The SSA-NAME3 Callable Routine can be invoked either through a socket interface using TCP/IP, or locally using shared (dynamic) libraries.
During development, this design allows the Workbench to be situated on a Windows client, accessing a Callable Routine on a server. It also means that the application program can be developed on a Windows client and access a single copy of the Callable Routine on the server.
This client/server architecture is not recommended to be the basis for a production implementation of an SSA-NAME3 application. It is recommended that the application that calls SSA-NAME3 and the SSA-NAME3 Callable Routine, even if the Callable Routine is started as a Server process, both reside on the same server in a production environment. Significant performance degradation may occur if this design is not followed.
Ssome applications will have to follow this architecture though. For example, the only way possible to use the Address Standardization Module from many platforms is to call the server callable routine on AIX, Solaris or Windows (all 32 bit). Another example is (even when co-resident on the same machine) the use of 64 bit client code with a 32 bit server or vice versa. A user’s production application only requires access to the Callable Routine and the particular Population rule-set(s) used.
The Workbench is a development only component. The Population Override Manager is also intended to be used in a development environment. However it is possible to use the Edit Rule Wizard in a production environment.