Re: Varying - Inz mandatory?

> I can see the reasoning behind this because you would have some pretty mad
> customers if they had to "fix" hundreds of programs, but I also think that
> the problems that could result from not throwing an error could be even
more
> troublesome than somebody having to go in and code a simple INZ for data
> structures.  Fixing errors based on the compiler listing is cake, fixing
> errors because a data structure wasn't initialized can be evasive.

And it's a good thought Aaron, but INZ is NOT mandatory.  In fact, I rely on
static variables (as in not initialised) in many of my procedures.  Whether
to initialise or not is really a design decision, rather than a compiler
decision.  I have no problem at all with a low severity warning because I
can always turn up the heat by changing the message file.
  --buck