CYBERPUNK: A Domain By Any Other Name

As parents of young children know, a lot of naive questions get asked because, to the asker, they seem like perfectly good ideas. Why can't we eat ice cream all the time? Why doesn't Santa Claus come every day? Why indeed? You don't have to be 4 years old to see the appeal behind such queries, even as our more rational selves recognize the answers.

This occurred to me the other day while I was watching TechTV. The cable channel was rerunning a news report about the new pack of Web-address domain suffixes that the Internet Corporation for Assigned Names and Numbers (ICAAN) recently sanctioned -- .biz, .aero, etc. As an addendum, the reporter mentioned the crusade of a California start-up company, RealNames (www.realnames.com), that replace URLs with "Internet keywords," much like AOL uses on its online service. Instead of typing "www.amazon.com" into your browser, you'd simply enter "Amazon." No more trying to remember if Dubya hides out at whitehouse.gov or whitehouse.com.

To this end, RealNames is building a central database of the words (called, in typical computerspeak silliness, "human-friendly identifiers") with which RealNames customers want their Web sites to be identified. For $100 a year, for instance, I can buy the words "Joab Jackson" from RealNames and when someone enters that into a browser, it would go to my Web page. The company is petitioning ICANN to make it an Internet standard of sorts and has partnered with Microsoft, which already has RealName capabilities embedded in its browsers.

The idea of using simple real-world words as Web addresses initially seems appealing, just like having ice cream for every meal -- until you start figuring out the consequences.

First of all, I'm not so sure we even need a simpler method of calling up Web pages. Human critters seem to do OK with the language of machines. Think about another machine language -- telephone numbers, for instance. There's nothing inherently "human-friendly" about strings of more or less random digits. In fact, using the telephone can be more difficult than using the Web -- I can guess Coca-Cola's Web address, but I couldn't guess its phone number. But with phones, we accept the complexity for the benefit of universal access -- using a small combination of numbers, I can reach the person next door or a one-room schoolhouse in the Australian Outback. That's the appeal of the current online standard, called the Domain Name System (DNS). It's global in scope, and it categorizes sites under different headings (.net, .org), much like area codes group phone numbers by region.

Truth is, people, as daft as they may be, are OK with a little bit of complexity, as long as it isn't one iota more complex than necessary. This is why I suspect RealNames is doomed -- its system will make Web navigation just a little bit more difficult than it needs to be.

Wait -- isn't RealNames out to simplify Web navigation? Well, only partially. From the start, the company decided only to register trademarks or brand names, not "generic" words, as the company calls them (generic word queries are instead directed to Microsoft's MSN search engine). This is a reasonable move. After all, it's obvious "Coke" should go to Coca-Cola and "Pepsi" to Pepsi. But which company should get "cola"? There would be big legal and political battles over issues such as this.

So by sticking to only brand names, RealNames makes its life easier. But it forgets one small but important detail: When seeking something on the Net, most people still think in "generic" words. They think "sex" and "MP3," not Playboy or Columbia House. RealNames may live in a magical universe where everyone thinks products, not needs -- and the rest of the corporate world may join them in this happy place -- but I doubt real people will start typing, say, "Nike" into their computers whenever they want to learn more about shoes.