New technologies appear in many ways, but there are two methods that are most common in the world of computers. Technologies are either developed by committees, which is to say in a very public process leading toward standardization and broad use, or they are developed in private and sprung on the world by folks hoping to make a lot of money for their proprietary work. Both routes can be equally effective and both have produced some noteworthy technologies. Linux is the quintessential open standard, developed by a group of folks working very much in public. Windows is the opposite — a proprietary standard. TCP/IP began as Bob Kahn and Vint Cerf noodling about the way internetworking ought to be, but then quickly turned into a committee exercise. In contrast, NetWare's IPX and SPX protocols have always been owned by Novell. And so it goes. The big mistake would be to say too loudly that one approach is better than another. It is better to say that each approach has its strengths.

One emerging technology that has gained a lot of press in recent months is Bluetooth, a method of networking devices at high speed over short distances with radio waves. The big idea behind Bluetooth is that one ought to be able to bring a new device into the room, and have it automatically become a part of the local network. It will identify all the other Bluetooth devices in the room and know which is the printer, which is the router and which is the color TV — all devices that promise to shortly have Bluetooth connectivity. Bluetooth is death to wires. Bluetooth is also very much a product of committees, with more than 100 companies aligned behind it and helping to shape the way it does and doesn't work.

What Bluetooth has right now is a lot of momentum. Bluetooth has the mo. And it would take something pretty extraordinary at this point to inhibit the success of Bluetooth. My friend Andy Viterbi, co-founder of Qualcomm and a god of wireless networking, thinks Bluetooth will be so successful that it will completely fill the 2.4-GHz frequency and the only hope for any competing solution is maybe to handle the overflow in some different frequency range.

Those are pretty high expectations, but Bluetooth does have the mo. It has been coming for several years and has almost no competition — until now.

This week at the Electronic Entertainment Expo (E3) video game trade show in Los Angeles, while Sony, Microsoft, and Nintendo duked it out in a competition I described in two earlier columns, a little company called Eleven Engineering from Edmonton, Alberta Canada introduced a worthy Bluetooth competitor. Called SPIKE, Eleven's product, while similar to Bluetooth in operation, is very different in its evolution. Bluetooth came from a hundred companies and SPIKE came from one. Bluetooth came with a broad mandate and high expectations. SPIKE comes from the world of video games and originally had very humble ambitions. Bluetooth has the mo, but SPIKE just might win the game.

What gives SPIKE a chance against Bluetooth is pure economics. Bluetooth chipsets right now cost around $100, while SPIKE costs $6.25. In a couple years, experts tell us Bluetooth will cost only $5.00, but by then SPIKE will be down to a buck. Whatever SPIKE costs, Bluetooth will always cost mo.

But wait, there's more. Bluetooth comes from the world of ultra-reliable data, which is to say that it is very careful, very precise, and somewhat slow because of required data retransmissions to make sure everything is just right. SPIKE comes from the world of high-speed gaming, where things keep moving and players never look back. As a result, SPIKE uses a true real-time protocol, where Bluetooth does not. SPIKE has to get it right the first time, so it is more resistant to interference than Bluetooth, using both frequency hopping and direct sequence spread spectrum techniques where Bluetooth only uses one. SPIKE uses less power than Bluetooth, has more range and, in the bargain, has more brains than Bluetooth will ever have. That's because in order to ensure real-time operation, SPIKE includes for your $6.25 a multithreaded (eight threads, total) RISC processor good for 50 million instructions per second. So while Bluetooth might be able to connect your PDA, SPIKE can connect your PDA AND PROVIDE ALL ITS PROCESSING at the same time, again for $6.25. If you just want the 50 MIPS RISC processor, that's called Hammerhead, and will set you back $2.00. That's processing power equivalent to an Intel i486DX2-66 for two bucks.

Now there is nothing wrong with Bluetooth. In fact, there is less wrong with Bluetooth than many in the press might claim. A few weeks ago there were press reports of the big Bluetooth interoperability demo failing at the CeBIT show in Germany. The truth is that the Bluetooth demo was running just fine an hour later, but by then, the reporter was gone. This kind of thing happens when coverage is so thin that one reporter's experience can shape a thousand Web accounts.

SPIKE isn't going to put Bluetooth out of business by any means. If you are a device manufacturer who wants to function in a Bluetooth world, then you'll use Bluetooth in order to interoperate. But what if you have an application that doesn't require interoperation, an application for which the manufacturer controls both ends of the wireless connection? Then Bluetooth makes no sense at all and SPIKE does. You might as well save the money. For wireless telephones, headsets, and many other devices that come with both ends of the connection in the same box, SPIKE shines.

Thomson Consumer Electronics — owner of the RCA brand — will be using SPIKE for a whole range of game peripherals, primarily for Microsoft's Xbox. Imagine up to 16 players in the same room all using wireless controllers that include voice. The players, wearing wireless headsets that share the SPIKE bandwidth and processing power, will be able to talk to each other, whether in the room or across the net, and every player will hear sound effects that go with his or her experience in the game. And because SPIKE's latency is less than the interval at which Xbox polls its controller ports, the game console can never get ahead of the player.

If it sounds like I'm a fan of SPIKE, the truth is that I am more properly a fan of little companies that do good work. I am also a fan of happy technical surprises — surprises that don't often emerge from technical committees or industry consortia. Bill Gates says the way to make money is by setting de facto standards. Maybe SPIKE — a good little technology at a fair price — has a chance to do just that.

Oh, one more point on a different topic. This column is supposed to appear on Thursdays and today, May 17, 2001, happens to be the tenth anniversary of the World Wide Web, when the first server running http and HTML was cranked up by Tim Berners-Lee at the European particle physics lab in Geneva. Credit Gene Watkins for pointing this out to me. One thing that is amazing to me about this story is that the original purpose of the World Wide Web was simply to allow Tim to get back to his desktop system from any of the more than 200 different types of computers running at CERN. As a systems guy, this helped him do his job.

That it changed the lives of the rest of us is just a delightful bonus.