Re: [GNU-COBOL] Re: Global files issues moot

["Glen Colbert" <gcolbert@uswest.net>]

>The length of variables is constrained by the Cobol language to be 1-160
for
>strings and 1-30 for numbers (right, Glen?).  All length issues will be
>caught by the parser.


How right depends on what we are implementing.  Numbers on most Cobol
implementations allow for numeric variables from 1-32 digits (plus
formatting characters which can add more).  The length limit on what we
would think of as string variables is not limited by the language, but is
frequently limited by the hardware.  PIC X(32000) is not out of the
ordinary.  This is particularly true when dealing with multidimensional
tables.  Compiler implementations tend to limit the length of data elements
depending on the addressability achievable by computing register offsets.
How much memory can we address with a single pointer?  Do we need to
implement multiple pointer processing for elements that go beyond this
limit?

The 160 character limit that you are thinking of is probabky the limit on
the size for string constants coded into continuation lines.

We should probably have a MEMMAX argument as a compiler option that would
allow the end-user to restrict the amount of memory that WORKING STORAGE can
consume with a default set to somewhere around 5K.

Glen


--
This message was sent through the gnu-cobol mailing list.  To remove yourself
from this mailing list, send a message to gnu-cobol@acm.cs.umr.edu with the
words "unsubscribe gnu-cobol" in the message body.  For more information on
the GNU COBOL project, send mail to gnu-cobol-owner@acm.cs.umr.edu.