Faster SDL Rendering

Question

I am trying to understand the whole process of 2D accelerated rendering using SDL 2.0.
So my question is which of the most effective ways to draw circles on the screen and why?

Some ways:

• First create a software surface, and then draw the desired pixels on that surface, then create a texture from that surface and finally copy this texture to the render target.

• Another implementation is to draw a circle using SDL_RenderDrawLine several times. And I think that this is how it is implemented in SDL 2.0 gfx

Or is there a more efficient way to do all this?

Take this question in the general case, if I wanted to manually draw other shapes, which probably could not be easily implemented using the 2D rendering API that provides SDL (using a draw line or a rectangle).

Answer 1

, , , . , , , , , 60 , , , , , , ( ). . SDL_RenderDrawLines , .

, , , . " " .

, API- , , , . - .

, , . , , .