Logo IMG
HOME > PAST ISSUE > Article Detail



Brian Hayes

An unfulfilled dream of computer science is the one true programming language, equally suited to all programs, all programmers and all computers. In years past committees and coalitions have tried to create or legislate such a language. Algol-60 was one of the first attempts; its greatest success, ironically, was in spawning dozens of new languages. A few years later PL/1 had the powerful backing of IBM. And the language called Ada had an even more imposing sponsor—the U.S. Department of Defense. And yet these languages too failed to conquer all.

The latest candidate for a computer Ursprache is a language called Java. This one is not the creation of a committee or an international standards board. Its origins are wonderfully humble. Java began as a language for programming "set-top boxes," the gadgets that are supposed to make TV interactive. So far, set-top boxes haven’t made much of a splash, but Java has become an extraordinary marketing phenomenon, with the kind of promotion and product tie-ins you might expect of a newly released Star Wars movie.

Do I exaggerate? Well, maybe McDonald’s will never give away Java trinkets, but Starbucks might. Barely two years after the language was introduced, there are Java magazines, Java conferences, Java videos. Usenet has a dozen Java newsgroups, with hundreds of messages every day. Java books have become an industry in their own right. (I review a few of them in this issue of American Scientist). Java businesses are springing up everywhere—startups and spinoffs and new divisions of established companies. And as far as I know, Java is the only programming language ever to have its own venture-capital fund (initial capitalization: $100 million).

Java's ambitions extend beyond becoming the one and only programming language. Java is being proposed as a new "computing platform," which could supplant the various alliances of hardware and operating-system software that now dominate the world of desktop computing. In this vision Java would elbow aside not only C, Pascal, Lisp and other programming languages but would also replace Windows and Unix and the Macintosh operating system. It would even be built into the "embedded" computers in cellular telephones and home thermostats (not to mention set-top boxes).

Before presenting my thoughts on the Java phenomenon, I have a couple of disclaimers to put on the record. In the first place, I have a favorite programming language of my own, and it is not Java. I prefer to code in a dialect of Lisp called Scheme. Anyone who has followed the computer-language wars of recent decades will immediately know that I come from an enemy camp. To make matters worse, I don't drink coffee! I am therefore unmoved by all the subtle appeals to caffeine craving that turn up in the names of Java products: Roaster, JavaBeans, C@fé, Mocha, etc. I will try to keep my prejudices in check, but the reader should bear in mind that the following comments come from someone who doesn’t know his latte from his cappuccino, and who sometimes dreads the prospect of a world overrun by undrinkable beverages and unthinkable languages.

comments powered by Disqus


Subscribe to American Scientist