Letters - letter to the editor

I don't think booking revenue (I call it revenue recognition) when a product is shipped is ever good practice. And I cannot imagine booking revenue simply because an order has been received.

What I've been doing for many years is to recognize software revenue when the software has been installed. If the software is a reasonably standard product, the mere act of installation should be sufficient for revenue recognition. If the software represents a specialized and custom solution, quite often the criterion applied is to require a sample problem to be solved.

This may sound conservative, but it's simply good business practice.

Scan-Optics sells mainly hardware, although we derive a reasonable percentage of our revenue from software. The practices noted above are not theoretical, this is what we do. Richard I. Tanaka Chairman and CEO Scan-Optics, Inc. East Hartford, Conn.

COBOL OBIT STILL PREMATURE

It seems as if Cobol bashing will never go away (June, Newsfront, p. 12). Mr. Martin, a respected "guru" in our industry, is certainly not the first to dismiss the Cobol language as if it were nothing more than a "wart" on the beast called Computers and Computing. A prestigious Dutch computer science professor, Edsgar Dykstra, also expressed contempt for Cobol when he said, "The teaching of Cobol should be made a criminal offense."

The modern Cobol language (a derivative of the original 1960 version) has survived Professor Dykstra and it will more than likely outlive Mr. Martin.

Cobol bashing is not new. The earliest known "Cobol bashing" was expressed in the first year of Cobol's existence.

In 1959, as the Codasyl Cobol Committee was working on the original draft of the Cobol language (borrowing heavily from Grace Hopper's "Flowmatic" compiler developed for Univac), one of the Codasyl members, Howard Bromberg, then of RCA, sent a tomstone to the chairman of the committee, Charlie Philips of the Department of Defense. (An interesting side to this story of which Grace Hopper Likes to remind people is that the tombstone was sent collect.) At the top of the tombstone, under a small sacrificial lamb, was the word "Cobol," engraved in gold leaf. Howard's message was clear. HE (and there were others) had little confidence that Cobol would survive very long and the tombstone was to "stand in the wings" to be ready for Cobol's demise at any time.

That tombstone was heard from again in 1985, the year the official American (and international) Cobol (85) standard was updated to reflect state-of-the-art programming techniques. It was found in Charlie Philips office (soon after he died). At a prestigious ceremony, in May 1985, at the Boston Computer museum, the tombstone was dedicated and forever written into the official annals of computing.

The Cobol language has turned out to be the most successful application development tool ever devised. The history of computers and computing in America (in the world) would have been quite different had Cobol (or its equivalent) not been introduced in the early 1960s. The modern Cobol languange, known as ANS Cobol 85, is a far cry from the original language, but clearly an outgrowth of the 1960 specification.

Despite being much maligned over the years, Cobol has developed into one of the most modern and popular application development tools. The current version of Cobol contains high-level functions and macro instructions. A new object-oriented version of Cobol is being developed by the (still functioning) Codasyl Cobol Committee and the Ansi (and ISO) Cobol Committees. It is likely that the main application development tools in use at the turn of this century will be Cobol-based or Cobol-compatible.

To Mr. Martin, as well as to Professor Dykstra and all of the other Cobol bashers and skeptics, I tell you Cobol is not dead or arcane--it's best years are still to come. Jerome Garfunkel Senior Consultant Micro Focus Palo Alto, CA

COPYRIGHT 1991 Wiesner Publications, Inc.
COPYRIGHT 2004 Gale Group