comments

Comments
 (0)