CodeVR's Vulkan Renderer
The following is the start of some discussion on the Rendering System Pull Request #5.
CodeVR's renderer is modeled after a centralized graphics state store, which is laid out similarly to the GLTF specification.
The renderer is an API specific class that follows the AbstractRenderer trait, that takes a GraphicsState object, and creates API specific data structures from that state from its own handles, and renders the state using the graphics API of choice.
Multiple Actors, 1 Graphics State
There's a large number of actors in a scene, each writing to the engine's state. Each could queue a request to the scene graph to add or remove an actor, each could add a construct to the graphics state, or a physics construct to the physics world, etc.
CodeVR's Vulkan Renderer
CodeVR's renderer is modeled after a centralized graphics state store, which is laid out similarly to the GLTF specification.
The renderer is an API specific class that follows the
AbstractRenderertrait, that takes aGraphicsStateobject, and creates API specific data structures from that state from its own handles, and renders the state using the graphics API of choice.Multiple Actors, 1 Graphics State
There's a large number of actors in a scene, each writing to the engine's state. Each could queue a request to the scene graph to add or remove an actor, each could add a construct to the graphics state, or a physics construct to the physics world, etc.