Return to Answer

I'm writing my current project entirely in WebGPU and WebAssembly. Prototype code in JS, using it for things like UI and assembling your shaders dynamically. As project gets heavier, replace performance-critical sections with C, Rust or AssemblyScript compiled to WASM modules, or push that load off to WebGPU compute. With Websockets, WebRTC for multiplay, one can go further.

If you want desktop support, compile to standalone executable using Electron or similar. Besides desktop platforms, WebGPU is also slated for Android and iOS (indeed there have already been some early releases). Arguably the most cross-platform non-propietary tech stack in existence.

WebGPU knowledge and code is portable to C, Rust etc via lib wgpu.

To me it's a modernised, pleasant GL environment that is a huge step up, organisationally and performance-wise, from WebGL. Not saying without it's issues, WebGPU's still a work in progress.

I'm writing my current project entirely in WebGPU and WebAssembly. WebGPU knowledge and code is portable to C, C++, Rust etc via lib wgpu. I prototype code in JS, using it for things like UI and assembling shaders dynamically. As project gets heavier, I replace performance-critical sections with C, Rust or AssemblyScript compiled to WASM modules, or push that load off to WebGPU compute. With Websockets, WebRTC for multiplay, one can go further.

If you want desktop support, compile to standalone executable using Electron or similar. Besides desktop platforms, WebGPU is also slated for Android and iOS (indeed there have already been some early releases). Arguably the most cross-platform non-propietary tech stack in existence.

To me it's a modernised, pleasant GL environment that is a huge step up, organisationally and performance-wise, from WebGL. Not saying without it's issues, WebGPU's still a work in progress.

I'm writing my current project entirely in WebGPU and WASM (WebAssembly). Prototype code in JS, using it for things like UI and assembling your shaders dynamically. As project gets heavier, replace performance-critical sections with C, Rust or AssemblyScript compiled to WASM, or push that load off to WebGPU compute. With Websockets, WebRTC for multiplay, one can go further.

If you want desktop support, compile to standalone executable using Electron or similar. Besides desktop platforms, WebGPU is also slated for Android and iOS (indeed there have already been some early releases). Arguably the most cross-platform non-propietary tech stack in existence.

WebGPU knowledge and code is portable to C, Rust etc via lib wgpu.

To me it's a modernised, pleasant GL environment that is a huge step up, organisationally and performance-wise, from WebGL. Not saying without it's issues, WebGPU's still a work in progress.

I'm writing my current project entirely in WebGPU and WebAssembly. Prototype code in JS, using it for things like UI and assembling your shaders dynamically. As project gets heavier, replace performance-critical sections with C, Rust or AssemblyScript compiled to WASM modules, or push that load off to WebGPU compute. With Websockets, WebRTC for multiplay, one can go further.

If you want desktop support, compile to standalone executable using Electron or similar. Besides desktop platforms, WebGPU is also slated for Android and iOS (indeed there have already been some early releases). Arguably the most cross-platform non-propietary tech stack in existence.

WebGPU knowledge and code is portable to C, Rust etc via lib wgpu.

To me it's a modernised, pleasant GL environment that is a huge step up, organisationally and performance-wise, from WebGL. Not saying without it's issues, WebGPU's still a work in progress.

I'm writing my current project entirely in WebGPU and WASM (WebAssembly). Prototype code in JS, use it UI and for things like assembling your shaders dynamically. As project gets heavier, replace performance-critical sections with C, Rust or AssemblyScript compiled to WASM, or push that load off to WebGPU compute.

If you want desktop support, compile to standalone using Electron or similar.

WebGPU knowledge is portable to C, Rust etc via lib wgpu.

  Besides desktop platforms, WebGPU is also slated for Android and iOS (indeed there have already been some early releases).

To me it's a modernised, pleasant GL environment that is a huge step up, organisationally and performance-wise, from WebGL. Not saying without it's issues, WebGPU's still a work in progress.

With websockets and WebRTC for multiplay, one can go a long way these days with web tech.

I'm writing my current project entirely in WebGPU and WASM (WebAssembly). Prototype code in JS, using it for things like UI and assembling your shaders dynamically. As project gets heavier, replace performance-critical sections with C, Rust or AssemblyScript compiled to WASM, or push that load off to WebGPU compute. With Websockets, WebRTC for multiplay, one can go further.

If you want desktop support, compile to standalone executable using Electron or similar. Besides desktop platforms, WebGPU is also slated for Android and iOS (indeed there have already been some early releases). Arguably the most cross-platform non-propietary tech stack in existence.

WebGPU knowledge and code is portable to C, Rust etc via lib wgpu.

To me it's a modernised, pleasant GL environment that is a huge step up, organisationally and performance-wise, from WebGL. Not saying without it's issues, WebGPU's still a work in progress.