I don't understand how OpenGL performs the blending of a texture on the screen. By default, the clear color is set to (0, 0, 0, 0). If we use the default blending equation GL_SRC_ALPHA and GL_ONE_MINUS_SRC_ALPHA, it means that if our texture is all (1, 0, 0, 0.3), the resulting RGB will be (1, 0, 0)×0.3 + (0, 0, 0)×0.7 = (0.3, 0, 0) BUT the alpha component will be 0.3×0.3 + 0×(1 - 0.3) = 0.3, so we then have a color of (0.3, 0, 0, 0.09) which is not correct, what we want to see on the screen is rather (0.3, 0, 0, 1).
So how does OpenGL manages the final blending with the framebuffer? Does it drop the alpha component (because the screen can only display RGB colors)?