Document LCC2 scenes and SPZ version 3 - #1338

* docs: Point CameraFrame example embeds at the camera-frame category

* Document devicerestorefailed for a failed device recovery

* docs: Document compute shader access to the scene depth and color maps

* docs: Add the Japanese translation of the gsplat vertex stage page

* docs: Document sceneDepthViewport for cameras rendering to a part of their target

* docs: Document that a compute instance is dispatched at most once in a frame

* docs: Describe the compute shader scene maps and parameters without version qualifiers

* docs: Document rendering to cubemap faces and texture array layers, and render target mipmaps

* docs: Add a Textures page, and document rendering to volume texture slices

* docs: Drop the pc. prefix from the code, and show the shader examples in GLSL / WGSL tabs

* docs: Point the render target example embeds to the render-targets category

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* docs: Document dispatching a compute shader for a number of threads

* docs: document shadow bounds, and point the light and shadow examples at lights-and-shadows

Adds a Shadow Bounds section to the Shadows page for LightComponent#shadowBounds, and updates the example links of the light and shadow examples, which moved from graphics to lights-and-shadows in the examples browser. Japanese translations updated to match.

* docs: update HTML-in-Canvas for the API shipping in Chrome 157

* docs: add the Frame Graph section on frame, render and compute passes

Explains physical render and compute passes and the frame graph, and documents the
engine's frame passes: the pass classes and their lifecycle, camera passes, RenderPass,
RenderPassShaderQuad, RenderPassForward and ComputePass. The CameraFrame custom render
passes page is folded into it, with a redirect to Camera Passes.

* docs: document CameraFrame effects, lens flare and dynamic sky

Replaces the compose shader and extending CameraFrame pages with two pages on
CameraFrameEffect: Custom Effects covers the composition and its compose slots,
the built-in effects with their ids, slots and order, writing, registering and
parameterizing an effect, debug views, chunk overrides and effect scripts for
the Editor. Effects with Passes covers the stages of the camera frame, the frame
resources, creating and rebuilding passes, per-frame values and resolving the
scene. Lens Flare and Dynamic Sky get pages of their own.

The camera frame overview lists the effects added since, and its Custom Post
Processing section points to the new pages and to Camera Passes, in English and
Japanese. The removed pages redirect to their replacements, and the sidebar and
the Engine LLM index link the new pages.

* docs: document the CameraFrame auto exposure effect and the effect scripts

* docs: link the translated pages by URL from the new pages, for the Japanese build

* docs: describe the dynamic sky in XR

The sky and the flat clouds render for each XR view, while the volumetric
clouds render once for the view of the camera and streak where the view of an
eye extends past it. Add an XR section recommending the flat clouds in XR, and
narrow the limitation to the volumetric clouds.

* docs: translate the frame graph and CameraFrame effects pages into Japanese

Adds Japanese versions of the Frame Graph section and of the Custom Effects,
Effects with Passes, Lens Flare, Dynamic Sky and Auto Exposure pages, with the
headings keeping their English anchors so links into them resolve. Brings the
Japanese camera frame overview and HDR Rendering pages up to date with the TAA
upscaling, SMAA, debug view and effect script sections, and labels the Frame
Graph and Render Passes sidebar categories.

* docs: move debugging into a Debugging and Optimization section and document the Inspector's Performance tab

* docs: describe the dynamic sky Editor script

The engine now has a dynamicSkyEffect script, from dynamic-sky.mjs. Add a
section on it to the Dynamic Sky page, with a script of your own animating the
time of day, list it in the effect scripts of the HDR Rendering page, and drop
the note that the dynamic sky has no Editor script, in English and Japanese.

* docs: Point example source links at the per-example folders

* docs: describe the dynamic sky haze and the sky in other shaders, in English and Japanese

* docs: Document planar reflections on the Environment Lighting page

* Deprecate the legacy post effects

Mark the legacy script-based post effects as deprecated on the post effects
landing page, the legacy index, the six effect pages and the four tutorials
built on them, in English and Japanese. The legacy index gets a table of the
CameraFrame replacement for each legacy script.

* Point the Inspector links at the moved debugging pages

* Give the Japanese Custom Post Processing heading its English anchor

* Document gaussian splat anti-aliasing (#1326)

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* Document the gaussian splat anti-aliasing filter size (#1327)

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* Use the renamed MeshInstance shadow, AnimComponent and pixel format APIs (#1329)

engine#9937 renames MeshInstance#castShadow/receiveShadow to castShadows/receiveShadows,
AnimComponent#findAnimationLayer to findLayer and PIXELFORMAT_111110F to PIXELFORMAT_RG11B10F,
keeping the old names as deprecated aliases. Use the new names in the manual, en and ja, and
link receiveShadows, which the API reference now documents.

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* Document the channel mixer of the CameraFrame color grading (#1330)

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* Document the mouse button actions and ground pan of CameraControls (#1331)

The Camera Controls page gets a Mouse Buttons section covering the actions of
the left, middle and right buttons, reserving a button for the application,
and dragging the ground like a map with groundPan. The page and the three
first-app tutorials said the right button pans, but the middle button pans and
the right button looks. Japanese pages are updated alongside.

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* Document the Gaussian mipmap filter and the scene color map mipmaps (#1328)

Documents the mipmap filters and the planar reflection changes of playcanvas/engine#9932, in English and Japanese: the Gaussian mipmaps of the planar renderer and its TAA jitter, the mipmaps of the scene color map, and the mipmapFilter option of render target textures.

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* Add an Add-on Libraries section to the engine manual (#1333)

* Add an Add-on Libraries section to the engine manual

A new Add-on Libraries section under PlayCanvas Engine collects the libraries
used alongside the engine: an overview page, and a page each for Spine, 3D Tiles
with 3DTilesRendererJS, tweening with playcanvas-tween, and loaders.gl.

The Spine page moves here from 2D, which keeps a link to it. The tweening
tutorial moves here as the Tween page, updated for the UMD and ESM builds of
playcanvas-tween. The old URLs redirect to the new pages. Japanese pages are
updated alongside.

* List the Add-on Libraries pages in the engine llms index

The index linked the Spine page at its old 2D URL, which the llms plugin
rejects, and did not list the new pages. A new Add-on libraries section lists
the overview and its four pages, with Spine moved there from Animation.

---------

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* docs: Update Gaussian splatting example ids after the examples category split (#1334)

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* docs: Document SOG version 3 and blind data (#1335)

The SOG format spec now covers versions 2 and 3, both supported, with a table of the differences
and the version-specific parts marked: the exponential-map orientation, squared opacity up to
1.5625, a 4th SH band, 24-bit palette labels and per-gaussian blind data.

The Work Buffer Format page shows how a copy modifier reads blind fields through
getBlind_<name>() and SOG_BLIND_<name>, with the Blind Data example, and Streamed SOG notes that
chunks may be either version. English and Japanese.

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

* docs: Document LCC2 scenes and SPZ version 3 (#1338)

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>

---------

Co-authored-by: Martin Valigursky <mvaligursky@snapchat.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>