How come you guys didn't know the difference between normal maps and bump maps. Don't you read up on anything?
Bump maps have been around forever in renderers but they never made it big in realtime game engines because they are very expensive. They are used to generate geometry at runtime which is used to figure out what direction a pixel on a face is "facing" (its normal) which is used to generate accurate lighting information.
A normal map uses three color channels to directly encode that normal into a single pixel of information so nothing has to be figured out at runtime. The pixel simply has to be lit based on whether it's "facing" a lightsource or not.
That's why there are PS filters to convert bump maps to normal maps, because if you can drop that step at runtime you gain much performance. There is really no difference in the results except for the performance gain in using normals over bumps.
It is not over for bump maps. It is very difficult to paint a normal map but it is much easier to paint a bump map. You will be painting bump maps and converting them to normal maps for game use.