I will never understand why people don't lead with the link to the live demo: <a href="https://suboptimaleng.github.io/slime-sim-webgpu/" rel="nofollow">https://suboptimaleng.github.io/slime-sim-webgpu/</a><p>It's very cool!
Very cool! I did a similar project last year. Really like how your result looks with the lighting setting on.<p>The most interesting results I got were from having multiple "variants" and giving each variant parameters for their attraction to agents of their own/other type. Implementation-wise it was a variant tag for each agent and a matrix of attraction coefficients as a uniform. Playing with the params resulted in interesting behaviors, like one variant chasing another around while they both try to avoid a third.<p>If you're looking for an addition to this project, I'd recommend giving it a try!
Super cool! I was experimenting with WebGPU and various simulations a couple months ago [1] and would love to learn more about your process! Do you have a blog or something similar?<p>[1] <a href="https://observablehq.com/collection/@miketahani/slime" rel="nofollow">https://observablehq.com/collection/@miketahani/slime</a>
I had to read that title several times to make sure that it wasn't the alcohol causing it to register as "Slime Mold."<p>Why someone would want to simulate a slime mold is something I don't understand, but more power to ya. =)