Re: In answer to RW - again (was: Sorts (revised)

From: Richard (riplin_at_Azonic.co.nz)
Date: 07/07/04 

Date: 7 Jul 2004 02:18:33 -0700

robert.deletethis@wagner.net (Robert Wagner) wrote

> >There *ARE* universal truths about "COBOL" about "programming" etc. Many
> >questions deserve answers that provide such.
>
> Anything in the Standard is a "universal truth" about Cobol.

There is not a 'the Standard'. The current standard_s_ in general use
include ANS'74, ANS'85, ANS'85 + amendment'89, X/Open. It may well be
that someday we may be able to use the ANS'02 standard.

Given that many issues differ between these various standards, none of
them form a 'universal truth'.

> On programming
> style in general, I can't think of any assertion that would go unchallenged
> here.

That is often because many of your assertions are simply, factually,
wrong, especially when unqualified.

> GO TO, periods and numbered paragraphs all have their champions on CLC.

Your 'assertions' are usually made as a religious crusade against
them. In the past you have been quite insulting about their use,
showing that your argument is emotional rather than rational.

No one requires that you use any of these items, I would not
'champion' them, but I would defend the freedom for anyone to use
whatever style they like (as long as they don't do it in my programs).

> -- Micro Focus, IBM, Fujitsu
> and (formerly) Realia -- which all began life as IBM clones.

The term 'IBM clone' is meaningless in this context. Which of the
several families of IBM compilers are you referring to that these may
be a 'clone' of. Some of these are: OSVS, AS/400, Visual Age.

MicroFocus 'began life' as very restricted subset on an 16Kb desktop
machine and then became CIS Cobol on CP/M and MP/M. It was never an
'IBM clone' but implemented much of X/Open. MicroFocus Level II Cobol
was almost a complete X/Open implementation. What part of X/Open does
'IBM' use ?

Perhaps you are confused because there was an 'IBM Cobol' on the PC
that was a rebadged version of Microfocus Cobol/2. MF Cobol/2
included many OSVS features, probably implemented at IBM's request for
this version, but also has MicroSoft, X/Open, and unique features that
are nothing to do any IBM extensions.

> Micro Focus has
> since gone beyond IBM-compatibility to become the 'gold standard' of Cobol,
> IMO.

When you start at the wong place you wind up at the wrong conclusion.
In this case it is: 'has since gone beyond ..'. MF never was an 'IBM
clone', it was always an interactive system.

> Thus, when an extension such as POINTER or GOBACK
> works the same on MF, IBM and Fujitsu, that's a universal truth in my
> universe.

Fujitsu does not document or support the particular MF POINTER
implementation. While the current compiler may work there is no
guarantee that a new version will continue to implement it.

> When a Standard-compliant(?) feature such as IF NUMERIC has tripped me up in
> the past, that's a universal falsehood in my universe.

But it is not a 'falsehood' of any Cobol, but of your knowledge of how
to use it correctly.

> I would advocate INSPECT ..
> CONVERTING as a more reliable alternative.

In what way is a 'INSPECT CONVERTING' an alternative for 'IF NUMERIC'
?

> We recently saw here an example of
> Class Test not working as expected, yet no one thought to recommend INSPECT as
> a solution.

The fault was the expectation, not the class test.

> In short, I think we'd profit more from commonalities than from differences.
> Rather than asking which compiler, I'd rather say 'this way will work on all
> compilers' .. now hastily qualified with '(in my experience)'.

Which is not useful at all when the particular compiler being used is
not one within your preference of a rather restricted subset.

In fact your '(in my experience)' qualification seems to be quite
wrong too. You just said:

> It's unlikely I'll ever again work on AcuCobol, RM, Kobol or Unisys,
> although I have used them in the past.

Your experience (if your statement about this use is actually true)
obviously includes these which _do_not_ work with many of the features
you choose to call 'universal'.

Relevant Pages

  • Re: exception handling for intrinsic functions
    ... past experiences) my GUESS is that IBM would implement COBOL features first ... I think that the VSE COBOL ... > implement the 2002 COBOL standard. ... > the benefit of the Intrinsic Functions. ...
    (comp.lang.cobol)
  • Re: Occurs Depending Memory Use
    ... differently as you don't really have job steps in the sense that IBM ... How familiar are you with IBM COBOL ... Their compiler conforms to the standard. ...
    (comp.lang.cobol)
  • Re: Revisiting an Old Prejudice: READ INTO/WRITE FROM
    ... > Standard" of always doing a READ INTO and WRITE FROM became very common in IBM ... As someone who coded varying flavors in varying flavors of IBM COBOL ... resulted if you attempted to address the record area for a file after ...
    (comp.lang.cobol)
  • Re: Please Explain about FACTORY
    ... IBM's OO implementation does NOT match the ANSI/ISO 2002 Standard, ... (Remember that IBM uses a Java-influenced and interacting OO solution for ... "Object-oriented Programming with COBOL" ... "Programming Factory Object Behavior" ...
    (comp.lang.cobol)
  • Re: Week Number in Fujitsu cobol
    ... reasonable to use the term "Julian date" as it is used in the COBOL ... '02 ISO Standard, ... the universe as it is understood by COBOL programmers. ...
    (comp.lang.cobol)