Untimed resolve ?

[Martin Vechev <martin at fadata dot bg>]

In the code below, If X.class is removed after Testx.java is compiled, then 
gij gives ClassNotFoundException.

class X extends Exception
{}

class Testx
{
	public static void main(String args[])
	{
		System.out.println("loaded.");
		new X();
	}
}

At the end of loading the X class it also resolves it, which seems 
incorrect. Shouldn't it rich the actual 'new' bytecode
and then throw NoClassDefFoundError. It is possible the user has a 
try-catch around the 'new X()'. I understand there
could be early resolving, but I thought it is correct to look like it is 
late, i.e. the exception must be thrown when
'new' bytecode is reached and the class must be loaded.