RE: Initialization of static variables

["Joe Pluta" <joepluta@xxxxxxxxxxxxxxxxx>]

> From: Joep Beckeringh
> 
> Most obvious solution would be to clean up when you leave. Perform
special
> calls that make the procedures initialize the static variables when
you
> leave the program with LR on.

Actually, Joep, the most "obvious" solution would be to make calls to
the procedure to initialize the variables when I enter the program, in
*INZSR.  This is doable, but it's really a pain in the butt.  When I'm
absolutely forced to do this, I simply set all my parameters to *NOPASS
and then execute the procedure with no parameters.  But it would be a
LOT easier to let the machine handle it.

The fact that procedures are in some ways different from RPG doesn't
mean that you should have to choose between one way of action and
another.  By specifying a keyword either at the procedure interface or
the d-spec level you would have fine-grained control over the actions of
the runtime, and you wouldn't have to code dozens of little what are in
effect "procedure initialization" routines.

And it's not necessary to tie it to the cycle; I think Booth's
suggestion of a RESET *GLOBAL command would actually be very powerful.
Or maybe RESET *LOCAL to reset just local static variables.

Joe