3 Jan
2012
3 Jan
'12
9:18 a.m.
On 3 Jan 2012 at 7:29, Ian King wrote:
I worked in a COBOL shop for a time as an analyst, and
found spurious
bugs based on a lack of understanding of the idioms of COBOL. When I
found a *real* bug, one that had tortured the cost accountants for a
couple of years, it was a challenge to get them to admit it was
legitimate, until our budget finally balanced.
My exposure to COBOL was from the inside-out. I worked on a
translator between dialects. Knowing the spec of the source and then
the CODASYL definition (the target) was essential. I probably didn't
write my first test COBOL program for several months.
It was surprising how many trouble reports were answered by a quote
(chapter and verse) from the CODASYL standard. The rules of the MOVE
verb were in particular not well understood. "I know how you expect
this to work, but it doesn't work that way." COBOL is a very
difficult language to *master*. Perhaps not as much as PL/I or
Ada, but close.
I can't imagine starting someone out with COBOL as a first
programming language--unless the student was a lawyer specializing in
contract law.
By comparison, the ANSI standard for FORTRAN 66 was 39 pages; for the
USA BASIC FORTRAN derived from that, perhaps 20--of big print.
Can you even imagine an ANSI standard description of a "modern"
language in 20 pages?
--Chuck