Preload Component Priority Ordering for Low-Traffic VR Scenes

Preload Component Asset Ordering

A-Frame's preload component caches external assets across entity lifetimes, preventing redundant network requests when the same asset is referenced multiple times. However, the component does not expose a priority or ordering mechanism for asset loading. In low-traffic VR workloads, this means low-priority assets may block high-priority ones during scene startup, creating unnecessary latency for critical resources.

The 'visible' state of an entity does not automatically cancel preload, requiring manual component disposal or lifecycle hooks to achieve cost reduction. Instanced meshes share geometry via BufferGeometry, and preload caching interacts when the same geometry is referenced across multiple entities, reducing duplicate loads. Changes to the preload component between major releases may affect asset caching expectations.

Which assets should receive loading priority when multiple entities reference the same external resource in a low-traffic VR scene? Should high-priority assets be loaded before lower-priority ones, or is there a configuration option to control this behavior?