Marshal-by-value objects are copied by the remoting system

and passed in their entirety to the caller's application

domain. Once copied to the caller's application domain (by

the marshaling process), all method calls and property

accesses are executed entirely within that domain. The

entire object exists in the caller's domain, so there is no

need to marshal accesses across domain boundaries. Using

marshal-by-value objects can increase performance and

reduce network traffic when used for small objects or

objects to which you will be making many accesses. However,

because the object exists entirely in the caller's

application domain, no state changes to the object are

communicated to the originating application domain, or from

the originator back to the caller. Marshal-by-value is not

a good choice for very large objects with many accesses. It

makes little sense to marshal an entire large object across

domain boundaries when all you need is access to a single

field, method, or property.