Understanding proxy indirection for one-to-one mappings

Question

The developer's guide states the following for proxy indirection:
To use proxy indirection, your domain model must satisfy all of the following criteria:
** The target class of the one-to-one relationship must implement a public interface.
** The one-to-one attribute on the source class must be of the interface type.
** If you employ method accessing ("Configuring Method Accessing"), then the getter and setter methods must use the interface.
My current descriptors all have get/set methods for their attributes. So does this mean that I need to create an interface for each descriptor that has corresponding get/set methods?
If so, this seems to add a little overhead to maintaining the descriptors. Anytime an attribute is added, removed or modified in a way that would change the method signature; the interface must also be updated to reflect the change. Is there any way to avoid this overhead?

Answer 1

The recommended approach for lazy loading one to ones is generally to use ValueHolders.
Proxy indirection is a solution for lazy loading one to ones, not requiring TopLink classes (i.e. ValueHolders) to be imported and referenced in the Domain model. It also doesn't require code generation, or static byte code enhancement.
In both solutions (regular ValueHolders or Proxy Indirection), TopLink needs to build the Domain objects with empty Place Holders. When the place holder is accessed, addition SQL and Object Building occurs. Coding to an interface in Proxy Indirection allows for the 'place holders' to be used under the interface without the user directly referencing ValueHolders. Without this interface TopLink couldn't do this. Proxy Indirection is not a perfect solution however.
Using TopLink's JPA support (10.1.3 in preview mode, or 10.1.3.1) ValueHolders are dynamically put into the byte codes of the classes at runtime, allowing the user to have lazy loading without proxy indirection or putting ValueHolders in the Object model.
Peter