On 25 Feb 2014 13:45:44 -0800, in bit.listserv.ibm-main you wrote:

>Yes, the LE routines and predecessor packages going back many years
>are available, but I have always found them all but unusable.  Perhaps
>if they were rewritten now that Enterprise COBOL supports the data
>type function pointer, something better could be produced; but as they
>stand they are both clumsy and preternaturally slow.

Simpler and better would be for Enterprise COBOL to support USAGE BIT
and the logical operations of AND, OR and exclusive OR that are in the
2002 standard.  There also now are standard ways to describe both
fixed and IEEE BINARY floating point fields of lengths starting at 1
byte for fixed length binary.  These would NOT conflict with usages of
COMP-1 and COMP-2 for hex floating point and would allow the two types
of floating point to co-exist in the same module.  There also has been
work to standardize language for decimal floating point and the
various rounding options associated with it.  Given the work done by
Mike Cowlishaw on floating point decimal and that Java, PL1 and C/C++
support it, it is about time that the business customers demand that
IBM COBOL support it.

Clark Morris 
>
>John Gilmore, Ashland, MA 01721 - USA

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to