Cobol - zero vs 0

Author

zero vs 0

strellnikof

2006-10-09, 7:56 am

when working with numeric data isn't it more effecient to use 0 (or +0
depending on whether the data item is signed) instead of the word zero?
I always thought a conversion took place when using the word zero and
that unusual things could happen when using zero.

Binyamin Dissen

2006-10-12, 6:55 pm

On 9 Oct 2006 07:06:23 -0700 "strellnikof" <bprice@humana.com> wrote:

:>when working with numeric data isn't it more effecient to use 0 (or +0
:>depending on whether the data item is signed) instead of the word zero?
:> I always thought a conversion took place when using the word zero and
:>that unusual things could happen when using zero.

Any compiler worth its salt should pre-convert the literal to a form matching
the comperand.

--
Binyamin Dissen <bdissen@dissensoftware.com>
http://www.dissensoftware.com

Director, Dissen Software, Bar & Grill - Israel

Should you use the mailblocks package and expect a response from me,
you should preauthorize the dissensoftware.com domain.

I very rarely bother responding to challenge/response systems,
especially those from irresponsible companies.

Richard

2006-10-12, 6:55 pm

strellnikof wrote:
> when working with numeric data isn't it more effecient to use 0 (or +0
> depending on whether the data item is signed) instead of the word zero?
> I always thought a conversion took place when using the word zero and
> that unusual things could happen when using zero.

Do you think that the compiler outputs machine code with the 4 letters
'ZERO' and sufficient code to recognise that and convert it to an
appropriate zero value at run time ?

The answer is no it doesn't. The compiler recognises ZERO or 0 or +0
and generates the same machine code whichever one is used.

Whether the compiler itself takes a few less or more machine cycles is:

a) dependant on how the compiler is written.

b) irrelevant.

William M. Klein

2006-10-12, 6:55 pm

As others have indicated, in GENERAL, this is not true.

I *do* believe that the "zero sign test" is (for SOME implementations) less
efficient that a test for a zero value, i.e.

05 Num-Field Pic 99999 (comp-whatever)

If Num-Field Zero (sign test)
is less efficient than

If Num-Field = Zero (or 0) (relational test)
may be more efficient

However, unless this were in a VERY tight (frequently performed) loop, I doubt
any run-time would ever be able to tell the difference. Also, you would need to
check your specific compiler -as the generated code may or MAY NOT be different
for the two types of source code.

--
Bill Klein
wmklein <at> ix.netcom.com
"strellnikof" <bprice@humana.com> wrote in message
news:1160402783.207211.9330@h48g2000cwc.googlegroups.com...
> when working with numeric data isn't it more effecient to use 0 (or +0
> depending on whether the data item is signed) instead of the word zero?
> I always thought a conversion took place when using the word zero and
> that unusual things could happen when using zero.
>

Donald Tees

2006-10-12, 6:55 pm

Michael Mattias

2006-10-12, 6:55 pm

> strellnikof wrote:
>
> you mean in code? Once compiled, what difference could there be?

If you are worried about 'efficiency' differences between "Zero," "0" and
"+0" in your source code, you have seriously misplaced priorities. Like,
first you should focus on learning what "compilation" means.

MCM

Sponsored Links