As a Java evangelist, I field many questions about the so-called new paradigm of developing and deploying Java/Internet-based business applications. Will we have problems in our development phase? Will the application perform? Will the Microsoft lawsuits ruin the platform? Will Bill Gates buy the Brooklyn Bridge?

New paradigms, old paradigms, paradigms, schmaradigms. Paradigm is consultant talk. Remember all the talk about GUI-based client/server back when it was touted as the latest and greatest paradigm? People asked the same paradigm questions, "How do we partition and deploy it? How do we get performance? And where are the standards?"

People never asked the most important question back then, which is "What are the gains from building and deploying a GUI client/server business application?" If somebody ever did ask such return-on-investment (ROI) questions and actually got an honest answer, nobody would have bothered with GUI client/server computing. The applications cost a fortune to build and another fortune to maintain, and what was the payback? A more pleasing user interface.

Spurred by the adoption of the Windows platforms, the GUI gave users a nicer look and feel. This is what organizations invested so many millions of dollars for? The bang from the investment in GUI client/server was questionable at best. Maybe the organization got some slight reduction in the user learning curve, but I doubt it. An entire, highly profitable industry emerged around GUI-Windows user training.

In the early years of this decade countless resources of time and labor, not to mention hard cash, were spent converting business applications to GUI client/server. Countless tool vendors jumped on the GUI client/server bandwagon. Many product generations and thousands of acronyms later, here we are and what do we have to show for it? At best, we have a generation of users who know how to use a mouse and click on an icon, and a new generation of developers who are finally beginning to understand event-driven programming. So much for the ROI of the GUI client/server development paradigm.

Now the same people are asking the same questions about the Java-Internet/Web development paradigm. But again, they are not asking the only question that really matters: the ROI question. As far as I'm concerned, it is the only question to ask.

When it comes to the Java-Internet/Web paradigm, ROI is the only raison d'tre for Java business applications. And its ROI can be spelled out in one word: accessibility. The Java-Internet/Web paradigm allows you to reach new users (new customers, suppliers, partners, anybody you want) as well as old users easily.

Accessibility translates into more business for you. Here is Java George's formula for Java-Internet/Web ROI: Customers + Accessibility + Product = Sale = Service = More/Better Business. You can plug in your own actuals. If they don't add up to more and better business for you, then skip Java and stick with what you are doing. (But check your math first because you don't want to make an error.)

The Internet is the primary catalyst behind the new paradigm of Java applications, just as Windows platform was the primary catalyst behind GUI client/server. With the Internet population running into the hundreds of millions worldwide and heading to the billions faster than the speed of Moore's Law, Java-Internet/Web is the Titanic of paradigms.

This means that a business providing products and services to people and organizations via Java-Internet/Web applications receives a material increase in prospective customers, and a potentially exponential increase in customers over time. Also, the organization that can interface with new and existing partners through business-to-business Java applications will reduce operating costs, shorten business cycles and increase its ability to respond to market conditions fast, further adding to the ROI.

Occasionally, people blame Java George for being overly optimistic towards the Java platform. But as far as the long term ROI for this new paradigm, I consider myself a conservative. Compared to the GUI client/server paradigm, Java-Internet/Web is a no-brainer.