RE: Structuring the Root

[Jim Fleming <JimFleming at doorstep.unety.net>]

On Wednesday, April 22, 1998 2:21 PM, Roeland M.J. Meyer[SMTP:rmeyer at mhsc.com] wrote:
@At 13:18 4/22/98 -0500, Jim Fleming wrote:
@>On Wednesday, April 22, 1998 12:32 PM, Roeland M.J.
@Meyer[SMTP:rmeyer at mhsc.com] wrote:
@>@At 07:22 4/22/98 -0500, Jim Fleming wrote:
@>@
@>@>In the IPv8 Plan there are 2,048 slots. This is like tracking the
@>@
@>@Here is a major problem I have with this, limited "slots", Why??!
@>@
@>@I have never ehard, or read, any valid reason for limiting TLDs, that
@>@wasn't pure speculative BS, or driven by some "hidden" commercial agenda.
@>@___________________________________________________ 
@>
@>There does not have to be any hard limit, but people
@>have to start someplace. At the present time, there
@>are almost 256 widely known TLDs in the world. To
@>increase that to a couple of thousand seems like a
@>large increase. Also, 2,000+ is twice as many categories
@>as Peter Roget decided to use in the 1800s. His
@>thesaurus has worked well for many years.
@
@This is a reason? Just becasue "Grand Pa did it that way" ? ... NOT.
@

What I said was that they used less in the past
and it worked for over 100 years and I have suggested
that we double the number to be safe and you call this
doing it the same way ?

You do not seem to be reading what is written...

@I can, of course, replay last months agruments from DOMAIN-POLICY, but I'll
@summarize instead.
@
@1) Country-Codes (CC) are brain-damaged. Any business getting on the
@Internet is, by definition, looking towards INTERNATIONAL markets. Local
@markets are already being served. The whole thrust of Internet business is
@to expand markets beyond what is available locally. There are also ample
@reasons already stated why CC do NOT work. If someone wishes to see
@corroboration, in detail, I can supply this from my archives.
@

In my opinion, TLDs are TLDs...they are just labels on slots...

@2) Catagories are brain-damaged. In the modern flux of business markets, it
@is quite possible for a corp to want to do business in multiple catagories.
@During the course of time, a corp may abandon one catagory in favor of
@another.
@
@Both of the above are shown to be inadequate over time.Notice that most of
@the growth is under .COM and .NET, withj some over-flow into .ORG. It is
@this over-flow which should give a REAL STRONG clue that there is something
@wrong with catagory designations.
@

I suggest that we support what humans come up with.
If they select TLDs for various reasons, let them. The
goal is to have structure that allows for creative and diverse
approaches. TLDs may end up being selected for mostly
marketing reasons. In other words, because people like them.

@>If you want the engineering reason for why IPv8 has
@>2,048 it is because there are 11 extra addressing bits
@>and 2 to the 11th power is 2,048. If you want to know
@>why there are only 11 extra addressing bits, I can
@>go into detail, but the bottom line is that this is all
@>that could be squeezed into the existing IPv4 header.
@>If you want more, use IPv6.
@
@Why are TLDs tied to IP numbers? That is the job of DNS and the root-server
@for the TLD. Direct mapping of TLD slots into DNS name-space would be a
@HUGE mistake. It will also kill the TLD registry business. Right now, there
@are no logical limits to the number of TLD registries. Each root-server is
@authoritative for a NAME-SPACE, not an IP address. The process of
@registration links an IP-BLOCK to a NAME-SPACE, but the mapping itself is
@virtual. To begin hard-coding it into the IP allocation scheme is a huge
@step backwards. I think it an unacceptable one.
@

This is not what is being proposed. I have explained many times
that the "stewardship" for business purposes is what ties the
IP blocks to the TLD authorities. It is simply a recognition that
people in the Registry Industry are mostly focused on domain name
mania. The idea is to leverage off of that business infrastructure.
Then you do not have IP registries that have to charge a fortune
just to pay their bills.

This is no different than having banks handle auto license plates.

@An IP-BLOCK registry assigns IP-BLOCK and a TLD registry assigns NAMES. DNS
@ties the two axis together dynamically. Logically, it's possible to have
@two incongruent IP-BLOCKS be referred to by the same DNS name-server and
@that name-server would be authoritative, for that domain name.
@
@>As for "hidden" commercial agendas...you can look
@>for those forever. As you look keep asking yourself who
@>gets paid for all of the work that goes into these debates
@>and who does not. I find it interesting that many of the
@>people involved object to people being paid when they
@>are assigned by their employer to work on this full-time.
@>In my case, that is not the case. How about you ?
@
@That wasn't the point. Most of those that I've seen argue for restricted
@TLD-space were trying to achieve a supply-side shortage which would
@increase market price for those TLDs allowed to exist. I see this as
@market-manipulation of the most evil kind. This applies especially to those
@same folk that allowed pre-registration, for their private buddies, at one
@price, and an order-of-magnitude price increase for everyone else. If we
@are to do open markets then let's do OPEN markets.
@

I agree. I would like to see an open market.

My goal with the IPv8 Plan is to develop an approach that
avoids having too many people chasing too few resources.
This is why I think that many TLDs need to be opened up
in the near future, as opposed to some slow start number.
I also want to kill several birds with one stone because far
too much time has been wasted on these discussions over
the past few years. It is easier to solve several problems at
this stage in one smooth motion to open up the entire
Registry Industry and not just domain names. If this does
not happen, people will have to debate the same issues
and waste more time.

In summary, I suggest that we create 2,048 IANAs and
allow them to work together to bring stability to the Registry
Industry. They can coordinate IPs, DNs, and ASNs for
starters and help to make sure that people around the
world have their fair share of Internet Resources.

-
Jim Fleming
Unir Corporation
IBC, Tortola, BVI