Staking a claim in the next gold rush

The percentage of people who actually struck gold and got rich during the California gold rush of 1849 was relatively small compared to the number of people that actually set off to strike it rich. The same will be true for the rush to market around web services and XML.

The percentage of people who actually struck gold and got rich during the California gold rush of 1849 was relatively small compared to the number of people that actually set off to strike it rich.

In fact, the people who made any real money during that period tended to sell mundane things such as shovels and blue jeans.

The same will be true for the rush to market around web services and XML. These technologies are crucial to the next wave of enterprise computing, but the number of startup and established vendors looking to strike the mother lode in this space far exceeds the number of companies that will actually succeed.

To understand who will benefit most from web services and especially XML, you need to look deeply into the arguments being made by the core proponents in the contest for the future of enterprise computing.

At the core of Sun's Java argument is the notion that networks are too fragile and the best way to advance computing is to distribute objects everywhere and execute them locally. The problem with this approach is that there is no framework in existence that can support a massive number of distributed objects that ultimately have to be tracked across everything and anything that has an IP address.

In contrast, the folks at Microsoft argue that the correct approach is to leverage Moore's Law and ship XML data all over the network. The core argument is that while XML requires additional network and server capacity to support, the cost/performance ratio of hardware infrastructure will easily keep pace thanks to Moore's Law.

Right now, Microsoft and the people backing XML-driven web services are winning the argument and the companies that stand to benefit most are Intel and Cisco, which are delivering the infrastructure needed to support web services. Intel and Cisco are selling the modern equivalent of shovels and blue jeans in the age of web services.

For Sun to win, it needs to create a Java framework capable of supporting a global network of distributed objects. Sun needs a miracle to ensure Java remains the dominant paradigm and that Sun itself does not implode. This "Death Star" scenario could very easily play out even if Sun is able to create a robust global framework for Java because even if the company can build it, there's no guarantee Sun can get it to market.

Former stalwart Sun allies such as IBM and BEA Systems are embracing XML and while they may not specifically embrace .Net, they understand the power of the paradigm. And then there are Sun customers such as Orbit that are abandoning Solaris in favour of Linux boxes interconnected by Jini and linked to the outside world via web services. If that architecture becomes the dominant model within Sun's customer base, the company simply won't have the financial capital necessary to fight a protracted software war.

It's too early to say who will win the war, but one thing is certain. No company will be in the same relative position it is today once the drama completely plays out. Intel's Andy Grove is fond of saying that only the paranoid survive in this business.

With the advent of web services, every vendor in the industry now has good reason to be paranoid.

Michael Vizard is editor in chief of US IDG publication InfoWorld. Send letters for publication in Computerworld New Zealand to Computerworld Letters.

Join the newsletter!

Error: Please check your email address.

Tags XML

More about BEABEA SystemsCiscoIBM AustraliaIDGIntelLinuxMichael VizardMicrosoft

Show Comments

Market Place

[]