Quite simply, IPv6 is a fix for a problem nobody has yet. Stopgap solutions such as NAT, while infuriating to network engineers, have proven effective. And IPv6 offers no compelling features to offset the headache of implementing it. In other words, until someone offers the equivalent of carbon credits for networking, IPv6 is one truth that's just too inconvenient.

11. Microsoft Passport. We all have too many online accounts and too many passwords to go with them. If Microsoft wants to save us some hassle by remembering them all for us, why not, right?

But Web users didn't flock to Microsoft's single sign-on service when it went live, and neither did partners. The idea was good, but the implementation was lousy. First came public debate over privacy concerns. The discovery of security holes in the Passport software was just the final straw.

To be fair, Microsoft's competitors had no luck with the concept, either. As it turns out, people think handing over their personal security information to a third party as a way to save effort is a bad idea. Who knew?

10. Itanium. For almost as long as its chips have powered PCs, Intel has had to compete with clones. So when Intel offered a high-powered 64-bit chip for server computing, it almost made sense that it should be a completely different design than the classic x86. Almost, that is, except for one thing: The reason Intel's rivals released clones was because people wanted x86s.

Correction: They wanted affordable x86, and Itanium was neither. But although it was a commercial flop, the Good Ship Itanic sails on, though to where we cannot say. It takes a brave captain, indeed, to follow into those waters.

9. Mac clones. Nobody dangled a carrot in front of its customers like Apple did in the late 1990s. For years, the Mac faithful had clamored for alternatives to Apple's expensive, proprietary hardware. Apple gave its blessing in 1995, and the market flooded with discount Mac clones from the likes of Power Computing, Motorola, and Umax. Then, just two years later, they were gone.

Fans of the clones fumed, but for Apple the only thing more damaging than this bait-and-switch was greenlighting the clones to begin with. Clone makers used generic PC components, an affront to the Mac's carefully cultivated image, but their popularity underscored how mediocre Apple's own offerings had become. As it turned out, what Apple really needed wasn't alternative manufacturers, but alternative management.

8. Electronic currency. If you're looking for the perfect complement to e-commerce, what else but e-money? If storefronts can go virtual, then why shouldn't customers' wallets?

The dot-com era spawned a cottage industry of alternative currencies and "loyalty programs." Many positioned themselves as solutions for micropayments -- the idea that online content providers could actually collect on the infinitesimally small fees that the market would bear. As it turned out, however, while employees of Internet startups might have been content to be paid in paper stock options, consumers preferred cold, hard cash -- proving once and for all that the only thing worse than a wooden nickel is a digital one.

7. The 64-bit desktop PC. We're all used to the major chipmakers wowing us with numbers, but as the gigahertz race slowed, life began imitating " Spinal Tap." How many bits does your processor support? 16? 32? Mine goes all the way up to 64!

Apple and IBM may have started the 64-bit craze in 2002, but AMD and Intel quickly followed suit. The elephant in the room, however, was that nobody could think of a serious reason why the average PC user might need to be able to access 16 exabytes of RAM.

64-bit OSes still lag behind the 32-bit versions and the applications aren't there, but who cares? After all, it's not how many bits you've got -- it's the cores that matter. Or something.