Re: [open-cobol-list] Latest tarball regression ?

[John Culleton]

On Wednesday 28 January 2009 12:02:02 pm Bill Klein wrote:
> "common" (and "initial") are valid in the program-id with the '85
> Standard. ("recursive" was added in the '02 Standard).
>
> For what they do, see your documentation <G>
>

All my documentation is old school, and I can't find a readable book 
at a reasonable price to upgrade to the new school. I have a text 
(Shelly Cashman Foreman, 2nd Edition) published in 2000 that 
allegedly covers COBOL 2000  but it doesn't cover this variation. 
Neither does Stern and Stern, 1985 edition.  

If someone will add a true GUI interface to the COBOL standard or an 
equivalent to Perl's way of handling cgi-bin chores then my ears will 
perk up. A standardized way to get to a generic RDBMS like Mysql or 
Postgres would be worth having.  But multiple variations on 
activities that already have standard solutions, like calling a 
subprogram and passing parameters back and forth,  simply muddy the 
waters. 

Frnkly I don't see a compelling reason to change from the 1972  way of 
dealing with subprograms. The newer style just adds confusion, as 
indicated by the start of this thread and my erroneous answer.   The 
great virtue of COBOL is that my program can be understood by you and 
your computer and vice versa.  The perpetual  game of standard 
fiddling just encourages a tower of Babel  approach.  Changes which 
increase the number of things a programmer must understand and keep 
track of without adding critical functionality simply add cost to the 
compiler and extend the learning curve for the new (or old!) 
programmer.


-- 
John Culleton
Resources for every author and publisher:
http://wexfordpress.com/tex/shortlist.pdf
http://wexfordpress.com/tex/packagers.pdf
http://www.creativemindspress.com/newbiefaq.htm
http://www.gropenassoc.com/TopLevelPages/reference%20desk.htm