ISPCON - Telage summary
Jeff Williams
jwkckid1 at ix.netcom.com
Sun Aug 24 14:16:40 EDT 1997
Ellen and all,
Notice: Official response to follow.
From: Jeff Williams Director and co-Founder of IEG. INC.)
Subject: IEG. INC. Official response to Mr. Don Telage NSI
To: Fellow Particapants at ISPCON
My response Starts here intermixed with Mr. Telage's comments.
Ellen Rony wrote:
ISPCON report, part 1 - Great conference again. Excellent,
informative
presentations. Overheard that it had 4,000 attendees (unconfirmed
number).
Here is a fairly lengthy summary of the comments made by Don Telage on
Friday.8/22. I thought the conference room would be packed, but such
was
not the case. It seems that the ISPs are focused on the bottom line and
do
not yet realize how governance issues will affect them or why they
should
be involved. More work needs to be done to raise that awareness.
Re: Dr. Telage: I expected he might a person who has hair on his teeth.
Instead, he is articulate, informational, personable, reasonable,
accessible, warm and likeable. So here are notes from his talk. No
spell
checker. All comments below are from Dr. Telage, not the messenger
(well,
one annotation). Questions from audience in brackets. A few full, exact
quotes included. Will try to make time to post notes from
Mockapetris,
Kashpureff and DNRC sessions later. For NSI response to NOI, see August
18
NTIA posting, but it will also be posted on netsol web site,
_________________________________________________
>From Dr Telage:
NSI - "we're the company that everyone loves to hate"
3 objectives of talk:
1. Address the question of why are we where we are in this governance
issue? Why should it not be expected? Hindsight point of view.
2. Look at some of the controversial issues. Stress areas where there
are
some misunderstandings. Where NSI has a view, want to present our
opinion.
Information/clarification phase.
3. Speculate on where whole process going? What hints do we have on
where
it needs to go? What should we as a community do to get a good outcome
for
the Internet. Find an outcome that respects the business and lets each
of
the stakeholders continue their function. Future direction.
Policy changes that we make today are going to have impacts down the
road.
You should take an interest in what is being done with your Internet,
because it is your Internet, no one else's.
Internet was done way it was done because it served an effective
purpose.
Right model for the functions it had in the first 20 years.
Consensus-driven work. No top down. Global skunkworks type of
arrangement. Marvelous quality to it.
ISPs responsible for why it needs to change. ISPs brought tens of
thousands of customers who had different mind set and objectives of
prior
years. Brought commercial interests and priorties, conflicting
interests,
business priorities. and that is the reason it needs to change. its
function is now changed. We have the old role, but the consensus model
isn't going to work when big money and rapid time elements are involved,
when laws are involved, when countries feel a stake in, that is what is
generating the change. You generated an unbelievable commercial
explosion.
I don't think we are close to the top curve. Perspective so small in
time
scale. In five years, this curve will be seen as flat. Growth will be
greater soon than we can ever imagine. Growth relentless. We have to
make
it more industrial strength to get where we need to go.
Commercial operations have impacted the operational stuff at bowels of
Internet, root server operators, IP registries, changed business models,
too.
IANA, IAB, IESG vulnerable, not indemnified not protected. "It's almost
inappropriate that individuals or small collections of individuals who
happen to have an historical name to them can operate in a governance
role
anymore." Too much to ask of these individuals. They must feel the
strain.]
We're in this role and it feels very threatening to us. Postel must
feel
very vulnerable.
I don't think Jon Postal feels threatened at all, form some
conversations that I have had and heard of with Jon, he feel more
betrayed personaly and profesionaly. So it seems like MR. Telage,
is trying to shore some things up here with the IANA, IA and the
IESG. Intreating shift of emphisis in this remark. Food for
thought possibly. You be the judge.
Internet basically grew up in U.S. From a national point of view, with
exception of RIPE and APNIC, most of critical elements are still within
continental U.S. Until month ago, everyone of root servers were in the
U.s. Month ago shipped a second server offshore. 2/3 of traffic volume
is
within U.S. Internet global in extent, but in terms of balance of
infrastructure, topology and activity, it is still US-centric. We
(U.S.)
should take leadership role to fix things.
First of all there have been two Root servers outside the US for
more than a year now, so I see that Mr. Telage is a bit innaccurate
with his facts. This is astounding in a way. He should know
better than this.
I agree that the US should take a leadership role baised on
his very reasons in his next sentance, but not a dominating one.
Even though the most usace and infrastructure in in the US
currently, this will only be true for a small amount of
time. Once China and most of asia comes online in a more
significant way, our dominance will not be there. Hence,
the reason for not dominating the situation now. Be open,
accepting, encourgaging and inviting to the whole world.
By doing this we better ourselves perportionaly.
Re: Cooperative agreement. NSI feels it doesn't have a technical
direction
role. Agreement almost implies that IANA has the responsibility and we
have
always treated that to be the case. That's where policy decisions,
appeals are made to IANA, network numbering policies and identifiers
originate there.
NSF trying to get out of its role for 2 years. IATF wouldn't let NSF
out
last April. Not no, but "hell no. One semblance of stability, wouldn't
allow NSF to leave. Legal activity. NSI sued by PGMedia for antritrust
activity. IANA told NSI it has no authority in these matters, do what
you
want. NSF letter on June 10 with a plan to introduce competition into
the
DNS, please approve or modify and respond. NSF said no, not until US
govt
has chance to review this entire matter. Aug 11 NSF said no again.
Mr. Telage, makes a very good point here. In accordance with
the Telecommunications act of 1996, NSF is mandated to allow for
private enterprise to be a working partner, unincumbered reguardless
of size or stature in the commercial sector. This puts NSF's
Aug. 11th answer in contrast with the US congress and federal
Law. SO it would seem that NSF is indeed in abayance of
antitrust laws as well.
More intrigue to issue: Justice Department conducting investigation
into
NSI antitrust behavior. Even though we are operating under govt
contract,
andappealed to both what we think are the authorities and have been
denied
approval. Being sued, denied an ability to defend ourselves (NSI), and
being investigated by Justice Department. Not well informed District
Court
judge could order a solution tomorrow which would overrrule the
contract.
Not the way to do it. Needs orderly process.
Well NSI and NSF had had more than a year to develope that process,
to little avail. Why? Becouse it is not in NSI's intrest? Maybe?
OR is it due to poor managment a feet dragging? Likely? I don't
pretend to know. Reguardless, if I know that I or We (IEG. INC.)
can do it, than surely NSI and the NSF could have or already should
have.
Airing dirty laundry. Recent snafu on Internet. Let bad root file go
out.
Tried to notify them. Couldn't notify root server operators because
they
are volunteers. Whole structure needs higher performance standers.
Needs
7x24 accountability. No fail safe in the system. That whole structure
needs to be put more under control.
I agree. And this should have been done more than two years ago.
Many warned, to no avail. So I guess, "Those who fail to manage,
manage to fail".
DNS Advantages: terrifically easy to use. Large name space. Easy to
understand. More self funded. Resilient to growth.
Disadvantages: Open to spoofing. Was a peer to peer system, not
designed
to deal with hostile activities.
DNS security is something that is already under development by
the IETF. In addition, this problem was and should have been known
long ago. Many, including myself warned of this weakness, and were
shruged off as being excentric. Well I guess, again failier to
realize the obvious is always a rude awakener.
Directory system is sorely lacking. Need to figure out how to get
information easily. Average user can't get WHOIS for Germany, etc.
Needs
to be put together in some unified way.
No uniform registration policies worldwide. Some treat it as holy grail
to
other extreme at NSI. We treat it as a brand, a registration identity.
The Internet needs to figure out what domains are, what registration is
and
how it should be dealt with registration in a uniform way.
This again has ben known or suspected for over two years now. Yet
NSI and NSF are just now realizing it. How short sited, which again
points out the poor judgment of NSI's managment and NSF's oversite.
Registries have to operate in legal vacuum are hugely vulnerable. In
order
to protect selves have to put in policies aren't popular. But
alternative,
with no indemnification from any legal body, risks going out of
business.
Other registries haven't hit volume points so haven't seen this yet.
They
will, soon.
"NSI supports competition of any form that the community decides is
necessary, either through the addition of new TLDS or through the
imposition of a shared registration concept if it is done with stability
and is done orderly and it respects the legacy positions of the
registries."
This is a very serious problem indeed. Replacment of the
lagacy DNS software is paramount to meet future demand. It
is still not too late to be proactive here, but somehow I
doubt this will happen in light of past history.
"Sharing .com, sharing .de, sharing .net, sharing .uk, that's not a
problem. Not easy to do, but we support it.
Relative to adding new TLDs, no technical limit. Not issue. Issue is
possible confusion in Internet space, implementation guidelines, pushing
the problem up one level. Community not in agreement on this. 3 camps:
1) want to get into business, want new TLDS ; 2) trademark owners or
those who got name , want no more TLDS, see this as intellectual
property
nightmare for them.; 3) group lost out in court battle and still lusts
after name...wants new TLD so they another shot at the name and
procedure.
This is where the curx of many of the discussions have centered
for nearly a year and a half now. I agree that there is no limit
to the possibilities of TLD's that could be created if a shared
system of registries in a Round Table fassion is accepted or
in place. This is already underway. SOme call these services
"Alternitive Registries". Alternitive or not, they exist, and
are not likely to go away. SO the IANA/IETF/ISOC/POC/CORE/IAB
need to find a way to work WITH these RSC's, not Flame them.
We have a long way to go here.
"Competition will one of the most important things that this industry
needs: it will induce people to invest."
NSI has invested 10s of millions of dollars into new facility. 34,000
sq
ft, state of the art facility.
Key question: do we want to make a change of DNS system that only
affects
top level. It is hierarchical system. Parallel in IP level,
portability.
Technically difficult to do. Issues of routability.
Well I agree. And first on that list is a modification or update
to RFC2050, RFC1917, RFC1918. In doint this in a open manner,
and getting on with either IPv6 or IPv8, we can shoot alot of
these problems very quickly. This doesn't seem to be the way
ARIN and the IANA seem to want to head however. So openess
and providing for the expandability in usefull way does not
seem to be in their intrest. One must question this logic.
If you do sharing, you will end up building another back office
facility.
That facility will wind up being a monopoly. Somewhere, all of this
stuff
needs to come together for reconciliation. Creates a wholesale monopoly
back office and retail front office.
I don't agree that the back office facility will end up being
an monopoly at all. Most of my coleague's do not either. This
sounds like another "Scare tactic" for some power freeks to control
as much of the central parts of the internet as they can and use
the ole "Medicare scare Tactic" argument to get tit done.
Whole issue of what a gTLD is has always baffled me. Policy change
should
affect all elements in that class. 60 TLDS where you can register no
matter you live or where your host is. What's the distinction between a
gTLD and one of these guys? Volume.
[comment from Rick Wesson....historical difference;
gTLDs do not have a historic value and do not have association with
ISO 3166]
Just above Mr. Telage stated that there is no limit to the
number of TLD's possible. He seems to almost conterdict himself
here with saying that 60 will allow you to register no matter
where you live or where your host is. This may be so, but does
not answer the brending concerns.
.COM is as big as it is for 2 reasons. .US not widely accepted. Ugly
ducking of TLDS. Growth in Internet has been in US and it doesnt have a
popular ISO 3166. .COM filled void.
Problem is volume, not geography. When other TLDs start experiencing
growth, they will have same problems.
Hence the need to offer alternitives. As many as there is a
precieved demand for. Not to mention splitting the load.
Re registration pricing graphs: Tremendous variability. Charge enuf to
make reasonable profit to invest and improve offerings. Other
registries
need to look on this as investment oriented activity.
I agree. And this is something that will need to be an ongoing
experiment. But I suspect as demand grows so will price, as well
as costs. Hence the need to act now.
[Q: could Poland start a new TLD?]
No, wouldn't get included in 13 root servers.
Were not taking any new applications based on NSF's directions.
Bad idea! Contredictory to his previous statments on this
critique as well. Mr. Telage seems to have no real direction
here, nor any suggestions as to what possible directions he
would like to see things go.
We have a fundamental problem with iPOC initiative. We feel changes
need
to be evolutionary. Stakes too high to do anything too radical until
we know impacts. Need managed transition period top make necessary
changes to stabilize infrastructure. Shipping the problem to
inexperienced
international body where complexity goes up is not a wise move. Needs
to
be market driven at some level, needs to be internationally governed,
needs
to be industrial strength. Needs period of transition to evolve to
another
body, some U.S. agency that has skills set perhaps.
Sounds like he is for NSF or some other government agency to
step in here. Regulation? Lets hope that is a misevaluation
on my part.
Broad objectives for transition: Overarching issues:
1) Establish legal framework. Who is in charge? What is in charge?
Everyone ducks the bullet.
We the stakeholders are in charge, all over the world. The
internet users is what is in charge, both commercial and otherwise.
2) Need to bring all the competition into the space. Only way will
spur
investment required to raise registries to performance levels to grow
Internet globally. Profit will motivate. Things that require investment
will respond to profit.
3) Fill in DNS trademark vacuum. How do domain names and trademarks
relate?
Must fix these first before moving whole infrastructure offshore. Will
take 18-24 months.
Too long. Not that big of a problem. Usage of international
Tradmark law should do it just fine. Make DOmain names a
tradmarkable entitity.
NSI PROPOSAL: 4-layer chart.
Legal layer - US government. Allows registry to derive authority from
someone. Similar to N.A. Numbering Council has same model. Intl
participation from stakeholders, does policy development.
Governance layer - IPAG (international policy advisory gorup) intl body
which accepts membership from interested stakeholder globally;
membership
based on usage or volume representation.
operational layer - manage critical functions. nonprofit business. not
highly capitalized. Need DNS manager: resp for managing the root, the
dot, failsafe operations. Should move from us and other volunteers.
Analog on DNS side to IP side.
business layer - litkely registries and ISPs willb e the same. capital
intensive. customer interface. needs incentive for investment. profit
layer.
To many layers here to be efectual and reactive to global stake
holders
needs not and certianly in the future. More streamlining is
needed. Remove unneeded redundance. WOrk this like a buisness,
not like a social club, from a structual point of view.
This process will never be affective or work long term. Just
creates a bunch of bearuacrats.
Cutover from us agency to international is easy if 4 layers are in
place.
"We want to be a registrar. That's all we want to be.'
Not realistic statment baised on this layering structure in this
statment.
Broad rush questions: How do we view names? Are DNS space and
trademarks
the same? What is business model that is needed at each layer? Does
model
we have chosen work operationally? What's the governance model here
that I
community wants?
Registries in a bind. No immunity from suit but must protect selves
from
financial instinction. Cannot wilfully and knowingly infringe
trademarks.
We registered 125K names last month. With no policy in place, we would
have gotten thousands of suits. As is, 30+ lawsuits naming us. 1,300
written disputes on 1.6 million registrations. Problem doesn't need to
be
solved by draconian means such as trademark pre-screening. Work on the
exceptions.
Good point here. I agree.
[remaining comments in response to audience questions]
NSI can't do a prepayment policy. Too many different payment
arrangements.
ISPs need to pay by check; companies need purchase orders.
I don't agree with this at all. But I understand his point. A
need to automate payment parctices is paramount to good financial
tracking and managment in todays electornic age. Old payment
methods can still be used, but should not be manditory. This
makes fudiciary responsibility to cumbersom for many and
very expensive as well. This is also anti small buisness.
Dispute policy prevents lawsuits for "contributory infringement".
Following advice of attorneys. Every domain registrant has signed a 2
year
contract. Cannot be prorated. If registrant wants to go another
registrar
when that becomes available, that is your choice, but you have signed a
2
year contract (my note: which means your money won't be refunded). This
is
a common business practice.
Good point. But what is in that two year contract. I hope it
is not like the gTLD-MoU contract. SIgning theirs IS NOT GOOD
BUISNESS! So I think the real story in this comment is in the
fine print.
In hindsight, NSI should have written the cooperative agreement
differently
so that NSF indemnified NSI from all policy related matters.
-end Telage comments-
Hindsite hell, NSI was told this by many befor they singed that
contract! <rofl> Pass the buck again, I see.
- End of our comments and official response -
--
Jeffrey A. Williams
DIR. Internet Network Eng/SR. Java Development Eng.
Information Eng. Group. IEG. INC.
Phone :913-294-2375 (v-office)
E-Mail jwkckid1 at ix.netcom.com
More information about the Naipr
mailing list