Columns, by their very nature, aren't the news. They're not reportage, not necessarily educational. Good columns, I believe, should provoke, cajole and maybe even irritate.

And that was, to say the least, the result of my last outing into the wild world of Bluetooth, published almost two years ago. Sure, you remember it-Bluetooth is dead. Contrary to popular belief, I didn't then and don't today hate Bluetooth (indeed, how can one hate a radio or a protocol stack?). My argument was that Bluetooth was a closed system, and nature abhors closed systems. Why limit all of those dandy (no, I mean it, really) Bluetooth profiles and applications to a single, obsolete PHY? Time marches on, and progress isn't always uniform across all seven layers at once. That's the whole point of designing network architectures today. With .11n, for example, there will soon be nine PHYs defined in 802.11. Pick the one you like.

And that was the point of the infamous column. I took issue with the idea of a 3-Mbit/second next-gen Bluetooth in an era of 100-Mbit/s WLANs and much more than that in UWB and the upcoming 802.15.3c.

What we need is a comprehensive set of applications to do all those common tasks, like copying files between devices, synchronization, device virtualization, service discovery and so on. This is where Bluetooth really shines, but few customers ever got to experience it. Bluetooth, like IrDA before it, has been a successful failure-huge installed based, but few users. Few people have ever copied a file or synchronized their phone directory with Bluetooth. The wireless carriers haven't helped the situation much, often disabling key Bluetooth capabilities so as to force their customers to use the WAN when a PAN would do. The word "shame" comes to mind here.

The Bluetooth Special Interest Group has decoupled the PHY from the upper layers, and you'll soon see Bluetooth on UWB. Sure, there are still challenges to overcome, not the least of which is worldwide regulatory approval. But that will happen; the market potential for UWB is just too great for the politicians to ignore. And the news gets better: The SIG tells me (yes, we're speaking again) that they're thinking about ports to near-field communications, ZigBee and even Wi-Fi. But why stop there? Why not support all Internet Protocol-based networks, including wire? Bluetooth on everything? Why not?

And this really is the point. All components of a given solution's value chain, from the PHY on up, need to click. Being able to pick the best pieces at each layer is practically as cherished as baseball and apple pie. With the Bluetooth SIG now realizing and supporting this, I'm forecasting great success ahead for Bluetooth. With their new positioning, they really don't have any competition. Smooth sailing from now on? Well, no, not by a long shot-there is still a lot of marketing damage to be overcome. But the Bluetooth community will do just that, and their best days are ahead.

Craig Mathias (craig@farpointgroup.com), principal at Farpoint Group (Ashland, Mass.)