C# 3.0 introduces the concept of automatic (or more correctly, auto-implemented) properties.  Being that the restrictions on this feature make it applicable to only few situations, I'm left wondering why the good folks at Redmond chose to make this a feature of the language (i.e. the compiler) instead of the IDE.

An automatic property looks like this:

  1   public int Id {
  2     get;
  3     set;
  4   }

Behind the scenes, the C# compiler converts this to something like:

  1   public int Id {
  2     get { return _id; }
  3     set { _id = value; }
  4   }
  5 
  6   // Compiler generated field
  7   private int _id;

Except that:

• The generated field isn't called _id; it has an internal name that you don't know.
• The generated field can't be accessed by your code.
• Automatic properties can't have attributes.

I understand the need for relieving the developer from manually writing hundreds of lines of code that look like this:

  1   // Public property
  2   public int Prop1 {
  3     get { return _prop1; }
  4     set { _prop1 = value; }
  5   }
  6   // Private field
  7   private int _prop1;
  8 
  9   // Public property
 10   public string Prop2 {
 11     get { return _prop2; }
 12     set { _prop2 = value; }
 13   }
 14   // Private field
 15   private string _prop2;
 16 
 17   ... etc ...

but the restrictions on automated properties make them less useful than I'd like.  Perhaps this generation of code could have been better purposed by the IDE?