Re: Waiting for Santa Claus? Privatize TLD's but *CHARGE* registrars for it !

[John Charles Broomfield <jbroom@manta.outremer.com>]

What you propose is something very similar to the way that many governments
"auction off" radio spectrum either to TV companies or to cell-phone
operators.
I think it is very easy to fall into the mistake of thinking that this can
be done in exactly the same way. However there are some important
differences that do not allow this.
Amongst other things, you have to have very clear what it is you are
auctioning off, and (of course) who it is that is actually doing the
auctioning.
In the case of a government auctioning off radio space, it's obvious that
they can sanction or not entities who decide to use the radio space in an
unofficial manner, basically because they have sovereignty in their
political territory.
In this case, you seem to be proposing auctioning off TLDs, so one would
presume you mean getting entries in the main set of root servers (the IANA
sanctioned ones?) to point at the company(ies) who "win" a TLD. In other
words, you are auctioning off entries in a few computers worldwide in the
hope that the rest of the world will continue to use them (and you certainly
can't force the world to do so).
You can't force down the worlds throat an idea of who should get what TLD for
a very simple reason. If it is generally perceived (and not just by a minority,
but GENERALLY) no to be correct, then a probable reaction would be that one
of the "alternate" (some say rogue) root server "confederations" would gain
momentum and people would decide to use that one instead of the "official"
one. Remember that after all, the mainly used set of root servers is only
so because they haven't made a big enough mess of it to warrant everybody
looking elsewhere to resolve (sure, there are alternate roots out there, but
as yet the support is very low in relative numbers). As a side note, one
VERY good thing which is happening today is that the worlds system
administrators are becoming more aware of what's going on, so they CAN make
a choice in what they use as root-servers -and they mainly seem to stick
with the IANA ones-.
Whatever the organisation that "profits" from the sale of the slots (and
someone has to organise the auction) will be no doubt seen as wanting to
make big bucks on the rest of us.
Sell enough TLDs to enough private companies, and you'll have enough people
"pissed off" at the rules that those companies put in on governing their
privately owned TLDs to create enough momentum for system administrators to
seek alternative root servers.

I think you've all together missed the point in its entirety. The whole
thing is NOT to somehow "milk the DNS cow as much as possible" (which is
what you propose), but rather try to get to a point where those who hold
domains are served better. This better service means:
-stability
-protection
-low price

In general, the granting of licenses for operating TLDs on a for-profit
basis will mean that those companies would have their own interest first and
foremost. They would want to a) recover their one-time payment and b) make a
profit (not unreasonable). If you're talking about making it a mega-bucks
scenario, then watch as the domain prices rise.

As many know, I am a supporter of the gTLD-MoU. In this scenario, it is the
interest of the domain holder which is first and foremost. Registrars *can*
make a profit, but that is the problem of the registrar, not the goal of the
system. It is up to the registrar to find how to earn money, not up to the
organisation to tell it "go forth and prosper".
The $10K that each prospective registrar has to pay now is *only* to get the
ball rolling. Assuming (and it's a lot to assume) that things work out as
planned, an initial consortium of registrars (called CORE) will operate the
TLD registries (or subcontract them out) on a cost-recovery basis (CORE is
non-profit). Once stabilised (probably 12-18 months), more registrars will
be added on a continuous basis, and the "joining" fee will probably *have* to
be lowered (remember, CORE is non-profit). Give it 2-3 years, and you will
find that there is a very large amount of registrars. During this time, more
TLDs will be added (again, as soon as the database part seems to work
seemlessly).
Try for a minute to picture the end scenario:
-aprox 1000 TLDs
-very low charges from CORE to each registrar for each registration
(non-profit!)
-thousands of registrars
The result is that domain holders do not have their domain names held to
ransom by the owner of a TLD (if a domain holder has a problem with a
registrar, it jumps to another one), and prices have all but disappeared.
My guess is that most ISPs would be registrars, and domain registration
would be an extra option as part of your account with the ISP.
Unrealistic you think? Think again, and take a look at nominet. That's
exactly what they're doing today (or very close). Nominet btw, is the
organisation that manages .uk (which unless I'm very much mistaken is the
largest ISO TLD right now, closely followed by .de . Hmmm, that probably
means that it's the 4th largest TLD overall after the big 3 currently
administered by NSI, and probably ahead of the .gov, .edu, .mil, though
anyone with better numbers can probably correct me).
Btw, the overall perception is that people in the uk are *generally* VERY
happy with how their TLD is run.
Yours, John Broomfield.

> Waiting for Santa Claus?        Privatize TLD's but *CHARGE* for it !    
> 
> Just to clarify a point, our proposal *is not to charge domain names 
> owners*, but rather to charge an appropriate value against registrars of 
> new TLD, as they will start their *businesses* with a lot of *gold 
> valued* domains names that may be sold.
> 
> To be fair, I think we should allow them to make a special 
> pre-registration process, similar to an auction, where, before regular 
> registration is open, those wanting special names would place bids on 
> their choice names.
> In this preliminary step, names would be *sold* by the highest bid. 
> We may even decide between open electronic auction or "blind" proposals, 
> nobody know other bids.
> 
> After this special registration, regular low price registration for any 
> name.
> 
> I think this is the best approach for pre-registration, as we know that 
> names have different value. 
> Better doing this way instead of having a lucky guy getting of of those 
> names for a nominal fee and selling it for a small fortune.  We will be 
> reflecting this in the price registrars would pay to get the rights over 
> a TLD.
> 
> Ask any major corporation how much would they pay to be the registrar of 
> a new TLD if in addition to regular reduced registration fees, they can 
> collect this initial pre-registration bids.
> 
> Benjamin Azevedo
>