I'm not the type to extract information from industry types through a series of in-depth, subtle and thought-provoking questions. I much prefer the brute-force approach. So I decided to stir the muck a bit by asking what type of technology (ATM or Ethernet) service providers should be using. Talking to a manufacturer of ATM switching equipment, I asked: "So if Ethernet is the technology of the future, why the heck are you still building this stuff?" Let's just say the look and subsequent answer I received were a bit animated. Ethernet is a simplistic technology geared for low-level applications, the vendor argued. But ATM! Well, there's an architecture that's proven to be rock-solid. The Ethernet bashing only got worse from there.
On the other side of the aisle, I talked to a manufacturer of Ethernet access equipment, a company that looks to compete in the last-mile transport world. This company's take was quite different, as you might guess. It painted ATM lovers as paranoid technology elitists who are looking only to protect their jobs. Ethernet solves all the problems ATM does without the complexity and cost--or so this company says.
In each case, I heard normally calm professionals spout off about how technology Y is the work of the devil. I couldn't help thinking how similar this conversation was to the many Ethernet-vs.-token ring and Microsoft-vs.-Sun/Linux/everyone/anyone discussions I've been part of over the years.
Why does it always need to be one or the other? Why can't we live with multiple viable technology solutions that a company can evaluate and choose based on its own requirements? Why not focus on making the technology stronger instead of wasting time undermining an alternate solution?
Sure, I'm an optimist. I think a technology's viability and functionality and customers' spending habits should dictate when it's time for that technology to go away. After all, many large carriers have a huge investment in ATM, so why would they toss away that money--money they don't have--just to be part of a new technology trend such as 10 Gigabit Ethernet? And carriers that have adopted native Ethernet for transport should focus on making sure that simplicity doesn't come at the expense of reliability.