RE: Subprocedures and files (was: Procedures are FUN!)

["Joe Pluta" <joepluta@xxxxxxxxxxxxxxxxx>]

> From: Scott Mildenberger
>
> I compiled your code and do get the same error you do.  It seems that the
> CALLB to the module doesn't get the correct value back from %open
> inside the
> procedure.  It is interesting that reversing the two calls in
> PGM1 makes the
> program work without error.  I'm at a loss to explain what is happening.

Yes, Scott, that's the part that really makes me a little crazy.  It seems
that as long as I call the subprogram first, everything is fine.
Unfortunately, in my environment I can't assure that sequence of events, and
in fact I call the procedure first, which is how I found the error.

It's definitely scary when the sequence of calls affects the operation of
the runtime.  And while I said at the very first that I know this isn't
exactly "cricket", it seems like a reasonable request.

Oh well, there's a workaround - separate the module out completely and then
have the old subprogram become nothing more than a stub to an external call
to the procedure.  Doable, but just a wee bit irksome.

Joe