Article 112714 of comp.os.vms:

At 18:31 24-11-96 GMT, Kevin wrote:
>On Sun, 24 Nov 1996 18:00:00 GMT, kevin@highveld.demon.co.uk (Kevin)
>wrote:
>[snip]
>>  Computers are used "in the real world" like finance,
>[snip}
>
>er, by this I don't mean that computer people don't live in the "real
>world", just that the end use of the product relies on a decimal base
>and so cannot be the reason for base 8 or 16 in a compiler ;-)
>

Huh?? Compilers working in base 8 or 16????

Kevin, the compiler does work in decimal, you type in 3.33, it will
understand it and store it in the varaable of your choice. But if you store
it in a floating point variable - and this is NOT a function of the compiler
- it will lose some precision because it is stored as a binary fraction.

If you want store STORE the number in decimal you should use the appropriate
datatype, floating point variables are not appropriate in a lot of
applications - like business apps where rounding is crucial and it sounds
like they are not appropriate for your app. If the language you are using
does not support decimal datatypes, then use one that does, COBOL, or DBL....

=======================================================
Nigel White  (nigel@synergex.com)
Systems Programmer, Synergex (Formerly DISC)
2330 Gold Meadow Way; Gold River, CA 95670
Phone: 916/8530366; Fax: 916/635-6549; WWW: http://www.synergex.com