Fixed point financial data versus floating point - Re: Spreadsheets (was Microsoft flamage) - test-drb@ccmp.vtda.org

24 Oct 2011

The same thing can happen in COBOL. All the guys when I started on IBM knew the native
data types and which ones to use for what. A few years ago somebody asked me to give a
training session to some guys who had 2 to 8 years with COBOL on IBM and nobody in the
room (about 15 people) had any idea about anything, not one of them ever opened an IBM
manual. Lucky for them and their employer, they never actually wrote any code, they just
cut and pasted and were using packed decimal for money. 

------Original Message------
From: Fred Cisin
Sender: cctalk-bounces at classiccmp.org
To: General Discussion: On-Topic and Off-Topic Posts
ReplyTo: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Fixed point financial data versus floating point - Re: Spreadsheets (was
Microsoft flamage)
Sent: 24 Oct 2011 19:38

On Mon, 24 Oct 2011, Vintage Coder wrote:
...
  Agreed. It's not floating point (no
mantissa/exponent) and the user has
 complete control over the decimal point, which is imaginary as far as
 the representation goes. There is no decimal point in the data, only
 digits and a sign. The name "packed decimal" seems like a safe way to
 differentiate it from floating point while giving a hint to the internal
 representation (two digits to a byte). 
As most everyohne here knows, the 80x85 family has some minimal support
for both packed and unpacked BCD.

But the lack of a widely known data type using those in C and BASIC
results in way too much software being written with inappropriate data
types, such as "float".