Jeff Lynch [MVP]

Hmmm... dunno if *anything* that generates code is RAD.

I would define RAD partly as anything that generates code that tightly couples itself to something else external to its own concerns has a good chance of being RAD. For example: using ASP .NET classes that make calls directly to a database by using command text persisted in the templates of UI controls. This kind of thing cripples testability which shortens the expected lifespan of an application or unnessessarily deflates the reasonable expectation for sustained developer productivity over the lifespan of the app.

For me, RAD is something that trades an artificially inflated perception of productivity for the expectation for sustainability inherent in contemporary test-centric development practices.

RAD is the soul of an application sold for a cheap, fleeting thrill.

RAD is sleeping with a really hot chick with the extremely poor personality, getting her pregnant, and having to spend the rest of your life listening to her.

Scott,

You kill me, man! Your last definition of RAD is by far, the best I've ever seen. Have you been talking with my wife again?

Seriously, in the old, old days we coded one line at a time on punch cards. When terminals were first introduced into the workplace, many "old school" programmers said "no way, not me. If you don't load it into a card-reader, it ain't real code". To them a terminal was RAD. As a degreed Mechanical Engineer (with six US patents by the way) I remember moving from pencil and paper drafting to CAD CAM and all the fuss over "real engineers design on paper". To them CAD CAM was RAD.

I believe the definition of RAD depends upon your perspective and your experience. I always write SQL stored procedures and call them from a business object class (BLL) but that's my "personal preference" at least for now. Until recently (ASP.NET 2.0) I'd never use drag & drop for web forms development and always wrote my HTML, JS and CSS by hand. After looking at the code generated by the new ASP.NET 2.0 server controls (including the new Atlas stuff) I'm not sure I could do better by hand.

Another great example is BizTalk Server (and now WF). All it really does is generate C#, xsd and xslt behind the scenes and host it in a runtime environment. A really good developer could write this code "by hand" (obviously someone did) but the question is "WHY BOTHER?".

I guess my point is that RAD (like beauty) is in the "eye of the beholder". Today's RAD may be tomorrow's mainstream technique. My (somewhat limited) perspective as an "end-user" developer is to use whatever tools best fit the job.

Jeff,

as I started with punch-cards as well (missed the tape though) : amen to that !

Well said, Jeff!

Embrace RAD in all its form when it makes perfect sense.

I *LOVE* this blogpost.

I need to make a small correction to my post. An old friend of mine (and hardcore mainframe programmer) reminded me that assembly language is a RAD tool for machine code. I guess I've been "RAD" for a long time!

--
Jeff

In a pleasantly provoking post Jeffrey warns us against the dangers of RAD. What RAD stands for in his...

In a pleasantly provoking post Jeffrey warns us against the dangers of RAD. What RAD stands for in his...

Some of my CodeBetter neighbors have been posting and arguing about the relative merits of Rapid Application...

[EDIT] Before I start, let me make this very clear.  RAD is not evil, it doesn't kill, and for a...

Author: <a href="blogs/Jeremy.miller">Jeremy D. Miller</a><br />Some of my CodeBetter neighbors have been posting and arguing about the relative merits of Rapid Application Development. There was also quite a bit of internal discussion, part of which was to stop being so opinionated in the blog posting and get back to technical posts. Let me get this one out of the way, then I'm back to coding.

If you've read many of the codebetter.com blogs over the past few weeks, you can tell that as a developer...