Glenn Fleishman, journalist to the stars and WiFi enthusiast, takes up the cudgels against the comments from Dan Bricklin I ran about the inadequacies of BlueTooth.

We’re in the early phases of Bluetooth which is now clearly a technology that will be here to stay. I used to be a Bluetooth denigrator, but having seen it survive some early nonsense (overhype by its backers, lack of hardware, lack of standards, a terrible 1.0 stack, etc.), it’s got a lot of advantages.

Advantages over 802.11:

1. No set up “cost”: ad hoc networking is and will continue to be a pain with Wi-Fi which is Ethernet sans fils (I mean without strings, but I might have said without parents). Bluetooth doesn’t have setup and breakdown costs, which reduces power use, network traffic, and user complexity.

2. Bluetooth is low power: it will work in devices in which battery life is at a premium.

3. Bluetooth chip costs are plummeting: yes, of course, it’s ridiculous that you pay $150 for a Bluetooth adapter, but, frankly, those are proofs of concept right now. In less than a year, you’ll buy a Bluetooth-equipped printer that may cost less than $25 more than a similar non-BT model.

4. Another point on ad hoc networking: Wi-Fi *must have* a base station to allow more than two devices to talk to each P2P. This is an inherent limitation of the protocol. If you want to have 3 devices exchange data, you need extra hardware, or one of the devices must turn itself into an infrastructure node which requires special software (available for Mac OS 9, for instance, but not Mac OS X yet or possibly ever).

Some of this will change as discovery technology improves. Apple’s great interest in offering discovery over several media (FireWire, wireless, wired, etc.) means that Bluetooth’s discovery protocol could be less important.

Wi-Fi can “discover” other devices, but it doesn’t know what to do then unless you join a network.

In a follow-up email, Glenn writes:

I thought Bluetooth was a lousy idea. However, they’re going to make the chips so small soon that they’ll be able to do some of the very cheap, very powerful concepts, like embed them in otherwise ridiculous devices (books, flexible drawings, e-paper), and make them super cheap and super low power.

Wi-Fi devices must conform to the certification process (otherwise, they’re just plain 802.11b), so they have to have minimum power and distance compliance.

Jonathan Peterson responds with an email supporting what seems to me to be Bricklin’s main problem with BlueTooth:

The real problem with bluetooth is that it isn’t an open transport standard that anyone can build on. You want to build an atomic clock wristwatch that acts as a timeserver to keep all your devices in sync (a cool idea if I say so myself)? Not only will your users have to dick around with it incessantly to convince everything to take time information from it, but the manufacturer will have to apply for certification to 7 Layers AG, you they won’t be able to sell the product until it has been approved.

How long would current WWW technologies have taken to mature if Tim Bremmers-Lee et al had to submit everything for ISO approval before moving forward with implementation?

Which among other thoughts I blogged of course:

Kevin Marks writes:

I know you have trouble with numbers, so here’s a handy guide to relative speeds in orders of magnitude

BlueTooth is 10 times faster than a 56k modem
USB is 10 times faster than Bluetooth
802.11b is as fast as USB
10baseT is as fast as USB
100baseT is 10 times faster than 10baseT
FireWire is 4 times faster than 100baseT and 40 times faster than USB
Gigabit Ethernet is 10 times faster than 100baseT

BlueTooth was designed to replace the slow things on USB (mice, keyboards, serial ports – the oens tht run at the slow speed).

It has no chance in hell of competing with 802.11b.

Categories: Uncategorized dw