Models. Models. Models. So What?

In 1985, in an interview for some then-popular magazine, I was asked when models and model-driven development would become commonplace. "In three years time", I replied confidently. In 1987, I was asked the same question, and my answer remained the same. And 1989. And '91. Were you to ask me the same question today, I would answer it in the same way. Perhaps I should have gone surfing instead. While my answer has the virtue of consistency, how could I---how could we?---have been so wrong? Of course, we didn't have the technology back then. And we didn't have the computer power that could allow us to ignore certain inefficiencies introduced by abstraction. But have things really changed that much in nearly a quarter of a century? Our tools and computers are certainly better, but it is clear that we have failed to convert the great unwashed to the benefits and wonders of models and model-driven engineering. This keynote will take a personal view of why we have (let's be positive, shall we?) yet to succeed. It will explore several technical, business and marketing issues that have impeded our progress. And because a keynote is intended to be positive, leaving delegates upbeat and energized, we shall also examine some encouraging indicators that could lead to model-driven engineering soon becoming commonplace. In, oh, let's say, three years?