Home / Blogs

Is the ICANN New gTLD Program a Draft for a V2 of the Internet?

A week ago I was discussing the alternative of a European Internet root with an ex-ICANN Board member. The idea that I like to develop in these discussions is to offer end-users a modern and cheaper naming system. Note that it does not mean that the ICANN root is a bad one, it means that a new and upgraded one could cohabit with the existing one.

One will wonder what this article has to do with new gTLDs but considering the creation of an Internet “version 2”, offers website and brand owners many advantages. Starting from scratch allows to avoid all mistakes made with domain names. Brands and domain name owners already know what they would like to get rid of with .COM domains and new gTLDs: an alternative root would allow that.

Alternative Roots do not make sense…unless…

As usual, discussing this subject often ends with a “come on, it is a non-sense” (in particular when you discuss this with a member of the actual Internet Governing body) or “many initiatives exist already, none works” or again: “users won’t understand and they will see more costs first”.

If I agree that an alternative root governed by a private company is not the good way to do it or has no chance to work, I believe a new alternative root Governed by an “International Body” starts to make sense. Here is why.

What exactly is an Alternative Root?

To make it simple, what we call an “alternative root” is the same as what ICANN does, unless it could be governed by someone else with different rules: it is the network on which ICANN works. If this network already makes Internet to function properly, thanks to ICANN, a body such as the European Commission could build its own network (with servers hosted in every member state) and create its own naming system (instead of “www.brand.com”, we could reasonably imagine that a Brand would want to use a simple “brand” as its name and get rid of what is not necessary around it: “www” for example). A classification created by the INTA could replace Top-Level Domains.

This Internet “version 2” would be Governed differently with its own rules (instead of ICANN’s). Protection mechanisms and the equivalent of new gTLDs would be studied seriously, Singular VS Plurals of a TLD (its equivalent) would not be allowed… anyway: who wants .HOTEL and .HOTELS domain names?

Why does it make sense now?

Such opportunity does not only make sense because Angela Merkel says so, but because there are many more advantages to start again with a blank copy and learn from mistakes of the ICANN new gTLD program.

The ICANN new gTLD program

The ICANN new gTLD program is a good idea and took long to launch. Now it is launched, some mistakes will remain and cannot be modified anymore. The problem with many of these issues is that they will keep bothering end-users for long… and they cost a lot. The more new domain name extensions launch, the more the level of infringement increases and as a brand, I cannot afford to protect myself from infringement registering all domain names or blocking someone to register my name in a different extension. Who can afford to be protected anyway?

It is easy to criticize the ICANN new gTLD program but could these mistakes have been avoided?

The list is long and one will wonder if the actual situations would have been different if these questions below had been asked to end-users and new gTLD applicants:

  • OK if ICANN allows to launch similar domain name extensions in different languages with different rules and with different launch dates and at different prices?
  • OK to pay each time for the same domain name in each of these extensions to block a third party from doing it?
  • Does it matter if your risk of infringement increases the more new domain name extensions are launched: do you mind paying for this?
  • OK to pay for the Trademark Clearinghouse?
  • OK to allow singular and plural versions of a domain name extension?
  • Do you see a problem if a Registry becomes a Registrar?
  • Do you agree for the domain name you are interested in to go to an auction if someone else is interested in it?
  • OK if it cost you to pay up to $1,000 a year for one single domain name?
  • OK with .SUCKS new gTLD?
  • OK with .SEX, .PORN and .ADULT if porn websites remain on .COM domain names?
  • OK if ICANN changes the accreditation procedure after you have applied for your domain name extension (and paid for it)?
  • OK if persons involved in ICANN operations have personal interests in certain TLDs?
  • OK to protect “Olympics games” and a few other lucky ones at the root level? What about other brands?
  • OK to protect wine Geographical Indications on all new TLDs and prior to launching the ICANN new gTLD program?

To all of these questions, I believe the answer would have been a “NO”, except for the last one.

Wine Geographical Indications

The example of protecting wine Geographical Indications is a good one: shouldn’t such issue have been solved prior to launching the ICANN new gTLD program and not after?

A new Internet root allows to start from scratch and learn from these errors

Of course launching a new European Root has a technical cost. There is one too for listening to end-users (and avoid service providers to “think it all” according to their interest). Building a new naming system requires to involve the INTA but end-users must be listened to first.

Copying the ICANN root and think that it will be enough to block a country from spying on another one is not true but if the security aspect of an Internet root is important, the real opportunity in Angela Merkel’s suggestion is to give Brands and Website owners an alternative which is not based on… paying.

What would you choose?

If you had the choice between using a version of the Internet:

  1. where pornography was uncontrolled, where receiving daily spam was natural, where the risk for brands to be squatted existed, where infringement was accepted, where parked domains for sale with useless content was a business, where it took weeks to take down a website showing child pornography, where the governing body didn’t listen to end-users ... and;
  2. where access to the network was controlled, where brands paid once to be protected, where the naming system allowed to enter a keyword to find a results, where pornography was located at a specific place, where spam did not exist and the sender of an email could be identified, where the governing body had legal power to act, where end-users were asked the question on how to improve the Internet ...

Does option two look like it is wonderland? Not necessarily if it required to log-in to a version 2 of the Internet prior to surfing: I personally see no problem with this.

How do you make this happen and what do you get in return?

Two versions of the Internet can cohabit: one on the old ICANN Root, the other on a new European Root. As a brand, I would want to exist on a controlled version of the Internet for many reasons:

  • Credibility is important. When a website is identified, the consumers’ level of trust increases;
  • If the price to exist on an European Internet allows anyone worldwide to come visit my website, I see no reason not to use such new address on my visit cards;
  • I see an advantage to create content on a version of Internet where I know there won’t be confusion with a similar website using my domain name with a mistype “to generate traffic”;
  • No spam? I take it;
  • One place to pay for my name? I take it. Even better if it is a way for our broke Governments to earn money;
  • One single directory? I want to be there and pay for this;
  • As a parent…I am able to offer my kids a safe place to surf;
  • ...

Search Engines now control the Internet: wouldn’t that be a problem for a second root?

Let’s say they are two (three?) American one to control browsing on the Internet. I am confident that they would adapt to a new root, this is just a matter of an upgrade. Wouldn’t they want to, I would download a browser promoted by a strong governing body if it were to develop one in exchange of guarantees that:

  1. it worked;
  2. it was strongly governed and able to take fast action against bad behaviors.

I am sure there are many Geeks in Europe able to develop such a browser for a strong governing body. A a brand and website owner, I would agree to change “for better” if this “Internet version 2” allowed anyone to visit my website and share electronic messages with me.

By Jean Guillon, New gTLDs "only".

Filed Under


JeanI'm sorry, but this article is a Michele Neylon  –  Dec 1, 2014 10:01 AM


I’m sorry, but this article is a rambling incoherent mess.
At a technical level it is fundamentally flawed. At an abstract level it is completely incoherent.
I could try to explain to you in detail why every single “proposal” in this is flawed, but I’d urge you instead to go off and learn a little bit about how the DNS actually works.
Hint: the root also includes ccTLDs, DNSSEC and other stuff ..



The DNS is one tool, there can be other alternatives Jean Guillon  –  Dec 1, 2014 10:23 AM

I understand the answer from a Registrar on a technical point of view but the technical aspect of building anoter root on another (equivalent of a) DNS infrastructure comes second, certainly not first in such a project. It would be like using this argument as a way to explain a .BRAND applicant the benefit of owning his TLD: end-users don't want to hear about the DNS infrastructure. I don't want to learn about how the DNS actually works...who does but service providers?

JeanI'm sorry but you cannot simply decouple Michele Neylon  –  Dec 1, 2014 10:25 AM

Jean I'm sorry but you cannot simply decouple technical reality from a "plan" involving the DNS simply because it doesn't fit with your view of reality. Regards Michele

What the idea is with this post Jean Guillon  –  Dec 1, 2014 10:36 AM

I know it is not so simple (and very expensive) to achieve but it is technically possible to build a second DNS infrastructure in the EU member States for the EC (European Commission) to Govern it. Governing such root on an "owned DNS" is what Angela Merkel has in mind in her speech, as a way to fight spying from abroad. If, added to her argument, a chance is made possible to build a better naming system (such as the one ICANN is actually offering end-users today), why not take it? A new root on a new DNS infrastructure is a start from scratch : it is the idea of the article (and sorry again for my poor English).

Neuland Volker Greimann  –  Dec 2, 2014 3:22 PM

Do not quote Mummy (Mrs. Merkels nom de guerre) on anything related to the net. She has absolutely no clue whatsoever. The internet is newland, i.e. an undiscovered country, to her. Her speech is what you get when politicians dabble in issues they have no clue about.

Politics Jean Guillon  –  Dec 2, 2014 3:45 PM

I faced that with .WINE ("an application I did not submit myself" :-). I think the interesting point is that the discussion is initiated by the Chancellor of Germany. I would probably have never posted about it if this had not been the case.

Just because it is possible does not mean it is a good idea. Volker Greimann  –  Dec 1, 2014 10:45 AM

government-controlled DNS is one of the worst ideas of recent years. It will only lead to segmentation or fracturing of the internet on national or political borders and destroy the one common internet for all.

By proposing it, you are aligning yourself with totalitarian states that have tried to get more control over the internet for years.

The DNS needs to stay as free from government control as possible for the internet to remain a tool for all of humanity.

How do you know? Jean Guillon  –  Dec 1, 2014 10:51 AM

Hello Volker, 1) segmentation does not mean blocking to anybody: many roots already cohabit, a new one on a different infrastructure and governed by a different body has never been done before. 2) this won't kill the Registrars' business model. The good thing about this is that users can be offered a different choice.

While there may be many roots already, Volker Greimann  –  Dec 1, 2014 11:11 AM

While there may be many roots already, there is only one official root and a few minor roots almost no-one uses. Segmentation is the first step to blocking and many regimes have already expressed their interest in this precisely because it means that they could control the flow of information even more tightly by shutting out traffic from outside their segment. It is a bad idea that would drag the internet back into the age of nation states.

Back? John Berryhill  –  Dec 2, 2014 5:49 AM

The internet cannot be dragged "back to the age of nation states". It's barely past feudalism.

Back to the Future Jean Guillon  –  Dec 2, 2014 7:46 AM

A root Governed by Member States does not mean that it has to be closed. The actual version of Internet is Feudalism to me: who is the authority to take down a website offering to build A-bombs (my list of examples is long)?

And here's a clear explanation as to Michele Neylon  –  Dec 1, 2014 11:55 AM

And here’s a clear explanation as to why your “proposal” is a disaster http://www.circleid.com/posts/breaking_the_internet_howto/

Segmentation or Monopoly? Jean Guillon  –  Dec 1, 2014 11:57 AM

I think it is a bad idea because it would block ICANN’s monopoly too. Also, it is a bad idea because it could stop the standardization on pornography on Internet. It is a bad idea because it could offer EU Governements to better secure transmission of information between themselves (ie: prohibiting the flow of info from using an different DNS infrastructure)...

October 2005 Jean Guillon  –  Dec 1, 2014 12:03 PM

I agree with the first line of the article. Now we have experience of the ICANN new gTLD program, or is it still too early to say? "Epensive": is a feed-back I receive quite often from Brands. A new root can change that. Excellent article by the way.

Replying to yourself now? Wow! Michele Neylon  –  Dec 1, 2014 1:12 PM

Replying to yourself now? Wow!

No, "October 2005" is my answer to Jean Guillon  –  Dec 1, 2014 1:14 PM

No, "October 2005" is my answer to your "And here's a clear explanation as to". Pressed the wrong button.

This is a very chinese proposal McTim  –  Dec 1, 2014 12:28 PM

Given a choice between #1 and #2 above, I will always choose what is behind door #2.

This is no proposal... Jean Guillon  –  Dec 1, 2014 7:51 PM

but I guess it re-opens old discussions. I received that link (dated 2000) as an answer: "IAB Technical Comment on the Unique DNS Root": https://www.ietf.org/rfc/rfc2826.txt Interesting I guess.

If you want censorship and fragmentation of Volker Greimann  –  Dec 2, 2014 9:33 AM

If you want censorship and fragmentation of the internet, by all means try to split up the root. If you prefer a unified and free’ish internet for all, with information flowing uninterrupted around the globe, you need a unified root.

De-unifying the root will always give rise to desires to shut out certain parts of the net.

It is a possibility Jean Guillon  –  Dec 2, 2014 9:42 AM

...but what if that speech is the wrong one? Actually, no one is able to tell if a different root won’t be better if it is Governed by someone else. Fragmentation can be competition: is that bad for business? And I am pretty much happy if some authority can turn off a website promoting terrorism… It makes sense. Who has authority for this today and how long does it take? And…how much does it cost?
I understand your position but different roots offer other alternatives: those bad ones we’ve kept repeating for years and new ones.

Sheesh John Levine  –  Dec 3, 2014 12:54 AM

The alleged quote from Mrs. Merkel doesn’t say what you claim it does, actually about routing European traffic through Europe rather than anything to do with the DNS, and I don’t see anything more perceptive anywhere else here.

Alternate roots are a WKBA (Well Known Bad Idea.) They have nothing to do with competition, merely fantasies by rentiers manqués who imagine they can find a bunch of suckers who will use their root and pay them instead of ICANN. There’s plenty not to like about ICANN, but it’s obvious to the reality based community that TLDs aren’t very important, and inconsistent DNS resolution will merely make the Internet work worse for people who want to use it.

My English is really poor... Jean Guillon  –  Dec 3, 2014 8:45 AM

What I read below the photo in the BBC article is: "German Chancellor Angela Merkel is proposing building up a European communications network to help improve data protection." My understanding of the article is that she wants to control it (or whoever would be in charge of it). My article says: "take this opportunity to create a new naming system and "upgrade the Internet". I disagree with this styatement: "TLDs aren't very important". They are the main reason recently why people have paid more interest in ICANN.

Several issues to consider Karl Auerbach  –  Dec 3, 2014 6:07 PM

I have long been an advocate of what I call “competing” roots.  I avoid the word “alternate” because I don’t believe that any one should have primacy.

The real issue is consistency.  Users do not like to be surprised.  What I mean by consistency is that if a root offers a TLD then then the contents of that TLD should be exactly the same as other roots that offer that TLD.  This is quite feasible technically and works with DNSSEC.  Any root operator creates inconsistency will face two coercive forces: legal fights and, more importantly, user rejection of that root because it does not give users the answers they expect.  Those forces will drive each root to be consistent with others else wither as users abandon it.

And competing roots allow for a far more natural selection of new TLDs - through mechanisms that are actually driven ‘bottom up’ by users rather than ICANN’s highly regulated system of central planning (and fees.)

Here is a note I wrote on this topic some years ago: http://www.cavebear.com/cbblog-archives/000331.html  The part about competing roots begins at the heading “The Alternate History”

Could consistency be delivered by Member States? Jean Guillon  –  Dec 3, 2014 6:20 PM

It the EC, in collaboration with 28 members States, created its own Root, what kind of legal fight would it face?

WKBA John Levine  –  Dec 3, 2014 6:34 PM

So long as nobody was required to use their fake root, probably none. If they demanded that businesses in the EU used the official EU government run root, I expect we could hear the screaming if we opened our window here on the other side of the Atlantic.

I don't get it. Jean Guillon  –  Dec 3, 2014 7:14 PM

John, I don't understand which fake root you are referring to. There are many small alternative roots. Are you referring to these?

Yes Karl Auerbach  –  Dec 3, 2014 6:31 PM

Any root operator that offers a TLD product that conflicts with an existing same-named TLD product offered by other roots will face legal arguments based on trademark and misrepresentation.

But even worse, if users decide that a new root system offers a product that gives surprising results, those users (or their agents, their ISPs) will simply abandon that root and move to one that gives more acceptable responses.

A variation of that is possible here in California: Here in California there was, and has been, an operating .web TLD registry since before ICANN’s year 2000 TLD expansion.  I have had it in my own name servers since then, and I purchased “cavebear.web” from them a long time ago - and I still have the paid invoice.  ICANN may discover that it can’t allow its (still pending) .web to operate in California because to do so would create a basis for legal action by the pre-existing operator or its customers.

What's their damages? John Levine  –  Dec 3, 2014 8:18 PM

This is, I gather, IOD's .web, the one where he tried to give Jon Postel a check and then claim Postal had accepting his application 15 years ago. The argument that he somehow had first dibs because he asked first never had a legal basis. Anyone who applied in a previous round waived remedies against ICANN (see the nitpicking in the Name Space suit) so there's no basis there. The term '.WEB' is a registered trademark of an unrelated company in Virginia, so he has no basis there. Whoever ICANN picks to run their .web won't claim to be IOD. Anyone who wants to register with him can still do so, anyone who wants to point at his name servers can still do so. In view of his notable failure to attract business over the past decade and a half, any argument that ICANN's .web will interfere with him is at best totally speculative. So what's the basis for legal action?

IOD/.web has rights to preemption without ICANN Karl Auerbach  –  Dec 4, 2014 12:25 AM

ICANN permission is not necessary to establish a business in California, or in any other location. IOD went into business. totally in accord with the law, with a product called .web. Federal trademark registration is not mandatory; common law protection of product names still exists. I was a customer. ICANN wasn't even yet asking for applications for the year 2000 round when we did this. That product has worked happily for me continuously for more than a decade. It's not my problem is some people can't use it, that's their choice; it's enough that it works for me. And I will be harmed if it stops working. Now if another company, with our without, ICANN's permission comes into California and starts selling a similar product with the same name, that may violate California rules against unfair competititon. If so, it is game over for the newcomer, at least in California, even if they have ICANN's blessing. This has nothing to do with ICANN's approval of IOD. It is rather that a newcomer is trying to displace an existing business and contractual relationships. So when you ask about damages you really ought to start by asking who has grounds for legal redress - IOD, yes. But also every customer, such as myself. Money damages? Perhaps, but because they are hard to measure and money doesn't really serve to compensate, injunctive relief against the newcomer might be more appropriate. So ICANN's designated choice for .web might be able to go into business elsewhere, but they better take care about operating in California.

So explain this to me ... John Levine  –  Dec 4, 2014 12:34 AM

At this point, on the order of 0.000001% of Internet users can resolve your .web name. When ICANN delegates .web, that won't change any of the configurations that point to IOD's servers, so those 0.000001% of Internet users will still see your domain. Where's your damages? IOD's claim that he'd be fantastically rich if only ICANN had accepted his application would not be persuasive to any court I know.

Calculation of damages happens after fact of violation is established Karl Auerbach  –  Dec 4, 2014 12:49 AM

Under legal procedures, the fact of violation of law is established. Then the question of remedy is addressed; it is at that stage that quantitive measures are attempted. But quantitive measures are not needed for relief - the law recognizes that there are things that can not be measured in money or repaired with money. Moreover, on the internet, which is a world of hyper expansive growth, the damages caused by denying one an opportunity to participate can be huge. Imagine the damages that would have accruded had AltaVista been able to deny existence to Yahoo or Google back in those formative years. That's IOD's damages. But there are damages to those of us who have existing business relationships and usages. Those are separate from any harms suffered by IOD. In my case, my damages would be a disruption of an ongoing use of the interenet. The cost of that to me is difficult to measure and also hard to compensate with money. In those situations equitable relief, such as an injunction is an appropriate remedy. Or to put it another way around - how much is the newcomer, that is not in established business anywhere else, damaged by not being able to enter a marketplace (California) in which it has no established presence or customers for an untested product for which there are many providers of similar, but differently named, products? This is not something that never actually happens - companies that move into new regions often find prior uses of company names or products. The newcomer is obligated to find an accomodation - often by buying out the incumbent - else risk prohibition from the region.

One of the many places that argument runs off the rails... John Berryhill  –  Dec 16, 2014 6:34 PM

...is in your use of the concepts "product" and "name of the product" as if, in this context, those were distinguishable concepts. They're not. In making this unexamined false distinction of yours, you don't see that you necessarily have to define the "product" as "registration services in a .web TLD", where ".web" being that string of characters is essential to the definition of what you are calling the "product" for the purpose of claiming unfair competition. To put it another way, putting "McDonald's" on a stand which sold "hamburgers" did not prevent anyone else from coming into California and selling hamburgers. Now, your objection here is along the lines of "no, I mean TLD registration services" is the "product" and .web is merely the "name" of the product. The problem is that in the relevant marketplace, someone wanting to register and use a ".web" name isn't looking for a brand of TLD service, the are looking for a domain name that ends in ".web". They want it as a functional thing, and couldn't give a hoot who is providing it or how the internet works. In other words, what they are looking for is the name qua name. The "name" is the "product". Levi Strauss sold pants to people who were looking for pants. He put his name on them to distinguish them from other pants. A .web TLD registry sells ".web names" to people who are looking for ".web names". A .web TLD registry does NOT sell "names" for people who are looking for "names", and then stick ".web" on it in order to distinguish them from other registries. You have drawn boxes around the concepts of the "product" and the "name of the product" as if those are two different things here, Karl, and as if it is somehow to be assumed that was a valid distinction. The point of ".web" here is not to serve as an distinctive identifier somehow selected by the registry to apply to the products they sell to customers, like Levi's pants, but the entire POINT here is to provide a registration service to other people so that THEY can put ".web" on whatever they want. In other words, nobody was showing up at Levi's shop looking to put a "Levi's" label on their own pants - they were looking to buy a pair of pants he made. In contrast, customers of a .web TLD registry are looking to stick ".web" onto the end of something of their own. It's inherent in the definition of the "product" here per se.

Think less Federal TM law and more California unfair business practice law Karl Auerbach  –  Jan 2, 2015 5:43 AM

I am aware that IOD was unable to establish a Federal trademark in .web - because of USPTO policies about DNS names and trademarks. The real issue, however, is the fact that there was a operating business in California of buying names in a .web top level domain. I am a one of the customers. California has its own laws about business acts or practices that are misleading or deceptive. Any business that comes into California and starts selling names in .web could be construed as running afoul of those laws - because they create confusion betwen the pre-existing sales (and existing contracts) for domain names in .web. http://www.leginfo.ca.gov/cgi-bin/displaycode?section=bpc&group=17001-18000&file=17200-17210 A business that comes into California selling domain names in .web ought to consider whether it wants to face a potential claim that is is in engaging in these unlawful practices. That's hardly a good way to launch a new business in a major market. A company that wants to sell domain names in .web here in California might find it prudent to consider how to quiesce these kinds of claims before they are raised. Also, ICANN itself might want to evaluate whether it is itself at risk of violating these statutes - ICANN can hardly claim any innocence or lack of knowledge about the pre-existance of IOD selling names in .web.

PS John Levine  –  Dec 3, 2014 8:21 PM

IOD has a trademark for .WEB, but it's for mouse pads and accessories. The .WEB trademark in Virginia is for domain name registry services.

Ummmm.... John Berryhill  –  Dec 16, 2014 6:49 PM

I'm not entirely sure what you mean by "the .web trademark", but if you are referring to Application No. 85/444,900 that is a pending registration application which is under refusal. The attorney prosecuting the application has made a reply to that refusal which is nearly comical in the "out of the frying pan into the fire" department, since they are now expressly arguing (to overcome a deceptiveness refusal) that they are an applicant to run a .web TLD. I gather this attorney is unfamiliar with how they have essentially been drawn in to make that point clear, so that the Examining Attorney can issue the standard USPTO refusal of "TLD strings for domain registry services". Straight from the Trademark Manual of Examining Procedure (the guide followed by the USPTO Examiners): --- 1215.02(d) Marks Comprised Solely of TLDs for Domain Name Registry Services If a mark is composed solely of a TLD for “domain name registry services” (e.g., the services of registering .com domain names), registration must be refused under Trademark Act §§1, 2, 3, and 45, 15 U.S.C. §§1051, 1052, 1053, and 1127, on the ground that the TLD would not be perceived as a mark. The examining attorney must include evidence from the LexisNexis® database, the Internet, or other sources to show that the proposed mark is currently used as a TLD or is under consideration as a new TLD. If the TLD merely describes the subject or user of the domain space, registration must be refused under Trademark Act §2(e)(1), 15 U.S.C. §1052(e)(1), on the ground that the TLD is merely descriptive of the registry services. ----- So, the spectacle of watching the applicant pay good money to pay an attorney to argue "But our proposed mark is not deceptive because we are intending to run the .web registry!" is just pure delight. It's like watching Wile E. Coyote just before he realizes he's run off a cliff.

What's funnier... John Berryhill  –  Dec 16, 2014 7:03 PM

...is that the attorney for the applicant in that application further amended the description of services to registration services "featuring the registration of domain names with the gTLD that appears in the mark". Doh! In other words, "yes, the mark is descriptive of the services". It would have cost less to just send in a Reply saying "Dear USPTO, Please refuse this application. Sincerely, Applicant."

I get this one Jean Guillon  –  Dec 3, 2014 7:23 PM

I think “Name Space” is an example of this situation and there was (is?) a lawsuit goign on. Thank you.

There have been some ill operators - but that does not invalidate the concept Karl Auerbach  –  Dec 3, 2014 7:47 PM

Over the years there have been several competing roots.  Many have been operated with disregard for written internet standards and broadly honored operational practices.  They created a bad odor.  However there have been quality operators, such as the ORSN in Europe.

The main problem is “why”?  The existing system of DNS roots operates very well and most people do not see any way of replicating that without spending a fair amount of $$ for little return.

However, if one starts to consider the value of marketing data that can be derived by monitoring the query stream to a root server one might begin to see some light, particularly if one has the means to link that data to other databases.  (If you need a concrete example think “Google”, or in a less positive light think NSA/CIA.)  And it may be somewhat naive to think that the root servers operated by the US government are not being data mined today or that companies that operate root servers are not covering costs that way.

In addition, as a recent Internet Draft in the IETF points out, many ISPs operate their own shadow root servers, either to give better performance to their customers or, as the draft indicates, to quickly short-circuit the huge volume of bad queries that would otherwise burden real root servers.

As I mentioned, the key issue is consistency - lack of surprising results - rather than there be one signular, catholic, set of servers to which one sends DNS query packets.

The consistency issue is interesting - there was another post on CircleID today that suggested that there should be means to pull the DNS records of names that are accused of misbehaving.

Comment Title:

  Notify me of follow-up comments

We encourage you to post comments and engage in discussions that advance this post through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can report it using the link at the end of each comment. Views expressed in the comments do not represent those of CircleID. For more information on our comment policy, see Codes of Conduct.

CircleID Newsletter The Weekly Wrap

More and more professionals are choosing to publish critical posts on CircleID from all corners of the Internet industry. If you find it hard to keep up daily, consider subscribing to our weekly digest. We will provide you a convenient summary report once a week sent directly to your inbox. It's a quick and easy read.

I make a point of reading CircleID. There is no getting around the utility of knowing what thoughtful people are thinking and saying about our industry.

Co-designer of the TCP/IP Protocols & the Architecture of the Internet




Sponsored byVerisign

Brand Protection

Sponsored byCSC

IPv4 Markets

Sponsored byIPv4.Global

Threat Intelligence

Sponsored byWhoisXML API


Sponsored byDNIB.com

Domain Names

Sponsored byVerisign

New TLDs

Sponsored byRadix