Monthly Meetings

August 18 2004:
Applying MDA in the Real World, Fact or Fiction?

You have probably heard or read of raging debates over Model Driven Architecture (MDA), UML's immaturity, code generation controversy, and modeling is for wimps.

We found Jon Kern (brief bio below), who takes a somewhat contrarian view. At the core, Jon is a fanatic about ensuring software development teams succeed by delivering business value. Jon's experiences stem from 15+ years of working in the field, leading his own teams to develop, numerous systems, from real-time flight simulation systems, to foreign exchange trading systems, to shrink-wrapped commercial UML modeling tools.

Jon will describe a pragmatic "distilled" usage of MDA concepts that has high project value when combined with visual tool support. He will discuss where UML and MDA overlap, and will dismiss the "Silver Bullet" theory that often accompanies design and development tools. Jon doesn't blow smoke. he is a pragmatic engineer at heart.

Along the way, Jon will discuss three simple keys to software development success that you should be practicing in every engagement:

• Separation of Concerns (It's the Business!)
• Duality of Architectures (Consistent Construction)
• Agile development

Understanding where these keys can be used in your development organizationBring along any favorite war story about a failed or stressed project and Jon will point out where the project failed to follow one of his three keys to success.

How to move your game to the next level when building non-trivial systems:

• If your software is not a strategic asset, could it be?
• Does technology rule your decisions, or does business?
• If you think there is a career to be had in mapping columns, in doing deployment descriptors, and banging out repetitive code think again.
• If your development does not include lots of repetitive, boring code creation you are doing something wrong.
• If you think learning the ins-and-outs of the latest and greatest technology is the only key to survival you're wrong.
• If you think clients are asking for cool infrastructure, stateless session beans, or even relational databases, think again. They want business value not technology!
• If you think that your development teams are providing optimal value, then you are among the rarified few. Could it be better?
• If you think that management is even allowing you to provide optimal value, then look around to see if you are given adequate management support and slack time to do the right thing for long-term business gain for the enterprise. Bring your manager, and your manager's manager!
• If you think agile development can only be achieved by starting at the code level, go ahead, keep typing. I'll race you!
• If you think using tools that reduce billable hours is a bad thing, you're kidding, right?
• If you think that modeling has nothing to do with agile development, step aside.

Model-Driven Development can bring incredible tactical and strategic value to your development organization if you understand the power. At the end of his presentation, Jon will provide a glimpse into how one tool can help accelerate development efforts and provide ways to do more with less.

Knowing Jon Kern, this is sure to be a bit raucous, stimulating, educational, controversial, and - hopefully lots of fun! Jon encourages lots of audience participation.

Jon is an author of The Agile Manifesto, Java Design (with Peter Coad), and others. After 15 years of DoD consulting, following running his own successful business, he joined Peter Coad as one of the original execs who started TogetherSoft. While there, he built the mentoring team (services), and he was a major contributor to ControlCenter's design and development success prior to its sale to Borland in 2003. Jon has 15+ years of experience from jet engine R&D, to real-time flight simulators, to developing IBM's manufacturing system (C++). He has led numerous teams, consulted with hundreds others, and led TogetherSoft's development teams (commuting to Russia).

Jon was recruited by Compuware's OptimalJ product team in May 2003. Once he saw the vision and power behind the product, he was hooked.