kzhsw

> Haven't tested it, but does this work: > > > chrome --js-flags="--max_old_space_size=8192" This could increase limit for ArrayBuffers, but not js objects, nor postMessage

Manually set the `tolerance` to `1.1920928955078125e-7` (the epsilon for float32) would help. ```js await document.transform( // Losslessly resample animation frames. resample({tolerance: 1.1920928955078125e-7}), // Remove duplicate vertex or texture data, if...

If running in a browser is not required, a natively-compiled opencascade-based app like [mayo](https://github.com/fougue/mayo) and [CAD Assistant](https://dev.opencascade.org/project/cad-assistant), for batched processing, mayo has a command-line mode. If running browser is required,...

Could the step file be shared? There seems to be a performance improvment here https://github.com/kovacsv/occt-import-js/issues/29#issuecomment-1596176140, but tests with large models are needed.

```js Reflect.defineProperty(globalThis, 'localStorage', {get() {return {setItem() {}, getItem() {}}}}) ``` Disable localStorage before using.

https://github.com/rustdesk/rustdesk/discussions/941

Some other features might worth considering: Interleaved animation data: https://github.com/KhronosGroup/glTF/issues/1864 EXT_animation_quantization: https://github.com/KhronosGroup/glTF/pull/1991 Resampling animation:

Where is the memory optimization now? Is it permanently out of the plan?

> Vis tests failed on "Baked Vertex Animation" because [PG#11](https://playground.babylonjs.com/?snapshot=refs/pull/14235/merge#14WJWW#11) calls `Matrix.toArray` and expects an array as the return value, which is incorrect behavior. This is fixed in [PG#13](https://playground.babylonjs.com/?snapshot=refs/pull/14235/merge#14WJWW#13). Seems...

A similar thing is the proposal of [EXT_texture_avif](https://github.com/KhronosGroup/glTF/pull/2235). Or maybe gltfpack can just keep all unknown extensions of texture for maximum compatibility.