Re: [Nano-devel] nanomiscbugs2 update

[Chris Allegretta]

On Wed, Nov 20, 2002 at 06:54:34PM -0500, David Benbennick wrote:
> I wonder whether this is a good idea.  Here are two reasons it might not 
> be:
> 
> 1) It's natural to distinguish three types of variables: booleans,
> numbers, and pointers.  Saying "if (x)" signals that x is a boolean, with
> only two values.  Indeed, Java enforces this style.  Blurring the
> difference between pointers and booleans doesn't make the code easier to
> understand.
> 
> 2) Lots of functions are documented as returning NULL for certain
> conditions, but I don't know of anything saying NULL is necessarily 0.  It
> just happens to be on every real system.
> 
> Anyway, in my opinion it's (slightly) better to explicitly test for NULL.

I would tend to agree.  I also like testing for == 0 or !=0 as well, 
just because it makes the program easier to read (I absolutely hate 
!variablename).  This may be overkill though ;-)

Chris A
-- 
Chris Allegretta        http://www.asty.org

"Share and Enjoy" - Douglas Adams, 1952 - 2001