The following article appearred on the cover of the Y2K Journal

COBOL
 Coca-Cola
and the Year 2000

by Jerome Garfunkel

Things Built On Core Principles Endure

Coke is the real thing; COBOL is clarity in computer systems. How wonderful to see COBOL programmers in such demand these days. COBOL has remained a technically relevant programming language for its entire 38 year history and will remain so well after January 1, 2000. The American and International COBOL Committees continue to assure this. COBOL has been fighting a public image problem for nearly all of its existence. COBOL programmers have been at the receiving end of abuse for too long. Academia, more interested in computer science than in commercial computing, turned its back on the COBOL language from the very beginning. It was left primarily to private industry to train employees in business computing subjects including COBOL programming. Many people have pointed to COBOL as one of the reasons for the Year-2000 debacle we are approaching,, citing the large number of COBOL programs that must now be "fixed."

The millennium problem facing the computer industry today is not caused by COBOL as many have asserted. If anything, the COBOL language will prove in the end to have helped in the solution of this problem. The millennium problem is caused by the shortsightedness of the computer application designers who chose to take shortcuts - using whatever languages were available - to represent dates with two digit years rather than 4 digit years (i.e. "97" rather than "1997"), and to ignore accurate rules for leap years.

The computer industry - indeed the entire international business community - is rediscovering the lessons learned 2 and 3 decades ago: for computer applications, clearness is a virtue, not cleverness. COBOL shines above all other programming languages in its clarity - always has, always will.

The perception that the 38 year old COBOL languages is obsolete and irrelevant for today’s computer systems, while being an accurate assessment of current opinion is nonetheless an incorrect conclusion. The notion that COBOL programmers will become "dinosaurs" again after January 1, 2000 is mistakenly presumed by the subsequent generation of computer programmers who grew up with the first home PC’s and who lack perspective of the computer industry’s history. COBOL is not irrelevant now nor will it be after 2000. It is true that the new generation of programmers (and managers) prefers C, C++, Java, etc. This is because these languages arose from the operating environments of the personal computers (DOS, Unix, Internet, etc.) with which they were most familiar. But this new generation of programmers (and managers) is ignoring the reasons that COBOL became so popular 30 years ago for programming business computers. COBOL is simply the clearest language around to develop complex computer applications. We need only look at the huge inventory of COBOL programs still running on computers today to observe that these applications, some of which were developed 10, 20 and 30 years ago, are so durable that rather than needing to be discarded and replaced over the years, they needed simply to be modified and updated. There are few things in the world today, so solid and durable, that can boast of such a legacy. Unfortunately the enormous popularity of more technical (lower-lever) languages such as C, C++, Java, etc. will result in an enormous computer application maintenance problem well into the next millennium. The computer industry is already seeing this emerging problem. In many ways it is a repeat of the problems that the computer industry faced 30 years ago as it struggled to replace its "assembler" (low level) applications with more maintainable systems developed in languages such as COBOL.

The move away from COBOL in the last decade has been felt very subtly by the entire commercial world. Many have observed that the world is becoming a little less "humanized" and more impersonal in its inter-relationships. Computers are often cited as the cause. The real problem is not the excessive use of computers. It is more often caused by the rigid design and clever development of modern computer applications. This is one consequence of replacing clear COBOL instructions with terse and clever C, C++, etc., instructions in the computers that run the commercial world.

So here’s to COBOL and Coca Cola. I expect they will both be around well after January 1, 2000.

The author is a member of the American COBOL Committee (ANSI X3J4), the International COBOL Committee (ISO WG4); as well as other computer industry standards committees.


Jerome Garfunkel
172 Tinker Street
Woodstock, NY   12498 USA
Tel/Fax:   +1 845 679 0121 
associates@jeromegarfunkel.com
www.jeromegarfunkel.com