From: Matthias Tschakert <>
To: NTIADC40.SMTP40("")
Date: 2/11/98 3:26pm
Subject: get out of the way!

Dear Ladies and Gentlemen,

since you issued a request for comments:

Let CORE rule and get out of the way!
These guys handled it well the last years, we can trust them.
It's outrageous what a mess you are doing here.

M. Tschakert
Free University of Berlin


From: Matthias Tschakert <>
To: NTIADC40.SMTP40("")
Date: 2/11/98 3:27pm
Subject: get out of the way!

Dear Ladies and Gentlemen,

since you issued a request for comments:

Let CORE rule and get out of the way!
These guys handled it well the last years, we can trust them.
It's outrageous what a mess you are doing here.

M. Tschakert
Free University of Berlin


From: Matthias Tschakert <>
To: NTIADC40.SMTP40("")
Date: 2/11/98 3:27pm
Subject: get out of the way!

Dear Ladies and Gentlemen,

since you issued a request for comments:

Let CORE rule and get out of the way!
These guys handled it well the last years, we can trust them.
It's outrageous what a mess you are doing here.

M. Tschakert
Free University of Berlin


From: "Paul Stahura" <>
To: NTIADC40.NTIAHQ40(dns)
Date: 2/11/98 3:56pm
Subject: Our Comments

Attached are our comments in response to the
Discussion Draft on Technical Management of Internet Domain Names.
It is a MS Word 6.0 document.

Paul Stahura
Syllogistics LLC
eNom, Inc.


In Opposition to Exclusive Ownership of Generic Top Level Domains

February 11, 1998

A letter in response to the Commerce Department's Green Paper:

"A Proposal to Improve Technical Management of Internet Names and Addresses

(Discussion Draft 1/30/98)"


To whom it may concern,

Although it is unlikely that anything close to a 80% or so consensus will be reached on the shared registry issue, we sincerely hope that this letter helps to tip the scale from 49% in favor of a single shared registry to 51%.

Since the green paper plan is already months late, and since it seems the government will take some action one way or the other, we urge that action in this matter be taken as soon as possible.

The green paper makes the valid distinction between a registrar and a registry.

In general we are strongly in favor of competition, but in the "competitive registry" case, exclusive ownership of gTLDs will *not* provide any additional diversity, efficiency, or consumer choice over the shared registry system. We believe the shared registry system will provide more efficiency, the same consumer choice and the same diversity when compared to the exclusive ownership model. Plus, a shared system does not have many of the problems that the exclusive ownership model has, such as: "domain portability", "who-gets-what", "windfalls", "gaming", "uneven SLD distribution" etc.

On balance, we believe the following policies will have more negative than positive consequences for the Internet:

1) Exclusive ownership of generic TLDs

2) Multiple generic TLD registries.

On balance, we also believe the following policies will have more positive than negative consequences for the Internet:

1) Competition among registrars.

2) A single, open, not-for-profit, shared registry.


The following eleven points help explain the above position

Point 1

Registries will require a good deal of oversight, which will effectively eliminate the benefit that free competition, would have provided.

In a competitive environment, what is to prevent a gTLD owner from auctioning *all* the SLDs in their TLD, selling them all in one shot to a single registrar on the TLD's opening day? Then that entity has complete control over all the names in that TLD, totally bypassing the guidelines and intent in the Green Paper or any other influence. In effect, the US Government granted a huge windfall to a "registry" that had no intention of actually providing a service to the public.

There will have to be strict controls and regulations placed on registries to prevent this and worse "gaming" behavior, thereby making all registries nearly the same and removing any benefits one can provide over another, further eliminating their competitive diversity.

Point 2

A multiple "competitive" (but not really) registry system is less efficient, and provides no greater consumer choice over a shared registry system.

A registrar operating under the green paper plan will need to interface to each and every registry. This will make the system inefficient. Since under the plan a registry must be open to registrations by any registrar, other organizations such as the IETF will need to be involved to help set the standard for registrar/registry communication. This has a significant cost and will take time. A more naturally occurring standard (via competition) will take even more time to "win", but the result will be the same: a standard. Since a standard, by definition, makes the interface the same for each registrar, the multiple registries will be indistinguishable, technically, from each other. Where is the diversity? Thus, again, the multiple registry system provides the registrars (the registries' only customers), and thereby the consumer, no additional choices or options when compared to a single shared registry system.

Point 3

Longer term, negative, ramifications such as: "dry well", "registry goes under", "nagging the TLD authority", and "who-gets-what-and-when"

A TLD is like an oil well, with the added disadvantage that the quality of the oil decreases the more you pump. NSI is busy pumping .com dry and at the same time the product quality is also decreasing. All the quality names are already taken, leaving names like: "". What happens when a registry operating under the green paper plan pumps their gTLD dry? Do they get another one? Or is the TLD just spent and the registry goes under? With multiple competing registries the "dry well" and "registry is out of business" issues will become major problems in the future. If there is no way for registries to drill their own wells, then by definition how can they grow? The registry will have to nag the authority for another gTLD. The nagging will begin as soon as the registry gets its first gTLD. With exclusive ownership, not only do we have the "who-gets-what" problem (see Point 10 below), but also the problem "when do they get it?"

Point 4

If the multiple registries are required to be uniform, then competition is minimized, if they are not required to be uniform, then huge inefficiencies and "gaming" possibilities will exist.

A "uniform trademark approach" and a "uniform registry interface" will likely be needed across all registries. If the registries are so uniform, they are not providing a differentiating service. Which leads us to our next point....

Point 5

Registries really only compete on the value of their "brand" and the quality of the SLDs that remain in the gTLD. Since the registry has little or no control over these, the registry is adding no value.

Since the value of a TLD is intrinsically built into the definition of the top level domain name (for example, ".web" and ".biz" more valuable than ".hello" and ".shoe"), it is nearly impossible for the registry to add value to the "brand" because they would need to change the definition of the word. The quality of the SLDs in the TLD is also not under the control of the registry. Therefore the registry is not really contributing any value to the public other than the access to the gTLD (if you call access a "contribution"). Access to the gTLD namespace can probably be provided at a lower cost and more efficiently in a non-profit shared system rather than in an exclusive ownership system.

Point 6

Some potential generic gTLDs will not be able to support a registry alone and therefore will not be offered as a choice to the public.

Many single gTLDs will not be valuable enough to support a registry. These gTLDs will have to be subsidized somehow to survive. It is likely that it is more efficient for one registry to handle more than one gTLD. Therefore, a more optimum situation is many gTLDs per registry, not one-to-one, and the most optimum, we believe, is all-gTLDs-to-one-registry. With multiple registries, who makes the decision which set of names go to which registry? This problem goes away with a single shared registry because all the new gTLDs would be assigned to a single registry where they would be competitively shared by all registrars.

Point 7

Domain Portability

If a name holder has registered in an exclusively owned TLD, then that registry has a "locked in" and customer and can dramatically increase the renewal fee for the customer year after year. The customer has no choice but to pay it because the customer can not switch to another registry and keep the same TLD. Think of apartments in New York city and rent controls, but the situation will be worse in this case, because with a non-shared TLD you can't just move to another apartment since what you are renting is the street address of the apartment, not just the space inside. For example, "Wall Street" may be more valuable than "132nd Ave". You need to take the land the apartment building is located on (the gTLD) with you in the domain name case. We do not know of a solution to this major problem of exclusive ownership besides switching to the shared registry framework.

Point 8


If the "competitive" (but not really) registry system proposed in the Green Paper stands, then why does NSI get an unfair advantage in having three TLDs when other registries get only one (or less)? NSI will not be able to support .org and .net forever because their revenue from .com will be decreasing, since that oil well is running dry, and since, if the domain portability issue is somehow solved, the renewal fees will drop.

So if a registry under the green paper plan is lucky and is granted a "good" name such as ".biz", it will likely get a windfall; if unlucky, and granted ".shoe" it may not be able to survive. The who-gets-what-and-how-many problem is extremely difficult. If the qualification bar is raised, limiting the number of eligible registrars, this problem becomes easier, but with just two or three registries the limited "competition" and decreased efficiency is not worth it. Why not have just one *shared* registry in this case? If the bar is lowered the problem is more difficult because there are more registry/gTLD combinations. Either way, the exclusive ownership system loses.

Point 9

The question "Who owns generic TLDs?" has the same answer as "Who owns the moon?" and the answer is everyone.

The country code TLDs are owned by their respective countries just as those countries own their land on the earth. The vast generic TLD namespace is like land on the moon. We don't believe that huge chunks of the moon should be given away and owned by anyone who looked up one night and saw it up there, even if they did have a telescope and "claimed" it first. One can argue how NSI should continue to own their 3 TLDs in perpetuity, because they were there first, and are the only one there now. Market pressures on their drying wells will bring them into the shared registry model over time. The moon is a shared resource of the earth, just as the "new lands" in the generic namespace ought to be.


Point 10

SLD name distribution

With a shared system, we believe the SLDs are more likely to be evenly and fairly distributed to the public at a lower cost. In an exclusive ownership model, we believe that SLDs are more likely to be distributed in "chunks", with groups of names, or even the entire namespace for a particular gTLD, going to domain name hoarders or insiders.

Point 11

Exclusive ownership of a TLD is monopoly ownership of a resource.

In a not-for-profit shared registry the registrars will put pressure on the shared registry to improve efficiency and to keep costs down. This is especially true if the registrars have a voice in the registry.

In the exclusive ownership model, the registrars have no voice

(by definition, they are separate), and the registries can set the price to anything they want because the registrar has no alternative than to accept the price if the registrar wishes to register a name in that particular TLD.


We support:

1) A single gTLD registry be set-up now or ASAP

2) That this registry be shared and open to all registrars

3) That this registry organization be not-for-profit.

4) That it has a respected, credible dispute resolution policy with authority from an entity other than the registry.

5) That this registry be granted a number of generic TLDs to start

6) NSI retaining ownership of .com, .net, and .org (no need to break-up NSI into a registry and registrar) until NSI wishes to join the shared system.

We also support the IANA/POC/CORE shared registry proposal: the gTLD-MOU.

We believe it is a fair, stable and reasonable way to move forward.

If we start with a single shared registry system, the gTLD authority can always add, or threaten to add, new registries if nasty problems occur. It would be nearly impossible to change the system from an initial state of many registries to a single shared registry system.

Thanks for your time in reading this letter and for considering the points made here.

If it can go "either way", we urge you to tip the scale to 51% in favor of a shared registry system.


Paul Stahura

Principal, Syllogistics LLC

President, eNom, Inc.

February 11, 1998

Redmond, WA




From: Electronics Direct <>
To: NTIADC40.NTIAHQ40(dns)
Date: 2/11/98 6:51pm
Subject: criticism of Discussion Draft

re: USG Discussion Draft on Domain Names dated January 30, 1998
attn: Commerce Department

The recently published Discussion Draft on internet domain names
is the Government's short-sighted attempt to introduce competition
to the Domain Name System.

The Discussion Draft recommends only 5 TLDs which is absolutely
ridiculous for a worldwide Internet population of more than 100
million users (currently) and potentially more than 200 million
users by the year 2000. Why should companies have to pay $100USD
for a name which only costs $5 dollars to create and enter into
a database?

Network Solution's (INTERNIC) exclusive monopoly has resulted in
an artificially high price for domain names which is a barrier
of entry for many businesses wishing to get on the Internet.
Every business eventually needs a domain name, just like every
person has a name and every house has an address. Domain names
should be an inexpensive commodity that is readily available.

There should be atleast 50TLDs because, as we all understand,
greater supply means lower prices for end users. By limiting
the number of TLDs to only five and by creating an exclusive
registry for each TLD, the registries are positioned to make
enormous sums of money at the expense or consumers and
businesses. The government should authorize the creation of
a minimum of 50TLDs immediately. Jon Postel of IANA has already
stated that there should be up to 150 new TLDs. Since it does
not actually cost anything to create a TLD, it is in the
best interests of the Internet community to create as many TLDs
as possible, atleast another 50 as a minimum. I would strongly
suggest that everyone involved take a course in Economics 100
and read the "Dummies Guide to the Internet" for starters.

Thanks for your time.

James Leong
Electronics Direct
423 S. Hindry Avenue, Unit D
Inglewood, CA


From: "Paul Lum" <>
To: NTIADC40.NTIAHQ40(dns)
Date: 2/11/98 7:39pm
Subject: domain names

Tell Bill Clinton and his buddies at Internic to stick
their sticky fingers somewhere else. Stop trying to
manipulate the Internet.

Get Your Private, Free Email at


From: jean david <>
To: NTIADC40.NTIAHQ40(dns)
Date: 2/11/98 9:23pm

The green paper has created doubts, uncertainty, and confusion by its
bleak description of how the Internet environment will operate.
It intentionally leaves vague answers to urgent and critical
questions, proving to be an almost complete waste of time.
While the US government decides to take it's time and sit on it's
Laurel's, the impending opportunities and the potential expansion
which has been created in the Internet community by the gTLD-MOU could
be caused to die or diminish due to the lack of duly needed attention
on this matter.
Does Ira Magaziner want to be responsible for the demise of the
Internet structure as we know it? Mr. Magaziner claims he wants to
get government out of the Internet, but since when has the US
government meddled in the Internet? One of the main reasons the
Internet has developed into the marvelous structure that it is today,
has been the lack of government intervention into it's policies. This
green paper changes all that.
Is he a man that is not open to new and creative idea's that have
sense and foundation? Is he a man who wants to sit and watch the world
pass by him? If he chooses to delay this subject any further, I would
wish not to be in his shoes. Not only will the governmental
authorities of other nations be quite perturbed due to the US
Government thinking they have the right to speak for all humanity, on
a world wide level, but he will go down as the person who forced
business to continue dealing with Network Solutions (InterNIC), and
their incompetent, sloth-like, bumbling service, that is known to the
many users who have tried to do business with them in the past.


Jean David

Get your free address at


From: <>
To: NTIADC40.NTIAHQ40(dns)
Date: 2/11/98 10:31pm
Subject: My Comments on "Discussion Draft"

As a web page developer, I am responsible for choosing and
Registering the domain names for my clients. I would like to add
my comments in regards to the recently published "Discussion
Draft" on internet names and addresses.

The newly proposed Domain Name System definitely limits the
choices available to my clients. My clients had been hoping to
register multiple domain names for marketing purposes. I had
already pre-registered over 500 domain names on behalf of
my clients with a domain registration company. The total cost
was supposed to be $20 for each domain name. Now that the
government wants to create EXCLUSIVE registries, I have
been told that I now have to pay $80 for each domain name!
Is this what you call free enterprise? We need a real solution
immediately, not six months or two years from now.

Anselm Lo
Web Design and Development