Rich kids are into COBOL

[Chuck Guzis]

On 02/18/2015 12:08 PM, Rich Alderson wrote:

<rant>
That's one of the things that bugs me today--there are too many 
incompatible versions of COBOL out there with IBM being one of the 
chiefest offenders, amazingly.

That's what COBOL 75, IIRC, was supposed to eliminate--no "extensions" 
unless specifically enabled.  The Navy audit tests of that time checked 
for "extended" behavior and would fail you if something sneaked through. 
  I mean, in the world of standard COBOL, what the hell is a COMP-5?

FORTRAN at about the same time was suffering similar "vendor enhancement 
creep".  CDC FORTRAN certainly had its share, including one of two 
"enhancements" that could lead to ambiguous statements.  Univac had 
FORTRAN V.  (Internally, IBM initially called PL/O "FORTRAN VI").  It 
was getting pretty bad.

Later FORTRAN/Fortran standards mandated that any standard FORTRAN had 
to have the capability of compiling and running the standard and 
flagging anything not conforming to the standard.

Which was pretty useless.the original author of a program set out from 
the beginning to write a portable program.  Otherwise, the typical 
programmer is going to enable all the bells and whistles and leave it to 
some poor sod in the future to handle cross-platform issues.

BASIC is perhaps the worst.  How many widely-used BASICs conform to 
X3.113-1987?  Visual BASIC?  Ha!

--Chuck