To stop your update function from becoming bloated. To make the code easier to read and maintain. Also, placing that code in your sprite class is a bad idea, sprites should be reusable and shouldn't be connected to input from the player.  Doing so will mean that you're unable to use the same class for other in game objects that ...

Haven't had the need to do it in any of my XNA projects yet but yes, that's how you do it.

After reaching 16.0f, each frame you're doing enemy.position.Z -= 0.5f then enemy.position.Z += 0.5f.  These will cancel each other out. Try something like: enemy.velocity = new Vector3( 0.0f, 0.0f, -0.5f ...

You can remove the limb from the original model by scaling all the bones in the limb to 0 length then having another "stub" limb inside the original that isn't skinned to those bones.  If you want the limb to be visable after being cut off that will need to be a seperate mesh, identical to the one on the main body.  Place ...

[quote user="Morsel"]For example: if I had an item class that requires a texture, the way I'm currently doing things is that my item.LoadContent method would have something like: texture = content.Load<Texture2D>("blah");  so if I had 50 items in my game it would load 50 textures even though there is ...

There's advantages to both methods of code design. Doing things using the method described in the book gives the advantages PinoEire pointed out, code is easier to maintain since there is only one version of the maths functions to maintain, code size is also reduced due to not requireing different implementation of them. Doing things the ...

No, you actually need to write the code to build a mesh from the height map.  There's alot of tutorials around to show you how to do this on most programming sites including this one.  Don't expect everything in life to be done for you.

Have it as a seperate class, keep any information about the player in there, things like screen or world position, information about movement, rendering information - either textures or information on what frame or state the player is in - this will depend how the rest of your game is setup, you could also store the GamePad.PlayerIndex if ...

I got this exact question at an interview once.

Randomising the next block for Tetris or checking for a complete line isn't AI, its game logic, similary so is the placement of mines in minesweeper.  I wouldn't say the fallling of blocks in Tetris or the "collision" that's done is implemented through physics although it could be done that way if want but you'd end ...