Wow, I'm impressed. Maybe I'm going to take some Cpp timeout to take a closer look on xna.

http://www.neoaxisgroup.com/index.htm
engine based on xna, looks pretty good so far but still in development.

As far as I know, neoaxisgroup is based on C# and Ogre technologies.

It looks interesting. But I am still not convinced. Most XNA-projects miss good scene-management, lights and shadows. There are the usual problems with collision and rendering big worlds.

I have no problem to program my own AI system, but programming everything else like scene-management, light-system and much more is simply too much.

I also wonder why there are no big projects made with XNA.

I figure it would be a lot easier to create scene-managment in C# and XNA, at least from what my experiences tell me, but hell that was quite a simple octree scene managment system, and thus not really efficiënt.


Maybe people still believe that C++ and directX will give them a great speed advantage (ok, the speed will go up a little, but it's not even that much anymore). I personally think that projects with C# and XNA would be more managable.

the big problem of XNA: it won't provide you all tools :
terrain,world editor , lightmapper ,folliage editor... caus a good game need these tools.
You'll have to reinvent the wheel once again.
Some people like lot more coding than doing 3D art , that's a choice.

it's not that simple but as machinery frank said you have to set up thing the way you want. For my part I'm using classes which allows me to set everything I need. Now a good game don't need the tool you mention. I mean world editor ? or Folliage editor ? This is not something important. Anyway the fact is that is just a matter of choice. I made mine.