Bob Frankston adds a critique of Bluetooth to the blogthread that began with my blogportage of a contretemps between the estimable Landry and the venerable Bricklin (or was it the venerable Landry and the estimable Bricklin?) and continued here (Bricklin) and here (Fleishman, Peterson, Marks).

I’ve written a lot about Bluetooth and don’t want to revisit it again but will risk some short and brief comments. Bluetooth reminds me of the story of the blind men and the elephant – each feels a part and presumes that it represents the elephant. People find some claim about Bluetooth they like and assume that it is the best way to get that particular aspect. But Bluetooth is like the days of dedicated word processors – they couldn’t compete with the ability to quickly evolve products on generic PCs.

Bluetooth would simply be a curiosity as the triumph of marketing over reality were it not for the damage of sucking the energy out of the wireless marketplace. Just a few quick bullets before I run down to the dinner waiting for me:

• Replacing a cable. This is a lie. If that were the goal, then I’d be able to get a generic cable replacement rather than a better way to get to my cell phones.
• Replacing a cable. It doesn’t just replace a cable, it requires I not use a cable since Bluetooth insists that the protocols work only over their transport and no other. I can’t even use a wire as way to go faster or use less power.
• Lower power. I don’t see any reason why the cost per bit for IP is any lower. The low power is an artifact of the higher level protocols which are designed to reduce traffic for trivial apps. If you use it like 802.11 you blow the design point.
• Low power? If it’s so important use IP over HomeRF.
• Low power? It is it so important then provide power management for 802.11
• Cheap. Cheap is 100% about volume. Color movies are now cheaper than black and white. I can make anything cheap by assuming a volume of a billion units per year. 802.11 is shipping and thus has the advantage.

• Bluetooth requires I use 802.11 Bluetooth’s design point is so limited it can’t be the only protocol so why bother? In order to do all the things a normal user wants, even if you have Bluetooth you will still need 802.11. [Later addition: Bluetooth’s profiles and applications and radios are all tied together and tethered by their accidental properties. Bluetooth’s competition is not 802.11, it is the concept of IP and the separation of transport from applications. If you like the Bluetooth profiles you can use them atop IP but, as you may notice, they don’t seem to have enough value for that to be considered worth doing.]

• It’s all about paying for connecting by making the cell phone the center of the universe. Even if you accept that I simply use a second cell phone in a PCMCIA slot in my laptop.
• Cell phones are bad designs anyway and date back to FM modulation tricks in the 1970’s and now are mainly ways  to pay for foolish 3G investments.
• Bluetooth is awfully slow.
• Bluetooth is IrDA with the same mistakes of being too clever and treating IP as a third class protocol over a simulated serial path

Enough – I could go much further in dissecting and trashing Bluetooth but this dead horse is already leather.

The real point, however, is that Bluetooth is a bad design that has the advantage of making expansive claims that have no basis in reality. Demos will work very well and that is a bad sign since it means that that which is not the demo is given short shrift.

Bluetooth – why should anyone care except insofar that it is a way to annoy us.

Mary Lu pulls together two threads by noting that Belkin has introduced a Bluetooth USB adapter.

Categories: Uncategorized dw