Re: [open-cobol-list] Posible difference in the precision of COMPUTE statements between OPEN COBOL 1.1 and mainframe COBOL

[Michael]

Not that it helps or answers any question, but: I've used Open Cobol, 
Fujitsu Cobol, MicroFocus Cobol, Acu Cobol, Kobol, and HP Cobol-II, and 
each have their own small differences and rounding has been a issue 
moving from one compiler to another. It was the root cause of a bug I 
chased down last month going from HP Cobol the AcuCobol. What I've 
started doing lately is to use DIVIDE with REMAINDER, and then ADD 1 or 
not depending on the value of the REMAINDER. That way I know that the 
results will be the same on other compilers, plus it will give you the 
ability to force a specific precision based on the value of the remainder.

FWIW: OpenCobol is a great thing for the Cobol community, and I hope 
that one day Open Cobol will set the standard for other Cobol compilers 
to follow!

Mike.

Wolfgang Westphal wrote:
> Hello,
>
> The following example gives different results in OPEN COBOL 1.1 and
> mainframe COBOL:
>
> 01  RESULT-1 PIC 9(6).99.
>     COMPUTE RESULT-1 ROUNDED = 46956.93 / 24 + 5308.00 / 24
>     DISPLAY RESULT-1
>
> OPEN COBOL displays 002177.71
> mainframe displays  002177.70
>
> Do I make any mistake?
> Could there be a fault in my installation?
> Is this normal behaviour of OPEN COBOL?
> Is there a posibility of forcing specific precision?
>
> I am using OPEN COBOL 1.1 with WINDOWS XP.
> For the installation I used a MINGW environment.
>
> Regards
> Wolfgang
>
>
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> open-cobol-list mailing list
> address@hidden
> https://lists.sourceforge.net/lists/listinfo/open-cobol-list
>
>