Skip to content

armgit / 67.js

Public Imported from GitHub
0

3D engine that AI agents drive

Add change notes for the mirror terrain levels (part of #316)

Nguyen Xuan Anbd75da33ae53
diff --git a/changes/316-mirror-terrain-lod.md b/changes/316-mirror-terrain-lod.md
new file mode 100644
index 00000000..5808cbe1
--- /dev/null
+++ b/changes/316-mirror-terrain-lod.md
@@ -0,0 +1,2 @@
+- `createTerrain({ mirrorLevels })` (default 0) builds up to that many extra coarser meshes for each chunk, used only by the water mirror draw. The renderer never picks a `mirrorOnly` level for the main camera, so main-camera pixels do not change; the mirror picks them by the screen error of the mirror's own camera. Terrain meshes of a chunk are not edge-matched across mirror levels, so a seam can show in the blurred reflection. Part of #316.
+- `lods` entries accept `mirrorOnly: true`: the level is skipped in a normal frame and in the shadow draw, and chosen by the water mirror draw like any other level.
diff --git a/changes/316-mirror-terrain-lod.size.md b/changes/316-mirror-terrain-lod.size.md
new file mode 100644
index 00000000..bf0ac02e
--- /dev/null
+++ b/changes/316-mirror-terrain-lod.size.md
@@ -0,0 +1,4 @@
+- entry:terrain: 279 bytes for the mirror meshes (coarse levels, their error and the option check); terrain is the only place that builds them (#316).
+- scene:agent-scene+lazy: 317 bytes, the terrain code above plus 37 bytes of core in the renderer, in a scene that bundles terrain.
+- goal:lit-box: asks for 37 bytes over the 50 KB goal; the core part is two lines in the renderer that let a `lods` level be skipped in a normal frame.
+- goal:agent-scene: asks for 36 bytes over the 100 KB goal, the same two lines.