System i developers are a relatively tight-knit group. Over the years, I've gotten to know quite a few RPG programmers who work in and around New York City. When I bump into one of them, the conversation invariably begins like this:

Me: "Hey, how's everything? Are you still working at XYZ Corporation?"

Acquaintance: "Yes. I'm coming up on my 5th/10th/15th anniversary!"

Me: "Congrats! Are you doing any web development on the System i yet?"

Acquaintance: "A little bit, but I'm mostly coding in RPG. We're still trying to get our arms around Java/WebSphere/PHP/RoR."

Me: "Wasn't your shop doing that when I saw you 1/3/5 years ago?"

Acquaintance: "Yeah, but these things take time. Our CIO/VP/IT Director is preparing us for the next decade. You wouldn't believe how much time we've spent evaluating options!"

Me: "Have you ever considered an application generator?"

Acquaintance: "Not really. My boss insists on a mainstream development environment."

Me: "How does anybody know what will be mainstream in a few years?"

Acquaintance: "Good question. He/she is also afraid of buying something that may limit us."

Me: "But you work at a bank/distributor/manufacturer. What's truly unique about your company?"

Acquaintance: "I can't think of anything offhand. But what happens if the software vendor goes out of business?"

Me: "The established ones have been around for more than 20 years. Doesn't that tell you something?"

Acquaintance: "Okay, okay, enough shop talk. Are you still dating that Brazilian supermodel?"

Sound familiar? In my opinion, more than a few System i shops have become hamstrung by overarching web development strategies. Instead of rolling out new applications, they're running in place on a treadmill of indecision and analysis. There's nothing wrong with carefully vetting options, mind you, but when months turn into years, it's only a matter of time before executives consider the System i on life support. And who can blame them? After viewing hundreds of Microsoft ads, CEOs and CFOs are bound to wonder why their companies don't just run everything on Windows. Mr. Ballmer knows how to market his wares. If we want the System i to remain relevant, we have to develop more web applications for it!

Application generators can help. They cost less than an enterprise-wide commitment to WebSphere and are easier to master than many languages, including Java. They also leverage the System i's core strengths and integrate seamlessly with RPG, Cobol, CL, and other programs. What's more, they shield programmers from the complexity of many lower level development tasks. Above all, application generators let programmers build sophisticated, eye-catching web solutions that run natively on the System i.

Don't get me wrong. I fully appreciate the value proposition of Java, WebSphere, et al. When I co-founded TimeTrade Systems with four friends in 2000, we set out to develop a web-based solution that would meet the scheduling needs of virtually any organization. Java was our primary development language from the start, and it was the right choice. Today, TimeTrade continues to grow (and recently raised $5 million in a Series C investment round). Many System i shops, however, do not have such demanding requirements. Take a moment to think about your organization's Internet projects. More likely than not, most require simply web-enabling legacy systems.

I've used an application generator to build web solutions on the System i for years. In my view, it's all about productivity. I don't want to spend hours writing relatively low-level code if a tool can do it for me. Today's generators feature source editors, compilers, debuggers, data dictionaries, page designers, and version control facilities. They've matured into full-blown integrated development environments. And because they support frameworks, they provide programmers with a structured, rapid-development environment as well.

Application generators aren't limited to churning out web pages, either. They generate programs to publish and consume web services, exchange XML documents, support PDAs, access data on other platforms, create Excel and PDF documents, and perform many other tasks. I have never felt handcuffed by my application generator.

Like many RPG programmers, you may not have the time to learn Java or WebSphere. Yet, the need to develop web applications has never been more acute. With the right application generator, you could deploy usable webfacing programs in a few weeks. To get your feet wet, I recommend building browser-based interfaces that interact with legacy programs. It's a baby step, but it lets you leverage your current skills and connect existing applications to the Internet. As you become more conversant with your tool of choice, you can use it to generate brand-new applications for the System i and for other platforms as well.

Of course, application generators are not a panacea. Developing software is an inherently complex process, and every tool has a learning curve. To be used effectively, they require a real commitment. Given the other options, though, why wouldn't you consider one? From what I've seen, fear and doubt are keeping these products off too many shortlists, and that is a shame.

Hopefully, I've piqued your interest in application generators. If you'd like to learn more, there's no better place to start than "The Twists, Turns, and Benefits of Application Generators" (October 2007, article ID 21070 at SystemiNetwork.com). This article covers the pros and cons of app generators and provides a comprehensive framework for evaluating products.

Scott Steinacher specializes in LANSA and RPG development, electronic document management, and business intelligence applications for the System i. He is a cofounder of TimeTrade Systems, Inc. (timetrade.com), a leading provider of web-based scheduling solutions. In addition to numerous magazine articles, Scott is author of "Data Warehousing and the AS/400" (29th Street Press, 1998). Scott holds an MBA in corporate finance from New York University’s Stern School of Business, and a B.S. in Information Systems from the State University of New York at Albany. He can be reached at scott.steinacher@timetrade.com.