Sorry should have selected only the text that I was referring too.
I code for a living but program in DX as a side hobby, and in terms of difficulty, the changes I've had to do to my code was minimal.
I can understand it being difficult for full blown development teams to modify an existing graphics engine or create an entirely new one but that is more of a man power / cost issue. If you structure your code properly from the start, the API calls you make from your code should be invisible to the operating system, i.e. I write a function to pixel on the screen, my code doesn't care what DX version is installed on the PC or what DX version the graphics card can use, the code automatically calls the appropriate DLLs to draw that pixel.
|