Editorial

Digg This!

Page 2 of 2   « previous page

However, in the 15+ years I've been using the product, I can't think of a project I've worked on that actually followed the RAD methodology. Instead, we generally used (and still use) a Waterfall model. In some cases it's augmented with feedback loops that result from our ability to produce early versions of the eventual product, so it's more of an incremental development approach. But no actual RAD methodology. That's probably (IMHO) something that was more widely accepted in the Visual Basic camp.

To the degree that we discuss the RAD development methodology, I think Michael Swindell (vice president of products for CodeGear - makers of Delphi) is correct when he indicates that it evolved into Agile Development. But where does that leave 4GL products like PowerBuilder that were originally developed in order to support RAD? Frankly, I don't think it affects them at all. The primary advantage of such 4GL tools - abstracting development away from low-level implementation details - is independent of the development methodology used, be it Waterfall, RAD, Agile, or something else.

While RAD may be dead - or at least living on in Agile - 4GL tools still make sense to a lot of developers. If that wasn't the case, why would folks like IBM's Bob Zurek (http://jdj.sys-con.com/read/281638.htm) indicate that what AJAX is missing is a DataWindow? Why would companies like WaveMaker bill their products as "PowerBuilder for the Web" (http://campustechnology.com/articles/57089/)? Or why would people be asking: "Who Will Become the PowerBuilder of the Web 2.0 Set?" Who compares themselves to "dead" technologies in order to indicate they can be or are a market leader?

If that is the case, why isn't there more demand for PowerBuilder? I think it boils down to a couple of reasons:

• While PowerBuilder is a great 4GL tool for client/server development, it (until recently) has remained a 3GL tool as far as Web development is concerned. The new WebForm capability introduced in PowerBuilder 11 is the first real 4GL offering in PowerBuilder for Web development. It still has a ways to go. Some AJAX-like methodology is needed to reduce postbacks for screen updates. While it facilitates generating Web applications that look like client/server applications, it also needs to have the ability to generate Web sites that don't look like client/server applications at all.

• One of the main critiques of 4GL tools is vendor lock-in. Many development shops are not as comfortable using proprietary languages as they once were. Sybase might make significant modifications to PowerScript in order to give it features similar to some of the market-leading languages (e.g., Java, C#), but they don't have the mindshare or marketing clout to declare it a non-proprietary language. In particular, there has to be other vendors willing to support the language. Instead, Sybase needs to open up the PowerBuilder IDE so that developers have a choice with regard to the scripting language used for development. And the other languages supported need to be well-known non-proprietary ones (e.g., C#, JavaScript) and not a variant of some lesser-known one (e.g., DynaScript, the Sybase embellished version of ECMAScript that PowerDynamo was based on).

Is RAD dead? Actually, it doesn't matter. What does matters is that 4GL is alive and well and being sought for the Web development arena. With the right kind of improvements, the answer to the question: "Who Will Become the PowerBuilder of the Web 2.0 Set?" might very well be "PowerBuilder."

Page 2 of 2   « previous page