Home / Blogs

The DNS Still Isn't a Directory

John Levine

Back in the mid 1990s, before ICANN was invented, a lot of people assumed that the way you would find stuff on the Internet would be through the Domain Name System. It wasn't a ridiculous idea at the time. The most popular way to look for stuff was through manually managed directories like Yahoo's, but they couldn't keep up with the rapidly growing World Wide Web. Search engines had been around since 1994, but they were either underpowered and missed a lot of stuff, or else produced a blizzard of marginally relevant results. (Brin and Page wouldn't publish their billion dollar PageRank idea until 1998.) Moreover, web browsers had started to do domain guessing, so if you entered, say, pickles in your browser's address bar, it would take you to http://pickles.com.

So when ICANN started in 1998, it was somewhat plausible to imagine that new TLDs could be directories for various areas. The majority of the TLDs that ICANN added during the following decade were supposed to be topic-specific, and a few tried to be directories for their topics. The .MUSEUM domain tried very hard with organized names like science.boston.museum and the .AERO domain reserved all of the three letter airport codes and two letter airline codes to be claimed by the respective airports and airlines.

What quickly became apparent is that the DNS makes a lousy directory. Since it only does exact name matches, minor variations in spelling make the lookups fail, and in any event, it's hard to encode more than one or two words into a domain name. Also, domain names only point to one place. If you type "pickles" and your browser takes you to pickles.com, you'll find pickles but if you wanted to find anything other than Kraft's Claussen brand, too bad. By 2000, Google's PageRank was providing very good results for searches, and in a search engine, spelling correction and multiple results are no problem.

These days, web browsers totally blur any distinction between search terms and domain names. Firefox has separate boxes for them, but you can type search terms into the address box and it'll work. Chrome doesn't even bother to have two boxes, just type something and it will guess, usually correctly, what you wanted.

What this means, is that ICANN's mandate to expand the TLD name space is now almost entirely pointless, since regardless of what domain someone's web site uses, most people will find it through a search engine and bookmark it, often not even looking at what the domain was. You can make an argument that TLDs can certify their registrants, e.g., all the registrants in .BANK would be real banks, but the one attempt to do that in the 2000s, .PRO, was a complete failure, and attempts to define high security TLDs for the new TLD program collapsed (I was there.)

Nonetheless, in some parts of the Internet, it'll always be 1998, which brings us to .NYC, which recently published a six month update on how the TLD for New York City is doing. Not so hot, it turns out. Three quarters of the 72,000 registrations are just parked, and of the unparked ones, they found only 458 indexed by Google. Most of the rest just redirect to existing names in .COM or elsewhere.

Many names are reserved for community organizations, by geography or activity. A few are active, like archives.nyc which is indeed the municipal archives, but the vast majority are unclaimed and are likely to stay that way. (This is consistent with .AERO's experience with the reserved airport and airline names, few of which are claimed and fewer of which do anything other than redirect to the airport or airline's real web site.) They're putting a brave face on it and describing the vast empty spaces as opportunities but really, don't hold your breath.

They note that the percentage of parked names in .LONDON, .PARIS, and .TOKYO domains are less, but since they're all considerably smaller than .NYC, the number of active domains in each is quite small. A little spot checking suggests most names that aren't parked redirect to other names.

I don't know why anyone should be surprised at this, but I expect the moaning and groaning about how people aren't using their wonderful community TLDs will continue until they run out of money and go bankrupt. After that, who knows what will happen.

By John Levine, Author, Consultant & Speaker
Follow CircleID on
Related topics: DNS, New TLDs, Web
SHARE THIS POST

If you are pressed for time ...

... this is for you. 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.

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

Share your comments

Question Jean Guillon  –  Mar 03, 2015 9:40 AM PDT

John, I don't understand what you want to say here: do you mean that the ICANN new gTLD program is a failure? I changed the ".com" domain name of my company to a ".consulting" because I provide consulting: isn't it a matter of time for the program to be a success?

Still 1998 John Levine  –  Mar 03, 2015 10:25 AM PDT

The words I used were "almost entirely pointless". No matter how badly the new gTLD program does, I'm sure ICANN and the speculators will move the goal posts and claim it was a thumping success.

How do you define success? Samantha Frida  –  Mar 03, 2015 11:34 AM PDT

I agree with Jean - that is a matter of time.

Also, how do you define success and what is the timeline set to judge what has been successful v. not? By registration numbers and by spot checking OR by really analyzing every domain name registered in that TLD zone and then judging (of course subjectively) on what is a success v. not and by whom? Parking v. non-parking is not necessarily the way to measure success - thats how our Industry looks at it because thats what we have defined in the past as a metric BUT I honestly, do think that is a metric that is no longer valuable given that domain name monetization is a different ball game than what it used to me.

I do think nTLD Registries have the bigger responsibility of creating awareness of their TLD and cannot necessarily depend on the registrar channel to do the heavy lifting. Reaching out to end-users to promote/make aware of the TLD is a good way to promote the TLD and get ideas of how to help the registrar channel grow their base.  Registrars, typically, would make more money selling value-added services attached to domain name, so if a TLD Registry can add value beyond just a registration, that would be a good thing for all involved from a renewal standpoint.

I cannot speak for any nTLD although I am sure anyone who spends the money they have to apply for a TLD, you'd think they would have a strategy behind growing that business. The back-end Registry providers are in a very influential position to guide many of these nTLD operators to succeed given they have years of experience that these nTLD operators can leverage from e.g. understanding respective registrar business models, leveraging data reporting to support effective targeted marketing programs etc

I joined Dataprovider.com (www.dataprovider.com) in October of 2014 (14 years in the domain name industry) to hopefully, impact the space in a positive way with the data/business intelligence tool we have. We index the Web and transform the Internet into a structured database providing valuable insights to many business verticals, including the domain name/hosting business. We ALSO, analyze, the zone files of any TLD Registry who wishes to use our technology to understand usage and meaning behind numbers, provide scorecard reporting to registrars, explore areas of opportunity to improve renewals etc

In my opinion, it would be wise to make statements of whether something is successful or not, supported by data beyond parking as a metric. We should also track the different metrics over a period of time to view growth. I am sure not all of us would interpret the data the same way but its a good starting point to take discussions to the next level.

Shouldn’t we, at least, look at historical Alex Tajirian  –  Mar 03, 2015 11:51 AM PDT

Shouldn’t we, at least, look at historical and theoretical models of new technology adoption rates? Otherwise it’s pure rhetoric with each side sticking to its guns based on feelings or self-interest.

New technology? John Levine  –  Mar 03, 2015 11:52 AM PDT

There's nothing technically different between new TLDs and old TLDs. It's just marketing.

Except old TLD's have experience in working Samantha Frida  –  Mar 03, 2015 11:57 AM PDT

Except old TLD's have experience in working with registrars and understand what marketing programs have worked v. not - best practices and learnings that can be shared with their TLD channel partners.

SamanthaThat's not entirely trueQuite a few of Michele Neylon  –  Mar 03, 2015 6:32 PM PDT

Samantha
That's not entirely true
Quite a few of the new TLDs are run by companies that have been in this space for years either as a registry or as a registrar
There are some new entrants who have approached things in a "different" fashion which has worked to a point in some cases and has failed badly in others ..
Michele

Michele,I am sure (and would agree) you Samantha Frida  –  Mar 03, 2015 6:46 PM PDT

Michele,

I am sure (and would agree) you have valid points about some - not calling them out here.  My comments were in general and in reference to the ones I know could be of help to their TLD channel partners given their experience.

Sam

My statements are more from a channel Samantha Frida  –  Mar 03, 2015 6:51 PM PDT

My statements are more from a channel distribution strategy/business development/marketing standpoint - not technical (some backend providers are purely providing the technology to power the nTLD's so no involvement in strategy) AND I am not referring to all nTLD's or legacy ones either.

JohnI'd agree.The only TLD I've seen that Michele Neylon  –  Mar 03, 2015 6:34 PM PDT

John

I'd agree.
The only TLD I've seen that tried to do something technically different was .tel and while it was fantastic (at a technical level) it wasn't easy for non-technical people to understand. Which, in many respects, backs up your point about marketing :)
Michele

.tel John Levine  –  Mar 03, 2015 6:40 PM PDT

I agree, it's technically clever, and I have a name registered there, but it doesn't seem to solve a problem that very many people have. For the most part, the reaction has been "why should I pay extra to get more spam?"

Let's look at history Ken Ryan  –  Apr 07, 2015 1:24 AM PDT

Introduced more than a decade ago, .INFO is the oldest of the new generic TLDs. At one point it counted more than 8.3 million second level names, today it registers only 5.4 million and continues to lose registrations month over month. The second oldest ngTLD is .BIZ with 2.3 million registrations, down from 2.68 million a year ago.  If it's a matter of time, how much time will you give the new TLDs?

To post comments, please login or create an account.

Related

Topics

Domain Names

Sponsored byVerisign

IP Addressing

Sponsored byAvenue4 LLC

New TLDs

Sponsored byAfilias

DNS Security

Sponsored byAfilias

Cybersecurity

Sponsored byVerisign