Triple Buffering with SDL_Surface

Directly controlling or enabling triple buffering when using the SDL_Surface API and SDL_UpdateWindowSurface() is not possible through standard SDL functions.

The SDL_Surface approach, as discussed in the lesson, primarily involves:

1. Getting a pointer to the window's surface (SDL_GetWindowSurface()). This represents the pixel data you'll manipulate.
2. Drawing onto this surface using CPU-bound functions (SDL_FillRect, SDL_BlitSurface, etc.).
3. Calling SDL_UpdateWindowSurface() to copy the contents of your modified surface to the visible window.

This API doesn't expose controls for managing multiple hardware buffers. While the underlying operating system or graphics driver might implement its own buffering (potentially even triple buffering) internally to handle the window updates smoothly after you call SDL_UpdateWindowSurface(), this is an implementation detail hidden from your SDL application code. You cannot request or manage a third buffer explicitly using this API.

Contrast with SDL_Renderer

This contrasts with the SDL_Renderer API. When creating a renderer, you can provide flags like SDL_RENDERER_PRESENTVSYNC.

1SDL_Renderer* renderer = SDL_CreateRenderer(
2  window,
3  -1,
4  SDL_RENDERER_ACCELERATED | SDL_RENDERER_PRESENTVSYNC 
5);

While SDL_RENDERER_PRESENTVSYNC primarily requests synchronization with the vertical retrace (VSync) to prevent screen tearing, graphics drivers often implement VSync using double or triple buffering automatically.

Enabling VSync with a renderer is the closest you get in SDL to influencing the use of triple buffering, although it's still typically an automatic decision made by the driver based on the VSync request.

Conclusion

If you need finer control over buffering strategies or want to potentially benefit from triple buffering (often used with VSync to reduce input lag compared to double-buffered VSync), you should use the hardware-accelerated SDL_Renderer API. The SDL_Surface API with SDL_UpdateWindowSurface() does not provide the means to manage or request triple buffering directly.