> ANSI does not require such a special case, and I really
> don't like the idea of masking bugs in applications by kluging up the
> C library.

In my experience, an implementation that masks or otherwise makes it 
difficult to detect errors is not a useful "reference platform" for 
writing portable software.  You'll never know you have a problem until 
you are sent to a vendor's porting center half-way across the country.
And you'll be forced to track it down and fix it with the primitive tools 
that are common on pre-release hardware. :-).

One of the most notable examples of this is AT&T's documenters workbench
(troff, etc.).  It was developed on machines that did not trap null
references.  So, it didn't work at all when it was sold (for lots of
$$$'s) to vendors whos machines did trap null references.

	--jtc


------------------------------------------------------------------------------