Re: [wg-c] .com vs. ccTLDs in Europe

[Kent Crispin <kent@songbird.com>]

Replying to a couple of messages:

On Thu, Oct 21, 1999 at 11:58:46AM -0400, Milton Mueller wrote:
> Kent Crispin wrote:
> 
> > Despite NSI's dreams to the
> > contrary, the explosive growth of the net was not driven by the
> > availability of domain names.
> 
> Another typically fact-free assertion.

Another hilarious unconscious self-reference.  :-)

> The expansion of commercial and
> private (non-educational) http:// -based Internet content could not have
> happened had there not been a way to get domain names quickly and easily
> and cheaply. Try setting up a web site without one. (Sure it can be
> done--but no one does it. Think there's a reason?)

No.  There is no reason, because what you say simply isn't true. 

There are *millions* of web sites with no private domain name.  How
do you think geocities makes money? Checked out AOL, Mindspring,
Earthlink, AT&T, and hundreds of other ISPs that offer websites. 
Many small businesses run quite comfortably under their ISPs domain
name. 

Let's contrast two scenarios:

1) Tim Berners-Lee does not invent the http protocol.  No web browsers, 
no web servers, no web sites.

2) Getting a domain name must be justified.  Most people have 
web sites hosted at their ISP.  

It's clear to anyone who thinks that the invention of the hypertext
transfer protocol was a primary impetus to the growth of the net;
people didn't *care* about getting a domain name until the web took
off.  That is, the growth of the net created a demand for domain
names, not the other way around.  The availability of domain names is
at best a secondary effect, and probably not even that. 

> > Also: arguably, if registrations in
> > .com, .net, and .org had been much more restricted, there would have
> > been growth in private third level domain registries.  Such a
> > development would have been far healthier.
> 
> By "healthier" Kent means - register where central authorities want you
> to register. 

[characteristic venomous rant deleted]

It's clear that you actually have no interest in what I mean.  
However, Joseph Friedman also questioned my statement, so I will 
take a moment to explain.

I base my assessment of "healthier" on several factors.  Whether
these factors would actually come in to play is of course hard to
say, since in some cases we are talking about a hypothetical
situation (What would have happened if Hitler hadn't been born?).  
But other effects are objectively "healthier":

[Note that I say "more restrictive", not "completely restricted".]

1) A DNS tree with more restrictive rules for .com, .net, and .org
would be far more balanced.  DNS is a distributed hierarchical
system, and such systems are designed with an expectation of a
reasonable branching factor at each node.  While DNS has proven
capable of handling the grossly distorted tree structure caused by
the bloated .com domain, it would work better if it were more
balanced. 

2) As should be obvious to all participants in this list, part of the
growth in the use of domain names has been driven by an attempt to
force DNS into the role of a directory service -- a means of looking
up information you want to find.  As is well known, DNS has *many*
failings as a directory service, and it was never designed to be one. 
Restricted growth in .com would impede this inappropriate use of dns,
and thus provide more incentive for the development of true directory
services. 

3) A more restrictive .com/.net/.org would mean that other 
registries would get more business.  This would create a more 
balanced competitive environment.

3) The TM situation would be vastly different.  Therefore the
politics of adding new TLDs would be vastly different.  That would
almost certainly be good, since it would not be hard to improve on
the current situation. 

4) Development of a competitive third level domain registry market
would mean that there would be many registry companies on hand.  
Having a set of functioning registries would be much better for the 
health of the internet -- there would be, for example, companies 
that could make a strong case to replace NSI in a competitive bid.  
================================================================

In another thread:

Tony Rutkowski mentions Mark Lottor's stats.  Karl Auerbach corrects 
him: 

> Mark Lotter/Netwizards has done many surveys over the years.  It's now
> done by ISC.  The counts are mainly of hosts and subdomains.
>   
> Check out http://www.isc.org/ds/             
>   
> I imagine that the "authoritative" number for .com could be derived simply
> by counting the .com zone file.  (I don't have a recent copy handy.)

The .com zone file does not contain the data of interest, which is
the physical location (if meaningful) of the registrant.

-- 
Kent Crispin                               "Do good, and you'll be
kent@songbird.com                           lonesome." -- Mark Twain