I've been having a discussion with another programmer about the code 
behind feature in .NET.  I've been treating it as a way to seperate 
presentation from logic.  IE.. I put objects on the page, assign a CSS 
sheet, and use code behind for binding to COM objects,etc..

He is going the route of everything is done in the code behind.  He does 
nothing in the presentation layer, and in fact only one page is ever 
created, with all of the objects being swapped in and out of the page to 
emulate other pages.

Is there any performance benefit one way or the other?  I know that .NET 
pages are compiled now, but I don't know if early vs late binding is an 
issue with it.  Any clarifiaction will be greatly appreciated.

-=3D John