I'm considering the use of statically bound SQL in an MS-Windows application (dotNet: ODBC and OLEDB) accessing DB2 for z/OS. Do some of you have experience with this? And comment on setup, complexity, advantages, disadvantages, ...

A first quick search leads me to the "db2cap" command and the page
IBM DB2 9.5 Information Center for Linux, UNIX, and Windows
This seems to indicate that dynamic ODBC queries are "catched" and turned into static in some way or another. Are there other ways to force "static" from within the application, something in the style of SQLJ?
Or did someone try the following setup: have a separately compiled (& precompiled) subprogram written in C (or C++) with embedded static SQL, and calling it from the dotNet program?

Thanks in advance for any comments!