Building Enterprise 3D Web Applications
summary
Uses Three.js over WebGL to render a two‑story demo house with thousands of parts; a single procedural tree consumes about six times the memory of the rest, and adding four fresh tree geometries spikes tab memory from ~120MB to ~268MB, while cloning geometry to reuse a single complex mesh lets dozens of trees appear instantly at ~122MB. Implements raycasting for precise object picking, uses CSG subtract to create cross‑sectional cuts for floor inspection, enforces pivot/position conventions and lazy asset loading to avoid heavy runtime recompute, and discusses server‑side streaming as a workaround for browser memory and IP leakage.
AI-written summary · about
chapters (10)
- 0:00Introduction to 3D on web
- 1:02Basics of 3D world
- 2:51Understanding 3D objects
- 5:21Optimizing 3D renderings
- 12:563D operations overview
- 15:27Identifying objects in 3D
- 16:56Dynamic object replacement
- 23:46Math in 3D applications
- 35:123JS and WebGL limitations
- 38:41Future of 3D web rendering
related talks
13mWebGL Demo with THREE.js
41mWebGL Is Easier Than You Think (feat. three.js)
37m3DGraphics with Threejs by Ruben Meschian (Eng) | JSConfAM19
34mNext-Generation 3D Graphics on the Web
20mEmbracing WebGPU and WebXR With Three.js
59mRendering 3D Like a Grown-Up
29mCreating 3D web maps with Cesium
45mThe visual web of today and tomorrow - WebGL and WebVR.