Why has 802.11 flourished and Bluetooth failed?

Why has 802.11 flourished while Bluetooth has essentially failed? Should we even care about Bluetooth?

The crux of the problem is that 802.11 represents the Internet and Bluetooth represents the faux Internet, and getting to know the difference gives us an understanding of what the Internet really is. On the surface, both promise connectivity without wires. That's really all 802.11 promises.

The Bluetooth story sounds much better. It promises to get rid of that tangle of wires on our desk and it will allow us to have a wireless headset that works with our cellular phone. We can connect our PDAs to our cellular phone and connect to the Internet and we can synchronize our cellular phone with our desktop and on and on. Unfortunately, Bluetooth synchronizes to the desktop only if you are at the desk--and that's the crux of the problem.

The Internet gives us opportunity to do all that and a lot more, but there isn't a promise that anything will work. Not all applications work well at first. The dial-up phone line, for example, is too slow for some applications-- we have to wait for the supply/demand cycle to give us more network capacity. In addition, we don't really know what applications will be most useful until after they've shipped.

Bluetooth harks back to an earlier time when the products were built for a single purpose. Even worse, since Bluetooth is baked into millions of chips before we get any experience, all of the initial design mistakes are locked in and we can't fix them. It's only now that people are discovering that Bluetooth's focus on eliminating wires means still having the limitations of wires in that you can only connect between nearby devices. 802.11 on the other hand takes advantage of the Internet and allows you to connect to any device, anywhere.

Since Bluetooth is emulating a wire, it has to have a protocol for pairing devices and, given the limited interfaces on the devices, this is very difficult. Bluetooth is very much associated with cellular phones because those companies have included Bluetooth capabilities in their chipsets, whereas other industries have the freedom to go with 802.11.

The protocols for 802.11 have had decades to prove themselves in the wired world. Bluetooth requires not having a wire, so we must have a radio to get the features.

We've ignored Bluetooth's predecessors, like the red IR (infrared) port on just about every laptop, so that we can care about Bluetooth. We shouldn't care too much. Unlike PC applications, we don't have many alternatives for the one Bluetooth application-- wireless headsets.-- that people care about.

If we decompose Bluetooth, it is just a radio and a set of applications called profiles. The Bluetooth radio is very much like the 802.11 radio, though with different tradeoffs. If we had only the radio, we could treat Bluetooth like 802.11 and decide which radio to use where, and we could bridge between the two worlds. One would probably win. But since Bluetooth has been deployed, we can only use it as long as we limit it to those badly designed profiles that interface the various devices. The question is whether the software in the cellphone can bypass the profiles and work more like an 802.11 radio. If so, then we have some salvage value.

Otherwise, what we have is a cellular industry that could change - if we have a few years and billions of dollars available. We don't.

So perhaps Bluetooth will survive as just a way to connect headsets to cell phones. I do expect some innovation and extension but, in the end, the value of creating general applications built on 802.11 gives it a major advantage. The Bluetooth SIG has to decide what its real purpose is and come to terms with existing in a world defined by 802.11.

Bob Frankston's background in computing and networking stretches across four decades during which he worked on the Multics projects; was co-founder (along with Dan Bricklin) of Software Arts,the company that shipped VisiCalc--the first spreadsheet; served time at Lotus Development where he created Lotus Express; and moved on to Microsoft where he initiated and championed home networking ("no new wires"). Frankston's blogs about everything from social trends to technology can be found at http://www.satn.org.