Much is being made of the meeting in Dubai happening now and ICANN’s opening of new TLDs in 2013. This has been a time to hyperventilate about the UN takeover of ICANN and the Internet, fraud, waste and mismangement at ICANN, the potential of 9 figures of “excess funds” in the non-profit ICANN who gives the aura of serving the Internet community. While these things are interesting to discuss, we should poetically look at the root issues and seek ways to sidestep many of the concerns.
The root issue is the control of the DNS root servers. Over the last two decades several have tried to become an alternative root to ICANN’s officially sanctioned one. Some have become a footnote in history, while others press on with only those in the know having any idea of their existence.
Buckminster Fuller famously wrote “You never change things by fighting the existing reality. To change something, build a new model that makes the old model obsolete.” As an alternative root, we should apply this concept to its fullest. Our solution is to abandon the root model for a rootless one. With the root model obsolete, and no central control, many issues partially or fully disappear.
A rootless architecture whose time has come
The root is where we go when we want to find the actual IP address of a domain like “ShofarDomain.com”. We, or our DNS software, go to the root and ask where to find the “.com” TLD. From the TLD we ask for “ShofarDomain.com”, and then we ask the authoritative server for the IP address. The root’s part of this is actually very small. When the DNS system and the root were first conceived, large computers measured their memory in a few megabytes. Today with gigabytes in our shirt pocket, we can look at things differently.
With the rootless architecture all TLDs are peers. Each TLD is in essence a root. As a DNS resolver I can go to any TLD to find the list of all TLDs. Just as we teach little kids to share, so do TLDs. Even growing to a million new TLDs, this is less than a gigabyte of data, not a concern for our phone, let alone a TLD server.
Internet freedom and the politics of control
As peers, no one can step in and say “you can’t do that”, and a given TLD still has the right to say “we don’t do that.” This is a fundamental of the freedom espoused in the Internet and something that should be present in the DNS.
Admittedly this opens the door for some to do things that are offensive. We must allow for this offense, which will be different depending upon who is looking, to maintain freedom. An instructive example is that Saudi Arabia is lobbing to prohibit the “.gay” and “.bible” TLDs. Much of the orthodox Christian community may want to side with Saudi Arabia in prohibiting “.gay” but not “.bible”. The rational solution, and the one that is consistent with freedom, is to allow them all. It is a simple matter for a country, organization or family home to prohibit TLDs they find offensive.
Just as Gutenberg’s movable type invention was first used to print the Bible, it is now used to print some of the foulest publications imaginable. We don’t stop all publication and destroy the printer because some use it in offensive ways. We must not limit TLDs to only those that don’t offend us, because virtually all will be prohibited. Otherwise we should be non-offensive to the Amish community and shut down the Internet completely.
Problem with simply using an alternative root
The alternative root idea seems good, but which one? ICANN’s root has a given set of TLDs. Alternative A has their set, plus ICANN’s or they are useless. Alternative B has their set, plus ICANN’s, but what about Alternative A’s. The only way this works is if everyone has everyone else’s TLD data. Why not eliminate the root and this problem?
Name.Space filed suit against ICANN on October 10, 2012 on many issues. One concern is that they already have 482 TLDs of which many overlap the new TLDs that ICANN is selling for $185,000 each, including the above mentioned “.gay”. What happens when a collision occurs and multiple roots have the same TLD? ShofarDomain’s software allows multiple solutions from first-come-first-serve at the TLD or the SLD level, or fixed priorities. The net result is that the market decides, not a closed door committee.
Proposal for alternative roots
We propose to the alternative roots not to seek a chair at the table with ICANN, but in keeping with Fuller’s quote let’s bring in a new table. Let’s eliminate the root and become peers to both remain compatible with ICANN while opening the door to innovative ways to bring more value to the table.
We anticipate the rootless TLD software fully functional by the end of 2012. We invite the alternative roots to join us in this design for our mutual benefit and for the cause of Internet freedom that we all espouse.
The rootless architecture means when any gain a marketplace win, or design innovation adds value and new users, a greater percentage of the world is connected to the rootless TLDs, and all gain.
If for years Jon Postel could run the early DNS essentially single handedly, why couldn’t an aggressive high school kid in Iowa, an enterprising family in Latvia, or even the IT department of a fortune 500 corporation run their own TLD? With the rootless architecture this is possible. In the world of the Internet where freedom is the hallmark, this should be common.