The stubcode interface offers ease of use with Chameleon at the expense of flexibility. The dynamic interface offers an alternative to using stubcode which gives more power, but requires more effort on the part of the programmer. This table summarizes the main differences between the approaches:

	Dynamic Interface	Static Interface
Needs code generation?
Can the compiler find out the spelling mistakes?
Advantages	Message tables can be explored dynamically. Application can be delivered without stubcode	Guaranteed to avoid the runtime problems caused by typos.
Disadvantages	Cannot avoid the problem caused by typos.	Application has to be delivered with stubcode
Changing messages and table grammar	Do not need to recompile	Need to recompile

To use the dynamic interface to parse a message, these steps must be followed:

• Instantiate a DynamicEngine object. This is required because the com.interfaceware.chameleon.Engine object has pure virtual methods and cannot be instantiated directly.
• Then load the message definition file
• Instantiate an empty Table object.
• Invoke parseMessageDynamically() to parse the message.
• The resulting Table object can then be explored using the methods defined on it.

The following code fragment shows these steps. For a complete example of the usage of the dynamic interface, look at the New Project Wizard labelled 'Dynamic', which makes a simple example project that uses the interface.

//Create a dynamic engine object.
DynamicEngine AEngine = new DynamicEngine();
         
//Load the  VMD file.
AEngine.load("project.vmd");

// Instantiate a table object.
try
{
   Table MessageTable = new Table();
   // Parse the message.
   int MessageIndex = AEngine.parseMessageDynamically(Message, MessageTable);
   // Extract information from the MessageTable
   // :
}
finally
{
   // We invoke the release() method on the Table to ensure it is released.
   // See the section on garbage collection with regard to Chameleon in this
   // manual
   MessageTable.release();
}

	Have a good read of Javadoc for the Table object to understand how to navigate through the table object. Methods like columnName(int ColumnIndex) are handy for finding the indexes of columns. Also, remember that you'll need to use the subTable method to access the actual tables defined in the message since the top most table is the message itself.