From: Roger Scudder
RS>  That is actually covered by the standard in section 7.20.3.2
RS>  regarding free:
RS>  "If ptr is a NULL pointer, no action occurs."

    I still test it, just to see if it is necessary to perform the
 call at all.  I like to keep a firm grip on my pointers so I don't
 accidentally poke my eye out while attempting to shoot myself in
 the foot.  I want them to be 0 when not in use, and not to be used
 when they are 0.

RS>  BTW, I didn't edit the test function in my reply to Bo, though
RS>  it definitely needs a test condition after the call to
RS>  malloc.  I was only focusing on the question, which was "how
RS>  do I free the memory?"... though in hindsight I probably
RS>  should have added testing to the entire snippet for the sake
RS>  of consistency.

    I put the test conditions in because, as Bob noted earlier,
 when you start out in DOS, you distrust anything untested.  I like
 to build and test any code I produce as an example, just to make
 sure I remember syntax and my C basics.  I haven't been doing
 enough coding to keep the rust off, unfortunately.  So, when I do
 the testing for compilability, I make sure I'm not playing
 pointer roulette at the same time.  Personal paranoia.  :)

>  kkuzba{at}centurytel.net   http://home.centurytel.net/kkuzba
>  'D`you mean to say you've been through this hole?' said Sam.

--- SoupGate-Win32 v1.05