Why do I get the impression that nobody pontificating on this patent has actually read it?  It’s number 7,337,910, available at uspto.gov.

It was filed in 2001 and, looking at the examples, was intended for the .CC ccTLD.  The patent is almost entirely about what to serve up on a default web page and says nothing about the DNS wildcard that was the key element in Sitefinder (and, the key reason that it caused such severe technical problems.)

Hi John,

As you say, .CC are only examples, so it can be extended to other TLDs : [In addition to the “.com” top-level domain ("TLD"), other TLDs are also illustrated including the “.cc” geographical TLD and the “.gov”, “.edu” and “.mil” organizational TLDs.]

And indeed the wildcard is also mentioned in the patent : [The last DNS resource record 395 in the file contains a wildcard expression “*” that matches any domain name that did not match an earlier entry in the file.]

Jean-Marie

Indeed it uses a DNS wildcard in one of the examples (and a sloppy one at that, since the figure is missing the labels), but as I presume you’re aware, what matters is the claims, which indeed say nothing about how you know the domain doesn’t exist.

If you look at the whole patent, this is about using a domain as a funky search engine for UPCs, phone number lookup, and so forth. Sitefinder was nominally about spelling correction, which I don’t see in this patent at all.

Well, spelling correction is there whether you see it or not, and it is certainly not limited to the .cc TLD, although Brian Cartmell was of course instrumental in the commercialization of that ccTLD.

Claim 1 reads:

1. A method for responding to received messages that identify domain names that are not registered as part of Domain Name System (DNS), the method comprising:
receiving a message that identifies a domain name that is not registered as part of DNS; detecting an identifier that is part of the identified domain name;

based on the user preference, responding to the received message by providing the determined type of information or performing the determined service.

Detecting a DNS query for an unregistered name, and then delivering a user-selected service in response to that detection is certainly inclusive of spelling correction, among other things, and is certainly not limited to any particular TLD.

However, not cited among the prior art considered during prosecution of this patent is:

United States Patent 6,332,158
Domain name system lookup allowing intelligent correction of searches and presentation of auxiliary information

...which has a 102(e) priority date of 1998, and is highly relevant to the validity of at least several of the claims of this later-filed patent.

VeriSign promised never to bring a SiteFinder-like service back

They made no such promise.  The settlement of the suit involved (a) the right to increase .com prices each year until 2012, and (b) a more definitely-structured approval process for “registry services”.

In fact, the day after the settlement was finalized, shareholders were asking on finance message boards, “When is SiteFinder being turned back on?”

John:

Thanks for the clarification. What I should have said was VeriSign agreed in the settlement that SiteFinder would not come back without prior approval by ICANN.

I think it’s safe to say that approval is highly unlikely. But yes they could try. Here’s a USA Today article from late 2005:

http://www.usatoday.com/tech/news/techpolicy/2005-10-24-icann-verisign_x.htm

Very interesting....