[At-Large] Fwd: [] Final Issue Report on New gTLD Subsequent Procedures

Karl Auerbach karl at cavebear.com
Thu Dec 10 09:56:20 UTC 2015

On 12/9/15 11:31 PM, Avri Doria wrote:
> On 09-Dec-15 17:56, Christopher Wilkinson wrote:
>> More generally, TLD policy should protect the Registrant. Apparently that is not part of the new gTLD deal. Or is it?
> To an extent.  ...
I very much agree with Avri's positions here.

ICANN was not constructed to be a consumer protection agency.  It was 
constructed to assure stability - in fact in the beginning the word 
"technical" was always pre-pended to "stability".  Stability is not 
permanency.  A DNS TLD is stable if it doesn't wobble within time frames 
measured in weeks and months.  But "stability" of anything related to 
the internet does not extend to terms measured by unbounded numbers of 

Now it is true, and remains true, that ICANN denies third party 
beneficiary rights to registrants.  Thus domain name registrants have no 
legal standing to compel TLD operators to meet existing ICANN 
established standards of operational quality.  That is something that 
would be very valuable and would be easy to put into place but which 
ICANN has resisted for its entire existence.  If anything should be 
fixed, it is that.

The question should be: "How much larger do we want ICANN's already 
large footprint to become?"

ICANN's regulatory footprint is already large and heavy.

ICANN required TLD applicants to set aside a full three years of 
operating costs - that was something I have not seen in any other area 
of business, even businesses that involve human health and safety.  That 
cost was an enormous burden to applicants, and prevented many people and 
groups to refrain from even bothering to apply for a TLD.  Isn't three 
years of funded operations "stable"? Should it be ten years, fifty 
years, a century?

And ICANN already requires the new TLD operators to have hired backup 
operators on hot standby.  That added no small amount of initial cost 
and creates a significant operational overhead cost.

Much of what is being proposed strikes me as a revival of centralized 
five-year style planning, but this time for the internet.  I can't say 
that that kind of centralized planning has had a great history of 
success.  And success is particularly unlikely for something that is so 
rapidly changing and evolving as the internet.

I remember back in the early 1970's when we were playing with packet 
switched networks that the telephone companies were raising the roofs 
claiming that we were playing with fire, that we would destroy the 
wonder-of-the-world which was the national and international telephone 
network.  They were right; we did, eventually.  But it was not the 
economic disaster that was predicted.

ICANN can not be, and ought not to be, a guarantor of anything more than 
the technical quality of upper tier DNS operations.  To do more risks 
strangling the golden goose of the internet or, more likely, sending the 
goose elsewhere.

The example of VW's cheating on emissions was raised as some sort of 
analogy to ICANN.  I do not see that analogy.  In the VW case, express 
statements were made to governments and buyers that vehicles would meet 
certain emission, power, and mileage numbers. Explicit promises to 
buyers were knowingly and intentionally broken - or in the words of the 
law, there was a misstatement of material facts upon which people relied 
to their detriment - in a word, fraud.  In most places fraud is unlawful 
and those harmed can bring actions to recover their losses.  And when 
really bad, fraud can rise to the criminal level and those responsible 
be punished.

No such promises are made by either ICANN nor TLD sellers beyond that 
TLD operations will meet certain minimum ICANN established 
requirements.  And no one has said that those requirements are not being 

ICANN's own long term requirement for many TLDs that domain names 
registrations be limited to no more than ten years undercuts arguments 
that domain name buyers some how believe that acquisition of domain 
names is something with permanency similar to a common law fee simple 
absolute right in real property.  Even the word "registration" rather 
than "purchase" tends to suggest impermanence.

I believe that some of the new TLD offerings did make it part of their 
business plans to provide very long term guarantees.  But if I remember 
correctly, those offerings were orders of magnitude more expensive per 
year than the more routine TLD offerings.  In other words, consumers 
have options to purchase long term guarantees by choosing those TLDs 
that make such warranties.

If one pushes ICANN to be a consumer protection agency than domain name 
prices will be forced to significantly rise.  Consider how much it would 
cost to acquire an business continuation insurance policy. The actuaries 
will crank the numbers in order to maintain 100% continuity for an 
indefinite period in the face of even unforeseeable risks and come up 
with a whopping dollar number that will strangle DNS innovation or drive 
people to do what has always been possible but always been dismissed 
because the existing system runs very well: competing DNS systems 
outside of the ICANN realm.

But if one looks at an even longer time scale - say ten years - it is 
becoming very clear that DNS names - which includes all TLDs - are 
slowly submerging to become internal internet machinery and not really 
of much interest to users.  Things like Facebook and Twitter handles and 
URL's (which don't need to contain domain names at all) are becoming the 
names and addresses of the internet.  In that future *all* TLDs are at 
risk of loosing revenue streams.

ICANN is already bloated to the degree that one fears that it is at risk 
from just one thin wafer (that's a Monty Python reference). There is not 
much that is funny when one begins to compare ICANN's staff size and 
budgets to that of the ITU (especially if one plots the growth of those 
numbers of each body over time and asks "when will they intersect?")


More information about the At-Large mailing list