Re: A comment by Esther Dyson -- Icann Hires Former Cybersecurity Chief as New C.E.O. [with comments]

[David Farber <dave () farber net>]

Begin forwarded message:

From: Thomas Narten <narten () us ibm com>
Date: July 2, 2009 10:31:19 AM EDT
To: dave () farber net
Cc: "ip" <ip () v2 listbox com>
Subject: Re: [IP] Re: A comment by Esther Dyson -- Icann Hires Former  
Cybersecurity Chief as New C.E.O. [with comments]

For IP, if you wish...

Truth in advertising: I'm the IETF liaison to the ICANN board and have
watched this space for about 4 years now. So perhaps I'm tainted. That
said, I find some of the recent discussion about ICANN to be almost
surreal.

Lauren Weinstein <lauren () vortex com>

> 1) The more TLDs that are introduced, the more confusion there is
>    among both domain name holders and the Internet user community in
>    general (that is, the population of the world).

This may be true. But I have to ask what does "confusion" mean
exactly, and how does one define "harm" associated with such
confusion. If I happen to have the name "google", does that mean any
other use of "google" in any domain name is confusing? Probably, at
least to somebody. But what exactly can anyone do about that, given
that there are already 250+ TLDs and there is almost no limit to how
names (and close variants like, say, ugoogle or googlesux) can be used
at the second level?

And, to turn things around, in the US, any name under ".com" is
recognized as being a domain name, which means ".com" has brand/market
value. Indeed, in the US, the value of the .com brand appears to be
substantial, if one just looks at the number of registrations in .com
vs. other TLDs (see http://www.domaintools.com/internet-statistics/)

Active  TLD
81M     .com
12M     .net
7,7M    .org
5,2M    .info
2M      .biz
1,6M    .us

One of the frequent criticisms I hear of ICANN is that there is no
alternative to .com, and thus ICANN is effectively supporting the
status quo. Seems to me that we'll never get away from .com's
dominance if there are no alternatives...

But the above is a US-centric view. Elsewhere in the world (e.g., .cn,
.de), ccTLDs have more brand value and are often preferred over names
such as .com.

> As such, the main
>    TLDs already in common use (com, net, org, edu + country TLDs)
>    gain in value and demand since they will increasingly stand out
>    amongst the clutter of MOBIes, SEXies, WIMPies, WACKies, and who
>    knows what else, most of which will quite rightly be treated by
>    consumers as confusing nonsense.

An interesting hypothesis: by creating more TLDs, we are actually
increasing the demand for names within existing TLDs, i.e., we are
increasing the total overall demand for names!

> The driving force behind the
>    introduction of new TLDs at this stage is creating new profit
>    centers through consumer confusion, and ICANN has become the
>    primary enabler of a domain name regime that we can charitably
>    categorize as just one notch short of a scam.

This is simplistic reasoning. Many people argue all the "good" names
in existing TLDs have already been taken (or are being held by
speculators and such on the secondary market).  Creating more TLDs
increases the total number of available second-level domain names. And
within ICANN, even ALAC is pushing for opening up the TLD space
quickly, to create more names for registrants.

Also, just last week, I had a conversation with a person very familiar
with the domain name market who mentioned that a particular domain
name under .com had recently been auctioned off for over $300K. For
that particular name, all the "good" variants related to that name
were already in use, and the company wanting the name couldn't find a
good (cheap) alternative to the name being auctioned.  He went on to
say that opening up the TLD space will change a lot of that, because
it will potentially increase the number of available names related to
"foo" (e.g., foo.phone, or phone.foo, etc.) This will decrease the
monetary value of many existing names on the secondary market,
presumably benefitting consumers/registrants.

"Bob Frankston" <Bob19-0501 () bobf frankston com> writes:

> Whatever the original mission of ICANN was we’ve learned a lot since
> then we need to do more than spawn lots of little NSIs – that’s  
> not
> competition, it’s just a feeding frenzy. ICANN isn’t addressing  
> the
> fundamental dysfunction and failures of the DNS:
>
> ·         The DNS cannot be a directory but adding support for more
>          languages only reinforces that misconception.

So, are you arguing that the DNS stay restricted to latin-only
characters? This, when the majority of people in the world don't
speak/understand English (or any *one* language, for that matter)? And
those same people may not even have keyboards or user input devices
capable of inputing characters in languages other than their own?

Given the importance of DNS names to end users (end users type them
in, put them on business cards, advertise DNS names/URLs on
television, etc.), it seems to me a non-starter to continue to
restrict DNS names to latin-only scripts.

> ·         We still don’t stable identifiers – the new gTLDs just
>          continue the tradition of creating billable events.
>         
>
> A fundamental principle of the Internet is that those outside the
> network create their own solutions. Yet the DNS has turned out to have
> been a failure – it keeps control firmly inside the network and itâ 
> €™s
> very existence frustrates efforts to move on.

The DNS a failure? That's a good one. The DNS is an *essential* part
of the Internet. The internet wouldn't exist without DNS names. Full
Stop.

And it does not keep control "inside the network". 99.9% of the DNS is
operated in a distributed fashion *without* a single point of
failure. And even at the root-level, there is quite a bit more
redundency than many people realize. E.g., thanks to anycasting, there
are over 100 root servers, not just the 13 many people think.  (see
https://prefix.pch.net/applications/ixpdir/summary/root-servers/)

Sure, I could imagine a nicer system than the DNS, that satisfied a
long wish list of nice-to-have requirements, but the reality is we
don't have such a naming system, and we really don't know how to build
one that scales to the size of the Internet. Really. (For those of you
who think there is a better system, please provide a reference.)

> ICANN should be doing
> all it can to deprecate the DNS. As an interim the DNS should
> immediately and without any further ado provide for stable handles
> that don’t have semantic baggage and thus have no need to be
> reused. It should then encourage others like WIPO and private
> companies like Google, Skype, Microsoft etc to provide their own
> directory and registry services

Nobody is preventing anyone from coming up with a new/better
system. But until we actually have one, and people actually agree its
an improvement over the existing DNS, and there is a realistic plan
for how to deploy it and transition to it, talk is cheap.

> Ultimately we mustn’t have to a fatal dependency on a single central
> point of failure and control like the DNS.

The DNS does have a centralized "root". But if you think about it,
it's hard to avoid having something like that. In any naming system
(especially a *global* naming system), one has to have a way of
arbibtrating disputes. If any two parties choose the same name, who
gets to have it? Someone or something has to be authoritative. If you
don't have a way of arbitrating such disputes, you get massive user
confusion, because names no longer have globally consistent
meaning. When I provide my email address to someone in china, they
want that email address (with its embedded DNS name) to refer to me,
and only me.

> To put it another way – ICANN is a finger in the dike. We shouldnâ 
> €™t
> treat it as a solution but rather a reminder that dike is in desperate
> need of repair. Can an outsider provide the kind of stronger
> leadership necessary to move ICANN beyond its original mission so it
> can do what has to be done to assure the continued vibrancy of the
> dynamic we call -Y΄The Internet‘?

In some ways, I view ICANN as the worst form of management of the DNS
space. Except for all the others that have been invented or could be
imagined.

Talk is cheap when it comes to finding faults in ICANN or the
model. But actually desiging a better system, one that actually
addresses the vast range of requirements, constraints and competing
interests any better than ICANN (and that people would support as a
better alternative) is a real challenge.

Thomas




-------------------------------------------
Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com