Logged In

Jeremy Allaire: Happy 10th Birthday, ColdFusion!

In the summer of 1994, my brother JJ and I, along with a friend and a colleague or two, were bent on making a run at the Internet as a business. We had spent the past several years close to the Internet, understanding it, feeling its pulse and power. When the web emerged through Mosaic in 1993, we knew it was a tipping point—that a new platform had emerged for media, communications, publishing, and commerce. It would be radically open, global, and accessible to anyone with a passion, an idea, and some creativity. Industries would be remade. We wanted to be part of it. The "it" we were not quite sure of yet.

What we knew for sure, though, was that the barely functional technology called HTML formed the cornerstone. When fused with interactivity and data, HTML held the promise of a new model of computing—one that blended documents and applications, that were dynamic and always changing, that could be easily and globally distributed to any computing device in the world, that would be easier to use than any past application or computing paradigm.

As children of the Internet, we also understood that the "killer app" of the web was "communications," that connecting people in rich ways, across geography, was deeply powerful, and we wanted to demonstrate this new computing model through this lens.

So in 1994 we sought out to build one of the very first web conferencing or forums systems for a Minneapolis weekly newspaper. What we learned may have changed web development forever.

While many smart people understood the power and potential of the web in late 1994, very few understood how to unlock its power in a simple way. What we learned in our first experiments was that everything we knew about programming as it applied to the web was wrong, that writing a lot of C code, or Perl code, or whatever, to create web applications was unnatural. We also learned that there was a massive wave of new people that would be needed to create the web, people who had never been trained in computer programming languages. We saw that the "role of software in society" as I used to say, would be dramatically broader than anyone could ever predict, and that what was needed was a new approach to programming that was simpler, and of course, molded into the mechanisms of the web, which would be the catalyst for this global revolution.

Thus was born ColdFusion, the first truly tag-based, declarative web programming language. If HTML documents could be crafted by hand with tags, their behavior changing through declarative instructions, why too couldn't software programs, programs that were wedded to these documents? If a new breed of developer was to be born—the web developer—one whom would be crafting and remaking these document applications, why not create a language suited for this work?

ColdFusion was the first, and never looked back. We got the “tag religion,” and continued to innovate by abstracting away common problems in web development into simple, declarative instructions that could be fused with pages---dynamically sending and receiving e-mail, uploading and managing files, distributing data across servers, indexing and searching documents and databases—everything could be expressed this way.

We were the first web development platform to add “tag extensibility” back in 1996, which allowed developers to write new tags using system programming languages, and then in 1997 using Java, COM and other systems. Most importantly, in 1998, we introduced the ability to write custom tags in CFML itself, ushering in an explosion in tag-based modules that abstracted nearly every common web development problem.

In 1998, we also introduced tag libraries that encapsulated common JavaScript client programming tasks, making DHTML controls available to the average web developer who did not have the appetite to master the browser DOM and JavaScript. It was also in 1998 that we introduced the Web Distributed Data Exchange, or WDDX, which when combined with custom tags and CFHTTP, enabled what we now know of as "web services." The ideas within that influenced Microsoft and others, who ultimately adopted their own flavor, known as SOAP. We went along with that, of course!

But a decade later, it is deeply gratifying to see that XML has become the de facto language of choice for expressing complex problems in simple, declarative languages. These tag-based languages now pervade the Java and .NET worlds, and not a vendor exists who is not working on some declarative interface to their complex system using an XML language.

Throughout the years there have always been deep skeptics—technologists who thumbed their noses at ColdFusion. ColdFusion is a toy language. ColdFusion can't scale. ColdFusion isn't open. ColdFusion is non-standard. The interesting thing is that along the way, at some point, all of these things were true—ColdFusion was a toy language, but one that worked for hundreds of thousands of people building important business applications. In its early years, ColdFusion couldn't scale without tremendous feats of human intervention. ColdFusion was a closed system that ran on limited and closed platforms. But today, ColdFusion is a rich multi-language environment that spans tags, scripting, and Java. It sits on top of J2EE and can scale as high as any application on the planet. And it's now built entirely on and with open standards, running on any operating system, with extensibility with any programming language, and with integration into nearly any type of system or device in the world.

At no time could anyone have predicted its success, and likewise no one can imagine what lies ahead.

Come celebrate ColdFusion’s 10th Birthday on July 13, 2005.

About the author