Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!virtech!cpcahil
From: cpcahil@virtech.uucp (Conor P. Cahill)
Newsgroups: comp.unix.sysv386
Subject: Re: Xenix 2.3.3 386 floating point problem
Keywords: Xenix floating point
Message-ID: <1990Oct03.042002.3473@virtech.uucp>
Date: 3 Oct 90 04:20:02 GMT
References: <1191@integow.uucp>
Reply-To: cpcahil@virtech.UUCP (Conor P. Cahill)
Organization: Virtual Technologies Inc., Sterling VA
Lines: 20

In article <1191@integow.uucp> hot@integow.uucp (Roland van Hout) writes:
>We are using an 25 Mhz Compaq 300 Mb no FPU 8 Mb ram,
>and Xenix V.2.3.3.
>I'm expiriencing troubles with the calculation of floats:
>I have a float which contains :90099090001018.000000

The problem isn't in xenix.  It isn't the compiler itself.  It is
that binary based hardware (like the computer you are running this
stuff on) does not acurately handle floating point digits.

This is a fact of life.  If you need exact precision, you must use
integers, find a BCD (binary coded decimal) implementation or provide
your own library to act on big floating numbers (can you remember how
to do long division by hand ? :-).


-- 
Conor P. Cahill            (703)430-9247        Virtual Technologies, Inc.,
uunet!virtech!cpcahil                           46030 Manekin Plaza, Suite 160
                                                Sterling, VA 22170