Bluetooth Faithful: I Want to Believe...
by David Sims
A discussion today on O'Reilly's editors list considered why Bluetooth has failed to take root year after year. I agreed with the camp that thought it was largely a failure of marketing (specifically, trying to market itself against 802.11b), and also that there has been a sort of chicken-and-egg problem, where Bluetooth in one gadget doesn't get you much; its power shows when it gets into all your gadgets. This is pretty different from 802.11b, where if you shell out for a wireless PC card, you're admitted to a glorious world of wireless networks in offices, conventions, airports and -- according to Cory Doctorow -- enough insecure and community-spirited wireless networks to give seamless coverage in a
cab riding through Manhattan.
But I'm an optimist, so I visited the Bluetooth Developers Conference at Moscone in San Francisco on Wednesday. Attendance was respectable; not as mobbed as last December's show in San Jose, but a whole lot more energy than the ghost town I walked into at the Annual Linux Showcase in Oakland last month. It was a lot like last year's show, right down to the demos of cool projects that one could be pretty sure aren't going to show up anytime soon: a Bluetooh-enabled pen that
is also a camera (!), thumb-sized Bluetooth radios that plug into the USB
port. And three times at three different booths, I saw the same demo: using a Bluetooth connection between a laptop and a mobile phone to use the phone as a modem that connects the laptop to the Internet. At 14 Kbps. Pretty unimpressive, and only slightly less so when you imagine it happening at GPRS speeds -- 40 Kbps realistically. Hardly a killer app to demo at shows.
The folks who seem furthest off base with this technology are those companies trying to sell it as a wireless LAN. Bluetooth developers have figured out that the original range of 10 meters is too small, so they're
calling that Class 2 Bluetooth now, and are developing a Class 1 that
supposedly will have a range of 100 meters. That would put it in (or
beyond) the range of 802.11 networks. But I still can't imagine any office
environment going for Bluetooth over 802.11, even if the price is about
the same, for at least two reasons: the throughput is a lot less (1 Mbps
against 802.11's theoretical 11 Mbps and actual 5.5 Mbps), and if you buy
802.11 cards for your laptops you get to use them in lots of places --
So why do I remain a Bluetooth optimist? I think once they get over the
idea that they are competing with 802.11 and begin to admit they are
competing with USB, things will move along more smoothly. IBM & Toshiba
are offering Bluetooth capability in laptops for about $100, and the top
three mobile phone handset manufacturers (Nokia, Motorola, & Ericsson) are
all promising to stick tens of millions of the chipsets into their phones
over the next few years. After two years of the world laughing at
Bluetooth's failure to arrive, they're beginning to learn to be
conservative in their estimates: one vendor told me they expect Bluetooth to be in 70% of the world's mobile phones by 2006. That's 2006. I expect we'll all be commuting on Segways by then, riding over soaring viaducts and
listening to MP3s streaming down from satellite radio...
Scaled back to these more modest expectations, the software applications
for Bluetooth come into view. Obviously there will be a need for good
synching software at all levels, from apps that attempt to provide lots of
"enterprise" information to workers, to programs that let your Palm
-- more optimism here, that there will still be a Palm platform in 2006 --
and your cell phone update each other. It ain't too sexy, but neither is
USB or ASCII text, and look how essential they've become. I met with one
vendor, Rococo Software out of Dublin, Ireland, that has an IDE for J2ME
applications and is porting it over to be a Bluetooth development
environment, presumably for the same types of applications.
I think putting Bluetooth into phones will make the biggest difference. We all know there are more folks carrying mobile phones than carrying laptops in the world. And here's a key difference: mobile phone consumers want everything to work automatically; they don't really need to know how or why, and most don't care about the technical details. It's a more passive (or at least casual) relationship than folks have with their laptops. I would guess that at least one third of laptop users could tell you what operating system they're running; I doubt that 1 percent of mobile phone users could tell you much about the software in those phones. That's how I think Bluetooth will work: quietly and automatically in the background, your phone and PDA updating each other's address books, your laptop talking with your printer -- about as sexy (and as necessary) as a USB port.
Am I kidding myself?
It will be for sure on every handheld by 2007.