Would you tolerate no options in who you buy connectivity and hardware from?
No? Then why do you tolerate it in the DNS business?
Because there can be only one authority for ".".
I don't like the one we have but I like your proposal even less.
IAHC's proposal fixes the problem folks are seeing with NS today.
No it doesn't.
If you register with someone, pay them, and they claim never to have
received payment you're just as screwed. They hold or delete your
record and you end up paying again.
What is it that you dislike about eDNS Paul? I'd love to see some actual
substantive criticism (as opposed to "IAHC is God") of the policy points.
you've seen it, you just won't admit it.
you are setting up additional monopolies. monopolies over critical
resources are to be avoided wherever possible. the problem for you, here,
of course, is that monopolies over critical resources can incur windfall
profits. you want to allow that. the iahc doen't.
you are attempting to coopt an established administrative structure
that has worked well for 10 years, rather than to work contructively on its
enhancements.
you are holding yourself beyond accountability
you are pretending that the DNS gTLD space is a US resource rather
than one that is global.
no doubt there are more substantive criticisms, but one grows weary
and the list is long enough.
d/
>
> Because there can be only one authority for ".".
I don't agree with this.
It is quite possible for everyone to maintain ".". Then, there would be no
need for root domain servers; HOWEVER, this would take massive cooperation
throughout the industy, which will never happen.
>What is it that you dislike about eDNS Paul? I'd love to see some actual
>substantive criticism (as opposed to "IAHC is God") of the policy points.you've seen it, you just won't admit it.
you are setting up additional monopolies. monopolies over critical
resources are to be avoided wherever possible. the problem for you, here,
of course, is that monopolies over critical resources can incur windfall
profits. you want to allow that. the iahc doen't.
Mr. Crocker, I really wish you'd stop posting material that just is not true
and that you *know* is in fact false.
eDNS enables *all* business models for registration of TLDs and SLDs.
Not one, not two, not three. It passes no judgement on which models are
appropriate, leaving that to the open marketplace.
Instead, it prevents any model or any organization from owning a "controlling
interest" in the namespace. THAT is the public policy portion of eDNS. It
is the only "policy" portion of eDNS which is actually enforced at the root
level.
you are attempting to coopt an established administrative structure
that has worked well for 10 years, rather than to work contructively on its
enhancements.
18 months of working "constructively" got nowhere. Eventually, the time
comes to change the structure. Remember, the Internet credo is "rough
consensus and working code". We have working code, and are building
consensus day by day.
you are holding yourself beyond accountability
On the contrary. I am one man, and the machine I run as a root is one
computer. It is the only one which I own or control in the entire eDNS
root structure, and will always be the only one. I also have publically
refused to take a position with an RA organization, and will do so again
if asked in the future.
Contrast this with the existing IANA roots, several of which are owned by
the existing monopoly registrar or have been partially or totally funded
by them. As an example, f.root-servers.net, which Paul Vixie has in his
control, he has admitted was partially or fully paid for by NSI.
Its tough to tell the person who pays the check every day "no". Very, very
difficult.
My accountability is simple. If I violate the process someone steps in and
my single machine gets replaced with another. I have no authority or control
over the root whatsoever. Only consent of the people who use it, and who
operate the RAs and registries make the structure work.
I don't pretend to hold in my hand that which is not mine.
you are pretending that the DNS gTLD space is a US resource rather
than one that is global.
Nonsense. The TLD namespace IS global. There is nothing preventing non-US
interests from registering TLDs, and in fact more than one has (proof positive
that this statement is ALSO false). There are currently registrars in Germany
and Japan -- pretty much opposite "ends" of the world.
Proof follows and can be reproduced by anyone who does not believe me (eDNS
delegations all have TXT records designating the RA and Registrar responsible
for their operation):
; <<>> DiG 2.2 <<>> txt jpn.
;; res options: init recurs defnam dnsrch
;; got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 6
;; flags: qr rd ra; Ques: 1, Ans: 2, Auth: 3, Addit: 2
;; QUESTIONS:
;; jpn, type = TXT, class = IN
;; ANSWERS:
jpn. 86400 TXT "2-23-1-1038 ,Yoyogi,Shibuya-ku, 151, Tokyo, Japan"
jpn. 86400 TXT "RA: Alternic / Shirokuma Publishing - Masafumi Yoshida <myoshida@po.iijnet.or.jp>"
;; AUTHORITY RECORDS:
jpn. 86400 NS aragorn.alternic.net.
jpn. 86400 NS nyc.alternic.net.
jpn. 86400 NS mx.alternic.net.
;; ADDITIONAL RECORDS:
nyc.alternic.net. 86400 A 207.51.48.15
mx.alternic.net. 86400 A 204.94.42.1
;; Total query time: 6 msec
;; FROM: Jupiter.Mcs.Net to SERVER: default -- 192.160.127.90
;; WHEN: Thu Mar 20 21:10:19 1997
;; MSG SIZE sent: 21 rcvd: 278
; <<>> DiG 2.2 <<>> txt ger.
;; res options: init recurs defnam dnsrch
;; got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 6
;; flags: qr rd ra; Ques: 1, Ans: 2, Auth: 3, Addit: 2
;; QUESTIONS:
;; ger, type = TXT, class = IN
;; ANSWERS:
ger. 86400 TXT "Kennedyallee 89 Frankfurt, D-60596 GERMANY"
ger. 86400 TXT "RA: Alternic / Callisto germany.net GMBH - Robert Hanke <robi@germany.net>"
;; AUTHORITY RECORDS:
ger. 172000 NS aragorn.alternic.net.
ger. 172000 NS nyc.alternic.net.
ger. 172000 NS mx.alternic.net.
;; ADDITIONAL RECORDS:
nyc.alternic.net. 86400 A 207.51.48.15
mx.alternic.net. 86400 A 204.94.42.1
;; Total query time: 6 msec
;; FROM: Jupiter.Mcs.Net to SERVER: default -- 192.160.127.90
;; WHEN: Thu Mar 20 21:11:29 1997
;; MSG SIZE sent: 21 rcvd: 264
These are new and not yet actually operational from my understanding, but
they ARE registered. These two just happend to pop immediately to mind.
no doubt there are more substantive criticisms, but one grows weary
and the list is long enough.
If you remove the blatently and easily proven false statements which you
have made from consideration, you haven't posted a single substantive
criticism here.
My understanding is that the precise protocol used to register a new name
has not been frozen yet. This issue can be addressed if CORE requires a
non-repudiable transaction from a registrar in order to register a domain
or to update its payment status.
And even if the IAHC were so foolish as to not consider this possibility
they certainly did create a Policy Oversight Committee that can change
procedures at any time.
http://www.iahc.org/draft-iahc-recommend-00.html
Michael Dillon - Internet & ISP Consulting
Memra Software Inc. - Fax: +1-250-546-3049
http://www.memra.com - E-mail: michael@memra.com
as soon as one person's root has a different .com (or whatever)
then you have parititioned the net.
there can be only one root.
d/
you are setting up additional monopolies. monopolies over critical
profits. you want to allow that. the iahc doen't.Mr. Crocker, I really wish you'd stop posting material that just is not true
and that you *know* is in fact false.
Karl, first of all, just because you are pissed at me is no reason
to revert to artificial formality. We've always called each other by first
names and there's no reason to stop now.
Second of all, you need to re-read the above statement to which you
took exception.
eDNS enables *all* business models for registration of TLDs and SLDs.
as I said, it sets up monopolies. the fact that you demure from
prohibiting or requiring them does not make the above false, since it DOES
mean that you permit them. Oh gee. Do we think that folks will take
advantage of this opportunity you are providing them to gain exclusive
control? Gosh, I wonder...
by them. As an example, f.root-servers.net, which Paul Vixie has in his
control, he has admitted was partially or fully paid for by NSI.Its tough to tell the person who pays the check every day "no". Very, very
difficult.
Now, you see. That's interesting, because I don't have any trouble
imagining Paul say no, even to someone who has been giving him money. It
has to do with the nature of the person. That's been one of the hallmarks
of the Internet, the personal integrity of those put into positions of
responsibility.
Why do you want to change that?
My accountability is simple. If I violate the process someone steps in and
my single machine gets replaced with another. I have no authority or control
Let's see. That means that you are offering the Root du Jour. And
tomorrow, it may be someone else. Personally, I rather have a system that
ensures rather more stability for the DNS root and TLD service than that.
Nonsense. The TLD namespace IS global. There is nothing preventing non-US
interests from registering TLDs, and in fact more than one has (proof positive
that this statement is ALSO false). There are currently registrars in
Germany
and Japan -- pretty much opposite "ends" of the world.
Well, since I've lost track of the number of "how dare you take
this outside the US" messages you've sent, your above declaration comes as
a bit of a surprise.
d/
So what?
The registrar claims you haven't paid and didn't submit the non-repudiable
transaction (because they claim it didn't happen). You have a cancelled
check.
This is precisely the situation people are claiming is happening to them.
Again, the problem isn't payments being posted that didn't happen, its
payments made which *didn't get posted*. Non-repudiation doesn't help
this situation; that's a control on *positive* events, not ones which
people claim didn't occur.
The choice of data base technology for the shared repsoitories will
be the choice of CORE, i.e., the registrars themselves. I believe it is a
matter strictly of operational issues, rather than policy. I suppose one
could imagine some bizarre decision by CORE which challenges policy issues,
but I doubt it.
Rick Wesson has organized a BOF for the Memphis IETF, to consider
requirements and technical issues, all of which are intended to jump-start
the background work for CORE.
d/
>eDNS enables *all* business models for registration of TLDs and SLDs.
as I said, it sets up monopolies.
No it doesn't. eDNS sets up NO business model. *RA*s set up business
models. Again, you are confusing things deliberately.
The eDNS model does not mandate, or prohibit, any model EXCEPT one which
seeks to stop other models from being born and tested in a free
marketplace. eDNS eschews monopolization of business models.
McDonalds does not have a "monopoly". They have a *BRAND*. You like to use
that word because it is emotionally charged and you get a "kick" from it
when you use it.
But the fact is that a company which develops a particular brand has certain
rights which come along with that development. McDonalds can stop people
from selling "Big Macs", unless they pay the appropriate license fee and
adhere to their standards.
We don't believe this is "bad" in any other line of work. In fact, the
United States and virtually every other country in the world honors these
principles.
Do you claim that McDonalds has a monopoly? Or do they have a brand of
hamburger?
>Its tough to tell the person who pays the check every day "no". Very, very
>difficult.Now, you see. That's interesting, because I don't have any trouble
imagining Paul say no, even to someone who has been giving him money.
Really? Well, let's see. So far they have all said no. So far NSI has
maintained the monopoly. So far *NSI* has not bought off on the IAHC model,
and in fact has issued press releases which pretty strongly indicate, at
least from how I read them, that they have no intention of doing so now or
in the future.
Yet the IANA roots remain closed.
>My accountability is simple. If I violate the process someone steps in and
>my single machine gets replaced with another. I have no authority or controlLet's see. That means that you are offering the Root du Jour. And
tomorrow, it may be someone else. Personally, I rather have a system that
ensures rather more stability for the DNS root and TLD service than that.
On the contrary. eDNS is a process. It is not a person. It will survive
if I get hit by a bus tomorrow, because the *process* is valid. It will
survive if I turn rogue tomorrow for the same reason.
eDNS isn't Karl Denninger. Its a model for recognition of the development
of competing models in a free marketplace of TLDs.
If the "all shared" model is the best one, then it will win on its own.
Nobody has to force it on anybody. The "brands" which aren't controlled by
huge numbers of registrars, all with equal access, will fail.
On their own.
I do *NOT* claim to be omniscient and know what is the "best" model. I *DO*
believe the market can figure that out for itself without my "help".
Meddling in what is fundamentally a free process inherently leads to
higher costs and poorer performance. History says that this is basically
always true, and I have no reason to believe that you, or anyone else,
myself included, is THAT good.
>Nonsense. The TLD namespace IS global. There is nothing preventing non-US
>interests from registering TLDs, and in fact more than one has (proof positive
>that this statement is ALSO false). There are currently registrars in
>Germany
>and Japan -- pretty much opposite "ends" of the world.Well, since I've lost track of the number of "how dare you take
this outside the US" messages you've sent, your above declaration comes as
a bit of a surprise.
Not in the least.
I've said that *I* want the right to register in a namespace which is
controlled by a US organization because if they screw me I want legal
recourse.
Others may not see it that way.
Under eDNS, they have that *choice*. Under the IAHC model, NOBODY gets to
make a free choice.
eDNS is about choice Dave. Its not about people, and its not about
dictators or monopolies. Its about users of the network choosing the
models of registration that they want, and the companies who provide
those models efficiently being the ones who "win" over time.
None of the other models can make that claim. All of them claim to know what
is best for everyone else. The arrogance displayed by the people making
those proclamations is, in many cases, transparent enough to see right
through -- most of those folks have quite a bit of self-interest driving
their conclusions, and the rest simply think they're smarter than everyone
else.
Virtually everyone who believes that of themself is eventually proven to be
foolish at best.
I like to use it because it fits. When a consumer is locked into
dependency on a particular vendor and does not have freedom to change
vendors, that vendor is in a monopoly position.
The difference between buying hamburgers and buying domain names is
that you are free to buy your next hamburger from someone else. While one
might counter that one is also free to buy the next domain name from
someone else, it ignores the continuing dependency for names already
purchased.
If the exclusive vendor of a name chooses to triple the fee next
year, the consumer is stuck. The cost of changing domain names, after
putting marketing collateral development and building up their own brand
equity in the domain name, is onerous.
d/
If CORE decides to hike the maintenance fee in Year 2, 3 or 4, all
registrars will have to pass that fixed cost of doing business on to their
customers, and not only are you stuck if you stay with one vendor, you're
stuck if you *MOVE*!
Ugh! Please, please, please, take this to another list?
CORE is required to be strictly cost-recovery. CORE is a creature
of the registrars, in the aggregate AND is subject to the POC which means
that there is public oversight. Whimsical and excessive increases in
charges won't happen.
eDNS cannot make any such assurances about domains subject to
monopoly control.
d/
ps. sorry, folks. I know you want this off the list, but misinformation
begs a response.