| Arnold Trembley 2005-08-25, 3:55 am |
|
William M. Klein wrote:
> "IKFCBL00" was certainly what one "executed" for OS/VS COBOL (and I think for
> ANS COBOL Vx). However, COBOL(F) was a specific release (version) - very
> early - of IBM's COBOL. It was (as I recall) one of the early "bundled in"
> software products that caused the original "IBM decree" starting "unbundling" of
> software from O/S (and hardware). Again, this was all before my time, so I
> could be mistaken on all/some of it.
>
When I got my first programming job in 1979, the company I worked for
had an IBM 370/145 (not perfectly sure of the model number) with one
meg of main memory, eight 3420 tape drives, and a mixture of 3340,
3344, and 3350 disk drives. The 3350's were the biggest, about the
size of washing machine for each drive, which could store about 350
megabytes.
The operating system was DOS/VS (not sure of the version) and the
compile JCL said something like "// EXEC FCOBOL". The old-timers in
that shop talked about converting from COBOL-D. They were just
starting to use CICS and most programmers had to use batch Panvalet to
edit programs. New programs were, of course, hand-written on coding
sheets and turned over to the keypunch department.
I remember one old guy who retired a couple of years after I started.
He kept a backup on punch cards in his cubicle of every program he
had ever worked on.
To get back on topic, I think having the compiler print out the
options actually used is a distinct benefit. Of course nowadays those
listings hardly ever get printed on paper.
My current shop uses Endevor, and the compile options are defaulted
based on the "Endevor Processor", which allows CICS to have NODYNAM,
and batch to have DYNAM. But we can override most options with
PROCESS or CBL statements embedded in the source code file, another
feature I think is a distinct benefit.
--
http://arnold.trembley.home.att.net/
|