Re: IBM Compilers (was RE: gnubol: How do we parse thislanguage,anyway?

[RKRayhawk@aol.com]

In a message dated 12/5/99 4:55:27 PM EST, wmklein@ix.netcom.com writes:

<< Note2: If anyone wants me to provide detailed examples of the NOT with
 combined abbreviated conditional differences - which was actually a
 difference between the '68 and '74 ANSI Standards - >>

That is another, potentially valuable subject. Actually there is a very 
specific question I am interested in that is different.  I find a list of 
migration issues in IBM documentation for those moving from OS/VS COBOL to VS 
COBOL II (early VS COBOL II that is not full COBOL '85 yet).

The documentation says that when the newer compiler of these two is regressed 
with COMPR2 the NOT _conditionals_ are still available.  My appreciation is 
that the NOT conditionals were not in any '68 or '74 standard, and I am 
assuming that these were vendor specific extensions.  Can you clarify that?

Did the OS/VS COBOL compiler have any NOT conditional alternates? Which verbs 
had them?

Presumably all arithmetic verb allowed  the positive ON SIZE ERROR. Even in 
the simplist case did the OS/VS compiler allow an arithmetic conditional 
clause to have an arithmetic statment? In any way could that inner arithmetic 
have a conditional clause?

Then narrowed to the specific, if all that applies, ((and I am not sure it 
did at all to the arithmetics; I just have docs that show NO conditionals 
croak when you migrated aggressively to the the earliest releases of VS COBOL 
II with NOCOMPR2, that is unregressed)).. but it it applied how did an inner 
conditional clause bind?

ADD a TO b
   ON SIZE ERROR
       ADD 1 TO error-count
   NOT ON SIZE ERROR
       DISPLAY 'bind me to the best one'
.

The NOT conditional clause syntax change is distinct from the combined 
abbreviated conditional expressions syntax change..  In one sense I am 
trolling for the source model for Micro Focus's extension that binds the 
clause to the inner arithemtic: who were they copying or was it original.  
But also it looks to me like IBM actual layout a migration path that cleansed 
the code of NOT conditional clauses (though because of market place inertia 
few went step by step through that manner of compilation).

Incidentally I am indebted to the poster who has encouraged us to avoid 
abbreviating in a similar manner Micro Focus and Mainframe. Seems like there 
are a number of reasons to avoid that.

Bob Rayhawk










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