DNSO Mailling lists archives


<<< Chronological Index >>>    <<< Thread Index >>>

RE: [ga] FYI: Staff Draft towards Mission Statement

Thank you Karl, for the validation.

For those of you on this list who's native language is not english, and
who may not be familiar with the north american idioms of speech, my
choice of words might have been better.

I know that much, if not all, of the key technical work in the Internet
is done by volunteers. Yes, it is true that some of those volunteers
have their salatries paid for by large multinational corporations who
benefit from Internet expansion, such as Cisco, 3com, etc.

The IETF work, the protocol work, and even most of the people who come
to ICANN meetings to try the "great experiment" do so "pro bono" that
is, unpaid outside of their normal commercial jobs. 

My comments were meant to emphasise that, US 10 million dollars or not-
those volunteers and that work will still be volunteer. 

IETF members will not get "new money" for working on standards. Protocol
developers will not participate in the US 10 million, 

In other words, all the things that happened in the pre-ICANN days will
still go on as before. 

What we WILL get is a better-funded bureaucracy, with job and contract
guarantees for more people working at ICANN, and more lawyers getting
legal fees. This does not insure or assure any better or different level
of Internet activity.

I am not saying that this new, well funded bureaucracy is a "bad thing".
It is just a new layer.

As for root server services- there are a number of commercial firms
providing robust DNS for web hosting companies, registrars, large ISPs
like SPRINT, and even major multinational corporations like Microsoft,
who chose to outsource this function by contract with known costs and
quality of service guarantees.

Most of those commercial DNS companies do NOT use BIND, for security and
other reasons.

It would be a "good thing" for us to get some proposals from some of
those commercial firms just to gauge the cost level of a 99.999 or
greater quality of service guarantee.

I am not saying officially, wearing ANY of my professional hats and
titles, that US $ 10 million is "too much". As a businessman, I would
seek several quotes.

Personally speaking though- I bet we could get the job done for less
than US $ 5 million, maybe even US 2.5 million.

Peter de Blanc

-----Original Message-----
From: owner-ga@dnso.org [mailto:owner-ga@dnso.org] On Behalf Of Karl
Sent: Friday, March 08, 2002 4:40 PM
To: Peter de Blanc
Cc: 'Alexander Svensson'; ga@dnso.org
Subject: RE: [ga] FYI: Staff Draft towards Mission Statement

On Fri, 8 Mar 2002, Peter de Blanc wrote:

> It is truely amazing that something once  done by volunteers,

In fact much is *still* done by volunteers, such as the running of the
root servers and thus isn't even on the current ICANN budget.

ICANN management likes to inflate the complexity of what it is doing - 
Stuart Lynn, for example brewed up an utterly fanciful statement that it

would cost ICANN $10,000,000/year to run 12 root servers.  By my own 
estimates, estimates that I doubled just to be on the safe side, I came
with about $3,000,000 in one-time costs and the same in yearly recurring


(I was appalled at the finacial naivete of ICANN's management insofar as
they were unable even to come up with the costs that ICANN/IANA incurrs
today to operate its root server and they didn't even break the
projected costs down into the basic one-time vs recurring structure.)

The competing root systems manage to do the same job for less than a
couple of percentage points of Lynn's estimates.  And I can assure you,
being a user of competing roots, that the quality of service and failure
rate they deliver is no worse than that of the ICANN root and, in fact,
has been better.

When the competition can deliver an apparently equivalent product (in
case DNS root services, including root zone preparation and
for pennies on the dollar there are clearly some hard questions that
to be clearly asked and clearly answered.

ICANN mangement's self-written job description is reminiscent of a
resume from someone who is trying to depict a job flipping burgers as
some sort of high-powered management position.

ICANN's management is obviously engaged in a game of empire building.

> ... Frankly, I do not see the need to pay so much for so little. Like 
> there is not even a QoS guarantee on the operation of the root server 
> system.

I agree.  Christian Huitma noted a couple of years ago that many DNS
queries are lost due to server congestion.  And that, in turn,
translates to significant delays for users.

Most of us depend upon the quality and speed of DNS name resolution
services - if name resolution slows down then the whole perception of
net speed goes down as well.

For those of us who build businesses on the Internet, it is hard to make
quality committments to our customers when one of the primary
ingreedients of that service, root name services, is not only outside of
our direct control but is, in fact, run by a bunch of volunteers.

It was only a couple of weeks ago when ICANN's management brought to the

board a Memorandum of Understanding - not a contract but a MoU - that 
retained nearly all the elements of the volunteer aspect of root server 
operations.  ICANN's management strongly advocated and fought for this 
loose and vague structure.

And now, when it comes to justifying ICANN, suddenly ICANN's management 
does a flip-flop and decides that it now must take a firm grasp of the 
root servers - something that three weeks ago was anathema to them - at
outrageous cost of $10,000,000 year.

(I personally believe that ICANN ought to either directly run its root
servers or have firm, clearly enforceable operations contracts.  But the
dollar figure quoted by ICANN management, dollars which largely
translate into head count and bigger org charts, are vastly out of line
with any reality except that of an empire building bureaucrat or a fee
grasping law


This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html

This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html

<<< Chronological Index >>>    <<< Thread Index >>>