Re: %EOF, %FOUND and the infamous Dow loop

[Douglas Handy <dhandy1@xxxxxxxxxxxxx>]

Scott,

>If you did the return value for a user-written function/procedure wouldn't
>it then have to re-run the function?

Yes.

>If that function updates something,
>or prints something, or does communication (etc) that might not be a good
>idea.

True, just like it is true in IDEs which do give you this capability.

>So how would you design this?

"Just" have it run the function and display the return value.  It would be up to
the programmer to consider the side efects, if any, of invoking the function
again.  Just like on other platforms.

>I agree that this would be handy, but it seems complicated :)

I don't know enough about the ILE architecture to know how easy it would be to
allow the debugger to dynamically call a prototyped function.  Obviously the
first obstacle is for it to find the correct prototype.

Even if this were possible, IBM may have other reasons for not supporting it.  A
recent post suggested one reason they don't support changing the instruction
pointer for the next statement to be executed is due to potential security
ramifications.

Doug