Re: [open-cobol-list] DATE-MODIFIED (in A-Margin)

[John R. Culleton]

On Friday 07 October 2005 07:50 pm, Bill Klein wrote:
> Just as a follow-up to the earlier thread.
>
> Can anyone provide me (or the group) with a CLEAN compile-listing from any
> IBM compiler that allows
>
>   DATE-MODIFIED.
>
> to appear in the A-margin?
>
> I am particularly interested in whether this works with "current" compilers
> vs. OS/VS COBOL (not supported by IBM for over a decade).
>
>   ***
>
> My best guess is that it does not (will not) compile cleanly - and that the
> "recent change" to OpenCobol should be "backed out".
>
> NOTE:
>   If coded in the B-margin and it appears after another Identification
> Division paragraph header, then it should work for ALL dialects.
>
>
>
I repeat my earlier recommendation. With the exception of
PROGRAM-ID no paragraphs in the IDENTIFICATION DIVISION affect
the compilation or operation of the program in any way. It
serves as documentaiton only. The stnadards writers have
whittled away at the allowable paragraphs over the years,
starting with the elimination of one of hte most useful
paragraphs, REMARKS. Therefore it makes no sense to check for
correct paragraph names. Very shortly they will all be
eliminated. 

So I propose that OpenCOBOL allow any kind of paragraph name
after PROGRAM-ID and before ENVIRONMENT DIVISION. If the other
gent likes to use DATE-MODIFIED let him. If I like to have
INSTALLLATION and REMARKS, let me.  

It is indeed a shame that the standards writers take such a
negative view of what Hopper hath wrought, but that is a fact
not likely to be changed. So simplify the compiler and let those
who like the traditional documentation format continue to use it.
If someone miskeys a paragraph header or even creates a new one
called FOOBAR let it go. The standard writers are forcing us
into a world of asterisked headers anyhow, and no compiler checks
those paragraph names. So what sense does it make to check for the
dwindling few that are still allowed? 

If we don't check for paragraph headers then old programs that
still use such things as REMARKS, INSTALLATION etc. will
compile cleanly. That is a virtue and not a defect. The compiler
will still be compliant in a positive sense, in that useful
features will be implemented and checked. But purely documentary
functions need not be checked for. If the standards folks were
truly interested in the documentation of programs they would not
have savaged the list of allowable paragraph names. None of them
did any harm. But they did and so in response we should just
allow any paragraph name in the IDENTIFICATION DIVISION. 
It makes compiler writing simpler and stops the endless discussion of
whether this or that paragraph name is worthy. IMO they are all
worthy if they help document the program.  

-- 
John Culleton
Books with answers to marketing and publishing questions:
http://wexfordpress.com/tex/shortlist.pdf

Book coaches, consultants and packagers:
http://wexfordpress.com/tex/packagers.pdf