Specifically, there's a user perception. The meaning of custom has "switched," evolving to encompass so-called modified-standard and full-solution modular designs. Fast-advancing technology leads users to think of ac/dc and dc/dc bricks and an infinite variety of tweaked subsystems to meet the job at hand. Fact is, however, there's a lot of discrete designing going on in this $6 billion market in North America, but by most accounts less than a few hundred million dollars in modules.

So on one side, we have those viewing custom as mostly an advanced sort of modified-standard solution, often believing that properly defined system standards do away with the need for custom. Others see custom as a low-margin business.

On the other side, the prevailing thought is that as long as there are high-volume requirements, custom will not only survive but thrive. That implies advances independent of the individual market segment. Some also note the misperception cast by the press that military COTS systems indeed utilize strictly "commercial off-the-shelf" products. It's a point well taken.

Ultimately, custom will survive because it's fueled by the human element of control: Customers want what they want when they want it. And to a good degree, at their price. They will retain control as long as the sheer number of users exceeds what a single vendor can deliver. Custom has far-ranging implications for the user, from the practical (single ac/dc modules for surface- or through-hole mounting on a pc board) to the architectural (detailed, user-interactive Web programs for designing and ordering prototype systems in as little as a few days). With that capability, custom won't likely go the way of a "paperless office" or a "plastic card society" anytime soon.