A warning patch, ptr assignment cleanup + error w/define after col 0

[Jeffrey A Law]

  In message < 199903222305.SAA24754@blastula.phys.columbia.edu >you write:
  > This is a good point.  I still think it's worth going through and
  > standardizing, however.  We want to use unsigned char everywhere
  > we possibly can, because that will reduce the likelihood of future bugs.
  > 
  > Is U_CHAR in system.h?  If so, we could start by s/unsigned char/U_CHAR/
  > everywhere.  That'd reduce human confusion if nothing else.  (I'd prefer to
  > downcase U_CHAR at the same time, but that's me.)
I'd prefer to just use "unsigned char".  I don't see the benefit of
indirection to use U_CHAR.


  > What determines whether char == signed char or char == unsigned char by
  > default?
It's system dependent.  Usually we try to be compatible with the
vendor's compilers.  And typically the vendor's compilers will choose
a signedness which is most efficient for their hardware.


jeff