🚀 go-pugleaf

Back in 1992, I was taught a "better and more portable" way of writing a
numerical algorithm I was working on. I was primarily interested in speed,
but I needed the precision of at least a double, and was not sure whether
to use a double or a long double, since my code may run on different
processors, and on one, long double was the fastest, and on another double
was the fastest. I was told use neither! Use double_t.

Specifically, I was told:

- double_t is guaranteed to be the fastest floating point type, for your
implementation, that is at least the size of a double.

- also there exists a typedef named float_t, which provides the fastest
floating point type, (float, double, long double), for your
implementation, that is at least the size of a float.

- start using the new standard header <fp.h>, which is where float_t and
double_t are defined.

This was then, I believe, the emerging standard from the Numerical C
Extensions Group.

Now it is 1996. Has there been any changes with respect to the three
points above? I did not see any references to float_t, double_t, or <fp.h>
in the April draft.

Please also reply to jzipnick@best.com.

Jay Zipnick


[ comp.std.c++ is moderated.  To submit articles: Try just posting with your
                newsreader.  If that fails, use mailto:std-c++@ncar.ucar.edu
  comp.std.c++ FAQ: http://reality.sgi.com/austern/std-c++/faq.html
  Moderation policy: http://reality.sgi.com/austern/std-c++/policy.html
  Comments? mailto:std-c++-request@ncar.ucar.edu
]