Re: J4 - presentation/discussion on "Future of the COBOL Standard"

On Fri, 14 Mar 2008 22:22:45 +1300, "Pete Dashwood"
<dashwood@[EMAIL PROTECTED]
>
wrote:

>I know the answer to this one, because it is one of the questions I asked
on 
>my first COBOL course in 1967...

Another neophyte. 

>It has nothing to do with variable strings being too complicated for 
>compiler writers to deal with.  Later I spent some time with people who
had 
>actually written a very popular COBOL compiler, and I even got to patch
the 
>compiler myself so it would do what was needed for the site where I was 
>employed. All of this experience confirmed that the 31 byte limit on a
COBOL 
>name was nothing to do with variable strings being too complicated. I'm
not 
>going to tell you what the real reason was, but I can give you this clue:
>
>At the time COBOL was developed the 8 bit byte had not yet become the 
>industry standard. Different architectures used 64, 32, or 24 bit words, 
>based on a 6 bit character. COBOL ran on these machines just as it did on
8 
>bit byte architectures.

So we're stuck with a 31 byte limit because we might wake one day to find
the last 40
years has been a bad dream, we're back to the ICL (whatever) with a 24 bit
word. 

Increasing the limit would have zero impact on legacy code. 

>> Last week, I passed up an op****tunity to do that -- wrap Cobol into SOA

>> services and give
>> them a pretty C#/AJAX face.
>
>It would be fun.

Yes, it would, and a tremendous resume enhancement. Available gigs would
increase by
50-100%. 

>> Think outside the Cobol box. You're not limited to Cobol functions, you

>> can call ANY
>> function that follows the XP/ELF/ABI standard. NetCOBOL under Win XP
has 
>> access to
>> hundreds, if not thousands, of useful functions. They're only a CALL
away.
>
>So what's the point in using COBOL, then? :-)

Ask Doc Dwarf what's the point in using English. Given that his postings
are 2% Latin, he
might as well write everything in Latin. And you should write everything
in German. :)