Yeah, what he said.

I'm reading the article thinking ambiguity is a remarkably unhelpful Dysonesque way to approach the issue.

The real problem--as Paul notes--is that people mistake the DNS for a directory service, when it's really just a naming service. It gives us portable and easy-to-remember mnemonics, so we don't have to memorize 192.108.105.37 or whatever arbitrary number is assigned to that computer in the future.

The Internet, unfortunately, lacks a good, standardized, deployed directory service. There are pieces out there (LDAP, WHOIS), but none fills the need. So people have glommed onto the DNS to fulfill that purpose, which it does in a crappy fashion. This misuse of DNS is why we have all these problems such as domain name scarcity and ICANN politics.

(Oh, and don't get me started on the proprietary solutions, such as Real Names.)

Nonetheless, it's good to see the issue brought forward, so people can begin looking beyond the DNS for a solution to the problem.

Bob mixes up some important ideas here, and Cory makes it worse with his title. The DNS is a lookup service which does not have ambiguity. What Bob wants (and what all of us want, really) is a good directory service that does intellgent matching based on what we ask, who we are, what languages we speak, where we are, and so on.

RFC 3254 gives good definitions and examples of directories and lookup services. This should be mandatory reading for anyone who wants to complain that the DNS doesn't do enough.

We need (at least) three kinds of locators on the Internet: machine locators (IP addresses), exact-match updatable service locators (DNS names), and human-friendly information locators (directory searches). The first two are completely covered, and the third is waaaay off, unfortunately.