ReclaimThe.Net.Forum

Paul Garrin Says Monopolies Choke the Web. Now Congress Is Starting to Listen. Casting a Wider Net by Sarah Ferguson From the Village Voice http://villagevoice.com/issues/0114/ferguson.shtml

Some time ago, this artist was pushing to reverse the direction of domain names, so that instead of la.indymedia.org, you'd have something like indymedia.la.web. And, he wanted $15 for the service. That's absurd. It's just profiteering.

I suppose this is an interesting symbolic protest, but it doesn't get at the heart of the matter, which is the process by which ICANN has screwed up namespace management. They screwed it up by not being democratic and by supporting a largely commercial special interest agenda.

Garrin's "solution" of a large number of top level domains doesn't work. The relatively small number of TLDs allows for a better distribution of name servers, which, in turn allows for better decentralized growth of names. It's all due to "delegation", which is the jargon describing how names are used by computers.

Domain names are a little like alphabetical phone numbers: 1-800-EAT-SHIT. The difference is, eat.shit.com becomes 128.4.23.24 (made up "IP number"). The name is converted or "resolved" to the "IP number" by a "name server". The top level domains, like indymedia.org, are "resolved" by the "root servers", and there aren't many of those.

The root servers don't know the exact number, but know of another name server that might have the number. It "delegates" the request to this other server, which is typically owned by an ISP. There are tens of thousands of these name servers at this level, and they take on most of the load of resolving addresses. Delegation allows the net to grow quickly, at a technical level.

Delegation also allows the registration to be decentralized. When someone wants a new indymedia.org site, you don't request it from Network Solutions, the .com company, you ask the IMC's sysadmins. Delegation, at some level, also allows for social decentralization.

The domain names are centralized, and they are decentralized. It's good to criticize the status quo, because it's unnecessarily narrow and biased, but this is more a piece of art than a feasible model for name resolution.

As art, name.space can be criticized as being Eurocentric and not much more interesting than the Usenet namespace. (A comparison of the two is merited.) The artist takes the position that the TLDs restrict freedom of expression, but ignores the fundamental technical limitations. He also wants to be the operator of a top level domain, by which he'd become fabulously wealthy. Perhaps he's unintentionally highlighting the social problem of turning language into commodified intellectual property. Despite this, Garrin makes an important point about TLD name spaces: they are spaces where free speech must be created and protected.

Some time ago, this artist was pushing to reverse the direction of domain names, so that instead of la.indymedia.org, you'd have something like indymedia.la.web. And, he wanted $15 for the service. That's absurd. It's just profiteering.

I suppose this is an interesting symbolic protest, but it doesn't get at the heart of the matter, which is the process by which ICANN has screwed up namespace management. They screwed it up by not being democratic and by supporting a largely commercial special interest agenda.

Garrin's "solution" of a large number of top level domains doesn't work. The relatively small number of TLDs allows for a better distribution of name servers, which, in turn allows for better decentralized growth of names. It's all due to "delegation", which is the jargon describing how names are used by computers.

Domain names are a little like alphabetical phone numbers: 1-800-EAT-SHIT. The difference is, eat.shit.com becomes 128.4.23.24 (made up "IP number"). The name is converted or "resolved" to the "IP number" by a "name server". The top level domains, like indymedia.org, are "resolved" by the "root servers", and there aren't many of those.

The root servers don't know the exact number, but know of another name server that might have the number. It "delegates" the request to this other server, which is typically owned by an ISP. There are tens of thousands of these name servers at this level, and they take on most of the load of resolving addresses. Delegation allows the net to grow quickly, at a technical level.

Delegation also allows the registration to be decentralized. When someone wants a new indymedia.org site, you don't request it from Network Solutions, the .com company, you ask the IMC's sysadmins. Delegation, at some level, also allows for social decentralization.

The domain names are centralized, and they are decentralized. It's good to criticize the status quo, because it's unnecessarily narrow and biased, but this is more a piece of art than a feasible model for name resolution.

As art, name.space can be criticized as being Eurocentric and not much more interesting than the Usenet namespace. (A comparison of the two is merited.) The artist takes the position that the TLDs restrict freedom of expression, but ignores the fundamental technical limitations. He also wants to be the operator of a top level domain, by which he'd become fabulously wealthy. Perhaps he's unintentionally highlighting the social problem of turning language into commodified intellectual property. Despite this, Garrin makes an important point about TLD name spaces: they are spaces where free speech must be created and protected.