Re: EXTPROC(*CL) and V6R1

hello David,

The argument is simple. Modify the parameter list on the called
procedure, recompile and no need to touch the callers that don't use
the parameter.

Given the example that you posted... if a new numeric field were added, you'd get a decimal data error, because it'd be initialized to blanks. So you'd have to touch the callers -- you'd have to change them to set the initial values of the DSes. Your example would make more sense if you import the DSes with INZ(*LIKEDS). That way you could set the initial values of your DSes in the copy books... then you might not have to touch the callers.

But for every field you're using as a "parameter", you'd have to have a "special value" that indicated that the parameter isn't used. Otherwise, the called routine would have no way of knowing whether the parameter was passed by the caller or not...

I think the real problem is wanting to still write
RPGIII. Procedures are being treated as OPM's and so they are
expected to bring everything back to the caller by parameters.

Yes, I agree.

I don't understand why someone with no experience in a new paradigm for programming would be allowed to set shop standards and force all of the other programs to use it. Why wouldn't they get someone with knowledge and experience to draw up their standards? Or wait until they've had experience with it?

Though, I would argue that even in RPG III this would be a bad way to do things...