Ricky Reusser

Ueda's Attractor

Jacobi Elliptic Functions in the Complex Plane

SDF Points with regl

Clifford and de Jong Attractors

Drawing Instanced Lines with regl

Instanced Line Rendering

Lawson's Klein Bottle

Interactive Multi-scale Turing Patterns

Selecting the Right Opacity for 2D Point Clouds

Adaptive Contouring in Fragment Shaders

Baker's Map

Fake Transparency for 3D Surfaces

GPU Boids

Double Compound Pendulums

3D Reaction-Diffusion

Mathematical Easter Egg Coloring

Toiletpaperfullerenes and Charmin Nanotubes

Quasicrystals

Observable + regl

2D (Non-physical) N-body Gravity with Poisson's Equation

Periodic Planar Three-Body Orbits

Quick (miterless) Lines in WebGL

Spin WebGL

icosphere

Finding Roots in the Complex Plane

glsl-complex

glsl-complex-auto-differentiation

Domain Coloring with Adaptive Contouring

gl-mat3

gl-mat4

gl-vec4

gl-quat

gl-mat2

gl-vec3

gl-vec2

Instanced WebGL Circles

regl-tools

Domain Coloring for Complex Functions

Contour Plots with regl Observable Plot

Drawing indexed mesh data as screen-space normals without duplicating data

GPUs, FFTs, and the 1-D Schrödinger Equation

WebGL

A series of WebGL tools and tutorials based around the regl WebGL library

Regl Tools and Tutorials

Why does backface culling remove part of the line? What determines the orientation of the faces?

Fabian Iwand

I'm drawing the geometry as a triangle strip, so—at least without duplicating *more* vertices—the winding needs to change depending on whether adjacent corners turn in the same or opposite directions. I've added a note below on this. In short, I don't think there's much value in preserving consistent winding order, so I didn't.

Here's the exact line which makes the decision: https://github.com/rreusser/regl-gpu-lines/blob/f3b0e57dacafaba9e65c2f3eff0f9b20440f8705/src/draw-segment.js#L143

This was driving me crazy, so I renumbered all of the indices. Instead of changing the geometry like I was (which flipped winding), it now repeats/doesn't repeat a vertex at the beginning of the fan to get the orientation correct. That flips winding when needed but without modifying geometry, which is way easier to work with. Culling now works more reasonably! There are still a couple stray pixels on mobile. (see: https://rreusser.github.io/regl-gpu-lines/docs/tests.html) Not sure where they're coming from, but it's still an improvement.

strip.png