Re: Parameter prototype question

Mark S. Waterbury wrote:

I am more concerned with checking the type, length and decimal places for "compatibility" across CALLs -- in other words, checking each parameter for the correct type of argument. Especially across separately compiled units (modules).

The compiler and/or runtime should provide for type checking of parameters / arguments for external procedures at bind-time, (for bind-by-copy or bind-by-reference), and compatibility checking at runtime for late-bound calls, to help ensure "type safety."

Mark, in David's case, the parameter that was actually passed to the procedure was the correct type, 9p2. At least, correct by RPG's rules. Not necessarily correct for the application.

I can imagine an RPG parameter-passing mechanism, say CONST(*EXACT) and VALUE(*EXACT), where the passed parameter would have to be a variable defined exactly the same as the prototype. Maybe a literal too, as long as it was the exact same as the prototype. (I see reading further in the thread that Scott Klement has already suggested this.)

It's a whole different thing to provide run-time type-checking on parameters between modules. That would be great, but I don't think it's currently feasible.