Re: Addresses and ports and taxes -- oh my!

[Dennis Glatting]

On Fri, 4 Aug 2000, Stephen Sprunk wrote:

> Thus spake "Mahadevan Iyer" <[EMAIL PROTECTED]>
> > At first glance, it seems sheer idiocy to use an open network like the
> > Internet to control critical matter-of-life-and-death public
> > infrastructure like power systems. What do you think?
> 
> Public power systems are not life-and-death.  Anywhere that AC service
> is even moderately important, the people responsible (assuming they're
> competent) will have contingencies for power loss.  Consider the ISPs
> that *voluntarily* go off-grid when there's a power crisis in the SF
> Bay.  Hospitals also make an interesting study.
> 

Food for thought (copied from Risks):

--

Date: Wed, 12 Jul 2000 18:18:53 -0500
From: Bill Higgins-- Beam Jockey <[EMAIL PROTECTED]>
Subject: Illinois man dies after utility cuts power

I found the following story at the *Chicago Sun-Times*. 


> Man dies after ComEd cuts power 
> 
> July 12, 2000
> 
> BY DAN ROZEK AND STEVE WARMBIR SUBURBAN REPORTERS 
> 
> An elderly Aurora man who used an electrically powered oxygen system
> to help him breathe died in his home several hours after ComEd shut
> off the power because he was behind in his bills.

In Aurora, Illinois, Eric Shackelford, an 81-year-old man, used oxygen 24
hours a day to help him breathe; he suffered from "severe heart disease."
His daughter, Renia Thomas of Chicago, claims that the power cutoff shut
his oxygen down, and may bring a wrongful-death lawsuit against the power
company, Commonwealth Edison.

The story reports, however, that a roommate says Shackelford had two
oxygen systems, one of which did not depend on electrical power.

The RISKS relevance is in the dispute over record-keeping.  The family
says that Shackelford's doctor had sent at least two letters to ComEd
asking that power not be shut off.

> A ComEd spokesman, however, said the utility had never received enough
> information to determine that Shackelford was entitled to be added to
> a list of about 1,000 customers who needed continuous electric power
> for medical equipment. ComEd files contain only one letter from a
> doctor regarding Shackelford, ComEd spokesman Don Kirchoffner said.
> 
> "We would never, ever cut the power to anyone we thought was on life
> support," Kirchoffner said.  [...]
> A final notice sent in June said
> Shackelford should notify ComEd if he had medical equipment that
> required electricity, and there's no record anyone contacted the
> utility, Kirchoffner said.  [...]
> Kane County Coroner David Moore said it was unclear whether the power
> shutdown caused or contributed to Shackelford's death.

It would be interesting to know more about the process by which a power
company keeps track of customers who are dependent on power.  How do you
make such a process fail-safe?

Bill Higgins  Fermi National Accelerator Laboratory  <[EMAIL PROTECTED]>

--

Re: Addresses and ports and taxes -- oh my!

[Robert Elz]

Date:Thu, 3 Aug 2000 13:53:25 -0500 (CDT)
From:Tim Salo <[EMAIL PROTECTED]>
Message-ID:  <[EMAIL PROTECTED]>

  | Geeks like us care about end-to-end transparency.  Refrigerator's don't.

Refrigerators don't care about connectivity at all.   They don't care
about anything at all.   Refrigerator owners care about connectivity,
and even geeks can be refrigerator owners.   Personally I want my
refrigerator to have end to end transparency (I'd care less about the
toaster, doing anything useful to that when not physically close is
harder to imagine).

  | And, if IPvX addresses cost money, a lot of households will pay money
  | for devices that enable them to operate with only one IP address.

True.   However there's no reason to assume that it will be possible to
get less than a (very large really) block of addresses, so if you pay for
the one, you're going to get a lot to go along with it.

There's no question but that address assignment and management costs, and
needs to be paid for, and as with anything, it is the end consumers who
pay in the end (everyone else just passes along the costs, whether they're
separately listed, or just bundled in an overall price for service).
However, with IPv6 there's no longer the need for the rationing push on
prices that exists in IPv4.  It costs a certain amount to perform an
assignment (and record it, etc) - it doesn't really matter how big the
assignment is, with IPv6 there should be no differentiated costs for
address assignments based upon size.

kre

Re: Addresses and ports and taxes -- oh my!

[Anthony Atkielski]

> They have not been overlooked by those who have been
> working on IPv6 address allocation policy.

What's the solution?  Hint:  No policy of advance address allocation will
help, and neither will any form of address-based routing, no matter how
clever.

Re: Addresses and ports and taxes -- oh my!

[Stephen Sprunk]

Thus spake "Mahadevan Iyer" <[EMAIL PROTECTED]>
> At first glance, it seems sheer idiocy to use an open network like the
> Internet to control critical matter-of-life-and-death public
> infrastructure like power systems. What do you think?

Public power systems are not life-and-death.  Anywhere that AC service
is even moderately important, the people responsible (assuming they're
competent) will have contingencies for power loss.  Consider the ISPs
that *voluntarily* go off-grid when there's a power crisis in the SF
Bay.  Hospitals also make an interesting study.

Additionally, just because two IP hosts are both connected to the Net
doesn't mean that they must use the Net to communicate.  Not all IP
networks are public.

> Or do you think, it is possible to build ultra-reliable secure
real-time
> communication channels in the Internet? Maybe..

A reasonable degree of cynicism tells us that this will be impossible;
"secure" and "real-time" I might buy, but the state of inter-provider
cooperation will need to advance significantly before I'll believe any
Net-based service will even approach "reliable".

S

 |  | Stephen Sprunk, K5SSS, CCIE #3723
:|::|:Network Design Consultant, GSE
   :|||:  :|||:   14875 Landmark Blvd #400; Dallas, TX
.:|||:..:|||:.Email: [EMAIL PROTECTED]

Re: Addresses and ports and taxes -- oh my!

[Steve Deering]

At 8:36 AM +0200 8/4/00, Anthony Atkielski wrote:
>I think we'll see IP addressable toasters and washing machines just after we
>all switch from automobiles to hovercars and from telephones to
>Picturephones.  According to predictions being made by futurists for the
>past few decades, all of these things are due to happen Real Soon Now.

The mere fact that some predicted technologies never materialized does not
mean that all predicted technologies will fail to materialize.

Steve

Re: Addresses and ports and taxes -- oh my!

[Steve Deering]

At 8:49 AM +0200 8/4/00, Anthony Atkielski wrote:
>Not relevant.  IPv6 will be exhausted by overly-generous allocation of
>address space, just like IPv4.  I've already explained in the past why this
>must be so.  In part, it comes from the subjective impression that any new
>address space is "more than we'll ever need" and the tendency to
>overallocate in consequence, until it's too late; this is probably the most
>common engineering mistake in IT history.  Another reason why IPv6 is not
>nearly as large as it appears to be is that IP addresses are closely linked
>to routing, instead of being randomly assigned; and this routing imposes
>severe restrictions on how the address space can be used.  Both issues are
>routinely and completely overlooked,...

They have not been overlooked by those who have been working on IPv6
address allocation policy.

Steve

Re: Addresses and ports and taxes -- oh my!

[Jim_Stephenson-Dunn]

I believe that we are looking at a wireless solution here, and the justification
for building alternate dimension networks could be difficult to justify unless
we have an overcrowding problem on earth. That in turn would open up all sorts
of other possibilities. You can see the advertisements now :

"Moving into a different reality, don't end up in the Dungeon Dimensions, Talk
to John Doe, you friendly neighbourhood inter-dimentional Realtor"

However the Network Engineers that implemented it, would make a killing, They
would be on the clock in several dimensions simultaneously, especially if you
think that you can have different dimensions occupying the same physical space.
this scenario presents interesting opportunities :

1. you could carry out a full days work by doing 1 hour in each of 8 realities,
OR

really bust the budget and

2. by working 8 hours in 8 different realities simultaneously, therefore
achieving 64 hours work in one work day (8 hours)
(1/3 of the time it takes for an orbital rotation of earth, assuming 24 hours is
the time it takes,( I believe the completely accurate time is 23h 56m to one
rotation))
and lets be honest here, how many Engineers do you know that work 8 hour days ?

3. based on the standard working model, in the course of one "normal" working
day (8 hours), you could do your standard 40 hour week plus 24 hours overtime
and take the rest of the week off to go hiking in the mountains.

The added double whammy is that in the state of California, anything over 40
hours is time and a half, so the bank balance looks healthy too. The only
possible problem is whether you can be considered to be in California and
simultaneously in 7 other dimensions and still charge time and a half.

The lawyers and politicians are going to have fun making a law for this work
ethic.

ah... Mr. President, welcome to quantum physics 101 ;->


Jim


***
My opinions are my own and do not represent the technical direction of 3Com or
any of it's subsidiaries

***





[EMAIL PROTECTED] on 04/08/2000 05:58:52

Sent by:  [EMAIL PROTECTED]


To:   Jim Stephenson-Dunn/C/HQ/3Com
cc:   [EMAIL PROTECTED]
Subject:  Re: Addresses and ports and taxes -- oh my!



On Thu, 03 Aug 2000 16:52:25 PDT, [EMAIL PROTECTED] said:
> why not consider all the dimentions, ever heard of polyfractal space ?

Fabricating the router connections would be interesting  WHat sort of
crimping tool would it take to make a 2.75D connector stay on the cable? ;)



 att1.unk

RE: Addresses and ports and taxes -- oh my!

[Parkinson, Jonathan]

>all we need is programmers that remember to actually CHECK that string
>lengths are in bounds.

Lets hope they are X-Microsoft programmers, That should keep all us support
staff busy and in secure jobs ;-)


I know, Its Friday !

Jon

-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Friday, August 04, 2000 5:35 PM
To: Mahadevan Iyer
Cc: [EMAIL PROTECTED]
Subject: Re: Addresses and ports and taxes -- oh my! 


On Fri, 04 Aug 2000 01:26:55 PDT, Mahadevan Iyer said:
> At first glance, it seems sheer idiocy to use an open network like the
> Internet to control critical matter-of-life-and-death public
> infrastructure like power systems. What do you think?

At first glance, it seems sheer idiocy to use something like a telco switch
to control critical matter-of-life-and-death public infrastructure like the
US 911 system.

Telco switches are hackable.  And I submit to you that the 911 system is
as life-and-death critical as it gets.  Hasn't seemed to have been a
problem so far, even though 911 systems *have* been hacked, subjected
to denial of service attacks, and all the other problems they are subject
to.

> Or do you think, it is possible to build ultra-reliable secure real-time
> communication channels in the Internet? Maybe..

It may be impossible to build ultra-reliable secure systems.  On the other
hand, remember that it's about *risk management*.  Nuclear launch codes
are one of the *very* few "zero failures acceptable" things we have.

We accept that on the order of 1 out of every million airplane takeoffs
ends badly.  We accept that the power grid fails in scattered areas
during the summer.  We accept that doctors, drug interactions,  and
hospitals accidentally kill a number of patients every year.  I don't
see the whole class of Aleve/Naprosyn painkillers being pulled off the
market, even though an amazing number of people die every year from
gastric bleeding.

We actually know almost all of what we need to build such systems. Now
all we need is programmers that remember to actually CHECK that string
lengths are in bounds. ;)

Valdis Kletnieks
Operating Systems Analyst
Virginia Tech

Re: Addresses and ports and taxes -- oh my!

[Valdis . Kletnieks]

On Fri, 04 Aug 2000 01:26:55 PDT, Mahadevan Iyer said:
> At first glance, it seems sheer idiocy to use an open network like the
> Internet to control critical matter-of-life-and-death public
> infrastructure like power systems. What do you think?

At first glance, it seems sheer idiocy to use something like a telco switch
to control critical matter-of-life-and-death public infrastructure like the
US 911 system.

Telco switches are hackable.  And I submit to you that the 911 system is
as life-and-death critical as it gets.  Hasn't seemed to have been a
problem so far, even though 911 systems *have* been hacked, subjected
to denial of service attacks, and all the other problems they are subject to.

> Or do you think, it is possible to build ultra-reliable secure real-time
> communication channels in the Internet? Maybe..

It may be impossible to build ultra-reliable secure systems.  On the other
hand, remember that it's about *risk management*.  Nuclear launch codes
are one of the *very* few "zero failures acceptable" things we have.

We accept that on the order of 1 out of every million airplane takeoffs
ends badly.  We accept that the power grid fails in scattered areas
during the summer.  We accept that doctors, drug interactions,  and
hospitals accidentally kill a number of patients every year.  I don't
see the whole class of Aleve/Naprosyn painkillers being pulled off the
market, even though an amazing number of people die every year from
gastric bleeding.

We actually know almost all of what we need to build such systems. Now
all we need is programmers that remember to actually CHECK that string
lengths are in bounds. ;)

Valdis Kletnieks
Operating Systems Analyst
Virginia Tech

RE: Addresses and ports and taxes -- oh my!

[Brijesh Kumar]

> -Original Message-
> From: Anthony Atkielski [mailto:[EMAIL PROTECTED]]
>
> Mahadevan Iyer writes:
>
> > At first glance, it seems sheer idiocy to use an open
> > network like the Internet to control critical matter=
> > of-life-and-death public infrastructure like power
> > systems. What do you think?
>
> I think there are lots of idiots out there preparing to do
> exactly this.

My logics have failed me due to the above text: which of the following
you mean.

  1) Any one who is preparing to do above will belong to a set called
"Idiots",
  2) Only people who already belong to a set called "Idiots" are
preparing to do it.

Hey - don't take it seriously it is friday.

On a change note, what about the use of secure tunnels using IPSEC?
Won't that solve problem. Perhaps not in all cases.

I think you need - security, performance guarantee, network path
reliability and ability to control restoration paths. Security is only
one dimension to the issue of who controls the end to end delivery
system in critical applications. It is hard to be sure of an event
outcome if you are not in control of all variables that can affect the
system. Basic control theory. Do you think people who design these
systems flunked their control theory course (may be they concentrated
too much on computer science :-))?

Cheers,

--brijesh
Ennovate Networks Inc.

RE: Addresses and ports and taxes -- oh my!

[Parkinson, Jonathan]

The world connected
Hmm Now I'm thinking Virus's

/Jon/div

-Original Message-
From: Evstiounin, Mikhail [mailto:[EMAIL PROTECTED]]
Sent: Friday, August 04, 2000 4:08 PM
To: Steven Cotton; [EMAIL PROTECTED]
Subject: RE: Addresses and ports and taxes -- oh my!


Toaster is much more quite, even it takes more time. And all your mail will
have brown-gold colour, while in blender you get everything mixed up:-)

> -Original Message-
> From: Steven Cotton [SMTP:[EMAIL PROTECTED]]
> Sent: Friday, August 04, 2000 5:05 AM
> To:   [EMAIL PROTECTED]
> Subject:  Re: Addresses and ports and taxes -- oh my!
> 
> On Fri, 4 Aug 2000, Anthony Atkielski wrote:
> 
> > The mere fact that something is technically possible doesn't mean that
> it
> > should be done.
> 
> Definitely - what benefit can I get from my toaster having Internet
> conectivity when I will be able to use my blender to read mail?
> 
> -- 
> steven
> 

Re: Addresses and ports and taxes -- oh my!

[Vernon Schryver]

> From: =?iso-8859-1?Q?M=E5ns?= Nilsson <[EMAIL PROTECTED]>

> > good point... but I do wonder how the border edge
> > router will handle a datagram with
> > TTL approx > 240 sec's
> > ( i.e min time required for msg to pass between earth <=> mars) ?
> > what about jitters, latency ,dropped packets, icmpv6 err msg well
> > whatever
>
> Well, back to UUCP, then :-) *nearly just kidding*

Which of the UUCP protocols would tolerate such RTT's?
Don't you remember why Telebit had to add "UUCP spoofing"?
The g protocol had a pretty small maximum sized window.
of course e and t don't count since they assume a loss-free wire.

TCP with large windows already works where UUCP never did.

That's not to disagree with the notion the UUCP has a lot to recommend
itself even today, or that you couldn't make a new and different UUCP
protocol that would tolerate the RTT to Pluto.


Vernon Schryver[EMAIL PROTECTED]

Re: Addresses and ports and taxes -- oh my!

[Andre-John Mas]

Dennis Glatting wrote:
> 
> On Fri, 4 Aug 2000, Andre-John Mas wrote:
> 
[SNIP]
> >
> > Though if the devices already use ethernet, then each device would
> > already have its own MAC address, and the IP address would be DHCP
> > assignable. As stated earlier once firewall/routers/DHCP server combos
> > come down in price, preferbly below $200, you will probably see these
> > devices as becoming standard. These devices would provide a level of
> > intrusion security and would be able to assign address to the devices
> > in the home, as is already done in most LANs.
> >
> 
> Firewalls do indeed provide a level of security but they are also
> vulnerable to attack and code and configuration entropy -- there are many
> examples of this. Also, if you have a trusted party managing your HVAC
> system and that trusted part is cracked, a firewall will probably provide
> no defense.
> 

I suppose what's needed here is some sort of EPROM, that stores an image
of
the system, and re-copies that image to main memory every day. The EPROM
could only be written to by pressing a hardware switch, otherwise it is
read-only. Since there are Linux based firewall solutions that fit on a
floppy, this sort of approach shouldn't be too difficult.

Andre

RE: Addresses and ports and taxes -- oh my!

[Evstiounin, Mikhail]

Toaster is much more quite, even it takes more time. And all your mail will
have brown-gold colour, while in blender you get everything mixed up:-)

> -Original Message-
> From: Steven Cotton [SMTP:[EMAIL PROTECTED]]
> Sent: Friday, August 04, 2000 5:05 AM
> To:   [EMAIL PROTECTED]
> Subject:  Re: Addresses and ports and taxes -- oh my!
> 
> On Fri, 4 Aug 2000, Anthony Atkielski wrote:
> 
> > The mere fact that something is technically possible doesn't mean that
> it
> > should be done.
> 
> Definitely - what benefit can I get from my toaster having Internet
> conectivity when I will be able to use my blender to read mail?
> 
> -- 
> steven
> 

Re: Addresses and ports and taxes -- oh my!

[Anthony Atkielski]

Mahadevan Iyer writes:

> At first glance, it seems sheer idiocy to use an open
> network like the Internet to control critical matter=
> of-life-and-death public infrastructure like power
> systems. What do you think?

I think there are lots of idiots out there preparing to do exactly this.

Re: Addresses and ports and taxes -- oh my!

[Dennis Glatting]

On Fri, 4 Aug 2000, Andre-John Mas wrote:

> 
> 
> Dennis Glatting wrote:
> > 
> > On Thu, 3 Aug 2000, Keith Moore wrote:
> > 
> > [snip]
> > 
> > > burning IP addresses into devices is a good way to give vendors the
> > > ability to control those devices, monitor their usage, and to lock
> > > their customers in to particular services.  not my idea of a desirable
> > > state.
> > >
> > 
> > It might also be a good way for script kiddies to efficiently scan the
> > Internet looking for a particular manufacturer's device to exploit a
> > discovered security flaw, such as turning off a stove's gas pilot and
> > turning on all burners. If that doesn't sound realistic, how about a
> > cracker inside a manufacturer's systems doing the same; or how about a
> > terrorist?
> 
> Though if the devices already use ethernet, then each device would
> already have its own MAC address, and the IP address would be DHCP
> assignable. As stated earlier once firewall/routers/DHCP server combos
> come down in price, preferbly below $200, you will probably see these
> devices as becoming standard. These devices would provide a level of
> intrusion security and would be able to assign address to the devices
> in the home, as is already done in most LANs.
> 

Firewalls do indeed provide a level of security but they are also
vulnerable to attack and code and configuration entropy -- there are many
examples of this. Also, if you have a trusted party managing your HVAC
system and that trusted part is cracked, a firewall will probably provide
no defense.


 

Re: Addresses and ports and taxes -- oh my!

[Måns Nilsson]

"Dawson, Peter D" wrote:
> 
> good point... but I do wonder how the border edge
> router will handle a datagram with
> TTL approx > 240 sec's
> ( i.e min time required for msg to pass between earth <=> mars) ?
> what about jitters, latency ,dropped packets, icmpv6 err msg well
> whatever

Well, back to UUCP, then :-) *nearly just kidding*

-- 
Måns NilssonDNS Technichian
+46 709 174 840 NIC-SE
+46 8 545 85 707MN1334-RIPE

Re: Addresses and ports and taxes -- oh my!

[Dennis Glatting]

On Fri, 4 Aug 2000, Mahadevan Iyer wrote:

> At first glance, it seems sheer idiocy to use an open network like the
> Internet to control critical matter-of-life-and-death public
> infrastructure like power systems. What do you think?
> 

I believe:

 * Engineers will build any possible feature into a
   product, install back doors, and not adequately 
   test their code. Of course complexity (IPsec, 
   policies, etc.) will probably not make adequate
   testing possible.
 * Marketing to demand all kinds of crazy features
   and convince the public at large they need them.
 * Lawyers to write liability limitations and 
   unintelligible wording into warrantee and 
   purchase contracts.
 * Consumers and providers to not properly configure
   devices and maintain them.
 * Congress (USA) to pass liability limiting laws.
 * Crackers to find holes and post cracking scripts.

Today we have remote control (not over the Internet, AFAIK) of alarm
systems and HVAC systems. The Internet provides a far greater reach and a
more cost effective model of managing these and other systems. So, I
believe these and many other systems will eventually find themselves
connected to the Internet. Additionally, there have been several articles
in the popular press in the past few years of the advertising community
salivating at the idea of displaying advertisements on home appliances
via the Internet -- imagine the "free Internet" equivalent of a "free
dishwasher."


> Or do you think, it is possible to build ultra-reliable secure
> real-time communication channels in the Internet? Maybe..
> 

No I do not believe so, but not so much for technological reasons; rather
for the reasons I previously listed.


>  
> 
> On Thu, 3 Aug 2000, Dennis Glatting wrote:
> 
> > 
> > 
> > On Thu, 3 Aug 2000, Keith Moore wrote:
> > 
> > [snip]
> > 
> > > burning IP addresses into devices is a good way to give vendors the
> > > ability to control those devices, monitor their usage, and to lock
> > > their customers in to particular services.  not my idea of a desirable
> > > state.
> > > 
> > 
> > It might also be a good way for script kiddies to efficiently scan the
> > Internet looking for a particular manufacturer's device to exploit a
> > discovered security flaw, such as turning off a stove's gas pilot and
> > turning on all burners. If that doesn't sound realistic, how about a
> > cracker inside a manufacturer's systems doing the same; or how about a
> > terrorist?
> > 
> > It is reasonable to assume that HVAC systems will someday soon be
> > controlled over the Internet by a maintenance firm (video surveillance
> > systems already are controlled and monitored over the Internet). It may
> > become possible, for example, to raise building temperatures across the
> > lower Manhattan area and shut down most financial centers, at least for a
> > short while.
> 
> 
> 
> 
> 
> 
> 

Re: Addresses and ports and taxes -- oh my!

[Andre-John Mas]

Dennis Glatting wrote:
> 
> On Thu, 3 Aug 2000, Keith Moore wrote:
> 
> [snip]
> 
> > burning IP addresses into devices is a good way to give vendors the
> > ability to control those devices, monitor their usage, and to lock
> > their customers in to particular services.  not my idea of a desirable
> > state.
> >
> 
> It might also be a good way for script kiddies to efficiently scan the
> Internet looking for a particular manufacturer's device to exploit a
> discovered security flaw, such as turning off a stove's gas pilot and
> turning on all burners. If that doesn't sound realistic, how about a
> cracker inside a manufacturer's systems doing the same; or how about a
> terrorist?

Though if the devices already use ethernet, then each device would
already
have its own MAC address, and the IP address would be DHCP assignable.
As
stated earlier once firewall/routers/DHCP server combos come down in
price,
preferbly below $200, you will probably see these devices as becoming
standard.
These devices would provide a level of intrusion security and would be
able
to assign address to the devices in the home, as is already done in most
LANs.

Andre

Re: Addresses and ports and taxes -- oh my!

[Steven Cotton]

On Fri, 4 Aug 2000, Anthony Atkielski wrote:

> The mere fact that something is technically possible doesn't mean that it
> should be done.

Definitely - what benefit can I get from my toaster having Internet
conectivity when I will be able to use my blender to read mail?

-- 
steven

Re: Addresses and ports and taxes -- oh my!

[Mahadevan Iyer]

At first glance, it seems sheer idiocy to use an open network like the
Internet to control critical matter-of-life-and-death public
infrastructure like power systems. What do you think?

Or do you think, it is possible to build ultra-reliable secure real-time
communication channels in the Internet? Maybe..

 

On Thu, 3 Aug 2000, Dennis Glatting wrote:

> 
> 
> On Thu, 3 Aug 2000, Keith Moore wrote:
> 
>   [snip]
> 
> > burning IP addresses into devices is a good way to give vendors the
> > ability to control those devices, monitor their usage, and to lock
> > their customers in to particular services.  not my idea of a desirable
> > state.
> > 
> 
> It might also be a good way for script kiddies to efficiently scan the
> Internet looking for a particular manufacturer's device to exploit a
> discovered security flaw, such as turning off a stove's gas pilot and
> turning on all burners. If that doesn't sound realistic, how about a
> cracker inside a manufacturer's systems doing the same; or how about a
> terrorist?
> 
> It is reasonable to assume that HVAC systems will someday soon be
> controlled over the Internet by a maintenance firm (video surveillance
> systems already are controlled and monitored over the Internet). It may
> become possible, for example, to raise building temperatures across the
> lower Manhattan area and shut down most financial centers, at least for a
> short while.

Re: Addresses and ports and taxes -- oh my!

[Anthony Atkielski]

Peter Dawson writes:

> v6 address space works out to about 1500 address
> per sq  mtr of the earth's surface...
> NOW..how many house fit on 1 sqm ?

Not relevant.  IPv6 will be exhausted by overly-generous allocation of
address space, just like IPv4.  I've already explained in the past why this
must be so.  In part, it comes from the subjective impression that any new
address space is "more than we'll ever need" and the tendency to
overallocate in consequence, until it's too late; this is probably the most
common engineering mistake in IT history.  Another reason why IPv6 is not
nearly as large as it appears to be is that IP addresses are closely linked
to routing, instead of being randomly assigned; and this routing imposes
severe restrictions on how the address space can be used.  Both issues are
routinely and completely overlooked, even though they are responsible for
the pending exhaustion of IPv4 today, even though we don't have four billion
computers on the Internet yet.

In the past, I've suggested open-ended IP addressing to permanently solve
this (similar to the way telephone numbers are assigned today), but there
wasn't much interest.  Perhaps a few years from now, when it becomes
necessary to go to IPv12, which will have "more space than we'll ever need,"
I'll try again.

Re: Addresses and ports and taxes -- oh my!

[Anthony Atkielski]

> The IPv6 address blocks allocated by ARIN are much
> much larger, so the price per address for an ISP
> is considerably lower.

And IPv6 will be exhausted just as quickly or more quickly than IPv4 in
consequence.

Doesn't anyone ever learn?  The same mistakes are being made over and over.
There's no point in going to IPv6 if blocks of wasted addresses will be
increased in size to go along with it.  You just end up in the same boat,
with no addresses left.

RE: Addresses and ports and taxes -- oh my!

[vinton g. cerf]

see www.ipnsig.org  

vint

At 09:46 PM 8/3/2000 -0400, Philip J. Nesser II wrote:
>-BEGIN PGP SIGNED MESSAGE-
>Hash: SHA1
>
>There is already a lot of work being done on the Interplanetary
>Internet problem.  Vint Cerf has lead pioneering work with people at
>JPL on the problem.  I don't remember the URL but it should be easy
>to find.  The current achitecture (last I checked) was for basically
>independent networks on each planet with gateways at the planetary
>borders.  There are some facinating problems that come up when you
>think about it.
>
>- --->  Phil

=
I moved to a new MCI WorldCom facility on Nov 11, 1999

MCI WorldCom
22001 Loudoun County Parkway
Building F2, Room 4115, ATTN: Vint Cerf
Ashburn, VA 20147
Telephone (703) 886-1690
FAX (703) 886-0047


"INTERNET IS FOR EVERYONE!" 
INET 2001: Internet Global Summit 
5-8 June 2001 
Sweden International Fairs 
Stockholm, Sweden 
http://www.isoc.org/inet2001

Re: Addresses and ports and taxes -- oh my!

[Anthony Atkielski]

> When household appliances begin becoming IP addressable,
> I think we will see a move towards assigning an Internet
> IP address per household (much like today's street address).
> The household will perform NAT for all devices within
> (one street address can house many people, not just one).

I think we'll see IP addressable toasters and washing machines just after we
all switch from automobiles to hovercars and from telephones to
Picturephones.  According to predictions being made by futurists for the
past few decades, all of these things are due to happen Real Soon Now.

The mere fact that something is technically possible doesn't mean that it
should be done.

Re: Addresses and ports and taxes -- oh my!

[vinton g. cerf]

At 05:48 PM 8/3/2000 -0400, Keith Moore wrote:
>The various proposals to burn IP addresses into devices are naive.
for the record, I was assuming that only a lower order unique end-id
was burned in and that the high order bits would be dynamically assigned
based on connectivity to the global Internet.
=
I moved to a new MCI WorldCom facility on Nov 11, 1999

MCI WorldCom
22001 Loudoun County Parkway
Building F2, Room 4115, ATTN: Vint Cerf
Ashburn, VA 20147
Telephone (703) 886-1690
FAX (703) 886-0047


"INTERNET IS FOR EVERYONE!" 
INET 2001: Internet Global Summit 
5-8 June 2001 
Sweden International Fairs 
Stockholm, Sweden 
http://www.isoc.org/inet2001

Re: Addresses and ports and taxes -- oh my!

[Valdis . Kletnieks]

On Thu, 03 Aug 2000 16:52:25 PDT, [EMAIL PROTECTED] said:
> why not consider all the dimentions, ever heard of polyfractal space ?

Fabricating the router connections would be interesting  WHat sort of
crimping tool would it take to make a 2.75D connector stay on the cable? ;)


 PGP signature

Re: Addresses and ports and taxes -- oh my!

[Valdis . Kletnieks]

On Thu, 03 Aug 2000 22:06:13 BST, Lloyd Wood said:
> in an ideal world, airlines don't _have_ differentiated seat pricing.

Of course, in reality, they split them into coach and tolerable. ;)
-- 
Valdis Kletnieks
Operating Systems Analyst
Virginia Tech


 PGP signature

Re: Addresses and ports and taxes -- oh my!

[vinton g. cerf]

I had been assuming that the aggregation would take place
at the "prefix" level of IPv6 addresses and that the unique
component would be in some lower order part of the 128 bit space.

If I have a bad model of that, I'd appreciate offline explanation.

thanks

vint

At 04:19 PM 8/3/2000 -0400, [EMAIL PROTECTED] wrote:
>You don't want to assign network addresses on a per-manufacturer basis.
>Network addresses have to be aggregable.
>
>That's why we have ARP because MAC addresses on an Ehternet are assigned
>per manufacturer.  But that's OK, because IPv6 has ARP ;)

=
I moved to a new MCI WorldCom facility on Nov 11, 1999

MCI WorldCom
22001 Loudoun County Parkway
Building F2, Room 4115, ATTN: Vint Cerf
Ashburn, VA 20147
Telephone (703) 886-1690
FAX (703) 886-0047


"INTERNET IS FOR EVERYONE!" 
INET 2001: Internet Global Summit 
5-8 June 2001 
Sweden International Fairs 
Stockholm, Sweden 
http://www.isoc.org/inet2001

RE: Addresses and ports and taxes -- oh my!

[Philip J. Nesser II]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

There is already a lot of work being done on the Interplanetary
Internet problem.  Vint Cerf has lead pioneering work with people at
JPL on the problem.  I don't remember the URL but it should be easy
to find.  The current achitecture (last I checked) was for basically
independent networks on each planet with gateways at the planetary
borders.  There are some facinating problems that come up when you
think about it.

- --->  Phil

> -Original Message-
> From: Dawson, Peter D [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, August 03, 2000 3:06 PM
> To: '[EMAIL PROTECTED]'
> Subject: RE: Addresses and ports and taxes -- oh my!
> 
> 
> good point... but I do wonder how the border edge
> router will handle a datagram with 
> TTL approx > 240 sec's 
> ( i.e min time required for msg to pass between earth <=> mars) ?
> what about jitters, latency ,dropped packets, icmpv6 err msg
> well whatever
> 
> ->-Original Message-
> ->From: Rick H Wesson [mailto:[EMAIL PROTECTED]]
> ->Sent: Thursday, August 03, 2000 2:50 PM
> ->To: Dawson, Peter D
> ->Cc: [EMAIL PROTECTED]
> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->
> ->
> ->
> ->peter,
> ->
> ->who said all the addresses were going to be used only on earth?
> ->
> ->-r
> ->
> ->On Thu, 3 Aug 2000, Dawson, Peter D wrote:
> ->
> ->> v6 address space works out to about 1500 address 
> ->> per sq  mtr of the earth's surface...
> ->> NOW..how many house fit on 1 sqm ?
> ->> 
> ->> ->-Original Message-
> ->> ->From: Parkinson, Jonathan
> [mailto:[EMAIL PROTECTED]] ->> ->Sent: Thursday, August
> 03, 2000 10:23 AM
> ->> ->To: 'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
> ->> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->> ->
> ->> ->
> ->> ->Err I think that would take some thinking about ? How many 
> ->> ->houses are there
> ->> ->in the world!
> ->> ->
> ->> ->-Original Message-
> ->> ->From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
> ->> ->Sent: Thursday, August 03, 2000 2:41 PM
> ->> ->To: 'Dennis Glatting'; [EMAIL PROTECTED]
> ->> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->> ->
> ->> ->
> ->> ->When household appliances begin becoming IP addressable, I 
> ->> ->think we will see
> ->> ->a move towards assigning an Internet IP address per household
>  ->> ->(much like
> ->> ->today's street address).  The household will perform NAT for 
> ->> ->all devices
> ->> ->within (one street address can house many people, not just
> one). ->> ->
> ->> ->
> ->> 
> ->
> 

-BEGIN PGP SIGNATURE-
Version: PGP 6.5.3

iQA/AwUBOYogWh8Cp2AdP9rUEQLV7ACfZqBq0j91GDO2DlQHKgb/efk+B8oAnijQ
dtfdvl+svrtvy7FlxVQ1u6Ip
=1QNv
-END PGP SIGNATURE-

Re: Addresses and ports and taxes -- oh my!

[Dennis Glatting]

On Thu, 3 Aug 2000, Keith Moore wrote:

[snip]

> burning IP addresses into devices is a good way to give vendors the
> ability to control those devices, monitor their usage, and to lock
> their customers in to particular services.  not my idea of a desirable
> state.
> 

It might also be a good way for script kiddies to efficiently scan the
Internet looking for a particular manufacturer's device to exploit a
discovered security flaw, such as turning off a stove's gas pilot and
turning on all burners. If that doesn't sound realistic, how about a
cracker inside a manufacturer's systems doing the same; or how about a
terrorist?

It is reasonable to assume that HVAC systems will someday soon be
controlled over the Internet by a maintenance firm (video surveillance
systems already are controlled and monitored over the Internet). It may
become possible, for example, to raise building temperatures across the
lower Manhattan area and shut down most financial centers, at least for a
short while.

Re: Addresses and ports and taxes -- oh my!

[Keith Moore]

> Geeks like us care about end-to-end transparency.  Refrigerator's don't.

NATs cause a lot more problems than the loss of transparency.
see http://www.cs.utk.edu/~moore/what-nats-break.html

> Most people are going to buy products based on the functions they
> perform (utility), not on their architecture.  If someone develops
> a nice household/Internet gateway that does something useful (and
> doesn't require a UNIX administrator), people will buy it, regardless
> of whether it performs some politically/religiously incorrect protocol
> transformations.

true enough...but they will insist that the product work as advertised.
and if households have NATs in them, the number of things that those 
products can do will be considerably reduced.  hence the number of
products available, and the utility of a household network connection,
will also be reduced.

> And, if IPvX addresses cost money, a lot of households will pay money
> for devices that enable them to operate with only one IP address.

true enough, at least if the addresses cost more than the NAT box.
but if you have to have a NAT box then the money you spend is
for the purpose of making your network less functional.
personally, I'd rather get something useful for my money.

folks who think this is a religious argument aren't paying attention.

but the bottom line is that we need to make sure that

a) IPv6 address blocks of reasonable size have near-zero cost
b) NATs aren't part of IPv6

Keith

RE: Addresses and ports and taxes -- oh my!

[Jim_Stephenson-Dunn]

why not consider all the dimentions, ever heard of polyfractal space ?

(sorry couldn't resist it ;-> )

Jim






"Evstiounin, Mikhail" <[EMAIL PROTECTED]> on 03/08/2000 19:33:50

Sent by:  "Evstiounin, Mikhail" <[EMAIL PROTECTED]>


To:   "Dawson, Peter D" , [EMAIL PROTECTED]
cc:(Jim Stephenson-Dunn/C/HQ/3Com)
Subject:  RE: Addresses and ports and taxes -- oh my!



Well, there are 8 computers in my house, and I don't count future IP enabled
appliances. At my work place (building take probably about 0.7 hectares or
less and this only one site) more than 700 computers:-). In any case, my
point is that we should consider 3rd dimension also:-)

> -Original Message-
> From:   Dawson, Peter D [SMTP:[EMAIL PROTECTED]]
> Sent:   Thursday, August 03, 2000 11:59 AM
> To: [EMAIL PROTECTED]
> Subject: RE: Addresses and ports and taxes -- oh my!
>
> v6 address space works out to about 1500 address
> per sq  mtr of the earth's surface...
> NOW..how many house fit on 1 sqm ?
>
> ->-Original Message-
> ->From: Parkinson, Jonathan [mailto:[EMAIL PROTECTED]]
> ->Sent: Thursday, August 03, 2000 10:23 AM
> ->To: 'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->
> ->
> ->Err I think that would take some thinking about ? How many
> ->houses are there
> ->in the world!
> ->
> ->-----Original Message-----
> ->From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
> ->Sent: Thursday, August 03, 2000 2:41 PM
> ->To: 'Dennis Glatting'; [EMAIL PROTECTED]
> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->
> ->
> ->When household appliances begin becoming IP addressable, I
> ->think we will see
> ->a move towards assigning an Internet IP address per household
> ->(much like
> ->today's street address).  The household will perform NAT for
> ->all devices
> ->within (one street address can house many people, not just one).
> ->
> ->

RE: Addresses and ports and taxes -- oh my!

[Alan Simpkins]

 NAT would definitely serve a purpose for those wishing to not pay a fee for Intert addresseable address space. It would seem though that if one pays for Internet access this should in fact be included in the price.
  "Evstiounin, Mikhail" <[EMAIL PROTECTED]> wrote: 
Wasn't avoiding NAT one of the goal of IPv6? I recall a pretty bigdiscussion here some time ago about NAT and IPv6. > -Original Message-> From: Rakers, Jason [SMTP:[EMAIL PROTECTED]]> Sent: Thursday, August 03, 2000 9:41 AM> To: 'Dennis Glatting'; [EMAIL PROTECTED]> Subject: RE: Addresses and ports and taxes -- oh my!> > When household appliances begin becoming IP addressable, I think we will> see> a move towards assigning an Internet IP address per household (much like> today's street address). The household will perform NAT for all devices> within (one street address can house many people, not just one).> > > -Original Message-> > From: Dennis Glatting> [SMTP:[EMAIL PROTECTED]]> > Sent: Thursday, August 03, 2000 8:32 !
 AM> > To: [EMAIL PROTECTED]> > Subject: Addresses and ports and taxes -- oh my!> > > > > > I've been thinking about the issue of ARIN fees from last night's> plenary> > and arrived at two philosophical questions.> > > > I run my business out of my home and my DSL link is an important part of> > my business. About six months ago my ISP started charging me a $20/mo.> fee> > for my /27 because "ARIN is now charging us." I am unhappy about this> fee> > but I understand its motivation -- conversation of IP space, though I> > believe fees do not really effect the true wasters of this space and the> > fee, or as it is called in some circles, a tax, is probably misguided.> > Nonetheless, with IPv6, I naively hoped, until last night, the> > conservation of space issues would go away, and thus the fees. Big duh!>!
  > > > If we look at today's marketing hype and think for

ward a bit there is a> > thrust to "Internet enable" appliances, such as dryers, ovens, and> > stereos. Assuming ARIN fees persist, my first philosophical question is> > whether any consumer of these appliances MUST periodically (e.g.,> monthly)> > drop coins in the ARIN fountain?> > > > Thinking laterally, the reserved port space (<1024) is tight. Using the> > same IP space conversation logic, should fees be charged to conserve> port> > space? If so, my second philosophiocal question is what is our role, as> > protocol designers and IETF volunteers, in creating, what is slowly> > becoming, an Internet consumption taxation model?> > > > Imagine for a moment the effect of a fee against the allocation or use> of> > port 80 or 443, maybe even port 25 or 53.> > > > > > > > > > AT would d3efinetDo You Yahoo!?
Kick off your party with Yahoo! Invites.

Re: Addresses and ports and taxes -- oh my!

[Keith Moore]

The various proposals to burn IP addresses into devices are naive.

IP addresses identify points in the network topology.  Hence they 
need to be assigned according to topology.  If you try to assign
IP addresses in some other fashion, you immediately need something
else to replace the IP address which is assigned according to network
topology.  And for various reasons that new thing will have a very 
similar structure to an IP address. 

(for reasons of efficiency, IP addresses are also sometimes used to name, 
or as parts of names for, lots of other things - including interfaces, 
hosts, and connection endpoints.  but it's the use of an IP address
to name a point in the network topology that drives their assignment.)

yes, you can add a layer of indirection to the network so that a
device has a constant 'virtual' IP address which gets forwarded to
the 'real' IP address which indicates the current location of the
device.  mobile-ip does this, for example, and it's quite useful
to be able to have a mobile device with a stable IP address.
but then you still need a 'current location' address.  you also need a 
home agent which sits on the net at the location corresponding to
your stable IP address, and either forwards your traffic to the
'current location' address or issues a redirect to that address.

if my toaster is going to be on the net, the last thing I want is for 
all of the traffic for that toaster to go to the toaster manufacturer's 
home agent just so it can get forwarded to my house.   I don't want the
toaster manufacturer to be able to see how often I'm checking on the
status of my toast.  nor do I want to have to pay the toaster manufacturer 
to get it to continue to route packets to my toaster.

burning IP addresses into devices is a good way to give vendors the 
ability to control those devices, monitor their usage, and to lock
their customers in to particular services.  not my idea of a desirable state.

personally, I'd rather have the address registries.

Keith

Re: Addresses and ports and taxes -- oh my!

[Valdis . Kletnieks]

On Thu, 03 Aug 2000 12:57:49 EDT, "Book, Robert" <[EMAIL PROTECTED]>  said:
> Vinton's idea has much merit. A scheme to allocate blocks of addresses to
> manufacturers would be much easier to support than an organization
> attempting to process individual email requests, or CGI scripted forms from
> a webpage, or a world-wide DHCP server for Amana ( and one for Maytag, etc.)
> to register a refrigerator. And easier administration should translate into

You don't want to assign network addresses on a per-manufacturer basis.
Network addresses have to be aggregable.

That's why we have ARP because MAC addresses on an Ehternet are assigned
per manufacturer.  But that's OK, because IPv6 has ARP ;)
-- 
Valdis Kletnieks
Operating Systems Analyst
Virginia Tech


 PGP signature

RE: Addresses and ports and taxes -- oh my!

[Scott Bradner]

as long as we do not care about fragmentation of the routing space this
idea is neat

Scott

---
Vinton's idea has much merit. A scheme to allocate blocks of addresses to
manufacturers would be much easier to support than an organization
attempting to process individual email requests, or CGI scripted forms from
a webpage, or a world-wide DHCP server for Amana ( and one for Maytag, etc.)
to register a refrigerator. And easier administration should translate into
lower cost.. :-) My apologies in advance for polluting the IETF announce
list. Btw, did we ever find out how many people were on that elevator? :-)

-Original Message-
From: vinton g. cerf [mailto:[EMAIL PROTECTED]]
Sent: Thursday, August 03, 2000 10:51 AM
To: Dennis Glatting; [EMAIL PROTECTED]
Subject: Re: Addresses and ports and taxes -- oh my!


Dennis

thanks for drawing attention to this question. One of the reasons
for fees, of course, is that the Address Registries also have responsibility
to support ICANN so they have some new costs in addition to their operating
costs (or if you like, their operating costs include support for ICANN).

It is a very good question whether one's internet-enabled household
appliances
will induce a monthly charges - do you suppose there would be a way to have
a
one-time charge to "pay" for some number of such addresses - perhaps built
into
the cost of the appliance (and paid by the manufacturer who "burns" an
address
into the device - at least the low order 64 bits or something to make it
end-to-end unique)? 

Please don't flame me for thinking out loud - Dennis' point is a good one
and
we ought to discuss - perhaps in a smaller group than the whole of ietf
announce
list!

Vint Cerf

At 05:32 AM 8/3/2000 -0700, Dennis Glatting wrote:

>I've been thinking about the issue of ARIN fees from last night's plenary
>and arrived at two philosophical questions.
>
>I run my business out of my home and my DSL link is an important part of
>my business. About six months ago my ISP started charging me a $20/mo. fee
>for my /27 because "ARIN is now charging us." I am unhappy about this fee
>but I understand its motivation -- conversation of IP space, though I
>believe fees do not really effect the true wasters of this space and the
>fee, or as it is called in some circles, a tax, is probably misguided.
>Nonetheless, with IPv6, I naively hoped, until last night, the
>conservation of space issues would go away, and thus the fees. Big duh!
>
>If we look at today's marketing hype and think forward a bit there is a
>thrust to "Internet enable" appliances, such as dryers, ovens, and
>stereos. Assuming ARIN fees persist, my first philosophical question is
>whether any consumer of these appliances MUST periodically (e.g., monthly)
>drop coins in the ARIN fountain?
>
>Thinking laterally, the reserved port space (<1024) is tight. Using the
>same IP space conversation logic, should fees be charged to conserve port
>space? If so, my second philosophiocal question is what is our role, as
>protocol designers and IETF volunteers, in creating, what is slowly
>becoming, an Internet consumption taxation model?
>
>Imagine for a moment the effect of a fee against the allocation or use of
>port 80 or 443, maybe even port 25 or 53.
>
>
>
>

=
I moved to a new MCI WorldCom facility on Nov 11, 1999

MCI WorldCom
22001 Loudoun County Parkway
Building F2, Room 4115, ATTN: Vint Cerf
Ashburn, VA 20147
Telephone (703) 886-1690
FAX (703) 886-0047


"INTERNET IS FOR EVERYONE!" 
INET 2001: Internet Global Summit 
5-8 June 2001 
Sweden International Fairs 
Stockholm, Sweden 
http://www.isoc.org/inet2001

RE: Addresses and ports and taxes -- oh my!

[John Day]

At 12:57 PM -0400 8/3/00, Book, Robert wrote:
>Vinton's idea has much merit. A scheme to allocate blocks of addresses to
>manufacturers would be much easier to support than an organization
>attempting to process individual email requests, or CGI scripted forms from

Good grief!  This is the worst possible solution.  We are allocating 
*addresses*, not serial numbers.

>a webpage, or a world-wide DHCP server for Amana ( and one for Maytag, etc.)

I presume that you assume that all Amana appliances will be in the 
same topological part of the network and all Maytag appliances will 
be a different part of the topology.

How do you suppose route aggregation will work with such an approach?

>to register a refrigerator. And easier administration should translate into
>lower cost.. :-) My apologies in advance for polluting the IETF announce
>list. Btw, did we ever find out how many people were on that elevator? :-)

John Day

RE: Addresses and ports and taxes -- oh my!

[Evstiounin, Mikhail]

I would say 12G * 4 ( at least ), but still OK. Reasoning behind this is
that
I'd like to have real IP for each computer in the house, plus mobile IPs 
for everybody, plus appliances, plus hotels, planes, cars, trains, etc

> -Original Message-
> From: Brian E Carpenter [SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, August 03, 2000 1:22 PM
> To:   [EMAIL PROTECTED]
> Subject:      Re: Addresses and ports and taxes -- oh my!
> 
> "Rakers, Jason" wrote:
> > 
> > Better question:  How many households are there in the world on the
> > Internet?
> 
> Wrong question. The correct question is how many should we plan for.
> Right now 12 billion people seems to be a reasoanble estimate -
> unthinkable for IPv4, but easily covered by IPv6.
> 
>Brian

Re: Addresses and ports and taxes -- oh my!

[John Stracke]

"Book, Robert" wrote:

> Vinton's idea has much merit. A scheme to allocate blocks of addresses to
> manufacturers would be much easier to support

Mmm...that's already there, isn't it? The low-order 64 bits he's talking about
are things like Ethernet addresses, for use in IPv6 autoconfig.  (We can't
allocate blocks of actual IP addresses to manufacturers, of course, or the
routers will go mad.  I don't want to wake up in the middle of the night to
find a zombie router, with its front panel glowing insanely, standing over me
with a Cat5 garrotte and screaming, "Revenge!".  :-)

--
/=\
|John Stracke| http://www.ecal.com |My opinions are my own.   |
|Chief Scientist ||
|eCal Corp.  |If God had not given us duct tape, it would have|
|[EMAIL PROTECTED]|been necessary to invent it.|
\=/

RE: Addresses and ports and taxes -- oh my!

[Dawson, Peter D]

good point... but I do wonder how the border edge
router will handle a datagram with 
TTL approx > 240 sec's 
( i.e min time required for msg to pass between earth <=> mars) ?
what about jitters, latency ,dropped packets, icmpv6 err msg well
whatever

->-Original Message-
->From: Rick H Wesson [mailto:[EMAIL PROTECTED]]
->Sent: Thursday, August 03, 2000 2:50 PM
->To: Dawson, Peter D
->Cc: [EMAIL PROTECTED]
->Subject: RE: Addresses and ports and taxes -- oh my!
->
->
->
->peter,
->
->who said all the addresses were going to be used only on earth?
->
->-r
->
->On Thu, 3 Aug 2000, Dawson, Peter D wrote:
->
->> v6 address space works out to about 1500 address 
->> per sq  mtr of the earth's surface...
->> NOW..how many house fit on 1 sqm ?
->> 
->> ->-Original Message-
->> ->From: Parkinson, Jonathan [mailto:[EMAIL PROTECTED]]
->> ->Sent: Thursday, August 03, 2000 10:23 AM
->> ->To: 'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
->> ->Subject: RE: Addresses and ports and taxes -- oh my!
->> ->
->> ->
->> ->Err I think that would take some thinking about ? How many 
->> ->houses are there
->> ->in the world!
->> ->
->> ->-----Original Message-
->> ->From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
->> ->Sent: Thursday, August 03, 2000 2:41 PM
->> ->To: 'Dennis Glatting'; [EMAIL PROTECTED]
->> ->Subject: RE: Addresses and ports and taxes -- oh my!
->> ->
->> ->
->> ->When household appliances begin becoming IP addressable, I 
->> ->think we will see
->> ->a move towards assigning an Internet IP address per household 
->> ->(much like
->> ->today's street address).  The household will perform NAT for 
->> ->all devices
->> ->within (one street address can house many people, not just one).
->> ->
->> ->
->> 
->

RE: Addresses and ports and taxes -- oh my!

[Tim Salo]

> From: "Dawson, Peter D" <[EMAIL PROTECTED]>
> Subject: RE: Addresses and ports and taxes -- oh my!
> Date: Thu, 3 Aug 2000 14:36:57 - 
> 
> ->The household will perform NAT for all devices
> ->within (one street address can house many people, not just one).
> 
> .. and lose out on e2e connectivity ?? 
>   [...]

Geeks like us care about end-to-end transparency.  Refrigerator's don't.

Most people are going to buy products based on the functions they
perform (utility), not on their architecture.  If someone develops
a nice household/Internet gateway that does something useful (and
doesn't require a UNIX administrator), people will buy it, regardless
of whether it performs some politically/religiously incorrect protocol
transformations.

And, if IPvX addresses cost money, a lot of households will pay money
for devices that enable them to operate with only one IP address.

(We now return you to discussions about how AOL can't claim to be
part of the Internet...)

-tjs

RE: Addresses and ports and taxes -- oh my!

[Rick H Wesson]

peter,

who said all the addresses were going to be used only on earth?

-r

On Thu, 3 Aug 2000, Dawson, Peter D wrote:

> v6 address space works out to about 1500 address 
> per sq  mtr of the earth's surface...
> NOW..how many house fit on 1 sqm ?
> 
> ->-Original Message-
> ->From: Parkinson, Jonathan [mailto:[EMAIL PROTECTED]]
> ->Sent: Thursday, August 03, 2000 10:23 AM
> ->To: 'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->
> ->
> ->Err I think that would take some thinking about ? How many 
> ->houses are there
> ->in the world!
> ->
> ->-Original Message-
> ->From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
> ->Sent: Thursday, August 03, 2000 2:41 PM
> ->To: 'Dennis Glatting'; [EMAIL PROTECTED]
> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->
> ->
> ->When household appliances begin becoming IP addressable, I 
> ->think we will see
> ->a move towards assigning an Internet IP address per household 
> ->(much like
> ->today's street address).  The household will perform NAT for 
> ->all devices
> ->within (one street address can house many people, not just one).
> ->
> ->
> 

Re: Addresses and ports and taxes -- oh my!

[Keith Moore]

> When household appliances begin becoming IP addressable, I think we will see
> a move towards assigning an Internet IP address per household (much like
> today's street address).  

make that an address *block* per household.

> The household will perform NAT for all devices
> within (one street address can house many people, not just one).

not likely - putting NAT in the path would make the household
net nearly useless (because a lot of the things for which you want 
a household net imply assigning an external address to devices
within your household) 

We are of course talking about IPv6 here - IPv4 doesn't have anywhere
nearly enough address space to give every household a single routable
address.  And NATs make no sense at all in a world where you have plenty 
of address space.

At best NAPTs trade off address bits for port number bits - and in IPv6 
port number bits are the ones that are scarce.  

Keith

FW: Addresses and ports and taxes -- oh my!

[David A. Higginbotham]

hmm... in Manahattan that may be more than one, in Silk Hope, North Carolina
it is less than one

-Original Message-
From: Dawson, Peter D [mailto:[EMAIL PROTECTED]]
Sent: Thursday, August 03, 2000 11:59 AM
To: [EMAIL PROTECTED]
Subject: RE: Addresses and ports and taxes -- oh my!


v6 address space works out to about 1500 address
per sq  mtr of the earth's surface...
NOW..how many house fit on 1 sqm ?

->-Original Message-
->From: Parkinson, Jonathan [mailto:[EMAIL PROTECTED]]
->Sent: Thursday, August 03, 2000 10:23 AM
->To: 'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
->Subject: RE: Addresses and ports and taxes -- oh my!
->
->
->Err I think that would take some thinking about ? How many
->houses are there
->in the world!
->
->-Original Message-
->From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
->Sent: Thursday, August 03, 2000 2:41 PM
->To: 'Dennis Glatting'; [EMAIL PROTECTED]
->Subject: RE: Addresses and ports and taxes -- oh my!
->
->
->When household appliances begin becoming IP addressable, I
->think we will see
->a move towards assigning an Internet IP address per household
->(much like
->today's street address).  The household will perform NAT for
->all devices
->within (one street address can house many people, not just one).
->
->

RE: Addresses and ports and taxes -- oh my!

[Evstiounin, Mikhail]

Well, there are 8 computers in my house, and I don't count future IP enabled
appliances. At my work place (building take probably about 0.7 hectares or
less and this only one site) more than 700 computers:-). In any case, my
point is that we should consider 3rd dimension also:-)

> -Original Message-
> From: Dawson, Peter D [SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, August 03, 2000 11:59 AM
> To:   [EMAIL PROTECTED]
> Subject:      RE: Addresses and ports and taxes -- oh my!
> 
> v6 address space works out to about 1500 address 
> per sq  mtr of the earth's surface...
> NOW..how many house fit on 1 sqm ?
> 
> ->-Original Message-
> ->From: Parkinson, Jonathan [mailto:[EMAIL PROTECTED]]
> ->Sent: Thursday, August 03, 2000 10:23 AM
> ->To: 'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->
> ->
> ->Err I think that would take some thinking about ? How many 
> ->houses are there
> ->in the world!
> ->
> ->-Original Message-
> ->From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
> ->Sent: Thursday, August 03, 2000 2:41 PM
> ->To: 'Dennis Glatting'; [EMAIL PROTECTED]
> ->Subject: RE: Addresses and ports and taxes -- oh my!
> ->
> ->
> ->When household appliances begin becoming IP addressable, I 
> ->think we will see
> ->a move towards assigning an Internet IP address per household 
> ->(much like
> ->today's street address).  The household will perform NAT for 
> ->all devices
> ->within (one street address can house many people, not just one).
> ->
> ->

Re: Addresses and ports and taxes -- oh my!

[Rick H Wesson]

Vint,

the ASO members don't support ICANN on a per block basis, in fact ICANN's
Task Force on Funding (TFF) observed that the IP Address Registries
operate on a non-profit business model from member fees and should foot
10% of ICANN's budget. (see 
http://www.icann.org/tff/final-report-draft-30oct99.htm)

If ICANN's budget grows the ASO's responsibility grows proportionally. If
the IP Address Registries (ASO Members) are doing cost-recovery then its
expected the price for membership will increase or the cost for a
delegation will increase. It is reasonable to assume that an address
maintance fee will eventually passed along to the consumer just to
maintain the 10% ICANN maintance.

I don't want to see IP Address space charges, no matter its version,
induce a monthly charges on the end-user side. Just this year the ASO's
10% responsibility will amount to $428,000 USD. If you don't think that
ICANN's budget will grow under stand tha ARIN's 2000-01 budget is 5.2M and
ICANN's budget for 2000-01 is 5.0M 

If you want to keep ICANN out of your pocket, ensure they stay lean and
focused on technical administration of Assigned Names and Numbers and not
inflating its own self worth.

I would prefer having ICANN set the ASO funding requirements in such a way
that it did not encourage ASO members to pass on the ICANN maintance fees
in a per assignment basis. In short Manageing ICANN's budget can have the
largest impact on the costs of future IP Address Registry operations.

I'd be happy to carry this conversation on in another forum, just let me
know where that is.

thanks,

-rick


On Thu, 3 Aug 2000, vinton g. cerf wrote:


> It is a very good question whether one's internet-enabled household appliances
> will induce a monthly charges - do you suppose there would be a way to have a
> one-time charge to "pay" for some number of such addresses - perhaps built into
> the cost of the appliance (and paid by the manufacturer who "burns" an address
> into the device - at least the low order 64 bits or something to make it
> end-to-end unique)? 

Re: Addresses and ports and taxes -- oh my!

[Harald Alvestrand]

At 05:32 03/08/2000 -0700, Dennis Glatting wrote:
>I run my business out of my home and my DSL link is an important part of
>my business. About six months ago my ISP started charging me a $20/mo. fee
>for my /27 because "ARIN is now charging us."

if the USD 2500/year fee Scott mentioned in the plenary is even 
approximately right, your ISP is recovering its ARIN costs from a base of 
around 10 customers.

Small ISP.

Harald

--
Harald Tveit Alvestrand, [EMAIL PROTECTED]
+47 41 44 29 94
Personal email: [EMAIL PROTECTED]

RE: Addresses and ports and taxes -- oh my!

[Kay, Rodney]

Why would internet-enable appliances necessarily have to incur a charge?
Although internet-enabled, they could also be defined as 10. address within
your LAN (a.k.a. home).  They would be visible to the outside only through
your personal server - for which you are already paying a charge.  Whatever
security you ran on your PC would protect you from someone, say, turning
your heat up to around 90 degrees F while you were gone on vacation.

Regards,

Rodney H. Kay
Department of Veterans Affairs
Puget Sound Health Care System
Systems Manager
Seattle, Washington

-Original Message-
From: vinton g. cerf [mailto:[EMAIL PROTECTED]]
Sent: Thursday, August 03, 2000 7:51 AM
To: Dennis Glatting; [EMAIL PROTECTED]
Subject: Re: Addresses and ports and taxes -- oh my!


Dennis

thanks for drawing attention to this question. One of the reasons
for fees, of course, is that the Address Registries also have responsibility
to support ICANN so they have some new costs in addition to their operating
costs (or if you like, their operating costs include support for ICANN).

It is a very good question whether one's internet-enabled household
appliances
will induce a monthly charges - do you suppose there would be a way to have
a
one-time charge to "pay" for some number of such addresses - perhaps built
into
the cost of the appliance (and paid by the manufacturer who "burns" an
address
into the device - at least the low order 64 bits or something to make it
end-to-end unique)? 

Please don't flame me for thinking out loud - Dennis' point is a good one
and
we ought to discuss - perhaps in a smaller group than the whole of ietf
announce
list!

Vint Cerf

At 05:32 AM 8/3/2000 -0700, Dennis Glatting wrote:

>I've been thinking about the issue of ARIN fees from last night's plenary
>and arrived at two philosophical questions.
>
>I run my business out of my home and my DSL link is an important part of
>my business. About six months ago my ISP started charging me a $20/mo. fee
>for my /27 because "ARIN is now charging us." I am unhappy about this fee
>but I understand its motivation -- conversation of IP space, though I
>believe fees do not really effect the true wasters of this space and the
>fee, or as it is called in some circles, a tax, is probably misguided.
>Nonetheless, with IPv6, I naively hoped, until last night, the
>conservation of space issues would go away, and thus the fees. Big duh!
>
>If we look at today's marketing hype and think forward a bit there is a
>thrust to "Internet enable" appliances, such as dryers, ovens, and
>stereos. Assuming ARIN fees persist, my first philosophical question is
>whether any consumer of these appliances MUST periodically (e.g., monthly)
>drop coins in the ARIN fountain?
>
>Thinking laterally, the reserved port space (<1024) is tight. Using the
>same IP space conversation logic, should fees be charged to conserve port
>space? If so, my second philosophiocal question is what is our role, as
>protocol designers and IETF volunteers, in creating, what is slowly
>becoming, an Internet consumption taxation model?
>
>Imagine for a moment the effect of a fee against the allocation or use of
>port 80 or 443, maybe even port 25 or 53.
>
>
>
>

=
I moved to a new MCI WorldCom facility on Nov 11, 1999

MCI WorldCom
22001 Loudoun County Parkway
Building F2, Room 4115, ATTN: Vint Cerf
Ashburn, VA 20147
Telephone (703) 886-1690
FAX (703) 886-0047


"INTERNET IS FOR EVERYONE!" 
INET 2001: Internet Global Summit 
5-8 June 2001 
Sweden International Fairs 
Stockholm, Sweden 
http://www.isoc.org/inet2001

Re: Addresses and ports and taxes -- oh my!

[Jon Crowcroft]

in an ideal world, this would be worked out thru a non linear dynamic
pricing model  the same way the airlines do differentiated seat
pricing  

we haev ipv4 and ipv6 addresses; they have different cost recovery
models and different utility functions and different marketing dweebs
selling the,em, and different incentiveve requirements...

so by then, the price would be very very low as the cost of dealing
with the possible future admin of assigning more addresses will be
vanishingly small



In message <[EMAIL PROTECTED]>, "Rakers, Jason" t
yped:

 >>Better question:  How many households are there in the world on the
 >>Internet?
 >>
 >>> -Original Message-
 >>> From:  Parkinson, Jonathan [SMTP:[EMAIL PROTECTED]]
 >>> Sent:  Thursday, August 03, 2000 10:23 AM
 >>> To:'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
 >>> Subject:   RE: Addresses and ports and taxes -- oh my!
 >>> 
 >>> Err I think that would take some thinking about ? How many houses are
 >>> there
 >>> in the world!
 >>> 
 >>> -Original Message-
 >>> From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
 >>> Sent: Thursday, August 03, 2000 2:41 PM
 >>> To: 'Dennis Glatting'; [EMAIL PROTECTED]
 >>> Subject: RE: Addresses and ports and taxes -- oh my!
 >>> 
 >>> 
 >>> When household appliances begin becoming IP addressable, I think we will
 >>> see
 >>> a move towards assigning an Internet IP address per household (much like
 >>> today's street address).  The household will perform NAT for all devices
 >>> within (one street address can house many people, not just one).
 >>> 
 >>> > -Original Message-
 >>> > From:Dennis Glatting
 >>> [SMTP:[EMAIL PROTECTED]]
 >>> > Sent:Thursday, August 03, 2000 8:32 AM
 >>> > To:  [EMAIL PROTECTED]
 >>> > Subject: Addresses and ports and taxes -- oh my!
 >>> > 
 >>> > 
 >>> > I've been thinking about the issue of ARIN fees from last night's
 >>> plenary
 >>> > and arrived at two philosophical questions.
 >>> > 
 >>> > I run my business out of my home and my DSL link is an important part of
 >>> > my business. About six months ago my ISP started charging me a $20/mo.
 >>> fee
 >>> > for my /27 because "ARIN is now charging us." I am unhappy about this
 >>> fee
 >>> > but I understand its motivation -- conversation of IP space, though I
 >>> > believe fees do not really effect the true wasters of this space and the
 >>> > fee, or as it is called in some circles, a tax, is probably misguided.
 >>> > Nonetheless, with IPv6, I naively hoped, until last night, the
 >>> > conservation of space issues would go away, and thus the fees. Big duh!
 >>> > 
 >>> > If we look at today's marketing hype and think forward a bit there is a
 >>> > thrust to "Internet enable" appliances, such as dryers, ovens, and
 >>> > stereos. Assuming ARIN fees persist, my first philosophical question is
 >>> > whether any consumer of these appliances MUST periodically (e.g.,
 >>> monthly)
 >>> > drop coins in the ARIN fountain?
 >>> > 
 >>> > Thinking laterally, the reserved port space (<1024) is tight. Using the
 >>> > same IP space conversation logic, should fees be charged to conserve
 >>> port
 >>> > space? If so, my second philosophiocal question is what is our role, as
 >>> > protocol designers and IETF volunteers, in creating, what is slowly
 >>> > becoming, an Internet consumption taxation model?
 >>> > 
 >>> > Imagine for a moment the effect of a fee against the allocation or use
 >>> of
 >>> > port 80 or 443, maybe even port 25 or 53.
 >>> > 
 >>> > 
 >>> > 
 >>> > 
 >>> > 
 >>

 cheers

   jon

Re: Addresses and ports and taxes -- oh my!

[Jim_Stephenson-Dunn]

Such a product is available already, check out :

www.maxgate.net/product_3200.htm

I believe the device does use NAT ( :-< ) but there are other devices in the
family, that do IPSEC. I am still waiting for mine to be delivered, sorry if I
am a little wooly about it, I haven't had chance to play with it yet, but it
looks promising.

At the moment the cost is about $300 USD, but as Andre points out the price will
drop eventually. Also I had believed that IPv6 had enough addresses for every
person on planet Earth and then some.

But here is an intresting question,

Would the address be assigned to a person or a property, on the basis that if
the person moved would the address go with them or stay with the house ?

What happened if the house was destroyed or the person died, would the address
be returned to the registry, from where it was issued, to be re-allocated to a
new property or person ?

Will we eventually get to the point where we are all as unique as our IPv6
addresses, and new born children are assigned an address automatically, when
they are born ?

Maybe george orwell and aldous huxley got it right ;->

Jim


**
This is my personal opinion only and does not represent the technical direction
of 3Com or any subsidury.
**







Andre-John Mas <[EMAIL PROTECTED]> on 03/08/2000 15:41:52

Sent by:  Andre-John Mas <[EMAIL PROTECTED]>


To:   "Rakers, Jason" 
cc:   "'Dennis Glatting'" ,
  [EMAIL PROTECTED] (Jim Stephenson-Dunn/C/HQ/3Com)
Subject:  Re: Addresses and ports and taxes -- oh my!



"Rakers, Jason" wrote:
>
> When household appliances begin becoming IP addressable, I think we will see
> a move towards assigning an Internet IP address per household (much like
> today's street address).  The household will perform NAT for all devices
> within (one street address can house many people, not just one).

The factor in all this is cheap firewalls/routers. With these coming
down in price they will be installed standard when asking for a cable
internet connection, or such. Also these boxes, probably being the size
of a phoneset would probably also include a DHCP server for addressing
the
various devices in the home, much like what is done for Sun's SunRay
line
of devices. If this does happen I just hope that they offer the option
for expert users to configure the devices themselves. The price factor
that will make these an option is arounf USD 100.

Andre

Re: Addresses and ports and taxes -- oh my!

[Brian E Carpenter]

"Rakers, Jason" wrote:
> 
> Better question:  How many households are there in the world on the
> Internet?

Wrong question. The correct question is how many should we plan for.
Right now 12 billion people seems to be a reasoanble estimate -
unthinkable for IPv4, but easily covered by IPv6.

   Brian

RE: Addresses and ports and taxes -- oh my!

[Book, Robert]

Vinton's idea has much merit. A scheme to allocate blocks of addresses to
manufacturers would be much easier to support than an organization
attempting to process individual email requests, or CGI scripted forms from
a webpage, or a world-wide DHCP server for Amana ( and one for Maytag, etc.)
to register a refrigerator. And easier administration should translate into
lower cost.. :-) My apologies in advance for polluting the IETF announce
list. Btw, did we ever find out how many people were on that elevator? :-)

-Original Message-
From: vinton g. cerf [mailto:[EMAIL PROTECTED]]
Sent: Thursday, August 03, 2000 10:51 AM
To: Dennis Glatting; [EMAIL PROTECTED]
Subject: Re: Addresses and ports and taxes -- oh my!


Dennis

thanks for drawing attention to this question. One of the reasons
for fees, of course, is that the Address Registries also have responsibility
to support ICANN so they have some new costs in addition to their operating
costs (or if you like, their operating costs include support for ICANN).

It is a very good question whether one's internet-enabled household
appliances
will induce a monthly charges - do you suppose there would be a way to have
a
one-time charge to "pay" for some number of such addresses - perhaps built
into
the cost of the appliance (and paid by the manufacturer who "burns" an
address
into the device - at least the low order 64 bits or something to make it
end-to-end unique)? 

Please don't flame me for thinking out loud - Dennis' point is a good one
and
we ought to discuss - perhaps in a smaller group than the whole of ietf
announce
list!

Vint Cerf

At 05:32 AM 8/3/2000 -0700, Dennis Glatting wrote:

>I've been thinking about the issue of ARIN fees from last night's plenary
>and arrived at two philosophical questions.
>
>I run my business out of my home and my DSL link is an important part of
>my business. About six months ago my ISP started charging me a $20/mo. fee
>for my /27 because "ARIN is now charging us." I am unhappy about this fee
>but I understand its motivation -- conversation of IP space, though I
>believe fees do not really effect the true wasters of this space and the
>fee, or as it is called in some circles, a tax, is probably misguided.
>Nonetheless, with IPv6, I naively hoped, until last night, the
>conservation of space issues would go away, and thus the fees. Big duh!
>
>If we look at today's marketing hype and think forward a bit there is a
>thrust to "Internet enable" appliances, such as dryers, ovens, and
>stereos. Assuming ARIN fees persist, my first philosophical question is
>whether any consumer of these appliances MUST periodically (e.g., monthly)
>drop coins in the ARIN fountain?
>
>Thinking laterally, the reserved port space (<1024) is tight. Using the
>same IP space conversation logic, should fees be charged to conserve port
>space? If so, my second philosophiocal question is what is our role, as
>protocol designers and IETF volunteers, in creating, what is slowly
>becoming, an Internet consumption taxation model?
>
>Imagine for a moment the effect of a fee against the allocation or use of
>port 80 or 443, maybe even port 25 or 53.
>
>
>
>

=
I moved to a new MCI WorldCom facility on Nov 11, 1999

MCI WorldCom
22001 Loudoun County Parkway
Building F2, Room 4115, ATTN: Vint Cerf
Ashburn, VA 20147
Telephone (703) 886-1690
FAX (703) 886-0047


"INTERNET IS FOR EVERYONE!" 
INET 2001: Internet Global Summit 
5-8 June 2001 
Sweden International Fairs 
Stockholm, Sweden 
http://www.isoc.org/inet2001

RE: Addresses and ports and taxes -- oh my!

[Dawson, Peter D]

v6 address space works out to about 1500 address 
per sq  mtr of the earth's surface...
NOW..how many house fit on 1 sqm ?

->-Original Message-
->From: Parkinson, Jonathan [mailto:[EMAIL PROTECTED]]
->Sent: Thursday, August 03, 2000 10:23 AM
->To: 'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
->Subject: RE: Addresses and ports and taxes -- oh my!
->
->
->Err I think that would take some thinking about ? How many 
->houses are there
->in the world!
->
->-Original Message-
->From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
->Sent: Thursday, August 03, 2000 2:41 PM
->To: 'Dennis Glatting'; [EMAIL PROTECTED]
->Subject: RE: Addresses and ports and taxes -- oh my!
->
->
->When household appliances begin becoming IP addressable, I 
->think we will see
->a move towards assigning an Internet IP address per household 
->(much like
->today's street address).  The household will perform NAT for 
->all devices
->within (one street address can house many people, not just one).
->
->

RE: Addresses and ports and taxes -- oh my!

[R A Lichtensteiger]

Rakers, Jason wrote:

<> When household appliances begin becoming IP addressable, I think we will see
<> a move towards assigning an Internet IP address per household (much like
<> today's street address).  The household will perform NAT for all devices
<> within (one street address can house many people, not just one).

Bad analogy, even assuming all parties involved agreed that NAT was
a good thing (tm).  Many don't and IPv6 was designed to alleviate
addr space problems, not work around them (whcih is what NAT is).

Using your analogy, consider that people do not have one single
address -- One may have home, work, vacation house and POBox. 
They're all valid and you'll find me at all of them (Well, except
work ). But not at the same time, and sometimes visitors need to
be rerouted from one location to another.

RL
-- 
R A Lichtensteiger  [EMAIL PROTECTED]
781 276 4500

Could not open /usr/bin/fortune. Lid on cookie jar sealed

Re: Addresses and ports and taxes -- oh my!

[Andre-John Mas]

"Rakers, Jason" wrote:
> 
> When household appliances begin becoming IP addressable, I think we will see
> a move towards assigning an Internet IP address per household (much like
> today's street address).  The household will perform NAT for all devices
> within (one street address can house many people, not just one).

The factor in all this is cheap firewalls/routers. With these coming
down in price they will be installed standard when asking for a cable
internet connection, or such. Also these boxes, probably being the size
of a phoneset would probably also include a DHCP server for addressing
the
various devices in the home, much like what is done for Sun's SunRay
line
of devices. If this does happen I just hope that they offer the option
for expert users to configure the devices themselves. The price factor
that will make these an option is arounf USD 100.

Andre

Re: Addresses and ports and taxes -- oh my!

[Bob Hinden]

Dennis,

>I run my business out of my home and my DSL link is an important part of
>my business. About six months ago my ISP started charging me a $20/mo. fee
>for my /27 because "ARIN is now charging us." I am unhappy about this fee
>but I understand its motivation -- conversation of IP space, though I

 From the ARIN web site I found that the minimum allocation is a /20 and 
the fee for that is $2,500/year ( 
http://www.arin.net/regserv/feeschedule.html ).  Larger allocations have 
higher charges but the price per address goes down quickly.  I would think 
that most ISP have larger allocations than a /20.
A /20 provides 128 /27 address blocks.  That works out to $19.53/year per 
/27.  You said that your ISP is charging you $20. per month.  Not a bad 
markup!  Interesting what happens when rationing starts.

>Nonetheless, with IPv6, I naively hoped, until last night, the
>conservation of space issues would go away, and thus the fees. Big duh!

The IPv6 address blocks allocated by ARIN are much much larger, so the 
price per address for an ISP is considerably lower.

Bob

Re: Addresses and ports and taxes -- oh my!

[Måns Nilsson]

"Rakers, Jason" wrote:
> 
> When household appliances begin becoming IP addressable, I think we will see
> a move towards assigning an Internet IP address per household (much like
> today's street address).  The household will perform NAT for all devices
> within (one street address can house many people, not just one).

IMNSHO, NAT is evil. While most end-luser applications work "fine" with
it, it does impose restrictions on what can be done as compared to
possibilities in the end-to-end universe. 

*Designing* something to be NAT in  a world that goes toward a larger
address space is broken and should be avoided att all practical costs. 

I do sense that this has been discussed before and will now stop. 

-- 
Måns NilssonDNS Technichian
+46 709 174 840 NIC-SE
+46 8 545 85 707MN1334-RIPE

Re: Addresses and ports and taxes -- oh my!

[Brian E Carpenter]

No, what will happen is one IPv6 prefix per household/car/whatever,
and therefore no reason for NAT.

  Brian

"Rakers, Jason" wrote:
> 
> When household appliances begin becoming IP addressable, I think we will see
> a move towards assigning an Internet IP address per household (much like
> today's street address).  The household will perform NAT for all devices
> within (one street address can house many people, not just one).

Re: Addresses and ports and taxes -- oh my!

[vinton g. cerf]

Dennis

thanks for drawing attention to this question. One of the reasons
for fees, of course, is that the Address Registries also have responsibility
to support ICANN so they have some new costs in addition to their operating
costs (or if you like, their operating costs include support for ICANN).

It is a very good question whether one's internet-enabled household appliances
will induce a monthly charges - do you suppose there would be a way to have a
one-time charge to "pay" for some number of such addresses - perhaps built into
the cost of the appliance (and paid by the manufacturer who "burns" an address
into the device - at least the low order 64 bits or something to make it
end-to-end unique)? 

Please don't flame me for thinking out loud - Dennis' point is a good one and
we ought to discuss - perhaps in a smaller group than the whole of ietf announce
list!

Vint Cerf

At 05:32 AM 8/3/2000 -0700, Dennis Glatting wrote:

>I've been thinking about the issue of ARIN fees from last night's plenary
>and arrived at two philosophical questions.
>
>I run my business out of my home and my DSL link is an important part of
>my business. About six months ago my ISP started charging me a $20/mo. fee
>for my /27 because "ARIN is now charging us." I am unhappy about this fee
>but I understand its motivation -- conversation of IP space, though I
>believe fees do not really effect the true wasters of this space and the
>fee, or as it is called in some circles, a tax, is probably misguided.
>Nonetheless, with IPv6, I naively hoped, until last night, the
>conservation of space issues would go away, and thus the fees. Big duh!
>
>If we look at today's marketing hype and think forward a bit there is a
>thrust to "Internet enable" appliances, such as dryers, ovens, and
>stereos. Assuming ARIN fees persist, my first philosophical question is
>whether any consumer of these appliances MUST periodically (e.g., monthly)
>drop coins in the ARIN fountain?
>
>Thinking laterally, the reserved port space (<1024) is tight. Using the
>same IP space conversation logic, should fees be charged to conserve port
>space? If so, my second philosophiocal question is what is our role, as
>protocol designers and IETF volunteers, in creating, what is slowly
>becoming, an Internet consumption taxation model?
>
>Imagine for a moment the effect of a fee against the allocation or use of
>port 80 or 443, maybe even port 25 or 53.
>
>
>
>

=
I moved to a new MCI WorldCom facility on Nov 11, 1999

MCI WorldCom
22001 Loudoun County Parkway
Building F2, Room 4115, ATTN: Vint Cerf
Ashburn, VA 20147
Telephone (703) 886-1690
FAX (703) 886-0047


"INTERNET IS FOR EVERYONE!" 
INET 2001: Internet Global Summit 
5-8 June 2001 
Sweden International Fairs 
Stockholm, Sweden 
http://www.isoc.org/inet2001

Re: Addresses and ports and taxes -- oh my!

[Magnus Danielson]

From: Dennis Glatting <[EMAIL PROTECTED]>
Subject: Addresses and ports and taxes -- oh my!
Date: Thu, 3 Aug 2000 05:32:04 -0700 (PDT)

> 
> I've been thinking about the issue of ARIN fees from last night's plenary
> and arrived at two philosophical questions.
> 
> I run my business out of my home and my DSL link is an important part of
> my business. About six months ago my ISP started charging me a $20/mo. fee
> for my /27 because "ARIN is now charging us." I am unhappy about this fee
> but I understand its motivation -- conversation of IP space, though I
> believe fees do not really effect the true wasters of this space and the
> fee, or as it is called in some circles, a tax, is probably misguided.
> Nonetheless, with IPv6, I naively hoped, until last night, the
> conservation of space issues would go away, and thus the fees. Big duh!
> 
> If we look at today's marketing hype and think forward a bit there is a
> thrust to "Internet enable" appliances, such as dryers, ovens, and
> stereos. Assuming ARIN fees persist, my first philosophical question is
> whether any consumer of these appliances MUST periodically (e.g., monthly)
> drop coins in the ARIN fountain?
> 
> Thinking laterally, the reserved port space (<1024) is tight. Using the
> same IP space conversation logic, should fees be charged to conserve port
> space? If so, my second philosophiocal question is what is our role, as
> protocol designers and IETF volunteers, in creating, what is slowly
> becoming, an Internet consumption taxation model?
> 
> Imagine for a moment the effect of a fee against the allocation or use of
> port 80 or 443, maybe even port 25 or 53.

Well, who would pay for allocating a new service?

It does not matter if it is 10 or 10 servers offering something on an
allocated port. It is allocated. There is also no good way to get ports back,
since how can you assure that there is no longer traffic on some port.

Neither the servers or the clients can be in a sufficient way charged for the
usage of a well-known port in order to acheive the same pressure as you can do
with (real) IP numbers. ISPs can naturally charge you for open up traffic to
and/or from a certain port, but that does not give the knowledge that you would
require. If someone is running some acient protocol in his/her network using
some well known port and would not be communicating it to the outside world, it
would still make this port occupied without any money changing hands.

Sadly enought I don't think money is the way to solve conservation of port
space, we have to rely on good engineering decissions, and boy, does we know
that these are not reliable ;)

Port numbers are as they are and to some degree we have run into a couple of
scaling issues with them. As allways, set a limit and we will (eventually)
outrun it.

Now, back to the elevators and their prime numbers!
(Interesting Elevator Task Force)

Cheers,
Magnus

Re: Addresses and ports and taxes -- oh my!

[Andre-John Mas]

"Rakers, Jason" wrote:
> 
> When household appliances begin becoming IP addressable, I think we will see
> a move towards assigning an Internet IP address per household (much like
> today's street address).  The household will perform NAT for all devices
> within (one street address can house many people, not just one).

The factor in all this is cheap firewalls/routers. With these coming
down in price they will be installed standard when asking for a cable
internet connection, or such. Also these boxes, probably being the size
of a phoneset would probably also include a DHCP server for addressing
the
various devices in the home, much like what is done for Sun's SunRay
line
of devices. If this does happen I just hope that they offer the option
for expert users to configure the devices themselves. The price factor
that will make these an option is arounf USD 100.

Andre

RE: Addresses and ports and taxes -- oh my!

[Dawson, Peter D]

->-Original Message-
->From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
->Sent: Thursday, August 03, 2000 9:41 AM
->To: 'Dennis Glatting'; [EMAIL PROTECTED]
->Subject: RE: Addresses and ports and taxes -- oh my!
->
->
  The household will perform NAT for 
->all devices
->within (one street address can house many people, not just one).

.. and lose out on e2e connectivity ?? 
imho, primary v6 address arch
was to negate the NAT bottleneck..and of course
v4 address exhaustion

RE: Addresses and ports and taxes -- oh my!

[Rakers, Jason]

Better question:  How many households are there in the world on the
Internet?

> -Original Message-
> From: Parkinson, Jonathan [SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, August 03, 2000 10:23 AM
> To:   'Rakers, Jason'; 'Dennis Glatting'; [EMAIL PROTECTED]
> Subject:  RE: Addresses and ports and taxes -- oh my!
> 
> Err I think that would take some thinking about ? How many houses are
> there
> in the world!
> 
> -Original Message-
> From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, August 03, 2000 2:41 PM
> To: 'Dennis Glatting'; [EMAIL PROTECTED]
> Subject: RE: Addresses and ports and taxes -- oh my!
> 
> 
> When household appliances begin becoming IP addressable, I think we will
> see
> a move towards assigning an Internet IP address per household (much like
> today's street address).  The household will perform NAT for all devices
> within (one street address can house many people, not just one).
> 
> > -Original Message-
> > From:   Dennis Glatting
> [SMTP:[EMAIL PROTECTED]]
> > Sent:   Thursday, August 03, 2000 8:32 AM
> > To: [EMAIL PROTECTED]
> > Subject:Addresses and ports and taxes -- oh my!
> > 
> > 
> > I've been thinking about the issue of ARIN fees from last night's
> plenary
> > and arrived at two philosophical questions.
> > 
> > I run my business out of my home and my DSL link is an important part of
> > my business. About six months ago my ISP started charging me a $20/mo.
> fee
> > for my /27 because "ARIN is now charging us." I am unhappy about this
> fee
> > but I understand its motivation -- conversation of IP space, though I
> > believe fees do not really effect the true wasters of this space and the
> > fee, or as it is called in some circles, a tax, is probably misguided.
> > Nonetheless, with IPv6, I naively hoped, until last night, the
> > conservation of space issues would go away, and thus the fees. Big duh!
> > 
> > If we look at today's marketing hype and think forward a bit there is a
> > thrust to "Internet enable" appliances, such as dryers, ovens, and
> > stereos. Assuming ARIN fees persist, my first philosophical question is
> > whether any consumer of these appliances MUST periodically (e.g.,
> monthly)
> > drop coins in the ARIN fountain?
> > 
> > Thinking laterally, the reserved port space (<1024) is tight. Using the
> > same IP space conversation logic, should fees be charged to conserve
> port
> > space? If so, my second philosophiocal question is what is our role, as
> > protocol designers and IETF volunteers, in creating, what is slowly
> > becoming, an Internet consumption taxation model?
> > 
> > Imagine for a moment the effect of a fee against the allocation or use
> of
> > port 80 or 443, maybe even port 25 or 53.
> > 
> > 
> > 
> > 
> > 

RE: Addresses and ports and taxes -- oh my!

[Parkinson, Jonathan]

Err I think that would take some thinking about ? How many houses are there
in the world!

-Original Message-
From: Rakers, Jason [mailto:[EMAIL PROTECTED]]
Sent: Thursday, August 03, 2000 2:41 PM
To: 'Dennis Glatting'; [EMAIL PROTECTED]
Subject: RE: Addresses and ports and taxes -- oh my!


When household appliances begin becoming IP addressable, I think we will see
a move towards assigning an Internet IP address per household (much like
today's street address).  The household will perform NAT for all devices
within (one street address can house many people, not just one).

> -Original Message-
> From: Dennis Glatting [SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, August 03, 2000 8:32 AM
> To:   [EMAIL PROTECTED]
> Subject:  Addresses and ports and taxes -- oh my!
> 
> 
> I've been thinking about the issue of ARIN fees from last night's plenary
> and arrived at two philosophical questions.
> 
> I run my business out of my home and my DSL link is an important part of
> my business. About six months ago my ISP started charging me a $20/mo. fee
> for my /27 because "ARIN is now charging us." I am unhappy about this fee
> but I understand its motivation -- conversation of IP space, though I
> believe fees do not really effect the true wasters of this space and the
> fee, or as it is called in some circles, a tax, is probably misguided.
> Nonetheless, with IPv6, I naively hoped, until last night, the
> conservation of space issues would go away, and thus the fees. Big duh!
> 
> If we look at today's marketing hype and think forward a bit there is a
> thrust to "Internet enable" appliances, such as dryers, ovens, and
> stereos. Assuming ARIN fees persist, my first philosophical question is
> whether any consumer of these appliances MUST periodically (e.g., monthly)
> drop coins in the ARIN fountain?
> 
> Thinking laterally, the reserved port space (<1024) is tight. Using the
> same IP space conversation logic, should fees be charged to conserve port
> space? If so, my second philosophiocal question is what is our role, as
> protocol designers and IETF volunteers, in creating, what is slowly
> becoming, an Internet consumption taxation model?
> 
> Imagine for a moment the effect of a fee against the allocation or use of
> port 80 or 443, maybe even port 25 or 53.
> 
> 
> 
> 
> 

RE: Addresses and ports and taxes -- oh my!

[Evstiounin, Mikhail]

Wasn't avoiding NAT one of the goal of IPv6? I recall a pretty big
discussion here some time ago about NAT and IPv6. 

> -Original Message-
> From: Rakers, Jason [SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, August 03, 2000 9:41 AM
> To:   'Dennis Glatting'; [EMAIL PROTECTED]
> Subject:      RE: Addresses and ports and taxes -- oh my!
> 
> When household appliances begin becoming IP addressable, I think we will
> see
> a move towards assigning an Internet IP address per household (much like
> today's street address).  The household will perform NAT for all devices
> within (one street address can house many people, not just one).
> 
> > -Original Message-
> > From:   Dennis Glatting
> [SMTP:[EMAIL PROTECTED]]
> > Sent:   Thursday, August 03, 2000 8:32 AM
> > To: [EMAIL PROTECTED]
> > Subject:Addresses and ports and taxes -- oh my!
> > 
> > 
> > I've been thinking about the issue of ARIN fees from last night's
> plenary
> > and arrived at two philosophical questions.
> > 
> > I run my business out of my home and my DSL link is an important part of
> > my business. About six months ago my ISP started charging me a $20/mo.
> fee
> > for my /27 because "ARIN is now charging us." I am unhappy about this
> fee
> > but I understand its motivation -- conversation of IP space, though I
> > believe fees do not really effect the true wasters of this space and the
> > fee, or as it is called in some circles, a tax, is probably misguided.
> > Nonetheless, with IPv6, I naively hoped, until last night, the
> > conservation of space issues would go away, and thus the fees. Big duh!
> > 
> > If we look at today's marketing hype and think forward a bit there is a
> > thrust to "Internet enable" appliances, such as dryers, ovens, and
> > stereos. Assuming ARIN fees persist, my first philosophical question is
> > whether any consumer of these appliances MUST periodically (e.g.,
> monthly)
> > drop coins in the ARIN fountain?
> > 
> > Thinking laterally, the reserved port space (<1024) is tight. Using the
> > same IP space conversation logic, should fees be charged to conserve
> port
> > space? If so, my second philosophiocal question is what is our role, as
> > protocol designers and IETF volunteers, in creating, what is slowly
> > becoming, an Internet consumption taxation model?
> > 
> > Imagine for a moment the effect of a fee against the allocation or use
> of
> > port 80 or 443, maybe even port 25 or 53.
> > 
> > 
> > 
> > 
> > 

Re: Addresses and ports and taxes -- oh my!

[Brian E Carpenter]

"Dawson, Peter D" wrote:
> 
> ->-Original Message-
> ->From: Dennis Glatting [mailto:[EMAIL PROTECTED]]
> ->Sent: Thursday, August 03, 2000 8:32 AM
> ->To: [EMAIL PROTECTED]
> ->Subject: Addresses and ports and taxes -- oh my!
> ->
> ->
> 
> ->Nonetheless, with IPv6, I naively hoped, until last night, the
> ->conservation of space issues would go away, and thus the
> ->fees. Big duh!
> 
> ARIN ..still needs to delegate/admin the space.. costs will be incured.

To be completely clear, the conservation argument has gone away as
a constraint of equal importance with route aggregation, and in IPv6
we can give route aggregation much more importance than conservation
in the allocation policies. But an allocation service still costs
money and somebody has to pay for it.

Some of the early deployment scenarios for IPv6 are intended to bypass
this issue; for example see draft-ietf-ngtrans-6to4-06.txt .

  Brian

RE: Addresses and ports and taxes -- oh my!

[Rakers, Jason]

When household appliances begin becoming IP addressable, I think we will see
a move towards assigning an Internet IP address per household (much like
today's street address).  The household will perform NAT for all devices
within (one street address can house many people, not just one).

> -Original Message-
> From: Dennis Glatting [SMTP:[EMAIL PROTECTED]]
> Sent: Thursday, August 03, 2000 8:32 AM
> To:   [EMAIL PROTECTED]
> Subject:  Addresses and ports and taxes -- oh my!
> 
> 
> I've been thinking about the issue of ARIN fees from last night's plenary
> and arrived at two philosophical questions.
> 
> I run my business out of my home and my DSL link is an important part of
> my business. About six months ago my ISP started charging me a $20/mo. fee
> for my /27 because "ARIN is now charging us." I am unhappy about this fee
> but I understand its motivation -- conversation of IP space, though I
> believe fees do not really effect the true wasters of this space and the
> fee, or as it is called in some circles, a tax, is probably misguided.
> Nonetheless, with IPv6, I naively hoped, until last night, the
> conservation of space issues would go away, and thus the fees. Big duh!
> 
> If we look at today's marketing hype and think forward a bit there is a
> thrust to "Internet enable" appliances, such as dryers, ovens, and
> stereos. Assuming ARIN fees persist, my first philosophical question is
> whether any consumer of these appliances MUST periodically (e.g., monthly)
> drop coins in the ARIN fountain?
> 
> Thinking laterally, the reserved port space (<1024) is tight. Using the
> same IP space conversation logic, should fees be charged to conserve port
> space? If so, my second philosophiocal question is what is our role, as
> protocol designers and IETF volunteers, in creating, what is slowly
> becoming, an Internet consumption taxation model?
> 
> Imagine for a moment the effect of a fee against the allocation or use of
> port 80 or 443, maybe even port 25 or 53.
> 
> 
> 
> 
> 

RE: Addresses and ports and taxes -- oh my!

[Dawson, Peter D]

->-Original Message-
->From: Dennis Glatting [mailto:[EMAIL PROTECTED]]
->Sent: Thursday, August 03, 2000 8:32 AM
->To: [EMAIL PROTECTED]
->Subject: Addresses and ports and taxes -- oh my!
->
->

->Nonetheless, with IPv6, I naively hoped, until last night, the
->conservation of space issues would go away, and thus the 
->fees. Big duh!

ARIN ..still needs to delegate/admin the space.. costs will be incured.

->
->If we look at today's marketing hype and think forward a bit 
->there is a
->thrust to "Internet enable" appliances, such as dryers, ovens, and
->stereos. Assuming ARIN fees persist, my first philosophical 
->question is
->whether any consumer of these appliances MUST periodically 
->(e.g., monthly)
->drop coins in the ARIN fountain?

what does the appliance have to do with a /32 or /28 ??

->Imagine for a moment the effect of a fee against the 
->allocation or use of
->port 80 or 443, maybe even port 25 or 53.

Does IANA charge for  port assignment numbers ?? 

Addresses and ports and taxes -- oh my!

[Dennis Glatting]

I've been thinking about the issue of ARIN fees from last night's plenary
and arrived at two philosophical questions.

I run my business out of my home and my DSL link is an important part of
my business. About six months ago my ISP started charging me a $20/mo. fee
for my /27 because "ARIN is now charging us." I am unhappy about this fee
but I understand its motivation -- conversation of IP space, though I
believe fees do not really effect the true wasters of this space and the
fee, or as it is called in some circles, a tax, is probably misguided.
Nonetheless, with IPv6, I naively hoped, until last night, the
conservation of space issues would go away, and thus the fees. Big duh!

If we look at today's marketing hype and think forward a bit there is a
thrust to "Internet enable" appliances, such as dryers, ovens, and
stereos. Assuming ARIN fees persist, my first philosophical question is
whether any consumer of these appliances MUST periodically (e.g., monthly)
drop coins in the ARIN fountain?

Thinking laterally, the reserved port space (<1024) is tight. Using the
same IP space conversation logic, should fees be charged to conserve port
space? If so, my second philosophiocal question is what is our role, as
protocol designers and IETF volunteers, in creating, what is slowly
becoming, an Internet consumption taxation model?

Imagine for a moment the effect of a fee against the allocation or use of
port 80 or 443, maybe even port 25 or 53.