In the client/server enterprise IT world, this plateau was hit within a year after the start of the Internet recession. Because of the predictable linear paths of Wintel and Linux architectures, and even the foreseeable topologies of grid computing and storage clusters, there's not a lot of surprises left in the IT community. Pick up a magazine on desktop and server computing these days, and you'll find that much of the content is thinly disguised analyses of the bureaucratic management of people. That's because most articles covering technology itself have to be prefaced with the unavoidable "Stop me if you've heard this one before."

If the recent NetWorld+Interop show is a guide, we're hitting that point in networks as well. The show was healthier than it's been in three years, but nothing disruptive was happening. Many would say that networking became predictable when the world coalesced on Ethernet frames at Layer 2 and IP at Layer 3, but plenty of interesting services and traffic discriminators were placed on top of IP between 2001 and 2003, even amid the recession.

But a second technology plateau has been hit this year. Metropolitan networks have been well-characterized in all seven layers; the edge demarcating public and private networks is described in granular fashion; and traffic on either side of the edge can be broken down, aggregated and analyzed to an extent never before seen. Even curveballs like mobile clients and wireless LANs fail to upset the managed-network applecart.

A skeptic might say that the modern communication network has been utterly dissected. If that were true, technology specialists could just pack it up and go home. But it's precisely when an environment becomes fully defined, predicted and controlled that unexpected revolutions take place from unknown vectors. The current malaise and ennui is merely a breather while we wait for the world to be turned upside-down anew.

Loring Wirbel is Communications editorial director for EE Times and its network publications.