For the half-year to 30 June 2014, the IPKat's regular team is supplemented by contributions from guest bloggers Alberto Bellan, Darren Meale and Nadia Zegze.

Two of our regular Kats are currently on blogging sabbaticals. They are David Brophy and Catherine Lee.

Friday, 15 March 2013

gTLDs and anticompetitive behaviour: is there a monopolistic trend on the web?

Remember when the dash for
dots was all to do with the
Morse code ...?
So much has been written about the new world of generic top level domains that it will hardly hurt to add a few more words on this weblog. Rather than do the usual business of telling readers what they have already read about and then giving half a dozen links to webpages they've already favourited or never want to visit, this Kat felt it might be better to offer a short, personal observation. Not so short but definitely personal, Katfriend Miri Frankel (Aegis Media Americas) has thus kindly provided an insight of her very own:
gTLDs: I Can … or I Can’t 
When the Internet Corporation for Assigned Names and Numbers (ICANN) announced, as early as 2007 or 2008, plans to create a system for the registration of generic top level domain names (gTLDs), say .food or .movie, concern abounded.  Chief among the issues were concerns from brand owners imagining a nightmare world in which they could never truly keep up with the vigilance needed to police a seemingly infinite number of new gTLDs that might host infringing sites or content. 

Now years later, and despite continuing concerns among brand owners, ICANN is nearly set to award gTLD registrations to applicants who have been approved through ICANN’s application process.  However, another issue has also come to the forefront.  Authors and publishers have expressed outrage at Amazon’s application to register, among dozens of other gTLDs, .book and .author.  Citing antitrust concerns, they claim that allowing Amazon to own these gTLDs creates an online monopoly for Amazon to direct consumers shopping for books to Amazon’s retail pages and to charge authors and publishers a fee to register branded secondary domain names (for example, randomhouse.book).  
According to this report from Thomson Reuters, the argument being raised by the literary world is summed up by famous crime fiction author and lawyer, Scott Turow:   
“Placing such generic domains in private hands is plainly anticompetitive, allowing already dominant, well-capitalized companies to expand and entrench their market power,” Scott Turow, a partner at SNR Denton, novelist and president of the Authors Guild, wrote in a letter to ICANN. “The potential for abuse seems limitless.” 
It is true that major retailers, brands, sports teams, service providers and the like will be the biggest beneficiaries of the gTLD application system.  Registering a new generic top level domain, alas, is not as simple as visiting whois or godaddy and paying a reasonably minor fee to register a generic secondary level .com domain.  I recall “American Dream” themed stories from the early days of the internet in which a John Doe had the good sense to register a generic site like shoe.com and then sell it to a shoe retailer for huge sums of money.  On the other hand, in the case of top level domains, ICANN’s registration process is sufficiently burdensome to make it appealing only to those possessing three traits with which I’m sure Scott Turow is quite familiar: motive, means and opportunity.  
Despite this potential bias towards the deep pocketed, someone is ultimately going to own generic top level domains offered for registration by ICANN.  If not Amazon, who?  In that vein, Amazon’s counterargument seems most intuitive to me:
"Why should a company be able to own 'widget.com' and not '.widget?"…. "Currently, .com may be considered more 'valuable' space, but that does not create a competition issue for the owner of the generic second level domain."
Is there any fairer way to assign gTLDs than a first-to-apply process, provided that the applicant meets ICANN’s registration requirements?  I can’t imagine what point there is to the creation of gTLDs if it were decided that private ownership was tantamount to anticompetitive behavior.  Are authors and publishers creating a work of fiction or highlighting a legitimate concern for monopolistic trends on the web?   
What do readers think, we wonder?

4 comments:

Gentoo said...

"Is there any fairer way to assign gTLDs than a first-to-apply process"

I would imagine it would depend upon what "fair" means. It's a constrained resource and therefore allocation brings its own problems.

I was less impressed by Amazon's widget example which seems to nothing more than resetting the exam question in a slightly different form

Let me reference a real example:

(If you're a lawyer, you'll probably have to have your under footman explain this one to you)

In an amazing bit of foresight many years ago B&Q registered diy.com for their website.


Now I understand that to mean, through custom and practice - you know, how semiotics is developed, by reading right to left that they are a company (as opposed to a not-for-profit .org, academic institution .ac, .edu, .sch or ISP .net, etc) from which I can get stuff to do DIY but, clever though they were, .com gives me a clue that there's more than one of them.

If B&Q had been allowed to own .diy then it's unclear what would be the use of them controlling it unless it were to prevent homebase.diy being registered?

Another example, on .cloud my understanding is that ICANN ran a suboptimal consultation, blink and you'll miss it and three companies want to take that into private ownership.

Similarly, what would be the point unless it were to prevent competitors using .cloud?

Actually, I'm not alone as a protest through the ICC has been lodged and that isn't cheap.

I'm unclear, unless you think monopolies are a societal good, what is the underslying understanding problem.

It's yet more enclosure of the commons and seems to run contrary to the philosophy of those that gave us the web.

W3C attribute the web's success to Sir Tim Berners-Lee's decision in 1993 to make his intellectual property rights to the web royalty-free. W3C members contributing to standards sign away royalties to avoid placing too great burden on a standard.

See http://www.computerweekly.com/blogs/public-sector/2011/10/open-standards-uk-dithers-whil.html

Roufousse T. Fairfly said...

Regarding the "shoe" example: people usually buy them in pairs, so I think that securing "shoeS.com" would have been a much better investment.

In addition to the anti-competitive behaviour concerns, I find arbitrary TLDs generally a bad idea for technical reasons.

Names can be resolved in many ways, and as astounding as it may sound to some, the internet isn't the only computer network on the planet, and the WWW isn't its only application.

The newly introduced GTLDs are for names that resolve globally through Internet Domain Name Servers (DNS).

Names can also resolve to a local resource, such as user machines, intranet servers, printers, or industrial devices (e.g.: process control, laboratory or CNC equipment).

An small facility (home, office, shop, factory) would typically have an in-house network, connected to the internet through a proxy server or a firewall router. The network could either be running internally on a single Ethernet logical segment with NETbios name resolution, or might rely on routers and an in-house DNS, but both varieties will feel the same to the end user.

Local names frequently look like a flat name space, even in larger organisations, whereas internet domain names use a hierarchical scheme (name+'.'+TLD). (Yes, I know that NetBIOS DOES allow in theory dots in names, but I don't remember the last time I saw some used).

The question arises as to how do you distinguish an internal from an external name.

Your operating system will normally first attempt to resolve a name typed in the URL bar of your browser locally
(by consulting your old-fashioned static HOSTS configuration file, or by shouting on the LAN "is there any machine around here with the XYZ?")

Imagine now that you named, for obvious reasons, your printers Canon and Brother, and that the laptop of your trusted employee Harry Potter is called "HP". Your mail server is reachable through the moniker "sodapop". What will/should happen when Canon, Brother or Hewlett-Packard register their own names as TLD? Or some conglomerate seizes all words connected to bubbly water? I suppose that they will want their servers to be reachable directly under the base name (e.g.: "http://canon" instead of "http://www.canon" ). Network administration and security just got more fun...

I already find it bad enough that browsers nowadays have the deplorable behaviour of forwarding the data entered in the URL bar to a search engine of their choice when the DNS request fails, when it isn't your internet provider.

Miri Frankel said...

Thank you for taking the time to write such in-depth comments! I think both of you have outlined reasons, among the many, against allowing gTLD registration in the first place.

So many arguments against this system have been raised and discussed over the years while ICANN developed its processes and procedures. And while I am generally of the mind that gTLDs may prove difficult for brand owners to monitor and cause confusion for consumers, the ICANN ship has sailed and gTLDs are coming whether we like it or not.

I don't see much reality in limiting ownership to public institutions rather than private stakeholders. It is thus imperative to distribute the gTLDs in the fairest way possible. Just as secondary level domains are first to register (and multiple parties may have legitimate interest in the same domain, e.g. Stanley tools v Stanley Steamer v Stanley thermos co all seeking stanley.com), it seems to me that the same process would be fair for gTLDs if the applicant has been determined to meet the application requirements and has a legitimate interest in the applied-for gTLD. This doesn't seem anti-competitive to me since competitors are still able to target consumers through other domain registrations at .com, .biz, .net, etc. Of course, it still remains to be seen whether the Internet-surfing public will embrace gTLDs as the norm!

CM said...

I believe the whole domain name system creates monopolies by design, and that it's up to competition authorities to regulate in case of abuse (for more on this: http://www.circleid.com/posts/20130222_questions_on_the_debate_on_closed_generic_gtld_applications/ )

Subscribe to the IPKat's posts by email here

Just pop your email address into the box and click 'Subscribe':