The issue with shader integration is that it, in order to be generic, needs to be merged with the other shaders. Otherwise you would get, either completely black shadows, or shadows with the wrong materials (shaders/textures) in the shadowed areas.
You can find examples on nVidia's developer website, which should be straightforward to integrate.
You should note that unless you do some filtering shadow maps generate pixelated shadows, which depending on rendertarget resolution can be severe. In addition point lights require either a very big rendertarget or six seperate rendertargets to produce good results. If you wish to use an A8R8G8B8, you need to pack and unpack the depth in the pixel shader (which isn't a biggy).
Yes, shadow mapping is mostly a better and faster choice, but it does have drawbacks.
You can find examples on nVidia's developer website, which should be straightforward to integrate.
You should note that unless you do some filtering shadow maps generate pixelated shadows, which depending on rendertarget resolution can be severe. In addition point lights require either a very big rendertarget or six seperate rendertargets to produce good results. If you wish to use an A8R8G8B8, you need to pack and unpack the depth in the pixel shader (which isn't a biggy).
Yes, shadow mapping is mostly a better and faster choice, but it does have drawbacks.