From jtk at depaul.edu  Fri Oct  8 14:39:32 2010
From: jtk at depaul.edu (John Kristoff)
Date: Fri, 8 Oct 2010 16:39:32 -0500
Subject: [ih] Historic DNS root zone files
Message-ID: <20101008213932.GA23315@condor.depaul.edu>

Hi folks,

I'm looking for copies of the root zone file (not HOSTS.TXT) prior to
July 1999.  DNS-OARC has copies as far back as then and I have access to
that data already.  If anyone can share a pointer or make what they have
available, that would be greatly appreciated.

Thank you very much,

John


From ernesto at cs.fiu.edu  Mon Oct 11 13:59:04 2010
From: ernesto at cs.fiu.edu (Ernie Rubi)
Date: Mon, 11 Oct 2010 16:59:04 -0400
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>

Hi folks,

I am in the process of writing a research paper for an Internet Law seminar at FIU Law and have a few questions for those of you with enough institutional memory to remember how IPv4 address allocations were first handed out:

1.  Who 'owned' IP addresses ab initio?  Were IP addresses 'property' of any one entity or person or agency?  What is the authority ICANN /  IANA had to allocate these addresses if they are not 'theirs.'

2.  Initially, were large blocks of IPv4 addresses 'handed out' with a complete ownership interest to their recipients?  For example, when you received an /16, was it yours to transfer to 	      	other entities if you pleased?  Could you have transferred sub-allocations of your /16 to other entities who weren't your customers/connectors?

3.  Were the initial IPv4 allocations rolled into RIRs/ICANN at any point? If so, under what legal framework?

Thanks all for your thoughts/comments,

Ernesto M. Rubi
Sr. Network Engineer
AMPATH/CIARA
Florida International Univ, Miami
Reply-to: ernesto at cs.fiu.edu

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20101011/5d38c906/attachment.htm>

From eric.gade at gmail.com  Mon Oct 11 14:52:38 2010
From: eric.gade at gmail.com (Eric Gade)
Date: Mon, 11 Oct 2010 22:52:38 +0100
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <AANLkTikk3cPTfNZfTUJ3rBvaCR9Gvv8Yy-rcJ8+AgR4e@mail.gmail.com>

I'm no expert on IP addressing issues, but you may want to consult Laura
DeNardis' new-ish book *Protocol
Politics<http://mitpress.mit.edu/catalog/item/default.asp?ttype=2&tid=11893>
.*
*
*
*I think, however, you're going to get into murky territory when it comes to
attributing the legal status of these things early on.  But I believe the
basic idea is this: because TCP/IP was created at (at first used by)
DDN/ARPA, the administrative authorities for those networks distributed the
addresses (IANA was really just Postel under contract with the DoD, though
it is my understanding that the NIC had this duty for a while).*
*
*
*In terms of issues with distribution of large IP address blocks of
different classes, again, DeNardis does a really good job breaking it down
in her book.  It's a quick read too -- you can probably pound through it in
a day.
*
On Mon, Oct 11, 2010 at 9:59 PM, Ernie Rubi <ernesto at cs.fiu.edu> wrote:

> Hi folks,
>
> I am in the process of writing a research paper for an Internet Law seminar
> at FIU Law and have a few questions for those of you with enough
> institutional memory to remember how IPv4 address allocations were first
> handed out:
>
> 1.  Who 'owned' IP addresses ab initio?  Were IP addresses 'property' of
> any one entity or person or agency?  What is the authority ICANN /  IANA had
> to allocate these addresses if they are not 'theirs.'
>
> 2.  Initially, were large blocks of IPv4 addresses 'handed out' with a
> complete ownership interest to their recipients?  For example, when you
> received an /16, was it yours to transfer to         other entities if you
> pleased?  Could you have transferred sub-allocations of your /16 to other
> entities who weren't your customers/connectors?
>
> 3.  Were the initial IPv4 allocations rolled into RIRs/ICANN at any point?
> If so, under what legal framework?
>
> Thanks all for your thoughts/comments,
>
> Ernesto M. Rubi
> Sr. Network Engineer
> AMPATH/CIARA
> Florida International Univ, Miami
> Reply-to: ernesto at cs.fiu.edu
>
>


-- 
Eric
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20101011/ded3838a/attachment.htm>

From ernesto at cs.fiu.edu  Mon Oct 11 15:01:59 2010
From: ernesto at cs.fiu.edu (Ernie Rubi)
Date: Mon, 11 Oct 2010 18:01:59 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTikk3cPTfNZfTUJ3rBvaCR9Gvv8Yy-rcJ8+AgR4e@mail.gmail.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<AANLkTikk3cPTfNZfTUJ3rBvaCR9Gvv8Yy-rcJ8+AgR4e@mail.gmail.com>
Message-ID: <791E426D-6BB7-45E9-BA42-490AE5A214AF@cs.fiu.edu>

Thanks, on my way to the library for the book.

ARIN has these Legacy RSA's that they want folks to sign; although I don't know who would if they were trying to make a case that IP addresses were "theirs."  ARIN also has told me that IP addresses 'don't belong to anyone.' 

That last statement I find funny - a legal entity (ARIN) contracts to allocate/assign IP addresses that it doesn't 'own' but says you can't give the IP addresses it assigns you to others.  

ARIN also says you can get in legal trouble for prefix hijacking, although those who have the IP address don't 'own' it and ARIN doesn't 'own' it.  It seems that the only legal claims could be brought by the US government (or whomever was the 'contractor' for IPv4 address development).

Wonder if Vint Cerf is around these parts to shed some more insight on the legal aspect of v4 'ownership.'  This research paper was inspired in part by the IGF talk/exchange on IPv4 'markets' that he was part of just last month.

Ernie

On Oct 11, 2010, at 5:52 PM, Eric Gade wrote:

> I'm no expert on IP addressing issues, but you may want to consult Laura DeNardis' new-ish book Protocol Politics.
> 
> I think, however, you're going to get into murky territory when it comes to attributing the legal status of these things early on.  But I believe the basic idea is this: because TCP/IP was created at (at first used by) DDN/ARPA, the administrative authorities for those networks distributed the addresses (IANA was really just Postel under contract with the DoD, though it is my understanding that the NIC had this duty for a while).
> 
> In terms of issues with distribution of large IP address blocks of different classes, again, DeNardis does a really good job breaking it down in her book.  It's a quick read too -- you can probably pound through it in a day.
> 
> On Mon, Oct 11, 2010 at 9:59 PM, Ernie Rubi <ernesto at cs.fiu.edu> wrote:
> Hi folks,
> 
> I am in the process of writing a research paper for an Internet Law seminar at FIU Law and have a few questions for those of you with enough institutional memory to remember how IPv4 address allocations were first handed out:
> 
> 1.  Who 'owned' IP addresses ab initio?  Were IP addresses 'property' of any one entity or person or agency?  What is the authority ICANN /  IANA had to allocate these addresses if they are not 'theirs.'
> 
> 2.  Initially, were large blocks of IPv4 addresses 'handed out' with a complete ownership interest to their recipients?  For example, when you received an /16, was it yours to transfer to 	      	other entities if you pleased?  Could you have transferred sub-allocations of your /16 to other entities who weren't your customers/connectors?
> 
> 3.  Were the initial IPv4 allocations rolled into RIRs/ICANN at any point? If so, under what legal framework?
> 
> Thanks all for your thoughts/comments,
> 
> Ernesto M. Rubi
> Sr. Network Engineer
> AMPATH/CIARA
> Florida International Univ, Miami
> Reply-to: ernesto at cs.fiu.edu
> 
> 
> 
> 
> -- 
> Eric


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20101011/f6dc5f13/attachment.htm>

From woody at pch.net  Mon Oct 11 15:14:35 2010
From: woody at pch.net (Bill Woodcock)
Date: Mon, 11 Oct 2010 15:14:35 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>


On Oct 11, 2010, at 1:59 PM, Ernie Rubi wrote:
> 1.  Who 'owned' IP addresses ab initio?

No one.  IP addresses are just addresses, they're not owned.  Their interesting property is their unique assignment, not some intrinsic value.

> Were IP addresses 'property' of any one entity or person or agency?

No.

> What is the authority ICANN /  IANA had to allocate these addresses if they are not 'theirs.'

There was a chain of responsibility that was handed down from DARPA / DODNIC / IANA / RIRs, by agreement between the USG and the Internet stakeholder community.

> 2.  Initially, were large blocks of IPv4 addresses 'handed out' with a complete ownership interest to their recipients?

No.  They were allocated based upon documented need, as described in RFC 2050.  The sizes of allocation blocks have changed over time, of particular note during the switch-over from classful to classless subnetting, and in a gradual increase in fineness of granularity since then.

> For example, when you received an /16, was it yours to transfer to other entities if you pleased?

No, that would violate the need-based principle upon which they were assigned.  You were always free to return the allocation to the allocating registry if and when it was no longer needed.  If someone else needs it, they document their need, present it to the registry, and get what they need.

> Could you have transferred sub-allocations of your /16 to other entities who weren't your customers/connectors?

You can sell crack on a street-corner, but that doesn't mean that people will think you're a respectable businessman.  The issue here is that if you sub-allocated to someone who you didn't actually route to, you (a) would have no documented need for it, and (b) wouldn't be keeping the delegation path following the routing path, so the prefix aggregation would break...  in short, there would be no point in it.  Like selling bicycles to fish.  Those other entities could, much more easily, just go their their own space directly from the registry, and it wouldn't come with all those problems, and everybody would be able to hold their heads up in polite society, and polite society would choose to route the prefix, because it wouldn't be violating aggregation, and wouldn't be chewing up resources on everybody else's router.

If you're really going to study these issues, you need to understand the economics of them, and not just gloss over it with the assumption that they're just numbers and any old random values can be attached to them.  The numbers aren't where the money is.  The routing is where the money is.  Just like buying and selling real-estate...  if you started making deals for street addresses, without looking at whether there was any land behind them, or where the land was, or how it could be reached, you'd lose your shirt very quickly.  The questions you're asking are about street addresses, rather than about land.

> 3.  Were the initial IPv4 allocations rolled into RIRs/ICANN at any point? If so, under what legal framework?

The Legacy Registration Services Agreement:   https://www.arin.net/resources/legacy/

                                -Bill








From cls at rkey.com  Mon Oct 11 15:38:02 2010
From: cls at rkey.com (Craig Simon)
Date: Mon, 11 Oct 2010 18:38:02 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <4CB391CA.1000509@rkey.com>

Hello Ernesto,

Much of the historical background is discussed in my Ph.D. dissertation, 
which you can find at http://www.rkey.com/essays/diss.pdf.

Craig Simon

On 10/11/2010 4:59 PM, Ernie Rubi wrote:
> Hi folks,
>
> I am in the process of writing a research paper for an Internet Law
> seminar at FIU Law and have a few questions for those of you with enough
> institutional memory to remember how IPv4 address allocations were first
> handed out:
>
> 1. Who 'owned' IP addresses ab initio? Were IP addresses 'property' of
> any one entity or person or agency? What is the authority ICANN / IANA
> had to allocate these addresses if they are not 'theirs.'
>
> 2. Initially, were large blocks of IPv4 addresses 'handed out' with a
> complete ownership interest to their recipients? For example, when you
> received an /16, was it yours to transfer to other entities if you
> pleased? Could you have transferred sub-allocations of your /16 to other
> entities who weren't your customers/connectors?
>
> 3. Were the initial IPv4 allocations rolled into RIRs/ICANN at any
> point? If so, under what legal framework?
>
> Thanks all for your thoughts/comments,
>
> Ernesto M. Rubi
> Sr. Network Engineer
> AMPATH/CIARA
> Florida International Univ, Miami
> Reply-to: ernesto at cs.fiu.edu <mailto:ernesto at cs.fiu.edu>
>


From craig at aland.bbn.com  Mon Oct 11 15:41:45 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Mon, 11 Oct 2010 18:41:45 -0400
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101011224145.2113728E137@aland.bbn.com>

> 1.  Who 'owned' IP addresses ab initio?  Were IP addresses 'property' of =
> any one entity or person or agency?  What is the authority ICANN /  IANA =
> had to allocate these addresses if they are not 'theirs.'

Back when I got started in 1983, you requested an IP address from the
DDN NIC.  You filled out a form explaining how big a network you were
creating and they gave you a class A, B, or C address (this is pre-CIDR).
As I recall, you were already warned at that time that getting a class A
(/8 in today's parlance) was very hard.  But getting a class B (/16) was
trivial.

> 2.  Initially, were large blocks of IPv4 addresses 'handed out' with a =
> complete ownership interest to their recipients?  For example, when you =
> received an /16, was it yours to transfer to 	      	other entities =
> if you pleased?  Could you have transferred sub-allocations of your /16 =
> to other entities who weren't your customers/connectors?

The concept of "ownership" never came up during the early 1980s.  You were
the "contact" for your IP networks.  And yes, if I was a contact for a
network, I could drop an email to the NIC and tell them to make someone
else the contact.  So I had some trappings of ownership -- but frankly it
didn't feel that way.

At one point I had three class C addresses (what are now called /24s)
for testing routing protocols.  After a few years I sent a note to
the NIC saying I no longer needed them and they went back into the address
pool.

> 3.  Were the initial IPv4 allocations rolled into RIRs/ICANN at any =
> point? If so, under what legal framework?

As I recall and others on this list were much closer to the process,
when ICANN got created, those who already had IP address blocks kept those
addresses outside the system -- squatters rights if you will.  And ICANN
and the RIRs spent some time trying persuade folks who already "possessed"
parts of the address space to please join within the agreements, with
limited success.

Thanks!

Craig

> 
> Thanks all for your thoughts/comments,
> 
> Ernesto M. Rubi
> Sr. Network Engineer
> AMPATH/CIARA
> Florida International Univ, Miami
> Reply-to: ernesto at cs.fiu.edu
> 
> 
> --Apple-Mail-1--714100781
> Content-Transfer-Encoding: quoted-printable
> Content-Type: text/html;
> 	charset=us-ascii
> 
> <html><head></head><body style=3D"word-wrap: break-word; =
> -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><span =
> class=3D"Apple-style-span" style=3D"font-size: 12px; ">Hi =
> folks,</span><div style=3D"font-size: 12px; "><br></div><div =
> style=3D"font-size: 12px; ">I am in the process of writing a research =
> paper for an Internet Law seminar at FIU Law and have a few questions =
> for those of you with enough institutional memory to remember how IPv4 =
> address allocations were first handed out:</div><div style=3D"font-size: =
> 12px; "><br></div><div style=3D"font-size: 12px; ">1. &nbsp;Who 'owned' =
> IP addresses ab initio? &nbsp;Were IP addresses 'property' of any one =
> entity or person or agency? &nbsp;What is the authority ICANN / =
> &nbsp;IANA had to allocate these addresses if they are not =
> 'theirs.'</div><div style=3D"font-size: 12px; "><br></div><div =
> style=3D"font-size: 12px; ">2. &nbsp;Initially, were large blocks of =
> IPv4 addresses 'handed out' with a complete ownership interest to their =
> recipients? &nbsp;For example, when you received an /16, was it yours to =
> transfer to&nbsp;<span class=3D"Apple-tab-span" style=3D"white-space: =
> pre; ">	</span>&nbsp;&nbsp; &nbsp; &nbsp;<span class=3D"Apple-tab-span"
>  =
> style=3D"white-space: pre; ">	</span>other entities if you pleased? =
> &nbsp;Could you have transferred sub-allocations of your /16 to other =
> entities who weren't your customers/connectors?</div><div =
> style=3D"font-size: 12px; "><br></div><div style=3D"font-size: 12px; =
> ">3. &nbsp;Were the initial IPv4 allocations rolled into RIRs/ICANN at =
> any point? If so, under what legal framework?</div><div =
> style=3D"font-size: 12px; "><br></div><div style=3D"font-size: 12px; =
> ">Thanks all for your thoughts/comments,</div><div style=3D"font-size: =
> 12px; "><br></div><div style=3D"font-size: 12px; "><div><span =
> class=3D"Apple-style-span" style=3D"border-collapse: separate; color: =
> rgb(0, 0, 0); font-family: Helvetica; font-size: 12px; font-style: =
> normal; font-variant: normal; font-weight: normal; letter-spacing: =
> normal; line-height: normal; orphans: 2; text-indent: 0px; =
> text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; =
> -webkit-border-horizontal-spacing: 0px; -webkit-border-vertical-spacing: =
> 0px; -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: =
> auto; -webkit-text-stroke-width: 0px; "><span class=3D"Apple-style-span" =
> style=3D"border-collapse: separate; color: rgb(0, 0, 0); font-family: =
> Helvetica; font-size: 12px; font-style: normal; font-variant: normal; =
> font-weight: normal; letter-spacing: normal; line-height: normal; =
> orphans: 2; text-indent: 0px; text-transform: none; white-space: normal; =
> widows: 2; word-spacing: 0px; -webkit-border-horizontal-spacing: 0px; =
> -webkit-border-vertical-spacing: 0px; =
> -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: =
> auto; -webkit-text-stroke-width: 0px; "><div style=3D"word-wrap: =
> break-word; -webkit-nbsp-mode: space; -webkit-line-break: =
> after-white-space; "><div><div>Ernesto M. Rubi</div><div>Sr. Network =
> Engineer</div><div>AMPATH/CIARA</div><div>Florida International Univ, =
> Miami</div><div>Reply-to:&nbsp;<a =
> href=3D"mailto:ernesto at cs.fiu.edu">ernesto at cs.fiu.edu</a></div></div><div>=
> <br></div></div></span></span></div></div></body></html>=
> 
> --Apple-Mail-1--714100781--
********************
Craig Partridge
Chief Scientist, BBN Technologies
E-mail: craig at aland.bbn.com or craig at bbn.com
Phone: +1 517 324 3425


From el at lisse.na  Mon Oct 11 16:21:17 2010
From: el at lisse.na (Dr Eberhard W Lisse)
Date: Tue, 12 Oct 2010 01:21:17 +0200
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>
Message-ID: <4CB39BED.6010007@lisse.na>

Bill,

this is a very interesting topic.

I would be very interested in reading documentation establishing
this chain.

greetings, el

On 2010-10-12 24:14 , Bill Woodcock wrote:
 >
 > On Oct 11, 2010, at 1:59 PM, Ernie Rubi wrote:
 >> 1.  Who 'owned' IP addresses ab initio?
 >
 > No one.  IP addresses are just addresses, they're not owned.  Their
 > interesting property is their unique assignment, not some intrinsic
 > value.
 >
 >> Were IP addresses 'property' of any one entity or person or agency?
 >
 > No.
 >
 >> What is the authority ICANN / IANA had to allocate these addresses if
 >> they are not 'theirs.'
 >
 > There was a chain of responsibility that was handed down from DARPA /
 > DODNIC / IANA / RIRs, by agreement between the USG and the Internet
 > stakeholder community.


From larrysheldon at cox.net  Mon Oct 11 16:58:39 2010
From: larrysheldon at cox.net (Larry Sheldon)
Date: Mon, 11 Oct 2010 18:58:39 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101011224145.2113728E137@aland.bbn.com>
References: <20101011224145.2113728E137@aland.bbn.com>
Message-ID: <4CB3A4AF.9080709@cox.net>

  On 10/11/2010 5:41 PM, Craig Partridge wrote:
>> 1.  Who 'owned' IP addresses ab initio?  Were IP addresses 'property' of =
>> any one entity or person or agency?  What is the authority ICANN /  IANA =
>> had to allocate these addresses if they are not 'theirs.'
> Back when I got started in 1983, you requested an IP address from the
> DDN NIC.  You filled out a form explaining how big a network you were
> creating and they gave you a class A, B, or C address (this is pre-CIDR).
> As I recall, you were already warned at that time that getting a class A
> (/8 in today's parlance) was very hard.  But getting a class B (/16) was
> trivial.
>
>> 2.  Initially, were large blocks of IPv4 addresses 'handed out' with a =
>> complete ownership interest to their recipients?  For example, when you =
>> received an /16, was it yours to transfer to 	      	other entities =
>> if you pleased?  Could you have transferred sub-allocations of your /16 =
>> to other entities who weren't your customers/connectors?
> The concept of "ownership" never came up during the early 1980s.  You were
> the "contact" for your IP networks.  And yes, if I was a contact for a
> network, I could drop an email to the NIC and tell them to make someone
> else the contact.  So I had some trappings of ownership -- but frankly it
> didn't feel that way.
>
> At one point I had three class C addresses (what are now called /24s)
> for testing routing protocols.  After a few years I sent a note to
> the NIC saying I no longer needed them and they went back into the address
> pool.
>
>> 3.  Were the initial IPv4 allocations rolled into RIRs/ICANN at any =
>> point? If so, under what legal framework?
> As I recall and others on this list were much closer to the process,
> when ICANN got created, those who already had IP address blocks kept those
> addresses outside the system -- squatters rights if you will.  And ICANN
> and the RIRs spent some time trying persuade folks who already "possessed"
> parts of the address space to please join within the agreements, with
> limited success.

My recollections as an obscure administrator at an obscure university, 
are that we owned our IP addresses, but that there was no market for them.



> Thanks!
>
> Craig
>
>> Thanks all for your thoughts/comments,
>>
>> Ernesto M. Rubi
>> Sr. Network Engineer
>> AMPATH/CIARA
>> Florida International Univ, Miami
>> Reply-to: ernesto at cs.fiu.edu
>>
>>
>> --Apple-Mail-1--714100781
>> Content-Transfer-Encoding: quoted-printable
>> Content-Type: text/html;
>> 	charset=us-ascii
>>
>> <html><head></head><body style=3D"word-wrap: break-word; =
>> -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><span =
>> class=3D"Apple-style-span" style=3D"font-size: 12px; ">Hi =
>> folks,</span><div style=3D"font-size: 12px; "><br></div><div =
>> style=3D"font-size: 12px; ">I am in the process of writing a research =
>> paper for an Internet Law seminar at FIU Law and have a few questions =
>> for those of you with enough institutional memory to remember how IPv4 =
>> address allocations were first handed out:</div><div style=3D"font-size: =
>> 12px; "><br></div><div style=3D"font-size: 12px; ">1.&nbsp;Who 'owned' =
>> IP addresses ab initio?&nbsp;Were IP addresses 'property' of any one =
>> entity or person or agency?&nbsp;What is the authority ICANN / =
>> &nbsp;IANA had to allocate these addresses if they are not =
>> 'theirs.'</div><div style=3D"font-size: 12px; "><br></div><div =
>> style=3D"font-size: 12px; ">2.&nbsp;Initially, were large blocks of =
>> IPv4 addresses 'handed out' with a complete ownership interest to their =
>> recipients?&nbsp;For example, when you received an /16, was it yours to =
>> transfer to&nbsp;<span class=3D"Apple-tab-span" style=3D"white-space: =
>> pre; ">	</span>&nbsp;&nbsp;&nbsp;&nbsp;<span class=3D"Apple-tab-span"
>>   =
>> style=3D"white-space: pre; ">	</span>other entities if you pleased? =
>> &nbsp;Could you have transferred sub-allocations of your /16 to other =
>> entities who weren't your customers/connectors?</div><div =
>> style=3D"font-size: 12px; "><br></div><div style=3D"font-size: 12px; =
>> ">3.&nbsp;Were the initial IPv4 allocations rolled into RIRs/ICANN at =
>> any point? If so, under what legal framework?</div><div =
>> style=3D"font-size: 12px; "><br></div><div style=3D"font-size: 12px; =
>> ">Thanks all for your thoughts/comments,</div><div style=3D"font-size: =
>> 12px; "><br></div><div style=3D"font-size: 12px; "><div><span =
>> class=3D"Apple-style-span" style=3D"border-collapse: separate; color: =
>> rgb(0, 0, 0); font-family: Helvetica; font-size: 12px; font-style: =
>> normal; font-variant: normal; font-weight: normal; letter-spacing: =
>> normal; line-height: normal; orphans: 2; text-indent: 0px; =
>> text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; =
>> -webkit-border-horizontal-spacing: 0px; -webkit-border-vertical-spacing: =
>> 0px; -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: =
>> auto; -webkit-text-stroke-width: 0px; "><span class=3D"Apple-style-span" =
>> style=3D"border-collapse: separate; color: rgb(0, 0, 0); font-family: =
>> Helvetica; font-size: 12px; font-style: normal; font-variant: normal; =
>> font-weight: normal; letter-spacing: normal; line-height: normal; =
>> orphans: 2; text-indent: 0px; text-transform: none; white-space: normal; =
>> widows: 2; word-spacing: 0px; -webkit-border-horizontal-spacing: 0px; =
>> -webkit-border-vertical-spacing: 0px; =
>> -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: =
>> auto; -webkit-text-stroke-width: 0px; "><div style=3D"word-wrap: =
>> break-word; -webkit-nbsp-mode: space; -webkit-line-break: =
>> after-white-space; "><div><div>Ernesto M. Rubi</div><div>Sr. Network =
>> Engineer</div><div>AMPATH/CIARA</div><div>Florida International Univ, =
>> Miami</div><div>Reply-to:&nbsp;<a =
>> href=3D"mailto:ernesto at cs.fiu.edu">ernesto at cs.fiu.edu</a></div></div><div>=
>> <br></div></div></span></span></div></div></body></html>=
>>
>> --Apple-Mail-1--714100781--
> ********************
> Craig Partridge
> Chief Scientist, BBN Technologies
> E-mail: craig at aland.bbn.com or craig at bbn.com
> Phone: +1 517 324 3425
>



From randy at psg.com  Mon Oct 11 17:05:29 2010
From: randy at psg.com (Randy Bush)
Date: Tue, 12 Oct 2010 09:05:29 +0900
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB39BED.6010007@lisse.na>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>
	<4CB39BED.6010007@lisse.na>
Message-ID: <m2ocb0l1h2.wl%randy@psg.com>

> I would be very interested in reading documentation establishing
> this chain.
>> There was a chain of responsibility that was handed down from DARPA /
>> DODNIC / IANA / RIRs, by agreement between the USG and the Internet
>> stakeholder community.

don't hold your breath waiting for it.  wishful thinking from someone
who wasn't there.

randy


From jeanjour at comcast.net  Mon Oct 11 17:14:49 2010
From: jeanjour at comcast.net (John Day)
Date: Mon, 11 Oct 2010 20:14:49 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101011224145.2113728E137@aland.bbn.com>
References: <20101011224145.2113728E137@aland.bbn.com>
Message-ID: <a06240823c8d956960d5e@[168.122.9.70]>

Craig has it pretty close to right.  And before that, it was even 
more informal.

But let me just say that this whole idea of "owning network 
addresses" is ludicrous. Clearly a misconception by people who have 
no clue.

Do you own the address where your house is?  When you move do you 
take it with you?  Do you get your mail at the address where you were 
born?  (the analog for the so-called MAC "address").

Network addresses are no different.  And if you think they are, then 
you need to think it about it more carefully.

At 18:41 -0400 2010/10/11, Craig Partridge wrote:
>  > 1.  Who 'owned' IP addresses ab initio?  Were IP addresses 'property' of =
>>  any one entity or person or agency?  What is the authority ICANN /  IANA =
>>  had to allocate these addresses if they are not 'theirs.'
>
>Back when I got started in 1983, you requested an IP address from the
>DDN NIC.  You filled out a form explaining how big a network you were
>creating and they gave you a class A, B, or C address (this is pre-CIDR).
>As I recall, you were already warned at that time that getting a class A
>(/8 in today's parlance) was very hard.  But getting a class B (/16) was
>trivial.
>
>>  2.  Initially, were large blocks of IPv4 addresses 'handed out' with a =
>>  complete ownership interest to their recipients?  For example, when you =
>>  received an /16, was it yours to transfer to	     	other entities =
>>  if you pleased?  Could you have transferred sub-allocations of your /16 =
>>  to other entities who weren't your customers/connectors?
>
>The concept of "ownership" never came up during the early 1980s.  You were
>the "contact" for your IP networks.  And yes, if I was a contact for a
>network, I could drop an email to the NIC and tell them to make someone
>else the contact.  So I had some trappings of ownership -- but frankly it
>didn't feel that way.
>
>At one point I had three class C addresses (what are now called /24s)
>for testing routing protocols.  After a few years I sent a note to
>the NIC saying I no longer needed them and they went back into the address
>pool.
>
>>  3.  Were the initial IPv4 allocations rolled into RIRs/ICANN at any =
>>  point? If so, under what legal framework?
>
>As I recall and others on this list were much closer to the process,
>when ICANN got created, those who already had IP address blocks kept those
>addresses outside the system -- squatters rights if you will.  And ICANN
>and the RIRs spent some time trying persuade folks who already "possessed"
>parts of the address space to please join within the agreements, with
>limited success.
>
>Thanks!
>
>Craig
>
>>
>>  Thanks all for your thoughts/comments,
>>
>>  Ernesto M. Rubi
>>  Sr. Network Engineer
>>  AMPATH/CIARA
>>  Florida International Univ, Miami
>>  Reply-to: ernesto at cs.fiu.edu
>>
>>
>>  --Apple-Mail-1--714100781
>>  Content-Transfer-Encoding: quoted-printable
>>  Content-Type: text/html;
>>	charset=us-ascii
>>
>>  <html><head></head><body style=3D"word-wrap: break-word; =
>>  -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><span =
>>  class=3D"Apple-style-span" style=3D"font-size: 12px; ">Hi =
>>  folks,</span><div style=3D"font-size: 12px; "><br></div><div =
>>  style=3D"font-size: 12px; ">I am in the process of writing a research =
>>  paper for an Internet Law seminar at FIU Law and have a few questions =
>>  for those of you with enough institutional memory to remember how IPv4 =
>>  address allocations were first handed out:</div><div style=3D"font-size: =
>>  12px; "><br></div><div style=3D"font-size: 12px; ">1. &nbsp;Who 'owned' =
>>  IP addresses ab initio? &nbsp;Were IP addresses 'property' of any one =
>>  entity or person or agency? &nbsp;What is the authority ICANN / =
>>  &nbsp;IANA had to allocate these addresses if they are not =
>>  'theirs.'</div><div style=3D"font-size: 12px; "><br></div><div =
>>  style=3D"font-size: 12px; ">2. &nbsp;Initially, were large blocks of =
>>  IPv4 addresses 'handed out' with a complete ownership interest to their =
>  > recipients? &nbsp;For example, when you received an /16, was it yours to =
>>  transfer to&nbsp;<span class=3D"Apple-tab-span" style=3D"white-space: =
>>  pre; ">	</span>&nbsp;&nbsp; &nbsp; &nbsp;<span class=3D"Apple-tab-span"
>>   =
>>  style=3D"white-space: pre; ">	</span>other entities if you pleased? =
>>  &nbsp;Could you have transferred sub-allocations of your /16 to other =
>>  entities who weren't your customers/connectors?</div><div =
>>  style=3D"font-size: 12px; "><br></div><div style=3D"font-size: 12px; =
>>  ">3. &nbsp;Were the initial IPv4 allocations rolled into RIRs/ICANN at =
>>  any point? If so, under what legal framework?</div><div =
>>  style=3D"font-size: 12px; "><br></div><div style=3D"font-size: 12px; =
>>  ">Thanks all for your thoughts/comments,</div><div style=3D"font-size: =
>>  12px; "><br></div><div style=3D"font-size: 12px; "><div><span =
>>  class=3D"Apple-style-span" style=3D"border-collapse: separate; color: =
>>  rgb(0, 0, 0); font-family: Helvetica; font-size: 12px; font-style: =
>>  normal; font-variant: normal; font-weight: normal; letter-spacing: =
>>  normal; line-height: normal; orphans: 2; text-indent: 0px; =
>>  text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; =
>>  -webkit-border-horizontal-spacing: 0px; -webkit-border-vertical-spacing: =
>>  0px; -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: =
>>  auto; -webkit-text-stroke-width: 0px; "><span class=3D"Apple-style-span" =
>>  style=3D"border-collapse: separate; color: rgb(0, 0, 0); font-family: =
>>  Helvetica; font-size: 12px; font-style: normal; font-variant: normal; =
>>  font-weight: normal; letter-spacing: normal; line-height: normal; =
>>  orphans: 2; text-indent: 0px; text-transform: none; white-space: normal; =
>>  widows: 2; word-spacing: 0px; -webkit-border-horizontal-spacing: 0px; =
>>  -webkit-border-vertical-spacing: 0px; =
>>  -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: =
>>  auto; -webkit-text-stroke-width: 0px; "><div style=3D"word-wrap: =
>>  break-word; -webkit-nbsp-mode: space; -webkit-line-break: =
>>  after-white-space; "><div><div>Ernesto M. Rubi</div><div>Sr. Network =
>>  Engineer</div><div>AMPATH/CIARA</div><div>Florida International Univ, =
>>  Miami</div><div>Reply-to:&nbsp;<a =
>>  href=3D"mailto:ernesto at cs.fiu.edu">ernesto at cs.fiu.edu</a></div></div><div>=
>>  <br></div></div></span></span></div></div></body></html>=
>>
>>  --Apple-Mail-1--714100781--
>********************
>Craig Partridge
>Chief Scientist, BBN Technologies
>E-mail: craig at aland.bbn.com or craig at bbn.com
>Phone: +1 517 324 3425



From jnc at mercury.lcs.mit.edu  Mon Oct 11 17:31:21 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Mon, 11 Oct 2010 20:31:21 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101012003121.B7B346BE5AD@mercury.lcs.mit.edu>

    > From: Ernie Rubi <ernesto at cs.fiu.edu>

    > have a few questions for those of you with enough institutional memory
    > to remember how IPv4 address allocations were first handed out:
    > 1. Who 'owned' IP addresses ab initio? Were IP addresses 'property' of
    > any one entity or person or agency? What is the authority ICANN / IANA
    > had to allocate these addresses if they are not 'theirs.'

Well, do remember that until 1992 or so, when the Internet was opened to the
general public, for commercial use, the only people who were supposed to be
connected to it (as with the ARPANET) were either US government entities, or
outside entities with a government contract which required use of the Internet
to perform. (The controls on who connected up to the ARPANET were more formal,
in large part because the technology made it much harder to delegate
connectivity control there.) So in some sense all Internet addresses
definitely 'belonged' to the US government before that.

In the very earliest days, of course, there were only class A network
numbers. I don't recall who applied for the MIT number (net 18 - which would
have been circa 1978) - it might have been me, it might have been Dave Clark
{... checks ...} RFC-790 says it was Dave. I suspect we just sent email to
Jon Postel, and he looked in the draft version of "Assigned Numbers" which he
kept in his disk directory on ISI{mumble}, gave you the next number, and
wrote it down there! :-) Jon was an employee of ISI, which was on a DARPA
contact, at the time.

The whole process was distinctly informal, compared to connecting to the
ARPANET, both then, and for years afterward. I remember getting Proteon
connected up (via a point-point serial link to MIT) - which would have been
circa 1986 or so - it was very informal: I got a number (from Jon, I think,
but I'm not positive - and maybe it was John Shriver, not me), stuck some
numbers in the appropriate MIT routers (which I was helping run then), and
that was it.

	Noel


From vint at google.com  Mon Oct 11 17:35:10 2010
From: vint at google.com (Vint Cerf)
Date: Mon, 11 Oct 2010 20:35:10 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <AANLkTinkmYv2cEeC0p=dqLrwnuVrOsUfBi95vfYA0oJS@mail.gmail.com>

the initial address space was decided on in 1977. initially just 256
blocks, but very soon  in 1981, this format was overtaken by the need
for more refined allocations and the "ABCD" format was adopted. Jon
Postel initially handled the tracking of assignments as the Internet
design evolved, expanding on his role as "numbers czar" for the
ARPANET becoming the Internet Assigned Numbers Authority (IANA) for
the Internet while serving at USC-ISI under contract to DARPA. Recall
that this was an experimental system.

In 1983, the ARPANET was split into the research ARPANET and the
military MILNET and all hosts were converted to run TCP/IP so they
needed IP addresses. Somewhere in the 1993 time frame, Classless
Inter-Domain Routing was introduced using a new version of Border
Gateway Protocol. All during this time, Postel continued to serve as
the IANA.

Around 1992, the European Internet contingent formed the RIPE-NCC to
support the Reseaux IP Europeen (European IP networks). The "Network
Control Center" undertook to make IP address assignments from blocks
supplied by Postel.

In 1997, the Americas Registry for Internet Numbers (ARIN) was formed.
They undertook to manage addreiss allocations from blocks provided by
Postel for the Americas (North/South/Central). Postel contnued to
manage allocations not handled by either ARIN or RIPE-NCC (which was
serving some non-European locations). In 1998, after considerable
turmoil, the Internet Corporation for Assigned Names and Numbers
(ICANN) and inherited the IANA function by way of the Department of
Commerce which had taken over from the National Science Foundation the
task of Address and Domain Name management.

Postel died in September 1998 just two weeks before the official
founding of ICANN.

through the ICANN Process, new Internet Registries were formed at Asia
Pac (APNIC), Latin American (LACNIC) and Africa (AFRINIC).

As to "ownership" I think the terms is more like "leaseholder" - The
authority went from DARPA to DCA/DISC, to NSF, to Department of
Commerce and then to ICANN via DOC.

Pre-ARIN and RIPE-NCC allocations are sometimes called "legacy"
assignments because the parties holding these addresses have the use
of them, free-of charge, until they do not need them any more. In
theory they should return the addresses to the pool but if there is
monetary value associated with IPv4 addresses space, you can be
assured that there will be a lot of interest in controlling these
identifiers. While in theory, unneeded address space is to be returned
to IANA (or possible to the allocating RIR), acquisitions of companies
have resulting in address space moving to the acquiring company.
Bankruptcy laws, at least in the United States, sometimes treat as
assets such things as domain names and address space. It is also worth
noting that if an address block is not in the routing tables, i may be
useless.

vint


vint


vint



On Mon, Oct 11, 2010 at 4:59 PM, Ernie Rubi <ernesto at cs.fiu.edu> wrote:
> Hi folks,
> I am in the process of writing a research paper for an Internet Law seminar
> at FIU Law and have a few questions for those of you with enough
> institutional memory to remember how IPv4 address allocations were first
> handed out:
> 1. ?Who 'owned' IP addresses ab initio? ?Were IP addresses 'property' of any
> one entity or person or agency? ?What is the authority ICANN / ?IANA had to
> allocate these addresses if they are not 'theirs.'
> 2. ?Initially, were large blocks of IPv4 addresses 'handed out' with a
> complete ownership interest to their recipients? ?For example, when you
> received an /16, was it yours to transfer to? ?? ? ? other entities if you
> pleased? ?Could you have transferred sub-allocations of your /16 to other
> entities who weren't your customers/connectors?
> 3. ?Were the initial IPv4 allocations rolled into RIRs/ICANN at any point?
> If so, under what legal framework?
> Thanks all for your thoughts/comments,
> Ernesto M. Rubi
> Sr. Network Engineer
> AMPATH/CIARA
> Florida International Univ, Miami
> Reply-to:?ernesto at cs.fiu.edu
>



From randy at psg.com  Mon Oct 11 17:50:02 2010
From: randy at psg.com (Randy Bush)
Date: Tue, 12 Oct 2010 09:50:02 +0900
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101012003121.B7B346BE5AD@mercury.lcs.mit.edu>
References: <20101012003121.B7B346BE5AD@mercury.lcs.mit.edu>
Message-ID: <m2k4lokzet.wl%randy@psg.com>

> I suspect we just sent email to Jon Postel, and he looked in the draft
> version of "Assigned Numbers" which he kept in his disk directory on
> ISI{mumble}, gave you the next number, and wrote it down there! :-)

not exactly.  the authoritative record was a written book.  oddly, that
book disappeared off his desk the morning after his death.  while many
folk know who stole it, folk have not worked up the guts to confront the
issue and get it back.

randy


From jcurran at arin.net  Mon Oct 11 18:05:12 2010
From: jcurran at arin.net (John Curran)
Date: Mon, 11 Oct 2010 21:05:12 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <m2ocb0l1h2.wl%randy@psg.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>
	<4CB39BED.6010007@lisse.na> <m2ocb0l1h2.wl%randy@psg.com>
Message-ID: <426A4D2D-060B-4D34-84F2-6661B7EABC88@arin.net>

On Oct 11, 2010, at 8:05 PM, Randy Bush wrote:

>> I would be very interested in reading documentation establishing
>> this chain.
>>> There was a chain of responsibility that was handed down from DARPA /
>>> DODNIC / IANA / RIRs, by agreement between the USG and the Internet
>>> stakeholder community.
> 
> don't hold your breath waiting for it.  wishful thinking from someone
> who wasn't there.

Randy was there, and is correct ** if by "chain of responsibility" you're 
looking for a proscriptive delegation of authority along those precise 
lines from DARPA to the RIRs of the present day **

I believe that there is a "chain of responsibility", if you accept
either Jon Postel's actual actions in this area or if you would prefer 
a series of USG cooperative agreements which state where IP address
administration is to be performed.

For example, the InterNIC registration services cooperative agreement 
(NCR-9218742) provide for the transition of IP address assignments 
from DISA NIC to NSI: <http://www.icann.org/en/nsi/coopagmt-01jan93.htm>  
The 06 and 07 amendments to NCR-9218742 in 1997 transfers responsibility 
for these IP Number assignment and related tasks to ARIN.

/John

John Curran
President and CEO
ARIN





From amyzing at talsever.com  Mon Oct 11 18:21:00 2010
From: amyzing at talsever.com (Amelia A Lewis)
Date: Mon, 11 Oct 2010 21:21:00 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <20101011212100478734.f6531f2f@talsever.com>

Hmmm.

On Mon, 11 Oct 2010 16:59:04 -0400, Ernie Rubi wrote:
> 1.  Who 'owned' IP addresses ab initio?  Were IP addresses 'property' 
> of any one entity or person or agency?  What is the authority ICANN 
> /  IANA had to allocate these addresses if they are not 'theirs.'

This seems shockingly ill-phrased, but then the whole "law school" 
thing seems to shine a light.

ISPs "sell" static IP addresses, as one of the services that they 
provide (I have one).  I don't "own" my IP address (it's the third or 
fourth ... maybe more than that? ... that I've had ... "static", 
right).  Well, if I owned it, I wouldn't pay for it every month, would 
I?  What I have, as a consequence of paying extra, is a routable 
address that is guaranteed not to change without notice (it does change 
*with* notice).

As someone else in the thread has pointed out, the map isn't the 
territory.  Likewise, the address isn't real estate.  It's even more 
imaginary than most imaginary property.  What *I* get, from a (single) 
static IP address, is the guarantee, from the 
person/organization/institution that "owns" the route into that block, 
is a route *out* of that block, into my machine (and the reverse route, 
from my machine, into their block, and from their block into the wider 
internet).  It's a lease; if I tried to sell my static IP to someone 
else, my ISP would whack me (and happily sell the lease to whoever it 
is I was trying to extract a rent from).

The principle extends to larger blocks.  I don't think that this has 
changed much from the early days (though I was too young, then, to do 
anything other than steal maintenance passwords to play adventure).  
The question *isn't* well-phrased as "who owns these IP addresses?"  It 
ought to be "who do I pay to get IP packets routed to me?"

Amy!
-- 
Amelia A. Lewis                    amyzing {at} talsever.com
The use of COBOL cripples the mind; its teaching should, therefore, be 
regarded as a criminal offence.
                -- Edsger Dijkstra


From randy at psg.com  Mon Oct 11 18:49:59 2010
From: randy at psg.com (Randy Bush)
Date: Tue, 12 Oct 2010 10:49:59 +0900
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <426A4D2D-060B-4D34-84F2-6661B7EABC88@arin.net>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>
	<4CB39BED.6010007@lisse.na> <m2ocb0l1h2.wl%randy@psg.com>
	<426A4D2D-060B-4D34-84F2-6661B7EABC88@arin.net>
Message-ID: <m2d3rgkwmw.wl%randy@psg.com>

> Randy was there, and is correct ** if by "chain of responsibility"
> you're looking for a proscriptive delegation of authority along those
> precise lines from DARPA to the RIRs of the present day **

indeed.  and, until this is decided in the courts, we can all opine as
much as we wish.  and, sad to say, with all the hard-ass positions folk
seem to take, it will pretty surely be in the courts.

randy


From jcurran at arin.net  Mon Oct 11 19:28:47 2010
From: jcurran at arin.net (John Curran)
Date: Mon, 11 Oct 2010 22:28:47 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <m2d3rgkwmw.wl%randy@psg.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>
	<4CB39BED.6010007@lisse.na> <m2ocb0l1h2.wl%randy@psg.com>
	<426A4D2D-060B-4D34-84F2-6661B7EABC88@arin.net>
	<m2d3rgkwmw.wl%randy@psg.com>
Message-ID: <1DA732D8-3295-4014-909E-1CBE8FD284BA@arin.net>

On Oct 11, 2010, at 9:49 PM, Randy Bush wrote:

>> Randy was there, and is correct ** if by "chain of responsibility"
>> you're looking for a proscriptive delegation of authority along those
>> precise lines from DARPA to the RIRs of the present day **
> 
> indeed.  and, until this is decided in the courts, we can all opine as
> much as we wish.  and, sad to say, with all the hard-ass positions folk
> seem to take, it will pretty surely be in the courts.

The only reason that a clearer proscriptive delegation did not appear
was that there was a large portion of the community that considered 
Jon-as-IANA-by-way-of-IAB as *the* authority, and another large part
of the community considered the USG (by nature of its role in creation)
as *the* authority.  When you have multiple prospective authorities, 
the result is not going to be nice clean delegation chain.

However, given the historical record of Jon's role in the formation of 
ARIN (including sitting ex-officio on the initial Board) as well as 
the USG direction via the referenced NSF cooperative agreement and its 
modifications, ARIN's responsibility to handle these tasks is quite clear.  
We continue to do these tasks, on behalf of whichever authority you care to 
recognize, since inception.  One hopes that the community-based processes 
we use for policy formation as well as adherence to the principles in RFC2050 
provides fidelity to our mission, but that's obviously of matter of personal
judgement.

/John

John Curran
President and CEO
ARIN




From larrysheldon at cox.net  Mon Oct 11 19:34:40 2010
From: larrysheldon at cox.net (Larry Sheldon)
Date: Mon, 11 Oct 2010 21:34:40 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <m2d3rgkwmw.wl%randy@psg.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>	<4CB39BED.6010007@lisse.na>
	<m2ocb0l1h2.wl%randy@psg.com>	<426A4D2D-060B-4D34-84F2-6661B7EABC88@arin.net>
	<m2d3rgkwmw.wl%randy@psg.com>
Message-ID: <4CB3C940.9050800@cox.net>

  Time for a muddy metaphor.....

Does the business at 18201 Wright Street own that address?  Who does.
Does the person that lives in apartment 3 own that address?  Who does?

What gives those addresses value?




From jack at 3kitty.org  Mon Oct 11 20:03:49 2010
From: jack at 3kitty.org (jfh)
Date: Mon, 11 Oct 2010 20:03:49 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <1286852630.2466.59.camel@localhost>

On Mon, 2010-10-11 at 16:59 -0400, Ernie Rubi wrote:

> Thanks all for your thoughts/comments,

Wow, I think you struck a nerve here - lots of comments.

Preface everything below with "I am not a lawyer...I'm an engineer"

As Noel said, in the early days, all of the numbers (IP nets, protocol
numbers, ports, etc.) were assigned by Jon, and you had to be a
contractor to get on the net.  I.E., everyone on the net was being in
some way paid to do a project which required net access, which was
provided along with all the necessary trappings like number assignments.
We didn't worry about ownership - whoever our project was funded by was
the owner.

As the net grew, it quickly became interconnected with other
organizations, e.g., in the UK, Norway, Italy, Germany, etc.  There were
also interconnects to some contractors.  Most of these probably had
gov't R&D contracts, but I'm not sure they all did.  There may have been
some MOUs involved there also - think PARC, DECWRL, FACC, etc.  I doubt
all those folks were getting paid by the US government, but there must
have been all sorts of MOUs which detailed the rules.  I'd look into
those to see how "ownership" was handled in the early days - e.g., when
the MOU ended, did the right to use the numbers also end?  Presumably
all of those MOUs are government documents, in a warehouse somewhere...

Somewhere along the line, commercial ISPs popped into being.  I'm not
sure who was first, but my recollection is that this happened as
spinoffs from NSFNet and/or CSNet.  Again, there must have been some
kind of agreements between those commercial entities and a piece of the
government, detailing the rules about ownership.  These first ISPs were
different because they were not research collaborators.  They were in
business simply to make money selling Internet service.  People like
Marty Schoffstall (PSINet) might be a good source.

Or, possibly more likely, nobody really thought about ownership and what
a big deal it might become, which is mostly what happened with domain
names in the beginning.  You just got a block of numbers by lobbying Jon
and then you ran with it.  Kind of like the old days of land grants and
railroads across the continent.  So the legal basis of ownership might
be really difficult to establish.  Good legal research project.

Last point - I think it's not really the specific number that might be
owned, but rather it's the right to use that number on a particular
Internet, which, as was pointed out, is possibly more of a rental than
an ownership.

I said "particular Internet" because I suspect that there are many quite
separate Internets out there.  In the 90s when I was at Oracle I was in
part responsible for Oracle's internal network, which was simply a clone
Internet, using the same equipment and software but the only connection
was an email bridge.  We got a block of numbers (I knew enough to ask
Jon) for our LANs and deployed into over a hundred countries.  We used
our official network numbers - but they were never visible on the public
Internet.  We also used other numbers that we didn't "own" for
convenience - i.e., numbers not assigned to us.  I remember that lots of
our customers did the same thing, as they set up their own corporate
Intranets.  

The visible public Internet, large as it is, is only a piece of the
larger Internet, when you include the "dark matter" that isn't visible
from the public space.

So, who owns the number?  Nobody.  Who owns the right to use a
particular number on a particular Internet?  The owner of the particular
Internet.  All you have to do is figure out who that is....

/Jack

PS - Perhaps I "own" some numbers too?!  Protocol # 15 (XNET) is mine,
and Port # 66 (SQLNet) is also mine.  Can I sell these?  Anybody want to
buy a vintage protocol number, only slightly used?

Hmmm, or perhaps Oracle owns port 66, since I got it assigned when I
worked there.

Hmmm, will there be a "portability" law, that allows you to take "your
number" from one Internet to another?  Or one ISP to another on the same
Internet?

Lots of good legal fodder here....have fun!





From randy at psg.com  Mon Oct 11 20:38:08 2010
From: randy at psg.com (Randy Bush)
Date: Tue, 12 Oct 2010 12:38:08 +0900
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <m2iq18jd27.wl%randy@psg.com>

since folk seem to like to indulge in conjecturbation <tm> on issues of
rights to address space, let me throw another span^h^h^h^hquestion in
the works.

some say the issue is not whether legacy holders own the space so much
as whether RIRs have the slightest rights in space the legacy holder
got from isi, sri, or other early allocators.  how did they suddenly get
in the middle anyway?

randy


From jmamodio at gmail.com  Mon Oct 11 20:43:42 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Mon, 11 Oct 2010 22:43:42 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <AANLkTi=eQxLhxFtLMF-fUUxKWBCF8kkvV9CnL94XWr1-@mail.gmail.com>

> 1. ?Who 'owned' IP addresses ab initio?

"ab initio" ?

It depends do you want the evolutionist or the creationism version ?

Cheers
-J



From ernesto at cs.fiu.edu  Mon Oct 11 20:43:15 2010
From: ernesto at cs.fiu.edu (Ernie Rubi)
Date: Mon, 11 Oct 2010 23:43:15 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1DA732D8-3295-4014-909E-1CBE8FD284BA@arin.net>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>
	<4CB39BED.6010007@lisse.na> <m2ocb0l1h2.wl%randy@psg.com>
	<426A4D2D-060B-4D34-84F2-6661B7EABC88@arin.net>
	<m2d3rgkwmw.wl%randy@psg.com>
	<1DA732D8-3295-4014-909E-1CBE8FD284BA@arin.net>
Message-ID: <414AFDB5-BA1A-4C3B-9820-C0A66CA84510@cs.fiu.edu>

Folks,

Thanks to all who've chimed here in these last few hours; your insight has been spot on.

From some of the uncomfortable replies to the original broadly-worded questions in my message I can tell I've asked the right questions.

I hope these questions challenge your assumptions about the issue as much as they've made me rethink some of my assumptions about networking and the legal framework of the Internet.

As a final thought, perhaps the 'home address' analogy is strained in this context - maybe we should compare this to the ownership of phone numbers (and their portability).

Finally, I must say I agree with Mr. Cerf in his comments at the IGF a few weeks back; I'm not sure it's does anyone any good to run around selling v4 space - what we likely should be doing is encouraging all who can to implement v6 to the extent possible.

Thank you all very much,

Ernesto M. Rubi
Sr. Network Engineer
AMPATH/CIARA
Florida International Univ, Miami
Reply-to: ernesto at cs.fiu.edu





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20101011/ada5a534/attachment.htm>

From jmamodio at gmail.com  Mon Oct 11 20:59:52 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Mon, 11 Oct 2010 22:59:52 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <m2iq18jd27.wl%randy@psg.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<m2iq18jd27.wl%randy@psg.com>
Message-ID: <AANLkTin4TcLO+sgBetYRw-yZmHNfWK67X30E36jYzjat@mail.gmail.com>

I believe the whole concept of "ownership" is a bad start and a source
of confusion for many folks when they mix
delegation/assignment/announcement with ownership.

Same thing with the name space, which is a total mess since "property
rights" got embedded into the system.

Cheers
Jorge


From vint at google.com  Mon Oct 11 21:12:24 2010
From: vint at google.com (Vint Cerf)
Date: Tue, 12 Oct 2010 00:12:24 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTi=eQxLhxFtLMF-fUUxKWBCF8kkvV9CnL94XWr1-@mail.gmail.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<AANLkTi=eQxLhxFtLMF-fUUxKWBCF8kkvV9CnL94XWr1-@mail.gmail.com>
Message-ID: <AANLkTimDPerLG2YVeX2hHWSLpG4KYTPdwFEHnRD+6WuR@mail.gmail.com>

the US Defense Advanced Research Projects agency had charge of the
addresses initially - it was their project that created the Internet.

vint cerf


On Mon, Oct 11, 2010 at 11:43 PM, Jorge Amodio <jmamodio at gmail.com> wrote:
>> 1. ?Who 'owned' IP addresses ab initio?
>
> "ab initio" ?
>
> It depends do you want the evolutionist or the creationism version ?
>
> Cheers
> -J
>
>



From jcurran at arin.net  Mon Oct 11 21:19:53 2010
From: jcurran at arin.net (John Curran)
Date: Tue, 12 Oct 2010 00:19:53 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <m2iq18jd27.wl%randy@psg.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<m2iq18jd27.wl%randy@psg.com>
Message-ID: <B02D6D19-27D4-4625-99E2-E1D953C7F5BD@arin.net>

On Oct 11, 2010, at 11:38 PM, Randy Bush wrote:
> some say the issue is not whether legacy holders own the space so much
> as whether RIRs have the slightest rights in space the legacy holder
> got from isi, sri, or other early allocators.  how did they suddenly get
> in the middle anyway?

Randy - 

All organizational assignments and allocations (except those located within
RIPE NCC and APNIC regions) were transferred to ARIN at inception. Reference
mods #6 & #7 of the previously noted NetSol/NSF cooperative agreement which 
transferred these registration services tasks to ARIN.  I don't know what 
"rights" per se you refer to, but know ARIN ended up with *responsibilities*
to manage these resources in accordance with its formative documents.

/John





From mfidelman at meetinghouse.net  Mon Oct 11 22:00:43 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Tue, 12 Oct 2010 01:00:43 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTimDPerLG2YVeX2hHWSLpG4KYTPdwFEHnRD+6WuR@mail.gmail.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>	<AANLkTi=eQxLhxFtLMF-fUUxKWBCF8kkvV9CnL94XWr1-@mail.gmail.com>
	<AANLkTimDPerLG2YVeX2hHWSLpG4KYTPdwFEHnRD+6WuR@mail.gmail.com>
Message-ID: <4CB3EB7B.2000804@meetinghouse.net>

Vint Cerf wrote:
> the US Defense Advanced Research Projects agency had charge of the
> addresses initially - it was their project that created the Internet.
>
>    
Vint,

Not to be pedantic or anything, but... had the name changed from ARPA to 
DARPA at that point? (ok, yes, I am being pedantic :-)

Miles Fidelman

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From el at lisse.NA  Tue Oct 12 00:53:05 2010
From: el at lisse.NA (Dr Eberhard Lisse)
Date: Tue, 12 Oct 2010 09:53:05 +0200
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <m2ocb0l1h2.wl%randy@psg.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>	<4CB39BED.6010007@lisse.na>
	<m2ocb0l1h2.wl%randy@psg.com>
Message-ID: <4CB413E1.1040604@lisse.NA>

Yes, indeed, I wish I had been there :-)-O.

But like many others, who were not, I am now, and though IPs are not
the primary concern for me, but perhaps for many others, I wish I
had more predictability :-)-O

el

on 2010-10-12 02:05 Randy Bush said the following:
>> I would be very interested in reading documentation establishing
>> this chain.
>>> There was a chain of responsibility that was handed down from DARPA /
>>> DODNIC / IANA / RIRs, by agreement between the USG and the Internet
>>> stakeholder community.
> 
> don't hold your breath waiting for it.  wishful thinking from someone
> who wasn't there.
> 
> randy

-- 
Dr. Eberhard W. Lisse  \        / Obstetrician & Gynaecologist (Saar)
el at lisse.NA el108-ARIN / *     |   Telephone: +264 81 124 6733 (cell)
PO Box 8421             \     /   Please do NOT email to this address
Bachbrecht, Namibia     ;____/        if it is DNS related in ANY way


From sbrim at cisco.com  Tue Oct 12 03:46:33 2010
From: sbrim at cisco.com (Scott Brim)
Date: Tue, 12 Oct 2010 06:46:33 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240823c8d956960d5e@[168.122.9.70]>
References: <20101011224145.2113728E137@aland.bbn.com>
	<a06240823c8d956960d5e@[168.122.9.70]>
Message-ID: <4CB43C89.7080005@cisco.com>

I never thought of it as "owning" a prefix.  I knew the prefixes were
"mine" in the sense that we were the only ones supposed to use them, but
it felt more collaborative.  We had to use something, someone had to
administer it, and the NIC told us what would work.


From vint at google.com  Tue Oct 12 04:41:17 2010
From: vint at google.com (Vint Cerf)
Date: Tue, 12 Oct 2010 07:41:17 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB3EB7B.2000804@meetinghouse.net>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<AANLkTi=eQxLhxFtLMF-fUUxKWBCF8kkvV9CnL94XWr1-@mail.gmail.com>
	<AANLkTimDPerLG2YVeX2hHWSLpG4KYTPdwFEHnRD+6WuR@mail.gmail.com>
	<4CB3EB7B.2000804@meetinghouse.net>
Message-ID: <AANLkTikzXiFnNH4YF0ADeUMW0PNQA6LMzwG=r2NZbejR@mail.gmail.com>

Well the address space was defined approx 1977; I don't think the name
change happened until a few years after that.

v


On Tue, Oct 12, 2010 at 1:00 AM, Miles Fidelman
<mfidelman at meetinghouse.net> wrote:
> Vint Cerf wrote:
>>
>> the US Defense Advanced Research Projects agency had charge of the
>> addresses initially - it was their project that created the Internet.
>>
>>
>
> Vint,
>
> Not to be pedantic or anything, but... had the name changed from ARPA to
> DARPA at that point? (ok, yes, I am being pedantic :-)
>
> Miles Fidelman
>
> --
> In theory, there is no difference between theory and practice.
> In<fnord> ?practice, there is. ? .... Yogi Berra
>
>
>



From richard at bennett.com  Tue Oct 12 04:41:44 2010
From: richard at bennett.com (Richard Bennett)
Date: Tue, 12 Oct 2010 04:41:44 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB43C89.7080005@cisco.com>
References: <20101011224145.2113728E137@aland.bbn.com>	<a06240823c8d956960d5e@[168.122.9.70]>
	<4CB43C89.7080005@cisco.com>
Message-ID: <4CB44978.8000209@bennett.com>

  Seems to me that a prefix is something like a SSN or a credit card #. 
I own these in the sense that I'm the only one entitled to use them, and 
anyone else who holds them out as theirs is committing fraud. These 
identifiers only have significance to the extent that one and only one 
party is entitled to their use.

We used to think that phone numbers belonged to the phone company and 
were leased to subscribers, but now think of them as personal property 
that we can take with us across carriers. I guess that's like a V6 PI 
prefix.

In any event, the original, historical understanding has been supplanted 
and no longer applies.

RB

On 10/12/2010 3:46 AM, Scott Brim wrote:
> I never thought of it as "owning" a prefix.  I knew the prefixes were
> "mine" in the sense that we were the only ones supposed to use them, but
> it felt more collaborative.  We had to use something, someone had to
> administer it, and the NIC told us what would work.

-- 
Richard Bennett



From jmamodio at gmail.com  Tue Oct 12 04:51:23 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Tue, 12 Oct 2010 06:51:23 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB43C89.7080005@cisco.com>
References: <20101011224145.2113728E137@aland.bbn.com>
	<a06240823c8d956960d5e@168.122.9.70> <4CB43C89.7080005@cisco.com>
Message-ID: <AANLkTikUjLrP+ruOiZ+8k27Yw8zWXKYsGxS_8w_FnV=D@mail.gmail.com>

On Tue, Oct 12, 2010 at 5:46 AM, Scott Brim <sbrim at cisco.com> wrote:
> I never thought of it as "owning" a prefix. ?I knew the prefixes were
> "mine" in the sense that we were the only ones supposed to use them, but
> it felt more collaborative. ?We had to use something, someone had to
> administer it, and the NIC told us what would work.

Agree, as far as I remember there was never a concept of "ownership",
you got assigned a prefix and others agreed to announce it and move
packets from/to that prefix, you can "own" a large chunk of IP space
but if nobody announces it what you "own" is binary smoke.

Cheers
Jorge



From dhc2 at dcrocker.net  Tue Oct 12 05:28:59 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Tue, 12 Oct 2010 05:28:59 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240823c8d956960d5e@[168.122.9.70]>
References: <20101011224145.2113728E137@aland.bbn.com>
	<a06240823c8d956960d5e@[168.122.9.70]>
Message-ID: <4CB4548B.9030405@dcrocker.net>



On 10/11/2010 5:14 PM, John Day wrote:
> Do you own the address where your house is?  When you move do you take it with
> you?  Do you get your mail at the address where you were born?


You mean like portable phone numbers, especially ones that are easier to remember?

Are you suggesting that folks find no value in such numbers and that their right 
to take them from one provider to another has no value?

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From jeanjour at comcast.net  Tue Oct 12 07:32:17 2010
From: jeanjour at comcast.net (John Day)
Date: Tue, 12 Oct 2010 10:32:17 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <414AFDB5-BA1A-4C3B-9820-C0A66CA84510@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<5B35165C-C6BF-4BC1-AFDA-F75A70E18068@pch.net>
	<4CB39BED.6010007@lisse.na> <m2ocb0l1h2.wl%randy@psg.com>
	<426A4D2D-060B-4D34-84F2-6661B7EABC88@arin.net>
	<m2d3rgkwmw.wl%randy@psg.com>
	<1DA732D8-3295-4014-909E-1CBE8FD284BA@arin.net>
	<414AFDB5-BA1A-4C3B-9820-C0A66CA84510@cs.fiu.edu>
Message-ID: <a0624082ac8da1f590f59@[168.122.9.70]>

At 23:43 -0400 2010/10/11, Ernie Rubi wrote:
>Folks,
>
>Thanks to all who've chimed here in these last few hours; your 
>insight has been spot on.
>
>From some of the uncomfortable replies to the original 
>broadly-worded questions in my message I can tell I've asked the 
>right questions.
>
>I hope these questions challenge your assumptions about the issue as 
>much as they've made me rethink some of my assumptions about 
>networking and the legal framework of the Internet.
>
>As a final thought, perhaps the 'home address' analogy is strained 
>in this context - maybe we should compare this to the ownership of 
>phone numbers (and their portability).

Not in the least.  I gave it some further thought after I sent my 
last email as to whether "ownership" was linked more to a "service" 
rather than the address per se and realized that no that was not it. 
Any "service" enabled by the address, should not be a property of the 
IP address at all.  The whole problem of ownership comes down to the 
Internet lacking a complete architecture;.

Your analogy to keeping your phone number is flawed.  Phone numbers 
are no longer addresses but application names.  Phone number ceased 
to network addresses 20+ years ago.

As I said, if you think addresses can owned, then you have not 
thought about the problem  carefully enough.

Take care,
John

>
>Finally, I must say I agree with Mr. Cerf in his comments at the IGF 
>a few weeks back; I'm not sure it's does anyone any good to run 
>around selling v4 space - what we likely should be doing is 
>encouraging all who can to implement v6 to the extent possible.


>
>Thank you all very much,
>
>Ernesto M. Rubi
>Sr. Network Engineer
>AMPATH/CIARA
>Florida International Univ, Miami
>Reply-to: <mailto:ernesto at cs.fiu.edu>ernesto at cs.fiu.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20101012/93f7ccef/attachment.htm>

From jnc at mercury.lcs.mit.edu  Tue Oct 12 08:28:00 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Tue, 12 Oct 2010 11:28:00 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>

    > From: Dave CROCKER <dhc2 at dcrocker.net>

    > On 10/11/2010 5:14 PM, John Day wrote:

    >> Do you own the address where your house is?  When you move do you take it with
    >> you? Do you get your mail at the address where you were born?

    > You mean like portable phone numbers, especially ones that are easier
    > to remember?

Neither one of those analogies is entirely applicable. One phrase for
everyone here: 'separation of location and identity'.

However, the former analogy is more applicable to IPvN addresses than the
latter. For one, portable phone numbers are (now) the equivalent of DNS
names. I.e. to actually use them for communication, they _have_ to go through
a binding layer, the output of which is the number's current, actual
location. (The fact that the _syntax_ of the output of that mapping looks
like the input, i.e. phone numbers, doesn't change the fact that the
_semantics_ of the output - i.e. the properties of that output number - are
different from the input.)

	Noel


From jack at 3kitty.org  Tue Oct 12 09:22:06 2010
From: jack at 3kitty.org (jfh)
Date: Tue, 12 Oct 2010 09:22:06 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
Message-ID: <1286900526.2518.38.camel@localhost>

This is a great discussion ... but I think it's way too logical and
engineering-based.  IMHO, legal structures are rarely logical to us
engineers.  But legal mechanisms will be applied to Internet
technologies and infrastructure, just like they were applied in
telephone, railroads, radio, shipping, etc. in the past.

So, for example, in legal property worlds there is a concept of
"prescriptive rights".  Basically if you use someone else's property for
long enough without their permission or objection, you acquire the
ongoing right to use that property.  I.e., for practical and legal
purposes you own it, or at least own a right to use it.

If I use a particular range of IP addresses long enough, without the
permission of whoever thinks they own it, do I now own those
addresses...?  It's clear to me what the engineering answer is, but I
have no idea what a court might decide.  Hence the question of ownership
- legal rights - is very relevant.

In fact, I wonder if all the laws currently governing portability of
telephone numbers also apply in some way to IP addresses, if someone
decides to litigate?  I know they're very different numbers from an
engineering perspective, but from a judge's bench...?

This is the kind of question that we can't answer by logical engineering
arguments...

/Jack


On Tue, 2010-10-12 at 11:28 -0400, Noel Chiappa wrote:
> > From: Dave CROCKER <dhc2 at dcrocker.net>
> 
>     > On 10/11/2010 5:14 PM, John Day wrote:
> 
>     >> Do you own the address where your house is?  When you move do you take it with
>     >> you? Do you get your mail at the address where you were born?
> 
>     > You mean like portable phone numbers, especially ones that are easier
>     > to remember?
> 
> Neither one of those analogies is entirely applicable. One phrase for
> everyone here: 'separation of location and identity'.
> 
> However, the former analogy is more applicable to IPvN addresses than the
> latter. For one, portable phone numbers are (now) the equivalent of DNS
> names. I.e. to actually use them for communication, they _have_ to go through
> a binding layer, the output of which is the number's current, actual
> location. (The fact that the _syntax_ of the output of that mapping looks
> like the input, i.e. phone numbers, doesn't change the fact that the
> _semantics_ of the output - i.e. the properties of that output number - are
> different from the input.)
> 
> 	Noel




From woody at pch.net  Tue Oct 12 09:44:23 2010
From: woody at pch.net (Bill Woodcock)
Date: Tue, 12 Oct 2010 09:44:23 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB4548B.9030405@dcrocker.net>
References: <20101011224145.2113728E137@aland.bbn.com>
	<a06240823c8d956960d5e@[168.122.9.70]>
	<4CB4548B.9030405@dcrocker.net>
Message-ID: <0871C0AA-B0B4-414D-B093-F9F7D7C16882@pch.net>


On Oct 12, 2010, at 5:28 AM, Dave CROCKER wrote:
> 
> On 10/11/2010 5:14 PM, John Day wrote:
>> Do you own the address where your house is?  When you move do you take it with
>> you?  Do you get your mail at the address where you were born?
> 
> Are you suggesting that folks find no value in such numbers and that their right to take them from one provider to another has no value?

On the contrary...  It has negative value.  Just as if everyone took their street numbers with them when they moved, and 125 Broadway were next door to 25337 El Camino Real and across the street from 17B Vine Terrace...  It would ruin the aggregation property of the system, which would remove all meaning.  "Easy to remember" is not an interesting property in an IP address except for your bootstrap recursive resolver.  "Uniquely allocated to you" and "aggregated with other like addresses so as to be findable through routing" are the two interesting properties of an address.

                                -Bill








From eric.gade at gmail.com  Tue Oct 12 09:50:11 2010
From: eric.gade at gmail.com (Eric Gade)
Date: Tue, 12 Oct 2010 17:50:11 +0100
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1286900526.2518.38.camel@localhost>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
	<1286900526.2518.38.camel@localhost>
Message-ID: <AANLkTikT3ex=hRYmr_3VR-ws1f7vbd=RPCcTfwLQLn+p@mail.gmail.com>

I want to chime in and suggest that a mobile phone number is much more like
a name than an address, primarily because an IP address is universally
explicit.  Telephone numbers are more like relative distinguished names,
especially if you consider their use internationally (dialing country codes,
etc).  The analogy of phone numbers was used quite often during both the
creation of DNS and the IFIP 6.5 discussions.

On Tue, Oct 12, 2010 at 5:22 PM, jfh <jack at 3kitty.org> wrote:

> This is a great discussion ... but I think it's way too logical and
> engineering-based.  IMHO, legal structures are rarely logical to us
> engineers.  But legal mechanisms will be applied to Internet
> technologies and infrastructure, just like they were applied in
> telephone, railroads, radio, shipping, etc. in the past.
>
> So, for example, in legal property worlds there is a concept of
> "prescriptive rights".  Basically if you use someone else's property for
> long enough without their permission or objection, you acquire the
> ongoing right to use that property.  I.e., for practical and legal
> purposes you own it, or at least own a right to use it.
>
> If I use a particular range of IP addresses long enough, without the
> permission of whoever thinks they own it, do I now own those
> addresses...?  It's clear to me what the engineering answer is, but I
> have no idea what a court might decide.  Hence the question of ownership
> - legal rights - is very relevant.
>
> In fact, I wonder if all the laws currently governing portability of
> telephone numbers also apply in some way to IP addresses, if someone
> decides to litigate?  I know they're very different numbers from an
> engineering perspective, but from a judge's bench...?
>
> This is the kind of question that we can't answer by logical engineering
> arguments...
>
> /Jack
>
>
> On Tue, 2010-10-12 at 11:28 -0400, Noel Chiappa wrote:
> > > From: Dave CROCKER <dhc2 at dcrocker.net>
> >
> >     > On 10/11/2010 5:14 PM, John Day wrote:
> >
> >     >> Do you own the address where your house is?  When you move do you
> take it with
> >     >> you? Do you get your mail at the address where you were born?
> >
> >     > You mean like portable phone numbers, especially ones that are
> easier
> >     > to remember?
> >
> > Neither one of those analogies is entirely applicable. One phrase for
> > everyone here: 'separation of location and identity'.
> >
> > However, the former analogy is more applicable to IPvN addresses than the
> > latter. For one, portable phone numbers are (now) the equivalent of DNS
> > names. I.e. to actually use them for communication, they _have_ to go
> through
> > a binding layer, the output of which is the number's current, actual
> > location. (The fact that the _syntax_ of the output of that mapping looks
> > like the input, i.e. phone numbers, doesn't change the fact that the
> > _semantics_ of the output - i.e. the properties of that output number -
> are
> > different from the input.)
> >
> >       Noel
>
>
>


-- 
Eric
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20101012/aa19b93a/attachment.htm>

From wmaton at ottix.net  Tue Oct 12 10:00:23 2010
From: wmaton at ottix.net (William F. Maton)
Date: Tue, 12 Oct 2010 13:00:23 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1286900526.2518.38.camel@localhost>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
	<1286900526.2518.38.camel@localhost>
Message-ID: <Pine.LNX.4.64.1010121249300.24321@iskra.ottix.net>

On Tue, 12 Oct 2010, jfh wrote:

> If I use a particular range of IP addresses long enough, without the
> permission of whoever thinks they own it, do I now own those
> addresses...?  It's clear to me what the engineering answer is, but I
> have no idea what a court might decide.  Hence the question of ownership
> - legal rights - is very relevant.

I've had a phone number for a very long time.  But, the Terms of Service 
say, in part:  "Subscribers do not have any property rights in telephone 
numbers assigned to them."  Oh dear, going down the bumpy road of the 
analogy....

> In fact, I wonder if all the laws currently governing portability of
> telephone numbers also apply in some way to IP addresses, if someone
> decides to litigate?  I know they're very different numbers from an
> engineering perspective, but from a judge's bench...?

Here in Canada, it's been noted that, "that telephone numbers are public 
resources, and as such are not owned by any person or entity." 
(quoting out of CRTC Decision 2004-52.)  OK, there are "prescriptive 
rights" as you point out, but like you also say, a judge may be convinced 
to try and take the analogy within the telecom sphere from phone numbers 
to IP numbers.

Yes, it's a limited analogy to use, ignoring all sorts of nuances as well.

Regarding portability, we have that today with PI-space, for those of us 
fortunate enough to have made it through the door early.  Those of us with 
PI can take our IP numbers with us to a new provider to route.  But the 
question of ownership still remains.

Good discussion all around.

wfms


From jeanjour at comcast.net  Tue Oct 12 10:00:31 2010
From: jeanjour at comcast.net (John Day)
Date: Tue, 12 Oct 2010 13:00:31 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
Message-ID: <a0624082ec8da361d6544@[168.122.9.70]>

Noel
Are you still wrapped around the pseudo problem of separating locator 
from identity?  Haven't you noticed the fundamental flaw in your 
reasoning?  This is why LISP ran into scaling problems.

You can't locate something without identifying it and vice versa. 
See Saltzer 1977. definition of "resolve."

The problem with the Internet architecture has never been separating 
locator from identifier, but separating physical "location" from 
logical "location" and having a "location-independent" identifier as 
well.  Actually, even these location-independent identifiers are 
location-dependent in some sense, it is just that the sense of 
"location" is much different than the other two.

Your interpretation of phone numbers is correct.  As I noted, they 
are currently location-independent identifiers not addresses.  As I 
also noted in an early note, MAC addresses are not addresses either 
for the same reason.

Take care,
John

At 11:28 -0400 2010/10/12, Noel Chiappa wrote:
>     > From: Dave CROCKER <dhc2 at dcrocker.net>
>
>     > On 10/11/2010 5:14 PM, John Day wrote:
>
>     >> Do you own the address where your house is?  When you move do 
>you take it with
>     >> you? Do you get your mail at the address where you were born?
>
>     > You mean like portable phone numbers, especially ones that are easier
>     > to remember?
>
>Neither one of those analogies is entirely applicable. One phrase for
>everyone here: 'separation of location and identity'.
>
>However, the former analogy is more applicable to IPvN addresses than the
>latter. For one, portable phone numbers are (now) the equivalent of DNS
>names. I.e. to actually use them for communication, they _have_ to go through
>a binding layer, the output of which is the number's current, actual
>location. (The fact that the _syntax_ of the output of that mapping looks
>like the input, i.e. phone numbers, doesn't change the fact that the
>_semantics_ of the output - i.e. the properties of that output number - are
>different from the input.)
>
>	Noel



From jeanjour at comcast.net  Tue Oct 12 09:59:44 2010
From: jeanjour at comcast.net (John Day)
Date: Tue, 12 Oct 2010 12:59:44 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1286900526.2518.38.camel@localhost>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
	<1286900526.2518.38.camel@localhost>
Message-ID: <a0624082fc8da43e7a084@[168.122.9.70]>

If I live in a house for 20 years and move, I still don't take the 
address with me.

There is no percentage in making excuses for sloppy thinking.  It is 
still sloppy and sooner or later will lead to problems.  As we see.

Take care,
John

At 9:22 -0700 2010/10/12, jfh wrote:
>This is a great discussion ... but I think it's way too logical and
>engineering-based.  IMHO, legal structures are rarely logical to us
>engineers.  But legal mechanisms will be applied to Internet
>technologies and infrastructure, just like they were applied in
>telephone, railroads, radio, shipping, etc. in the past.
>
>So, for example, in legal property worlds there is a concept of
>"prescriptive rights".  Basically if you use someone else's property for
>long enough without their permission or objection, you acquire the
>ongoing right to use that property.  I.e., for practical and legal
>purposes you own it, or at least own a right to use it.
>
>If I use a particular range of IP addresses long enough, without the
>permission of whoever thinks they own it, do I now own those
>addresses...?  It's clear to me what the engineering answer is, but I
>have no idea what a court might decide.  Hence the question of ownership
>- legal rights - is very relevant.
>
>In fact, I wonder if all the laws currently governing portability of
>telephone numbers also apply in some way to IP addresses, if someone
>decides to litigate?  I know they're very different numbers from an
>engineering perspective, but from a judge's bench...?
>
>This is the kind of question that we can't answer by logical engineering
>arguments...
>
>/Jack
>
>
>On Tue, 2010-10-12 at 11:28 -0400, Noel Chiappa wrote:
>>  > From: Dave CROCKER <dhc2 at dcrocker.net>
>>
>>      > On 10/11/2010 5:14 PM, John Day wrote:
>>
>>      >> Do you own the address where your house is?  When you move 
>>do you take it with
>>      >> you? Do you get your mail at the address where you were born?
>>
>>      > You mean like portable phone numbers, especially ones that are easier
>>      > to remember?
>>
>>  Neither one of those analogies is entirely applicable. One phrase for
>>  everyone here: 'separation of location and identity'.
>>
>>  However, the former analogy is more applicable to IPvN addresses than the
>>  latter. For one, portable phone numbers are (now) the equivalent of DNS
>>  names. I.e. to actually use them for communication, they _have_ to 
>>go through
>>  a binding layer, the output of which is the number's current, actual
>>  location. (The fact that the _syntax_ of the output of that mapping looks
>>  like the input, i.e. phone numbers, doesn't change the fact that the
>>  _semantics_ of the output - i.e. the properties of that output number - are
>>  different from the input.)
>>
>>	Noel



From ajs at crankycanuck.ca  Tue Oct 12 10:12:18 2010
From: ajs at crankycanuck.ca (Andrew Sullivan)
Date: Tue, 12 Oct 2010 13:12:18 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0871C0AA-B0B4-414D-B093-F9F7D7C16882@pch.net>
References: <20101011224145.2113728E137@aland.bbn.com>
	<a06240823c8d956960d5e@[168.122.9.70]>
	<4CB4548B.9030405@dcrocker.net>
	<0871C0AA-B0B4-414D-B093-F9F7D7C16882@pch.net>
Message-ID: <20101012171218.GA49035@shinkuro.com>

On Tue, Oct 12, 2010 at 09:44:23AM -0700, Bill Woodcock wrote:

> system, which would remove all meaning.  "Easy to remember" is not
> an interesting property in an IP address except for your bootstrap
> recursive resolver.  "Uniquely allocated to you" and "aggregated
> with other like addresses so as to be findable through routing" are
> the two interesting properties of an address.

Surely, though, there are people to whom the interesting property is,
"One can assert property rights in it and therefore make money from
its scarcity," no?  Given the situation with IPv4, I predict
interesting times ahead for the assertion that the addresses are not
owned by anyone (although I happen to agree with the proposition).

A

-- 
Andrew Sullivan
ajs at crankycanuck.ca


From sbrim at cisco.com  Tue Oct 12 10:45:52 2010
From: sbrim at cisco.com (Scott Brim)
Date: Tue, 12 Oct 2010 13:45:52 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a0624082ec8da361d6544@[168.122.9.70]>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
	<a0624082ec8da361d6544@[168.122.9.70]>
Message-ID: <4CB49ED0.3070106@cisco.com>

On 10/12/2010 13:00 EDT, John Day wrote:
> The problem with the Internet architecture has never been separating
> locator from identifier, but separating physical "location" from logical
> "location" and having a "location-independent" identifier as well. 

I'm always willing to follow this digression, although perhaps we should
change the subject.  Yes, loc/id separation was never a problem in
routing or just in layer 3 (our fault for looking under the street
lights).  Essentially, the problem is to wean any and all identification
functions, wherever they may be, away from using topology-dependent
inputs.  Routing and forwarding can use whatever they want.

Scott


From ernesto at cs.fiu.edu  Tue Oct 12 10:45:57 2010
From: ernesto at cs.fiu.edu (Ernie Rubi)
Date: Tue, 12 Oct 2010 13:45:57 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTinkmYv2cEeC0p=dqLrwnuVrOsUfBi95vfYA0oJS@mail.gmail.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<AANLkTinkmYv2cEeC0p=dqLrwnuVrOsUfBi95vfYA0oJS@mail.gmail.com>
Message-ID: <D4AFEE0E-FB02-4275-ABB9-FA2B5A5F4C12@cs.fiu.edu>

Sorry to belabor the point, but:

Do you remember who decided the space and who employed that person(s), where it was decided (at work, at home)?  

Were you and Mr. Postel's actions under the control/direction of the US Government or a agency of said gov't at the time?  

I know these questions seem silly and perhaps intrusive but I'm trying to tie some of this to the law of agency.

On Oct 11, 2010, at 8:35 PM, Vint Cerf wrote:

> the initial address space was decided on in 1977

Ernesto M. Rubi
Sr. Network Engineer
AMPATH/CIARA
Florida International Univ, Miami
Reply-to: ernesto at cs.fiu.edu





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20101012/9ac55fb4/attachment.htm>

From jnc at mercury.lcs.mit.edu  Tue Oct 12 10:51:35 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Tue, 12 Oct 2010 13:51:35 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101012175135.C64E06BE54F@mercury.lcs.mit.edu>

    > From: John Day <jeanjour at comcast.net>

    > separating locator from identity?

Actually, I said "location from identity". "location", as in, the abstract
concept, not "locator", a particular mechanism.

    > This is why LISP ran into scaling problems.

Something that's not deployed yet "ran" (past tense) into scaling problems?
But this is the wrong mailing list for a discussion about that.

	Noel


From ernesto at cs.fiu.edu  Tue Oct 12 11:01:13 2010
From: ernesto at cs.fiu.edu (Ernie Rubi)
Date: Tue, 12 Oct 2010 14:01:13 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a0624082fc8da43e7a084@[168.122.9.70]>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
	<1286900526.2518.38.camel@localhost>
	<a0624082fc8da43e7a084@[168.122.9.70]>
Message-ID: <EE7297DD-1A2B-4E58-A8B4-39325802B229@cs.fiu.edu>

Mr. Day - I recognize you have strong feelings regarding this topic.  I'm glad you've voice them and we've all noted them.

However, I'll take some objection to your implying sloppy thinking on my part or on others who chimed in.  

The purpose of my original email is to encourage discussion - you seem intent on quashing it.  The more discussion, the less sloppy thinking and the more fully the issues can be explored.

Thanks,

Ernie

On Oct 12, 2010, at 12:59 PM, John Day wrote:

> 
> There is no percentage in making excuses for sloppy thinking.





From jklensin at gmail.com  Tue Oct 12 11:01:15 2010
From: jklensin at gmail.com (John Klensin)
Date: Tue, 12 Oct 2010 14:01:15 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTikzXiFnNH4YF0ADeUMW0PNQA6LMzwG=r2NZbejR@mail.gmail.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<AANLkTi=eQxLhxFtLMF-fUUxKWBCF8kkvV9CnL94XWr1-@mail.gmail.com>
	<AANLkTimDPerLG2YVeX2hHWSLpG4KYTPdwFEHnRD+6WuR@mail.gmail.com>
	<4CB3EB7B.2000804@meetinghouse.net>
	<AANLkTikzXiFnNH4YF0ADeUMW0PNQA6LMzwG=r2NZbejR@mail.gmail.com>
Message-ID: <AANLkTi=fkktzqqSjNNfKPMd3C0KC0Yv2=B4ogYpY29Af@mail.gmail.com>

On 10/12/10, Vint Cerf <vint at google.com> wrote:
> Well the address space was defined approx 1977; I don't think the name
> change happened until a few years after that.

Vint,

My vague recollection is that the first name change occurred as a
byproduct of the 1973 Mansfield Amendment (although perhaps a bit
later). but that the name was changed back and forth once or twice
before the "D" settled in permanently.  A final name change somewhat
after 1977 would be consistent with that, but I''m reasonably sure
that the first one occurred earlier.  Should be that hard to track
down if it were really important.

   john


From ernesto at cs.fiu.edu  Tue Oct 12 11:01:13 2010
From: ernesto at cs.fiu.edu (Ernie Rubi)
Date: Tue, 12 Oct 2010 14:01:13 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a0624082fc8da43e7a084@[168.122.9.70]>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
	<1286900526.2518.38.camel@localhost>
	<a0624082fc8da43e7a084@[168.122.9.70]>
Message-ID: <EE7297DD-1A2B-4E58-A8B4-39325802B229@cs.fiu.edu>

Mr. Day - I recognize you have strong feelings regarding this topic.  I'm glad you've voice them and we've all noted them.

However, I'll take some objection to your implying sloppy thinking on my part or on others who chimed in.  

The purpose of my original email is to encourage discussion - you seem intent on quashing it.  The more discussion, the less sloppy thinking and the more fully the issues can be explored.

Thanks,

Ernie

On Oct 12, 2010, at 12:59 PM, John Day wrote:

> 
> There is no percentage in making excuses for sloppy thinking.





From jack at 3kitty.org  Tue Oct 12 11:33:25 2010
From: jack at 3kitty.org (jfh)
Date: Tue, 12 Oct 2010 11:33:25 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101012171218.GA49035@shinkuro.com>
References: <20101011224145.2113728E137@aland.bbn.com>
	<a06240823c8d956960d5e@[168.122.9.70]> <4CB4548B.9030405@dcrocker.net>
	<0871C0AA-B0B4-414D-B093-F9F7D7C16882@pch.net>
	<20101012171218.GA49035@shinkuro.com>
Message-ID: <1286908405.2518.68.camel@localhost>

On Tue, 2010-10-12 at 13:12 -0400, Andrew Sullivan wrote:
> I predict
> interesting times ahead 

>From a technical perspective, The Internet is decades old.  There's lots
of established rules, mechanisms, and technologies.  It works.

>From a legal perspective, The Internet is in its infancy.  Few laws,
little precedent from prior court rulings, etc.

In a way, the Legal Internet today is at a stage the Technical Internet
was back in the 80s.  There's an established base of the traditional
legal framework, just like the 80s framework of PTTs etc.  But most of
the critical legal concepts - ownership, liability, etc. - might have
some informal structure of gentlemen's agreements with contracts and
MOUs in today's Internet, but little if any legal foundations that I'm
aware of.  We put together the technology, defined the protocols, and
did the engineering.  But we mostly ignored legal issues.

I've heard people ask "Who runs the Internet?"  but never heard "Who
owns the Internet?" or "Whom do I sue when the Internet does me harm?"
Yet.

As in other similar technology revolutions, most laws were written when
The Internet wasn't around, or at least wasn't understood in legislative
circles - and we may not be there yet either.  So laws, even recent ones
concerning things like telephone numbers, aren't obviously applicable.
But sooner or later some judge will decide how existing laws apply.  And
legislators will make laws, which may or may not make sense technically.

Of course, there are jurisdictional issues.  What court, in what
country, hears Internet disputes, and what countrys' laws do they use?

Back in the 90s, I half-jokingly promoted the notion that The Internet
should be treated as a Country, complete with its own government, laws,
judicial system, monetary system, treaties, etc.  Parts of that
structure exist - ICANN et al - but they're not treated as a country
(whatever that means).  Given the rise of cyber-crime and cyber-wars, a
Police department and cyber-military might be needed too.  From what I
read in the news, it sounds like they may already exist.

I'm an engineer, not a lawyer or diplomat**.  So what do I know?

But maybe looking at the Internet as a country is not such a joke after
all...

Yes, interesting times ahead...

/Jack

** In the interest of full disclosure, I *am* a government official -
well sort of - an elected member of the board of our local ambulance
district.  It was an accident...



From jack at 3kitty.org  Tue Oct 12 11:33:25 2010
From: jack at 3kitty.org (jfh)
Date: Tue, 12 Oct 2010 11:33:25 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101012171218.GA49035@shinkuro.com>
References: <20101011224145.2113728E137@aland.bbn.com>
	<a06240823c8d956960d5e@[168.122.9.70]> <4CB4548B.9030405@dcrocker.net>
	<0871C0AA-B0B4-414D-B093-F9F7D7C16882@pch.net>
	<20101012171218.GA49035@shinkuro.com>
Message-ID: <1286908405.2518.68.camel@localhost>

On Tue, 2010-10-12 at 13:12 -0400, Andrew Sullivan wrote:
> I predict
> interesting times ahead 

>From a technical perspective, The Internet is decades old.  There's lots
of established rules, mechanisms, and technologies.  It works.

>From a legal perspective, The Internet is in its infancy.  Few laws,
little precedent from prior court rulings, etc.

In a way, the Legal Internet today is at a stage the Technical Internet
was back in the 80s.  There's an established base of the traditional
legal framework, just like the 80s framework of PTTs etc.  But most of
the critical legal concepts - ownership, liability, etc. - might have
some informal structure of gentlemen's agreements with contracts and
MOUs in today's Internet, but little if any legal foundations that I'm
aware of.  We put together the technology, defined the protocols, and
did the engineering.  But we mostly ignored legal issues.

I've heard people ask "Who runs the Internet?"  but never heard "Who
owns the Internet?" or "Whom do I sue when the Internet does me harm?"
Yet.

As in other similar technology revolutions, most laws were written when
The Internet wasn't around, or at least wasn't understood in legislative
circles - and we may not be there yet either.  So laws, even recent ones
concerning things like telephone numbers, aren't obviously applicable.
But sooner or later some judge will decide how existing laws apply.  And
legislators will make laws, which may or may not make sense technically.

Of course, there are jurisdictional issues.  What court, in what
country, hears Internet disputes, and what countrys' laws do they use?

Back in the 90s, I half-jokingly promoted the notion that The Internet
should be treated as a Country, complete with its own government, laws,
judicial system, monetary system, treaties, etc.  Parts of that
structure exist - ICANN et al - but they're not treated as a country
(whatever that means).  Given the rise of cyber-crime and cyber-wars, a
Police department and cyber-military might be needed too.  From what I
read in the news, it sounds like they may already exist.

I'm an engineer, not a lawyer or diplomat**.  So what do I know?

But maybe looking at the Internet as a country is not such a joke after
all...

Yes, interesting times ahead...

/Jack

** In the interest of full disclosure, I *am* a government official -
well sort of - an elected member of the board of our local ambulance
district.  It was an accident...



From benc at hawaga.org.uk  Tue Oct 12 11:41:54 2010
From: benc at hawaga.org.uk (Ben Clifford)
Date: Tue, 12 Oct 2010 18:41:54 +0000 (GMT)
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
Message-ID: <Pine.LNX.4.64.1010121837520.21970@dildano.hawaga.org.uk>


The law recognises degrees of ownership, rather than absolute ownership. 
(for example: you won a house but have a mortgage on it, so you don't own 
it in the sense that you may without the consent of the mortgage holder 
demolish it, for example; or you are a lord off on a crusade and have 
handed over your manor to your brother to own in trust for you; or you own 
land but now The Powers That Be acquire it without your consent to build a 
freeway over it). Maybe some of this discussion (esp in light of this 
being for a law seminar) could be framed in who gave what rights to who, 
rather than who gave this abstract "ownership" to who.

-- 
http://www.hawaga.org.uk/ben/



From cos at aaaaa.org  Tue Oct 12 12:00:10 2010
From: cos at aaaaa.org (Ofer Inbar)
Date: Tue, 12 Oct 2010 15:00:10 -0400
Subject: [ih] Commercial ISPs (Re:  Some Questions over IPv4 Ownership)
In-Reply-To: <1286852630.2466.59.camel@localhost>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<1286852630.2466.59.camel@localhost>
Message-ID: <20101012190010.GV18254@mip.aaaaa.org>

On Mon, Oct 11, 2010 at 08:03:49PM -0700,
jfh <jack at 3kitty.org> wrote:
> Somewhere along the line, commercial ISPs popped into being.  I'm
> not sure who was first, but my recollection is that this happened as
> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
> kind of agreements between those commercial entities and a piece of
> the government, detailing the rules about ownership.  These first
> ISPs were different because they were not research collaborators.
> They were in business simply to make money selling Internet service.

Software Tool & Die's (aka "the World") was the first commercial
consumer ISP, and UUNET was the first commercial IP network provider -
who sold IP service to STD.

STD's primary business was actually software consulting, they just
sold Internet service because they wanted to make it available to
the public, IIRC.  It didn't become the focus of their business until
some years later, by which point there were a lot of other ISPs.
I could be mistaken about their business and goals, but maybe someone
from STD is on this list to comment.

Although UUNET was willing to route commercial traffic, the NSFnet
"backbone" at the time was not; blocks of IPs assigned to non-research
entities did not get routed across the NSFnet, which meant they
couldn't get to a lot of the Internet.  So early users of the World
had partial IP connectivity; they could send email anywhere, through
UUNET's relays, but they couldn't telnet or ftp everywhere.  They could
connect to some but not all IRC servers, but of course as long as
some IRC servers they connected to were connected to some of the rest,
they could chat live with anyone on IRC anywhere on the Internet.

[ IIRC they started selling access to TCP/IP accounts before http
  began, though it may have been the other way 'round, but either way,
  there was not much of a "web" at the time.  Real TCP/IP access
  mattered mainly for telnet, ftp, talk/ytalk, irc, and gopher, more
  or less in that order of importance I believe. ]

Certainly, the main issue back then about IP address allocations was
not "what addresses do you own" but rather "how can you get your
traffic routed".  Addresses were plentiful and you were going to get
enough of them to suit what you were doing without trouble.
  -- Cos


From sbrim at cisco.com  Tue Oct 12 12:12:05 2010
From: sbrim at cisco.com (Scott Brim)
Date: Tue, 12 Oct 2010 15:12:05 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTi=fkktzqqSjNNfKPMd3C0KC0Yv2=B4ogYpY29Af@mail.gmail.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>	<AANLkTi=eQxLhxFtLMF-fUUxKWBCF8kkvV9CnL94XWr1-@mail.gmail.com>	<AANLkTimDPerLG2YVeX2hHWSLpG4KYTPdwFEHnRD+6WuR@mail.gmail.com>	<4CB3EB7B.2000804@meetinghouse.net>	<AANLkTikzXiFnNH4YF0ADeUMW0PNQA6LMzwG=r2NZbejR@mail.gmail.com>
	<AANLkTi=fkktzqqSjNNfKPMd3C0KC0Yv2=B4ogYpY29Af@mail.gmail.com>
Message-ID: <4CB4B305.5080501@cisco.com>

On 10/12/2010 14:01 EDT, John Klensin wrote:
> On 10/12/10, Vint Cerf <vint at google.com> wrote:
>> Well the address space was defined approx 1977; I don't think the name
>> change happened until a few years after that.
> 
> Vint,
> 
> My vague recollection is that the first name change occurred as a
> byproduct of the 1973 Mansfield Amendment (although perhaps a bit
> later). but that the name was changed back and forth once or twice
> before the "D" settled in permanently.  A final name change somewhat
> after 1977 would be consistent with that, but I''m reasonably sure
> that the first one occurred earlier.  Should be that hard to track
> down if it were really important.
> 
>    john

I believe I remember one in about 1987 (??)


From sbrim at cisco.com  Tue Oct 12 12:45:43 2010
From: sbrim at cisco.com (Scott Brim)
Date: Tue, 12 Oct 2010 15:45:43 -0400
Subject: [ih] Commercial ISPs (Re:  Some Questions over IPv4 Ownership)
In-Reply-To: <20101012190010.GV18254@mip.aaaaa.org>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>	<1286852630.2466.59.camel@localhost>
	<20101012190010.GV18254@mip.aaaaa.org>
Message-ID: <4CB4BAE7.4090800@cisco.com>

On 10/12/2010 15:00 EDT, Ofer Inbar wrote:
> On Mon, Oct 11, 2010 at 08:03:49PM -0700,
> jfh <jack at 3kitty.org> wrote:
>> Somewhere along the line, commercial ISPs popped into being.  I'm
>> not sure who was first, but my recollection is that this happened as
>> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
>> kind of agreements between those commercial entities and a piece of
>> the government, detailing the rules about ownership.  These first
>> ISPs were different because they were not research collaborators.
>> They were in business simply to make money selling Internet service.
> 
> Software Tool & Die's (aka "the World") was the first commercial
> consumer ISP, and UUNET was the first commercial IP network provider -
> who sold IP service to STD.

The way I remember it:

STD provided Internet services but they were not a network.

UUNET was a non-profit exchange, not a network, until 1990 (?)

PSINet was started as a network, with NYSERNet members as an instant
installed base, in 1989.



From spencer at mcsr-labs.org  Tue Oct 12 13:01:49 2010
From: spencer at mcsr-labs.org (Spencer Dawkins)
Date: Tue, 12 Oct 2010 15:01:49 -0500
Subject: [ih] Some Questions over IPv4 Ownership
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
	<1286900526.2518.38.camel@localhost>
Message-ID: <CD97CF20340A4C738F337A713F400CE0@china.huawei.com>

I would suggest that this question:

> If I use a particular range of IP addresses long enough, without the
> permission of whoever thinks they own it, do I now own those
> addresses...?  It's clear to me what the engineering answer is, but I
> have no idea what a court might decide.  Hence the question of ownership
> - legal rights - is very relevant.

really needs to be framed in terms of routability.

In a gentler time (pre-RFC 1918, et al), people just picked addresses at 
random for networks that "we know will never be connected to the Internet" 
to use internally, and when these networks were connected to the Internet, 
they were either renumbered (my team helped roll over a few thousand nodes 
one weekend) or NATted, because no one would send us packets to the prefix 
we were using (the prefix was already in the routing table, pointing 
somewhere else).

My understanding from the mid-1990s was that some prefixes were considered 
"polluted" - you didn't assign them, because someone was already "squatting" 
on them - but if the routing tables are updated to send traffic somewhere 
else, squatting wasn't interesting any more. Squatting became even less 
interesting when people started doing ingress filtering. At that point you 
couldn't receive packets sent to a squatted prefix, AND your ISP was 
dropping your packets at ingress time if they came from a squatted prefix.

If you "use" that address range, but no one sends you traffic for it, and no 
one accepts traffic from you that uses it, what does "use" actually mean?

Spencer 



From craig at aland.bbn.com  Tue Oct 12 13:10:00 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Tue, 12 Oct 2010 16:10:00 -0400
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
Message-ID: <20101012201000.3F1B228E137@aland.bbn.com>

> On 10/12/2010 15:00 EDT, Ofer Inbar wrote:
> > On Mon, Oct 11, 2010 at 08:03:49PM -0700,
> > jfh <jack at 3kitty.org> wrote:
> >> Somewhere along the line, commercial ISPs popped into being.  I'm
> >> not sure who was first, but my recollection is that this happened as
> >> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
> >> kind of agreements between those commercial entities and a piece of
> >> the government, detailing the rules about ownership.  These first
> >> ISPs were different because they were not research collaborators.
> >> They were in business simply to make money selling Internet service.
> > 
> > Software Tool & Die's (aka "the World") was the first commercial
> > consumer ISP, and UUNET was the first commercial IP network provider -
> > who sold IP service to STD.
> 
> The way I remember it:
> 
> STD provided Internet services but they were not a network.
> 
> UUNET was a non-profit exchange, not a network, until 1990 (?)
> 
> PSINet was started as a network, with NYSERNet members as an instant
> installed base, in 1989.

CSNET started in 1981 with a goal of achieving profitability within a few
years (which it achieved).  NSFNET modeled the NSF regional networks, in
part, on the CSNET model.

Thanks!

Craig


From mfidelman at meetinghouse.net  Tue Oct 12 13:30:53 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Tue, 12 Oct 2010 16:30:53 -0400
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
In-Reply-To: <20101012201000.3F1B228E137@aland.bbn.com>
References: <20101012201000.3F1B228E137@aland.bbn.com>
Message-ID: <4CB4C57D.3000702@meetinghouse.net>

Craig Partridge wrote:
>> On 10/12/2010 15:00 EDT, Ofer Inbar wrote:
>>      
>>> On Mon, Oct 11, 2010 at 08:03:49PM -0700,
>>> jfh<jack at 3kitty.org>  wrote:
>>>        
>>>> Somewhere along the line, commercial ISPs popped into being.  I'm
>>>> not sure who was first, but my recollection is that this happened as
>>>> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
>>>> kind of agreements between those commercial entities and a piece of
>>>> the government, detailing the rules about ownership.  These first
>>>> ISPs were different because they were not research collaborators.
>>>> They were in business simply to make money selling Internet service.
>>>>          
>>> Software Tool&  Die's (aka "the World") was the first commercial
>>> consumer ISP, and UUNET was the first commercial IP network provider -
>>> who sold IP service to STD.
>>>        
>> The way I remember it:
>>
>> STD provided Internet services but they were not a network.
>>
>> UUNET was a non-profit exchange, not a network, until 1990 (?)
>>
>> PSINet was started as a network, with NYSERNet members as an instant
>> installed base, in 1989.
>>      
> CSNET started in 1981 with a goal of achieving profitability within a few
> years (which it achieved).  NSFNET modeled the NSF regional networks, in
> part, on the CSNET model.
>    
And then there was NEARnet, that had commercial members almost from day 
one.  As I recall, since NEARnet didn't have any NSF funding, it didn't 
have to follow the limit on commercial users (except at the gateway to 
the NSFnet backbone).

Miles Fidelman

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From louie at transsys.com  Tue Oct 12 13:36:08 2010
From: louie at transsys.com (Louis Mamakos)
Date: Tue, 12 Oct 2010 16:36:08 -0400
Subject: [ih] Commercial ISPs (Re:  Some Questions over IPv4 Ownership)
In-Reply-To: <4CB4BAE7.4090800@cisco.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>	<1286852630.2466.59.camel@localhost>
	<20101012190010.GV18254@mip.aaaaa.org> <4CB4BAE7.4090800@cisco.com>
Message-ID: <C70C1E56-FD73-42CE-9693-D8949EF190C6@transsys.com>


On Oct 12, 2010, at 3:45 PM, Scott Brim wrote:

> On 10/12/2010 15:00 EDT, Ofer Inbar wrote:
>> On Mon, Oct 11, 2010 at 08:03:49PM -0700,
>> jfh <jack at 3kitty.org> wrote:
>>> Somewhere along the line, commercial ISPs popped into being.  I'm
>>> not sure who was first, but my recollection is that this happened as
>>> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
>>> kind of agreements between those commercial entities and a piece of
>>> the government, detailing the rules about ownership.  These first
>>> ISPs were different because they were not research collaborators.
>>> They were in business simply to make money selling Internet service.
>> 
>> Software Tool & Die's (aka "the World") was the first commercial
>> consumer ISP, and UUNET was the first commercial IP network provider -
>> who sold IP service to STD.
> 
> The way I remember it:
> 
> STD provided Internet services but they were not a network.
> 
> UUNET was a non-profit exchange, not a network, until 1990 (?)

More accurately, UUNET Communication Services was started by USENIX 
to provide paid UUCP and USENET news service for those that
couldn't find a friendly neighbor site for a UUCP connection.  UUNET
became a major UUCP hub for email and news.

Sometime later, UUNET Technologies, Inc. was started as a separate
enterprise, picking up the customers of the previous entity (and
paying a royalty back to the original company for some period of time)
that helped fund various good works.  It also started a commercial
IP backbone service, and then later sought venture funding to expand
the IP part of the business.  That was about the time that I joined
in 1995; the nascent backbone was already running by then.

It was a bit later that along with PSI and others, started an early
multi-party Internet exchange operated by MFS Datanet.  UUNET was a
customer of that exchange and it certainly wasn't non-profit as far
as MFS was concerned.

For quite some time, UUNET was a major email gateway interconnect
between UUCP-based email and the Internet; as well for USENET news
via UUCP and over the Internet with NNTP.

Louis Mamakos



From jack at 3kitty.org  Tue Oct 12 13:37:24 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Tue, 12 Oct 2010 13:37:24 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <CD97CF20340A4C738F337A713F400CE0@china.huawei.com>
References: <20101012152800.AF4AD6BE56B@mercury.lcs.mit.edu>
	<1286900526.2518.38.camel@localhost>
	<CD97CF20340A4C738F337A713F400CE0@china.huawei.com>
Message-ID: <1286915844.2518.72.camel@localhost>

On Tue, 2010-10-12 at 15:01 -0500, Spencer Dawkins wrote:
> If you "use" that address range, but no one sends you traffic for it,
> and no 
> one accepts traffic from you that uses it, what does "use" actually
> mean? 

Don't know.  Perhaps it means that you sue someone to enforce your
rights?   The legal world is often very confusing to us engineers...time
will tell.

/Jack




From wmaton at ottix.net  Tue Oct 12 13:35:56 2010
From: wmaton at ottix.net (William F. Maton)
Date: Tue, 12 Oct 2010 16:35:56 -0400 (EDT)
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
In-Reply-To: <20101012201000.3F1B228E137@aland.bbn.com>
References: <20101012201000.3F1B228E137@aland.bbn.com>
Message-ID: <Pine.LNX.4.64.1010121635000.24321@iskra.ottix.net>

On Tue, 12 Oct 2010, Craig Partridge wrote:

> CSNET started in 1981 with a goal of achieving profitability within a few
> years (which it achieved).  NSFNET modeled the NSF regional networks, in
> part, on the CSNET model.

Ah, now if only we all had a copy of The Matrix by John Quarterman at 
hand, which was updated to include the 90s.

wfms


From cos at aaaaa.org  Tue Oct 12 13:38:20 2010
From: cos at aaaaa.org (Ofer Inbar)
Date: Tue, 12 Oct 2010 16:38:20 -0400
Subject: [ih] Commercial ISPs (Re:  Some Questions over IPv4 Ownership)
In-Reply-To: <4CB4BAE7.4090800@cisco.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<1286852630.2466.59.camel@localhost>
	<20101012190010.GV18254@mip.aaaaa.org> <4CB4BAE7.4090800@cisco.com>
Message-ID: <20101012203820.GX18254@mip.aaaaa.org>

Scott Brim <sbrim at cisco.com> wrote:
> STD provided Internet services but they were not a network.
>
> UUNET was a non-profit exchange, not a network, until 1990 (?)
> 
> PSINet was started as a network, with NYSERNet members as an instant
> installed base, in 1989.

STD didn't have TCP/IP connectivity until UUNET started providing
that, which was either 1990 or 1991, I'm pretty sure.  Somewhere in my
email archive is a message from a friend - who I'd gotten to sign up
for a World account - telling me that he could now telnet or get on
IRC or something, and my disbelieving response (because I "knew" that
actual Internet connectivity was not allowed for public accounts like
that), so if I find that I could get the approximate date.  Again, if
there's anyone from STD on this list they can probably get the exact
date.

Saying that STD was "not a network" is ambiguous in the current context.

It was "not a network" but neither were the rest of the consumer ISPs
that followed them in the next few years.  IP to the home was very
rare.  Mostly, people got dialup accounts to Unix servers where they
logged in to a shell from which they could run stuff, and those places
were generally called "pubnix" accounts (public unix), and STD was the
first pubnix to offer such a shell account with IP connectivity.  Soon
many of them also started offerring PPP so your dialup account would
give your home computer a temporary IP address, but the ISPs were
still just offering dialup accounts to their servers; the "network"
portion was piggybacked on the existing circuit-switched telephone
network.

So in the early/mid 90s, ISPs and network providers were mostly
separate things.  An ISP was who you dialed into, whose server was in
turn connected to the net via a network provider.  But the ISP still
needed to be on the network, so it had its own IP allocation, and in
the early 90s it was routability of those blocks that was the issue.
And if course if you used PPP dialup, your public IP address came out
of the ISP's block too.  So in the context of IP addresses, the ISP
was your network.
  -- Cos


From craig at aland.bbn.com  Tue Oct 12 13:39:16 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Tue, 12 Oct 2010 16:39:16 -0400
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
Message-ID: <20101012203916.6003928E137@aland.bbn.com>


It is sitting at my elbow, what question do you wish to ask?

> On Tue, 12 Oct 2010, Craig Partridge wrote:
> 
> > CSNET started in 1981 with a goal of achieving profitability within a few
> > years (which it achieved).  NSFNET modeled the NSF regional networks, in
> > part, on the CSNET model.
> 
> Ah, now if only we all had a copy of The Matrix by John Quarterman at 
> hand, which was updated to include the 90s.
> 
> wfms
********************
Craig Partridge
Chief Scientist, BBN Technologies
E-mail: craig at aland.bbn.com or craig at bbn.com
Phone: +1 517 324 3425


From jack at 3kitty.org  Tue Oct 12 13:39:44 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Tue, 12 Oct 2010 13:39:44 -0700
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
In-Reply-To: <20101012201000.3F1B228E137@aland.bbn.com>
References: <20101012201000.3F1B228E137@aland.bbn.com>
Message-ID: <1286915984.2518.74.camel@localhost>

But when all these initial non-research-project ISPs and customers were
starting out, how were the legal issues framed?  Maybe the old contracts
or licenses or MOUs or whatever would shed some light.  /Jack


On Tue, 2010-10-12 at 16:10 -0400, Craig Partridge wrote:
> > On 10/12/2010 15:00 EDT, Ofer Inbar wrote:
> > > On Mon, Oct 11, 2010 at 08:03:49PM -0700,
> > > jfh <jack at 3kitty.org> wrote:
> > >> Somewhere along the line, commercial ISPs popped into being.  I'm
> > >> not sure who was first, but my recollection is that this happened as
> > >> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
> > >> kind of agreements between those commercial entities and a piece of
> > >> the government, detailing the rules about ownership.  These first
> > >> ISPs were different because they were not research collaborators.
> > >> They were in business simply to make money selling Internet service.
> > > 
> > > Software Tool & Die's (aka "the World") was the first commercial
> > > consumer ISP, and UUNET was the first commercial IP network provider -
> > > who sold IP service to STD.
> > 
> > The way I remember it:
> > 
> > STD provided Internet services but they were not a network.
> > 
> > UUNET was a non-profit exchange, not a network, until 1990 (?)
> > 
> > PSINet was started as a network, with NYSERNet members as an instant
> > installed base, in 1989.
> 
> CSNET started in 1981 with a goal of achieving profitability within a few
> years (which it achieved).  NSFNET modeled the NSF regional networks, in
> part, on the CSNET model.
> 
> Thanks!
> 
> Craig




From craig at aland.bbn.com  Tue Oct 12 13:53:17 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Tue, 12 Oct 2010 16:53:17 -0400
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
Message-ID: <20101012205317.E2F4828E137@aland.bbn.com>

> But when all these initial non-research-project ISPs and customers were
> starting out, how were the legal issues framed?  Maybe the old contracts
> or licenses or MOUs or whatever would shed some light.  /Jack

CSNET was initially funded as an NSF project in cooperation with DARPA with
the project requiring that CSNET be self-supporting within a certain number
of years.

CSNET had paying customers ("members") from the start -- the NSF money was
used to cover the shortfall in the initial years.

So there was a contract with NSF (actually, two contracts, because I believe
NSF contracted with UCAR and UCAR contracted operations to BBN), and an
agreement between DARPA and NSF that CSNET could attach to ARPANET and
relay traffic.

Not sure that gives any insight into IP addressing...

Thanks!

Craig

> 
> On Tue, 2010-10-12 at 16:10 -0400, Craig Partridge wrote:
> > > On 10/12/2010 15:00 EDT, Ofer Inbar wrote:
> > > > On Mon, Oct 11, 2010 at 08:03:49PM -0700,
> > > > jfh <jack at 3kitty.org> wrote:
> > > >> Somewhere along the line, commercial ISPs popped into being.  I'm
> > > >> not sure who was first, but my recollection is that this happened as
> > > >> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
> > > >> kind of agreements between those commercial entities and a piece of
> > > >> the government, detailing the rules about ownership.  These first
> > > >> ISPs were different because they were not research collaborators.
> > > >> They were in business simply to make money selling Internet service.
> > > > 
> > > > Software Tool & Die's (aka "the World") was the first commercial
> > > > consumer ISP, and UUNET was the first commercial IP network provider -
> > > > who sold IP service to STD.
> > > 
> > > The way I remember it:
> > > 
> > > STD provided Internet services but they were not a network.
> > > 
> > > UUNET was a non-profit exchange, not a network, until 1990 (?)
> > > 
> > > PSINet was started as a network, with NYSERNet members as an instant
> > > installed base, in 1989.
> > 
> > CSNET started in 1981 with a goal of achieving profitability within a few
> > years (which it achieved).  NSFNET modeled the NSF regional networks, in
> > part, on the CSNET model.
> > 
> > Thanks!
> > 
> > Craig
> 
********************
Craig Partridge
Chief Scientist, BBN Technologies
E-mail: craig at aland.bbn.com or craig at bbn.com
Phone: +1 517 324 3425


From jack at 3kitty.org  Tue Oct 12 14:05:50 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Tue, 12 Oct 2010 14:05:50 -0700
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
In-Reply-To: <20101012205317.E2F4828E137@aland.bbn.com>
References: <20101012205317.E2F4828E137@aland.bbn.com>
Message-ID: <1286917550.2518.76.camel@localhost>

Sure doesn't give any insight to me.  But I'm not a lawyer.  Maybe it's
kind of like real estate, where there's a whole industry - Title
Companies - that figure out who owns what.  At least enough to issue
insurance.

/Jack

On Tue, 2010-10-12 at 16:53 -0400, Craig Partridge wrote:
> > But when all these initial non-research-project ISPs and customers were
> > starting out, how were the legal issues framed?  Maybe the old contracts
> > or licenses or MOUs or whatever would shed some light.  /Jack
> 
> CSNET was initially funded as an NSF project in cooperation with DARPA with
> the project requiring that CSNET be self-supporting within a certain number
> of years.
> 
> CSNET had paying customers ("members") from the start -- the NSF money was
> used to cover the shortfall in the initial years.
> 
> So there was a contract with NSF (actually, two contracts, because I believe
> NSF contracted with UCAR and UCAR contracted operations to BBN), and an
> agreement between DARPA and NSF that CSNET could attach to ARPANET and
> relay traffic.
> 
> Not sure that gives any insight into IP addressing...
> 
> Thanks!
> 
> Craig
> 
> > 
> > On Tue, 2010-10-12 at 16:10 -0400, Craig Partridge wrote:
> > > > On 10/12/2010 15:00 EDT, Ofer Inbar wrote:
> > > > > On Mon, Oct 11, 2010 at 08:03:49PM -0700,
> > > > > jfh <jack at 3kitty.org> wrote:
> > > > >> Somewhere along the line, commercial ISPs popped into being.  I'm
> > > > >> not sure who was first, but my recollection is that this happened as
> > > > >> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
> > > > >> kind of agreements between those commercial entities and a piece of
> > > > >> the government, detailing the rules about ownership.  These first
> > > > >> ISPs were different because they were not research collaborators.
> > > > >> They were in business simply to make money selling Internet service.
> > > > > 
> > > > > Software Tool & Die's (aka "the World") was the first commercial
> > > > > consumer ISP, and UUNET was the first commercial IP network provider -
> > > > > who sold IP service to STD.
> > > > 
> > > > The way I remember it:
> > > > 
> > > > STD provided Internet services but they were not a network.
> > > > 
> > > > UUNET was a non-profit exchange, not a network, until 1990 (?)
> > > > 
> > > > PSINet was started as a network, with NYSERNet members as an instant
> > > > installed base, in 1989.
> > > 
> > > CSNET started in 1981 with a goal of achieving profitability within a few
> > > years (which it achieved).  NSFNET modeled the NSF regional networks, in
> > > part, on the CSNET model.
> > > 
> > > Thanks!
> > > 
> > > Craig
> > 
> ********************
> Craig Partridge
> Chief Scientist, BBN Technologies
> E-mail: craig at aland.bbn.com or craig at bbn.com
> Phone: +1 517 324 3425




From wmaton at ottix.net  Tue Oct 12 14:28:53 2010
From: wmaton at ottix.net (William F. Maton)
Date: Tue, 12 Oct 2010 17:28:53 -0400 (EDT)
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
In-Reply-To: <20101012203916.6003928E137@aland.bbn.com>
References: <20101012203916.6003928E137@aland.bbn.com>
Message-ID: <Pine.LNX.4.64.1010121727590.24321@iskra.ottix.net>

On Tue, 12 Oct 2010, Craig Partridge wrote:

>
> It is sitting at my elbow, what question do you wish to ask?

Mine is a simple but broad question:  Can you take a quick peek and see 
how 'recent' the historical coverage was?  up to '88?  '90?

Please and thanks,

>> On Tue, 12 Oct 2010, Craig Partridge wrote:
>>
>>> CSNET started in 1981 with a goal of achieving profitability within a few
>>> years (which it achieved).  NSFNET modeled the NSF regional networks, in
>>> part, on the CSNET model.
>>
>> Ah, now if only we all had a copy of The Matrix by John Quarterman at
>> hand, which was updated to include the 90s.
>>
>> wfms
> ********************
> Craig Partridge
> Chief Scientist, BBN Technologies
> E-mail: craig at aland.bbn.com or craig at bbn.com
> Phone: +1 517 324 3425
>

wfms


From jnc at mercury.lcs.mit.edu  Tue Oct 12 14:40:22 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Tue, 12 Oct 2010 17:40:22 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101012214022.F0E476BE54F@mercury.lcs.mit.edu>

    > From: Jack Haverty <jack at 3kitty.org>

    >> If you "use" that address range, but no one sends you traffic for it,
    >> and no one accepts traffic from you that uses it, what does "use"
    >> actually mean?

    > Perhaps it means that you sue someone to enforce your rights? 

Don't forget that due to the international nature of the Internet, you might
have to sue people in all sorts of jurisdictions if you resort to this
approach...

Suits over DNS entries have often worked when filed just in the US, because
ICANN, IANA, the main registry (VeriSign), many of the top-level servers, etc
are all here.

Whether the 'sue someone' approach would work for addresses (i.e. whether ISPs
around the world would voluntarily follow the lead of a court order in any
particular country, when it comes to entering routes in their routers) is an
interesting question...

	Noel


From lyndon at orthanc.ca  Tue Oct 12 14:45:09 2010
From: lyndon at orthanc.ca (Lyndon Nerenberg)
Date: Tue, 12 Oct 2010 14:45:09 -0700 (PDT)
Subject: [ih] Commercial ISPs (Re:  Some Questions over IPv4 Ownership)
In-Reply-To: <C70C1E56-FD73-42CE-9693-D8949EF190C6@transsys.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<1286852630.2466.59.camel@localhost>
	<20101012190010.GV18254@mip.aaaaa.org> <4CB4BAE7.4090800@cisco.com>
	<C70C1E56-FD73-42CE-9693-D8949EF190C6@transsys.com>
Message-ID: <alpine.BSF.2.00.1010121442270.79772@legolas.orthanc.ca>

> More accurately, UUNET Communication Services was started by USENIX
> to provide paid UUCP and USENET news service for those that
> couldn't find a friendly neighbor site for a UUCP connection.  UUNET
> became a major UUCP hub for email and news.

My memory says it was started by Rick Adams with a loan from Usenix.


From craig at aland.bbn.com  Tue Oct 12 14:59:25 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Tue, 12 Oct 2010 17:59:25 -0400
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
Message-ID: <20101012215925.C156B28E137@aland.bbn.com>


Don't forget the UUCP gateway via SEISMO (run by Rick Adams) before UUNET
Technologies.

At one point, I think CSNET's and SEISMO's were the two most used
IMP port on ARPANET (wish I had the old traffic reports...).

I do distinctly remember that shortly after 5pm Eastern, CSNET and
SEISMO would light up the ARPANET with traffic -- in the old days,
5pm was when telephone rates dropped and they could dial out to get
email.  And if SEISMO's IMP had trouble, it meant a call to a CSNET
techie (e.g. yours truly) in the middle of the night because the CSNET
disks would begin to overflow with traffic for SEISMO.

This is c. 1986.

Thanks!

Craig



> 
> On Oct 12, 2010, at 3:45 PM, Scott Brim wrote:
> 
> > On 10/12/2010 15:00 EDT, Ofer Inbar wrote:
> >> On Mon, Oct 11, 2010 at 08:03:49PM -0700,
> >> jfh <jack at 3kitty.org> wrote:
> >>> Somewhere along the line, commercial ISPs popped into being.  I'm
> >>> not sure who was first, but my recollection is that this happened as
> >>> spinoffs from NSFNet and/or CSNet.  Again, there must have been some
> >>> kind of agreements between those commercial entities and a piece of
> >>> the government, detailing the rules about ownership.  These first
> >>> ISPs were different because they were not research collaborators.
> >>> They were in business simply to make money selling Internet service.
> >> 
> >> Software Tool & Die's (aka "the World") was the first commercial
> >> consumer ISP, and UUNET was the first commercial IP network provider -
> >> who sold IP service to STD.
> > 
> > The way I remember it:
> > 
> > STD provided Internet services but they were not a network.
> > 
> > UUNET was a non-profit exchange, not a network, until 1990 (?)
> 
> More accurately, UUNET Communication Services was started by USENIX 
> to provide paid UUCP and USENET news service for those that
> couldn't find a friendly neighbor site for a UUCP connection.  UUNET
> became a major UUCP hub for email and news.
> 
> Sometime later, UUNET Technologies, Inc. was started as a separate
> enterprise, picking up the customers of the previous entity (and
> paying a royalty back to the original company for some period of time)
> that helped fund various good works.  It also started a commercial
> IP backbone service, and then later sought venture funding to expand
> the IP part of the business.  That was about the time that I joined
> in 1995; the nascent backbone was already running by then.
> 
> It was a bit later that along with PSI and others, started an early
> multi-party Internet exchange operated by MFS Datanet.  UUNET was a
> customer of that exchange and it certainly wasn't non-profit as far
> as MFS was concerned.
> 
> For quite some time, UUNET was a major email gateway interconnect
> between UUCP-based email and the Internet; as well for USENET news
> via UUCP and over the Internet with NNTP.
> 
> Louis Mamakos
********************
Craig Partridge
Chief Scientist, BBN Technologies
E-mail: craig at aland.bbn.com or craig at bbn.com
Phone: +1 517 324 3425


From jklensin at gmail.com  Tue Oct 12 15:38:29 2010
From: jklensin at gmail.com (John Klensin)
Date: Tue, 12 Oct 2010 18:38:29 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <D4AFEE0E-FB02-4275-ABB9-FA2B5A5F4C12@cs.fiu.edu>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<AANLkTinkmYv2cEeC0p=dqLrwnuVrOsUfBi95vfYA0oJS@mail.gmail.com>
	<D4AFEE0E-FB02-4275-ABB9-FA2B5A5F4C12@cs.fiu.edu>
Message-ID: <AANLkTikRWoaqCqAcGhU8AbRZ9Y7ZT=BThnRA95ZRMR7r@mail.gmail.com>

On 10/12/10, Ernie Rubi <ernesto at cs.fiu.edu> wrote:
> Sorry to belabor the point, but:
>
> Do you remember who decided the space and who employed that person(s), where
> it was decided (at work, at home)?
>
> Were you and Mr. Postel's actions under the control/direction of the US
> Government or a agency of said gov't at the time?
>
> I know these questions seem silly and perhaps intrusive but I'm trying to
> tie some of this to the law of agency.

Ernie,

I won't claim this answer is authoritative in any way, but we've had
this sort of discussion too many times, with now several generations
of law students or scholars wanting to generate theses, write books,
promote particular theories, or otherwise find retroactive
explanations for a lot of the Internet's origins and mechanisms.  From
my point of view (and I am not an engineer), they have all been
unsatisfactory to a greater or lesser degree because someone ultimate
seized on a theory and tried to force the facts and circumstances to
fit.   Maybe you will be the one who gets this completely right, but
you should forgive a certain level of skepticism and impatience based
on prior experience.

There are a few things you should probably understand to put things in context.

First, [D]ARPA almost never (perhaps never) handled its own
contracting and contract administration.  They issued "orders" to
other agencies to set up and manage the contracts.  Most technical
oversight  (such as it was) came from [D]ARPA staff, but there was
generally not very much of that at any sort of detailed level.

Second, at least with regard to topics related to Information
Technology (including the ARPANET/Internet work) during that period
[D]ARPA was very much in the "advanced research" business.  Contracts
and grants usually more closely resembled "take these resources, apply
them in this general direction, and see if you can produce something
useful" rather than "execute this particular set of tasks".

In that context, IP addresses (and ARPANET IMP and host numbers before
them) aren't assets in any sense that would align with the questions
you are trying to ask.  They are just identifiers that are
consequences of the design of the protocols (IPv4 in this case).  They
are necessary to reaching one host from another because one has to
have some uniform mechanism for talking about those hosts (and the IP
specification defined that mechanism in terms of addresses and
specified the format of the addresses).  That, in turn, created a
situation in which it was necessary to have a registry of addresses in
use on what we would now call the public Internet in order to preserve
uniqueness.  Creating that registry, and assigning addresses on
request, didn't need to be a big deal, a separate contract, or a
transfer of authority: it just needed to be done as a necessary part
of having IP[v4] work in practice.

>From that perspective, "who owns IPv4 addresses" can end up sounding a
lot like "who owns SMTP (or email address syntax) and grants me the
'right' to use it". I'll come back to that analogy in a moment.

It is also relevant that once one starts using an address that will be
used by other hosts, changing it ("renumbering") is a PITA.  How much
of a PITA has varied over time as different mechanisms have come into
being and the complexity of the networks making up the Internet has
evolved.   But it has never been completely trivial and doing so
imposes costs on whomever is doing the renumbering.  That observation
creates a situation that you may find unusual at first glance: someone
can say "ownership is irrelevant and not an issue" in one sentence and
"no one has the right to take this address away from me and force me
to renumber, at least without compensating me.for the costs" in the
next and be perfectly consistent.

Coming back to the email example,  if you obtain email service from me
and ask for a particular local-part of the mailbox name (the stuff
left of the "@") and I register it for you, you probably don't "own"
the address.  Maybe I'm leasing it to you (certainly some service
agreements have been drawn that way), maybe not, but, if you stop
using my service, your rights to receive mail at that address vanish.
Nonetheless, if I then reassign that address to someone whose goal is
to collect and use mail that was intended for you, you may have some
cause for complaint so I can't really claim I completely own it
either.

I hope that helps... at least with your understanding of why you
aren't going to get the crisp and simple answer you are looking for.

    john


From spencer at mcsr-labs.org  Tue Oct 12 19:00:19 2010
From: spencer at mcsr-labs.org (Spencer Dawkins)
Date: Tue, 12 Oct 2010 21:00:19 -0500
Subject: [ih] Some Questions over IPv4 Ownership
References: <20101012214022.F0E476BE54F@mercury.lcs.mit.edu>
Message-ID: <C4591A98D4D347EE95E846F602B7B4C0@china.huawei.com>

Noel actually pointed to the interesting middle-of-the-road case, when I was 
looking at the extreme case, which is probably less interesting.

I was saying "nobody will route to you/nobody will accept traffic with your 
source addresses", and you could probably sue your first-hop Internet 
connection and make them route to you, and make them accept traffic with 
your source addresses, so problem solved ...

... until you send an HTTP request to a server in another country. The 
request gets there, the server does its thing and sends a response, and the 
server's first-hop Internet connection, who has never been sued and doesn't 
think your address range goes where you think it goes, drops the response on 
the floor.

You'd have to decide how well you want the address range to work, to decide 
whether this is worth chasing, I think.

Spencer

>    >> If you "use" that address range, but no one sends you traffic for 
> it,
>    >> and no one accepts traffic from you that uses it, what does "use"
>    >> actually mean?
>
>    > Perhaps it means that you sue someone to enforce your rights?
>
> Don't forget that due to the international nature of the Internet, you 
> might
> have to sue people in all sorts of jurisdictions if you resort to this
> approach...
>
> Suits over DNS entries have often worked when filed just in the US, 
> because
> ICANN, IANA, the main registry (VeriSign), many of the top-level servers, 
> etc
> are all here.
>
> Whether the 'sue someone' approach would work for addresses (i.e. whether 
> ISPs
> around the world would voluntarily follow the lead of a court order in any
> particular country, when it comes to entering routes in their routers) is 
> an
> interesting question...
>
> Noel 



From spencer at mcsr-labs.org  Tue Oct 12 19:24:26 2010
From: spencer at mcsr-labs.org (Spencer Dawkins)
Date: Tue, 12 Oct 2010 21:24:26 -0500
Subject: [ih] Some Questions over IPv4 Ownership
References: <20101012214022.F0E476BE54F@mercury.lcs.mit.edu>
	<C4591A98D4D347EE95E846F602B7B4C0@china.huawei.com>
Message-ID: <56AB71ABCF6747138B47ED0E89763281@china.huawei.com>

Sorry, I got this wrong.

> I was saying "nobody will route to you/nobody will accept traffic with 
> your source addresses", and you could probably sue your first-hop Internet

this should have been "... you could conceivably sue ..."

> connection and make them route to you, and make them accept traffic with 
> your source addresses, so problem solved ...

My apologies. I was thinking "sue your first-hop guy with equipment in the 
same country" as opposed to "sue every ISP on earth".

Spencer 



From dhc2 at dcrocker.net  Tue Oct 12 17:08:13 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Tue, 12 Oct 2010 20:08:13 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <0871C0AA-B0B4-414D-B093-F9F7D7C16882@pch.net>
References: <20101011224145.2113728E137@aland.bbn.com>	<a06240823c8d956960d5e@[168.122.9.70]>	<4CB4548B.9030405@dcrocker.net>
	<0871C0AA-B0B4-414D-B093-F9F7D7C16882@pch.net>
Message-ID: <4CB4F86D.3020101@dcrocker.net>



On 10/12/2010 12:44 PM, Bill Woodcock wrote:
> On the contrary...  It has negative value.  Just as if everyone took their
> street numbers with them when they moved, and 125 Broadway were next door to
> 25337 El Camino Real and across the street from 17B Vine Terrace...  It would
> ruin the aggregation property of the system, which would remove all meaning.


While the technical aspects of your points are, of course, quite correct, I was 
trying to raise a different point.  (In general, I think that this discussion 
has been sidetracked by its overly-nuanced technical focus, given that the 
question was about legal and, therefore, business and marketing, issues.)

The point is that these tags have two kinds of functions, which can make 
discussion confusing.  One is the downward, technical side.  IP Addresses and 
street addresses really are addresses, in terms of providing topological 
addresses.  Telephone numbers used to be, but no longer are.  But this was not 
my point.

The upward aspect of these tags is that they are referents and it's often true 
that some referents are more appealing than others.  Even when telephone numbers 
were true addresses, some were more desirable than others and companies tried to 
get preferable ones.  And by the way, there was a degree of number portability 
even back then, albeit with serious geographic limitations.

That this latter aspect is in the 'identifier' realm and the form is in the 
'locator' realm is no doubt also true, but wasn't all that interesting to the 
specific point I was making, which was that these things tend to have 'value' 
and that features like portability are a virtue to that value.  (That they might 
be negatives in terms of the downward, locator-ish function is reality... for 
some of them).

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From dhc2 at dcrocker.net  Tue Oct 12 21:56:55 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Wed, 13 Oct 2010 00:56:55 -0400
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
In-Reply-To: <20101012205317.E2F4828E137@aland.bbn.com>
References: <20101012205317.E2F4828E137@aland.bbn.com>
Message-ID: <4CB53C17.7060600@dcrocker.net>



On 10/12/2010 4:53 PM, Craig Partridge wrote:
> CSNET had paying customers ("members") from the start


It did?

While I was but a mere techie doing the email services, I had the impression 
that the initial roll-out was sufficiently challenged, just to find willing 
adopters for this crazy, telephone-based relaying service.  Given how especially 
crappy the software was that they initially received for this, I'd be even more 
chagrined to hear that they were paying for it.

On the other hand, they DID pay for each telephone call... when they initiated it.

d/

-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From craig at aland.bbn.com  Wed Oct 13 03:57:46 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Wed, 13 Oct 2010 06:57:46 -0400
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
Message-ID: <20101013105746.CB38028E137@aland.bbn.com>


Hi Dave:

You preceeded me, but by the time I joined CSNET technical team in late 1984
or early 1985, all CSNET members paid dues and my understanding was that had
been true from inception.

Thanks!

Craig
> 
> 
> On 10/12/2010 4:53 PM, Craig Partridge wrote:
> > CSNET had paying customers ("members") from the start
> 
> 
> It did?
> 
> While I was but a mere techie doing the email services, I had the impression 
> that the initial roll-out was sufficiently challenged, just to find willing 
> adopters for this crazy, telephone-based relaying service.  Given how especia
> lly 
> crappy the software was that they initially received for this, I'd be even mo
> re 
> chagrined to hear that they were paying for it.
> 
> On the other hand, they DID pay for each telephone call... when they initiate
> d it.
> 
> d/
> 
> -- 
> 
>    Dave Crocker
>    Brandenburg InternetWorking
>    bbiw.net
********************
Craig Partridge
Chief Scientist, BBN Technologies
E-mail: craig at aland.bbn.com or craig at bbn.com
Phone: +1 517 324 3425


From jnc at mercury.lcs.mit.edu  Wed Oct 13 08:24:53 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Wed, 13 Oct 2010 11:24:53 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101013152453.E8C676BE566@mercury.lcs.mit.edu>

    > From: Ernie Rubi <ernesto at cs.fiu.edu>

    > On Oct 11, 2010, at 8:35 PM, Vint Cerf wrote:

    >> the initial address space was decided on in 1977

While Vint will have to answer this definitively, let me give you a sense of
what quadrant the answers will be in....


    > Do you remember who decided the space and who employed that person(s)

In answering this, I think it best to give you a little bit of background to
the development of IPv4 - because your question does not have a simple
answer. The choice of a 32-bit space was only the final stage in a longer
design/decision process, and prior versions of the design had a very similar
design,

To start with, you need to understand that in the beginning, there was no
separate IP protocol: TCP (the protocol that provides a reliable stream
service) had not been split apart from IP.

With that in hand, here is a table of the various versions preceeding IPv4,
with their dates, and what their address spaces looked like:

TCP 1 (documented in RFC 675)
December 1974 - Variable length (network number given in a field of up to 15
'nibbles' (i.e. 4-bit chunks), plus a 2-byte 'host number' on that network)

TCP 2 (documented in IEN-5)
March 1977 - Fixed 32 bits (1 byte network number, plus a 3 byte 'host number')

TCP 3 (documented in IEN 21)
Jan 1978 - Variable length (given in a field of up to 15 bytes)

TCP/IP 3.1 (documented in IENs 26/27/28)
Feb 1978 - Fixed 32 bits (as for TCP 2)

TCP/IP 4 (first documented in IENs 40/41)
Jun 1978 - Fixed 32 bits (as for TCP 2)

Today's IPv4 addresses (fixed 32 bits, but without any fixed boundary between
'network number' and 'host') came about in two even later steps; the
provision of class A/B/C addresses (circa 1980-81: they are not in RFC 760,
January 1980, and are in RFC 791, September 1981), and the adoption of CIDR
(circa September 1993 - RFC 1338 and RFC 1519).


So if you're asking i) 'who first decided on _an_ Internet host address
space', the answer is Cerf et al, in TCP 1, and they were all at Stanford at
the time. If you're asking ii) 'who first decided on _a_ 32-bit namespace',
it was again Cerf et al, in TCP 2 - but note that this was later replaced
with a different design for a while! If you're asking iii) 'who first decided
on the _current_ 32-bit namespace', it came in several stages: the first
(iiia) was Cerf et al, in TCP 3.1, going back to a 32-bit space - but it
didn't take today's form (with semi-arbitrary-sized chunks of that space)
until CIDR (which we can call iiic, with iiib being A/B/C).

When the particular engineering decision to go with a 32-bit address space
was taken, for both forms ii) and iii) of the question, Vint (whom I believe
has the largest share of the credit/blame for iiia), at least :-) was
employed by the USG, at DARPA (IIRC he was there from 1976-1982, after being
at Stanford previously). Others involved at that time were employed by a
variety of private entities, I think all under Government contract from DARPA.

As for all the others (e.g. the later CIDR stuff) - that would take a certain
amount of research to figure out (e.g. who was funding Cerf/Sunshine/Dalal at
Stanford in the early 70s).


    > where it was decided (at work, at home)?

It certainly wasn't a private decision (if that's what this question is
about). I would think that most of these decisions were taken at meetings
(most of the early ones all at DARPA-funded Internet project meetings; the
CIDR one at IETF meetings, for which the funding situation is infinitely more
complex).

    > Were you and Mr. Postel's actions under the control/direction of the US
    > Government or a agency of said gov't at the time? 

Vint was a US Govt employee during the ii)/iiia)/iiib) decisions. Jon, as an
employee of USC-ISI (from March of 1977 until his untimely death; before that
at UCLA, Mitre and SRI) was always, I _believe_, acting in the context of a
contract between the USG and ISI.

However, as others have noted, these contracts at that point in time did not
specifically say 'do this specific task, then that specific task', but were
rather more general - 'do good stuff in this area' kind of thing.

About i), I don't know. About iiic), most of those involved were neither
government employees, nor acting at USG direction under a contract from the
USG.

	Noel


From jmamodio at gmail.com  Wed Oct 13 08:36:32 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Wed, 13 Oct 2010 10:36:32 -0500
Subject: [ih] Commercial ISPs (Re: Some Questions over IPv4 Ownership)
In-Reply-To: <4CB4C57D.3000702@meetinghouse.net>
References: <20101012201000.3F1B228E137@aland.bbn.com>
	<4CB4C57D.3000702@meetinghouse.net>
Message-ID: <AANLkTi=hXPoG14Qaw64QiPqN5a8nMM+dW0hAenL1Og3_@mail.gmail.com>

> And then there was NEARnet, that had commercial members almost from day one.
> ?As I recall, since NEARnet didn't have any NSF funding, it didn't have to
> follow the limit on commercial users (except at the gateway to the NSFnet
> backbone).

JvNCNet became a commercial ISP under thenaem  Global Enterprise Services
in the early 90's, we provided services in eastern PA, NJ, NY (some overlap
with NYSERNet), CT, RI and MA, and international gateway for  IN (the first
two Software Technology Parks, PE, and VE), since we transitioned from the
network managed by Princeton University and had many academic institutions
we still had some funding from NSF and transit via NSF contract with MCI.

VERIO acquired it in 1997 as several others of the post NSF spin offs, such
as PREPNet, NorthwestNet, Sesquinet, etc.

Jorge



From craig at aland.bbn.com  Wed Oct 13 09:28:43 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Wed, 13 Oct 2010 12:28:43 -0400
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101013162843.CD58028E138@aland.bbn.com>

> As for all the others (e.g. the later CIDR stuff) - that would take a certain
> amount of research to figure out (e.g. who was funding Cerf/Sunshine/Dalal at
> Stanford in the early 70s).

I always viewed CIDR as a logical extension of subnets (the ability to own
a A, B or C address and break it up into smaller networks within your
organization).  And it is clear who invented subnets -- that was Jeff
Mogul (that's my memory and RFC 917 [published 1984] supports that
recollection).

In the context of Noel's note, Jeff was at Stanford at the time (I can't
recall in what capacity).

Thanks!

Craig


From dot at dotat.at  Wed Oct 13 09:28:50 2010
From: dot at dotat.at (Tony Finch)
Date: Wed, 13 Oct 2010 17:28:50 +0100
Subject: [ih] Commercial ISPs (Re:  Some Questions over IPv4 Ownership)
In-Reply-To: <20101012203820.GX18254@mip.aaaaa.org>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<1286852630.2466.59.camel@localhost>
	<20101012190010.GV18254@mip.aaaaa.org> <4CB4BAE7.4090800@cisco.com>
	<20101012203820.GX18254@mip.aaaaa.org>
Message-ID: <alpine.LSU.2.00.1010131719100.31035@hermes-2.csi.cam.ac.uk>

On Tue, 12 Oct 2010, Ofer Inbar wrote:
>
> Saying that STD was "not a network" is ambiguous in the current context.
>
> It was "not a network" but neither were the rest of the consumer ISPs
> that followed them in the next few years.  IP to the home was very
> rare.  Mostly, people got dialup accounts to Unix servers where they
> logged in to a shell from which they could run stuff,

The situation in the UK was rather different. IP connectivity to the home
was normal for ISPs based on the way Demon Internet pioneered consumer
Internet connectivity (starting in 1992). I think part of the reason was
per-minute call charges made it undesirable for users to sit on the phone
while reading mail and news on a remote shell server; instead they
downloaded it for reading offline.

(PIPEX was founded earlier, in 1990, but it concentrated more on
commercial customers, one of which was Demon.)

Tony.
-- 
f.anthony.n.finch  <dot at dotat.at>  http://dotat.at/
HUMBER THAMES DOVER WIGHT PORTLAND: NORTH BACKING WEST OR NORTHWEST, 5 TO 7,
DECREASING 4 OR 5, OCCASIONALLY 6 LATER IN HUMBER AND THAMES. MODERATE OR
ROUGH. RAIN THEN FAIR. GOOD.


From jnc at mercury.lcs.mit.edu  Wed Oct 13 10:03:15 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Wed, 13 Oct 2010 13:03:15 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101013170315.E9A056BE566@mercury.lcs.mit.edu>

    > From: Craig Partridge <craig at aland.bbn.com>

    > I always viewed CIDR as a logical extension of subnets (the ability to
    > own a A, B or C address and break it up into smaller networks within
    > your organization).

Also (and even more so) Roki's then-well-known work on supernetting, which
everyone keeps forgetting (hint, hint :-)... See this message:

  http://www.ietf.org/mail-archive/web/ietf/current/msg24111.html

and following.

    > And it is clear who invented subnets -- that was Jeff Mogul (that's my
    > memory and RFC 917 [published 1984] supports that recollection).

Ahem. IP subnets had been in use at MIT for many years - see, for instance,
IEN 82, February 1979, "LCS Net Address Format". Jeff was an undergrad at MIT,
and 'merely' (an invaluable service, to be sure) documented basically what we
were already doing, as one entry in the 'subnet sweepstakes' (see also RFC's
925, 932, etc), once it became clear that other sites were going to run into
the same situation that MIT had been in for many years (too many LANs at the
site, the 'main' routing tables could not - and needed not - have an entry for
each one).

I should give a further 'shout out' to David Moon, who is the origin of the
'subnet mask' part of subnetting (he mentioned the idea in email to me, and I
promptly glommed onto it, and passed it on to Jeff).

    > In the context of Noel's note, Jeff was at Stanford at the time (I
    > can't recall in what capacity).

Grad student.

	Noel


From craig at aland.bbn.com  Wed Oct 13 10:05:14 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Wed, 13 Oct 2010 13:05:14 -0400
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101013170514.8863C28E137@aland.bbn.com>

> Ahem. IP subnets had been in use at MIT for many years ...

Pardon me!  Always fun to learn these details.

Thanks!

Craig


From craig at aland.bbn.com  Wed Oct 13 10:07:15 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Wed, 13 Oct 2010 13:07:15 -0400
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101013170715.372AB28E137@aland.bbn.com>


By the way...

Wearing my hat as an editorial for the IEEE Annals of the History of Computing,
it is clear from these notes that there's the potential for a good paper on
the technical evolution of Internet addressing here (Noel's progression of
formats, then subnets and supernets, then CIDR...) if someone feels up to
taking on the effort.

Thanks!

Craig

>     > From: Craig Partridge <craig at aland.bbn.com>
> 
>     > I always viewed CIDR as a logical extension of subnets (the ability to
>     > own a A, B or C address and break it up into smaller networks within
>     > your organization).
> 
> Also (and even more so) Roki's then-well-known work on supernetting, which
> everyone keeps forgetting (hint, hint :-)... See this message:
> 
>   http://www.ietf.org/mail-archive/web/ietf/current/msg24111.html
> 
> and following.
> 
>     > And it is clear who invented subnets -- that was Jeff Mogul (that's my
>     > memory and RFC 917 [published 1984] supports that recollection).
> 
> Ahem. IP subnets had been in use at MIT for many years - see, for instance,
> IEN 82, February 1979, "LCS Net Address Format". Jeff was an undergrad at MIT
> ,
> and 'merely' (an invaluable service, to be sure) documented basically what we
> were already doing, as one entry in the 'subnet sweepstakes' (see also RFC's
> 925, 932, etc), once it became clear that other sites were going to run into
> the same situation that MIT had been in for many years (too many LANs at the
> site, the 'main' routing tables could not - and needed not - have an entry fo
> r
> each one).
> 
> I should give a further 'shout out' to David Moon, who is the origin of the
> 'subnet mask' part of subnetting (he mentioned the idea in email to me, and I
> promptly glommed onto it, and passed it on to Jeff).
> 
>     > In the context of Noel's note, Jeff was at Stanford at the time (I
>     > can't recall in what capacity).
> 
> Grad student.
> 
> 	Noel
********************
Craig Partridge
Chief Scientist, BBN Technologies
E-mail: craig at aland.bbn.com or craig at bbn.com
Phone: +1 517 324 3425


From louie at transsys.com  Wed Oct 13 11:30:07 2010
From: louie at transsys.com (Louis Mamakos)
Date: Wed, 13 Oct 2010 14:30:07 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101013170315.E9A056BE566@mercury.lcs.mit.edu>
References: <20101013170315.E9A056BE566@mercury.lcs.mit.edu>
Message-ID: <695DEE37-FC87-47F7-AE65-F66F3511E2AB@transsys.com>

Two other random thoughts on this topic:

For a while, the generalized subnet architecture "supported" non-continguous
subnet masks to distinguish the network part from the host part.  I don't
know what the driver for this was.  Maybe IP address <-> ARPANET host/imp
mapping [10.host.X.imp] where the X octet was used as a "logical host" behind
some sort of multiplexing device or gateway/router.   A subnet mask 
might be 255.255.0.255 in that scenario.

Of course, when CIDR came on the scene, a strict prefix approach was adopted.
Actual deployment of CIDR was serialized behind the BGP3->BGP4 deployment
activities on the major backbones at that time.

Louis Mamakos

On Oct 13, 2010, at 1:03 PM, Noel Chiappa wrote:

>> From: Craig Partridge <craig at aland.bbn.com>
> 
>> I always viewed CIDR as a logical extension of subnets (the ability to
>> own a A, B or C address and break it up into smaller networks within
>> your organization).
> 
> Also (and even more so) Roki's then-well-known work on supernetting, which
> everyone keeps forgetting (hint, hint :-)... See this message:
> 
>  http://www.ietf.org/mail-archive/web/ietf/current/msg24111.html
> 
> and following.
> 
>> And it is clear who invented subnets -- that was Jeff Mogul (that's my
>> memory and RFC 917 [published 1984] supports that recollection).
> 
> Ahem. IP subnets had been in use at MIT for many years - see, for instance,
> IEN 82, February 1979, "LCS Net Address Format". Jeff was an undergrad at MIT,
> and 'merely' (an invaluable service, to be sure) documented basically what we
> were already doing, as one entry in the 'subnet sweepstakes' (see also RFC's
> 925, 932, etc), once it became clear that other sites were going to run into
> the same situation that MIT had been in for many years (too many LANs at the
> site, the 'main' routing tables could not - and needed not - have an entry for
> each one).
> 
> I should give a further 'shout out' to David Moon, who is the origin of the
> 'subnet mask' part of subnetting (he mentioned the idea in email to me, and I
> promptly glommed onto it, and passed it on to Jeff).
> 
>> In the context of Noel's note, Jeff was at Stanford at the time (I
>> can't recall in what capacity).
> 
> Grad student.
> 
> 	Noel
> 




From jklensin at gmail.com  Wed Oct 13 11:50:21 2010
From: jklensin at gmail.com (John Klensin)
Date: Wed, 13 Oct 2010 14:50:21 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTikRWoaqCqAcGhU8AbRZ9Y7ZT=BThnRA95ZRMR7r@mail.gmail.com>
References: <0FB6929C-D791-481D-9D1C-B434349B513B@cs.fiu.edu>
	<AANLkTinkmYv2cEeC0p=dqLrwnuVrOsUfBi95vfYA0oJS@mail.gmail.com>
	<D4AFEE0E-FB02-4275-ABB9-FA2B5A5F4C12@cs.fiu.edu>
	<AANLkTikRWoaqCqAcGhU8AbRZ9Y7ZT=BThnRA95ZRMR7r@mail.gmail.com>
Message-ID: <AANLkTinWxpVq3JYRg5ZcAO_XALH_AzsPpmpY8WSvvbaG@mail.gmail.com>

On rereading, I realized that one aspect of my long note yesterday
might easily be misinterpreted.  To clarify, when I said that there
was generally not very much technical oversight from [D]ARPA staff at
any sort of detailed level, I did not intend to imply any lack of
diligence or attention.  Instead, my impression is that the people
involved were doing exactly what they should have been doing -- paying
careful attention to design, protocol, and strategic issues (even in
considerable detail), but, except for periodic high-level reviews,
trusting that the people doing administrative and registry work were
doing those jobs.   In particular, what I was referring to when I
mentioned "detailed level" in the context of address allocations was
that I don't believe there was any required sign off process at
[D]ARPA before a given allocation became final (nor, unlike top-level
domain names), is there any such sign off process by US DoC about
individual allocations from IANA to the RIRs today.

Sorry for any confusion I may have caused by not being more precise.

   john


From jnc at mercury.lcs.mit.edu  Wed Oct 13 15:39:48 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Wed, 13 Oct 2010 18:39:48 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>

    > From: Louis Mamakos <louie at transsys.com>

    > For a while, the generalized subnet architecture "supported"
    > non-continguous subnet masks to distinguish the network part from the
    > host part. I don't know what the driver for this was.

Pure serendipity. The basic mechanism used (the bit-mask) allowed
non-continguous masks as a natural consequence of the mechanism; we would
have had to specifically disallow them, and we didn't bother to. I seem to
recall we thought we'd leave them in, and see if people found some good use
for them.

	Noel


From randy at psg.com  Wed Oct 13 21:22:27 2010
From: randy at psg.com (Randy Bush)
Date: Thu, 14 Oct 2010 07:22:27 +0300
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
Message-ID: <m2k4lle73w.wl%randy@psg.com>

> Pure serendipity. The basic mechanism used (the bit-mask) allowed
> non-continguous masks as a natural consequence of the mechanism; we would
> have had to specifically disallow them, and we didn't bother to. I seem to
> recall we thought we'd leave them in, and see if people found some good use
> for them.

paul tsuchiya's still unfindable campei non-contiguous subnet mask
architecture c. 1992

randy



From vint at google.com  Wed Oct 13 21:50:10 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 14 Oct 2010 00:50:10 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101013152453.E8C676BE566@mercury.lcs.mit.edu>
References: <20101013152453.E8C676BE566@mercury.lcs.mit.edu>
Message-ID: <AANLkTinpK3ZT-_KaYmtu9iD6FOm8GHy-zxyTAec6V7kH@mail.gmail.com>

Noel's summary is accurate as far as I know.

the question:
As for all the others (e.g. the later CIDR stuff) - that would take a certain
amount of research to figure out (e.g. who was funding Cerf/Sunshine/Dalal at
Stanford in the early 70s).

the answer: Yogen Dalal and Carl Sunshine were graduate students at
Stanford and paid out of my ARPA contract.

vint


From benc at hawaga.org.uk  Wed Oct 13 23:59:03 2010
From: benc at hawaga.org.uk (Ben Clifford)
Date: Thu, 14 Oct 2010 06:59:03 +0000 (GMT)
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
Message-ID: <Pine.LNX.4.64.1010140658150.25630@dildano.hawaga.org.uk>



> Pure serendipity. The basic mechanism used (the bit-mask) allowed
> non-continguous masks as a natural consequence of the mechanism; we would
> have had to specifically disallow them, and we didn't bother to. I seem to
> recall we thought we'd leave them in, and see if people found some good use
> for them.

did anyone find a good use for them?

-- 


From vint at google.com  Thu Oct 14 03:03:09 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 14 Oct 2010 06:03:09 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <Pine.LNX.4.64.1010140658150.25630@dildano.hawaga.org.uk>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<Pine.LNX.4.64.1010140658150.25630@dildano.hawaga.org.uk>
Message-ID: <AANLkTikScXSQ4t8QAsdOPe9TRJeN0QbMmwzUdtiU_jyH@mail.gmail.com>

the guys at stanford have invented an "open flow" router that works this way.

On Thu, Oct 14, 2010 at 2:59 AM, Ben Clifford <benc at hawaga.org.uk> wrote:
>
>
>> Pure serendipity. The basic mechanism used (the bit-mask) allowed
>> non-continguous masks as a natural consequence of the mechanism; we would
>> have had to specifically disallow them, and we didn't bother to. I seem to
>> recall we thought we'd leave them in, and see if people found some good use
>> for them.
>
> did anyone find a good use for them?
>
> --
>


From sbrim at cisco.com  Thu Oct 14 04:51:34 2010
From: sbrim at cisco.com (Scott Brim)
Date: Thu, 14 Oct 2010 07:51:34 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <m2k4lle73w.wl%randy@psg.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com>
Message-ID: <4CB6EEC6.2050107@cisco.com>

On 10/14/2010 00:22 EDT, Randy Bush wrote:
>> Pure serendipity. The basic mechanism used (the bit-mask) allowed
>> non-continguous masks as a natural consequence of the mechanism; we would
>> have had to specifically disallow them, and we didn't bother to. I seem to
>> recall we thought we'd leave them in, and see if people found some good use
>> for them.
> 
> paul tsuchiya's still unfindable campei non-contiguous subnet mask
> architecture c. 1992
> 
> randy
> 

kampai and rfc1219?

As for actual use, I recall some planned uses of discontiguous mask but
I don't remember anyone actually putting them in production.


From dot at dotat.at  Thu Oct 14 04:59:30 2010
From: dot at dotat.at (Tony Finch)
Date: Thu, 14 Oct 2010 12:59:30 +0100
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <Pine.LNX.4.64.1010140658150.25630@dildano.hawaga.org.uk>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<Pine.LNX.4.64.1010140658150.25630@dildano.hawaga.org.uk>
Message-ID: <alpine.LSU.2.00.1010141238330.29333@hermes-2.csi.cam.ac.uk>

On Thu, 14 Oct 2010, Ben Clifford wrote:
> Noel Chiappa <jnc at mercury.lcs.mit.edu> wrote:
> >
> > Pure serendipity. The basic mechanism used (the bit-mask) allowed
> > non-continguous masks as a natural consequence of the mechanism; we would
> > have had to specifically disallow them, and we didn't bother to. I seem to
> > recall we thought we'd leave them in, and see if people found some good use
> > for them.
>
> did anyone find a good use for them?

I played around with them a bit in about 1998 at Demon. The "homepages"
service had a /16 and two /18s for HTTP virtual hosts (the service
predated widespread support for the Host: header). These address ranges
were routed to a number of reverse proxy caches running Squid on FreeBSD,
which was patched so each one could accept connections to any of the 96K
IP addresses. (http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/12071)

Addresses were allocated from these networks sequentially, and the oldest
web sites tended to get the most traffic, so a straightforward setup that
spread the six /18s across the reverse proxies didn't balance the load
particularly well. I toyed with using 0xffff0003 netmasks to split the /16
so that successive addresses could be routed to each of the four London
reverse proxies in turn.

This worked in testing but I didn't deploy it because it broke my
colleagues' brains and non-contiguous netmasks were an unsupported
feature.

Tony.
-- 
f.anthony.n.finch  <dot at dotat.at>  http://dotat.at/
HUMBER THAMES DOVER WIGHT PORTLAND: NORTH BACKING WEST OR NORTHWEST, 5 TO 7,
DECREASING 4 OR 5, OCCASIONALLY 6 LATER IN HUMBER AND THAMES. MODERATE OR
ROUGH. RAIN THEN FAIR. GOOD.


From ernesto at cs.fiu.edu  Thu Oct 14 06:05:56 2010
From: ernesto at cs.fiu.edu (Ernesto Rubi)
Date: Thu, 14 Oct 2010 09:05:56 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB6EEC6.2050107@cisco.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
Message-ID: <2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>

As a follow up - ::wearing lawyer hat::

Language in law is important, and nowhere is that more evident than in
contracts and contract clauses.

To give you an idea of the pervasive 'misunderstanding' that legal folk
currently have with regards to IP addresses and their 'property
interest'/status, here is an actual subpart from a contract Qwest signed
with salesforce.com which is *listed on Westlaw as a model contract for
telecommunications services*

I.8. Rights and Obligations of Qwest; Disclaimer of Warranties


(a) As may be set forth in the Addendum, Qwest will secure domain names
and assign IP address space (subject to reasonable availability) for the
benefit of Customer during the Term, and Qwest will route those addresses
on Qwest's network; it being understood and agreed that neither Customer
nor any of its ?Users? (as defined in the AUP) shall have the right to
route these addresses. Customer understands and agrees that it shall have
no ownership interest in any IP address which Qwest obtains on Customer's
behalf and that Qwest shall retain ownership of all such IP addresses, and
upon termination of the Agreement, Customer's access to and utilization of
such IP addresses shall terminate.

"ownership" is not defined anywhere else in the contract.

Maybe we should put together some workshops.

=)

Ernesto Rubi
Network Engineer
FIU / AMPATH
Email:  ernesto at cs.fiu.edu




From jeanjour at comcast.net  Thu Oct 14 06:49:56 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 09:49:56 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
Message-ID: <a06240834c8dcb842b520@[10.0.1.21]>

I see nothing wrong with this language.  Qwest is 
the one assigning addresses so that it can use 
them to route.  If anyone "owns" the addresses, 
it is the entity assigning them to facilitate 
routing in its domain.

To take our street analogy earlier, the "owner" 
of street addresses is the City.  It administers 
the registry and when a new building is created 
it decides what number it should get to be 
consistent with previous assignments.  It "owns" 
the address space.

It is administering it so that people and those 
providing services can find the building with the 
address.  In rural Illinois a few years ago, 
there was a major effort to assign addresses to 
homes in the "country" i.e. not in any town, for 
911 purposes so that police and fire depts could 
be directed to an emergency.  Previous to this 
there had been no formal addressing.  (The locals 
just knew where everyone lived. "O, that is the 
old Garrett place" that sort of thing went on. 
;-) But with county wide dispatching it needed to 
be regularized.)  The county owned the address 
space.

If Qwest at some point wanted to re-shuffle its 
use of addresses that make routing "better" that 
would be its prerogative.  Of course it would 
want to weigh disruption to customers, but it is 
still its business to do that.

In the Internet case, since we have to do 
provider based addressing, Qwest corresponds to 
the City or County.

I see nothing inconsistent with the language of 
the contract and it seem completely in keeping 
with what we would intend.

Take care,
John

>As a follow up - ::wearing lawyer hat::
>
>Language in law is important, and nowhere is that more evident than in
>contracts and contract clauses.
>
>To give you an idea of the pervasive 'misunderstanding' that legal folk
>currently have with regards to IP addresses and their 'property
>interest'/status, here is an actual subpart from a contract Qwest signed
>with salesforce.com which is *listed on Westlaw as a model contract for
>telecommunications services*
>
>I.8. Rights and Obligations of Qwest; Disclaimer of Warranties
>
>
>(a) As may be set forth in the Addendum, Qwest will secure domain names
>and assign IP address space (subject to reasonable availability) for the
>benefit of Customer during the Term, and Qwest will route those addresses
>on Qwest's network; it being understood and agreed that neither Customer
>nor any of its ?Users? (as defined in the AUP) shall have the right to
>route these addresses. Customer understands and agrees that it shall have
>no ownership interest in any IP address which Qwest obtains on Customer's
>behalf and that Qwest shall retain ownership of all such IP addresses, and
>upon termination of the Agreement, Customer's access to and utilization of
>such IP addresses shall terminate.
>
>"ownership" is not defined anywhere else in the contract.
>
>Maybe we should put together some workshops.
>
>=)
>
>Ernesto Rubi
>Network Engineer
>FIU / AMPATH
>Email:  ernesto at cs.fiu.edu




From ernesto at cs.fiu.edu  Thu Oct 14 07:04:27 2010
From: ernesto at cs.fiu.edu (Ernesto Rubi)
Date: Thu, 14 Oct 2010 10:04:27 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240834c8dcb842b520@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<a06240834c8dcb842b520@[10.0.1.21]>
Message-ID: <2d3f01ca98dd8b4c71953183a0b432d9.squirrel@webmail.cs.fiu.edu>

I understand, it's just that when 'own' is used; it implies the same
bundle of rights that you'd have as a real or intellectual property
'owner.'

So, routing policy aside, Qwest can't 'own' in the same fee-simple
absolute way you 'own' your real property.

For example, there are severe limitations as to what Qwest can and cannot
do with regards to alienation of the 'address' property it purpotedly
'owns.'


 If anyone "owns" the addresses,
> it is the entity assigning them to facilitate
> routing in its domain.
>
-- 
Ernesto Rubi
Network Engineer
FIU / AMPATH
Email:  ernesto at cs.fiu.edu




From el at lisse.NA  Thu Oct 14 07:08:10 2010
From: el at lisse.NA (Dr Eberhard Lisse)
Date: Thu, 14 Oct 2010 16:08:10 +0200
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
Message-ID: <4CB70ECA.8070202@lisse.NA>

Ernesto,

I have always wondered what an IP address (and of course a domain name)
actually is, legally speaking.

But then, I am not a laywer, only an elderly Obstetrician (though, to
quote Eric "Otter" Stratton: "What's the difference?" (Yes, I know,
wishful thinking :-)-O)), but

greetings, el

on 2010-10-14 15:05 Ernesto Rubi said the following:
[...

> (a) As may be set forth in the Addendum, Qwest will secure domain names
> and assign IP address space (subject to reasonable availability) for the
> benefit of Customer during the Term, and Qwest will route those addresses
> on Qwest's network; it being understood and agreed that neither Customer
> nor any of its ?Users? (as defined in the AUP) shall have the right to
> route these addresses. Customer understands and agrees that it shall have
> no ownership interest in any IP address which Qwest obtains on Customer's
> behalf and that Qwest shall retain ownership of all such IP addresses, and
> upon termination of the Agreement, Customer's access to and utilization of
> such IP addresses shall terminate.
> 
> "ownership" is not defined anywhere else in the contract.
[...]


-- 
Dr. Eberhard W. Lisse  \        / Obstetrician & Gynaecologist (Saar)
el at lisse.NA el108-ARIN / *     |   Telephone: +264 81 124 6733 (cell)
PO Box 8421             \     /   Please do NOT email to this address
Bachbrecht, Namibia     ;____/        if it is DNS related in ANY way


From jmamodio at gmail.com  Thu Oct 14 07:16:27 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Thu, 14 Oct 2010 09:16:27 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
Message-ID: <AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>

> Language in law is important, and nowhere is that more evident than in
> contracts and contract clauses.
...

> "ownership" is not defined anywhere else in the contract.

There is a BCP (RFC2008) that talks about "address ownership"

http://tools.ietf.org/html/bcp7

Jorge


From ernesto at cs.fiu.edu  Thu Oct 14 07:30:49 2010
From: ernesto at cs.fiu.edu (Ernesto Rubi)
Date: Thu, 14 Oct 2010 10:30:49 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
Message-ID: <2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>

Yup, saw/read RFC 2008, but there are two things with respect to the
contract language;

It doesn't incorporate RFC 2008 or a similar concept into the contract
itself.

RFC 2008 is circa 1996, so if you base an ownership theory just on RFC
2008, that still leaves a gap from 1977-1996.

-----
As to Dr. Lisse's question, it would appear, at least in the US, that
domain names have a very strong ownership interest.

In fact, through the application of trademark law (anti-dilution, etc) and
later through the Anti-Cybersquatting Protection Act (ACPA) Congress gave
owners of marks (and especially famous marks) the ability to go to a U.S.
court and get at least an injunction and possibly monetary damages against
the person(s) using the mark as a domain name.

Importantly, ACPA also gave 'in rem' (~ explained: under English common
law reserved only for real property) jurisdiction to actions pertaining to
domain names in the district court covering the registrar of the domain
name.  So a plaintiff could now come into a US court and obtain
jurisdiction over ford.com in Virginia (NSI) and if successful could have
the court command NSI transfer registration/point NS to the proper IP.

Internationally, ICANN has adopted UDRP, which broadly speaking allows
folks to do the same as ACPA.

As a general statement, DNS case law is very well established, while IP
case law is non-existent (except for a one-off ex parte order in N.D.
california, which didn't even really decide ownership issues) and what you
do find is court decisions repeatedly saying things like "the owner of the
IP address" when refering to individuals whose IP has been tracked/logged,
etc.  Just plain odd.

Also part of my motivation to write the paper.

I'm taking 3 more courses during the evenings and work full time during
the day, so I have limited time and my paper is limited to 25 pages. 
Someone, somewhere really ought to spend a whole semester locked in a room
researching this.  If only I had graduate assistants...hehehe


----
Jorge Amodio wrote:
>> Language in law is important, and nowhere is that more evident than in
>> contracts and contract clauses.
> ...
>
>> "ownership" is not defined anywhere else in the contract.
>
> There is a BCP (RFC2008) that talks about "address ownership"
>
> http://tools.ietf.org/html/bcp7
>
> Jorge
>


-- 
Ernesto Rubi
Network Engineer
FIU / AMPATH
Email:  ernesto at cs.fiu.edu




From jklensin at gmail.com  Thu Oct 14 08:09:02 2010
From: jklensin at gmail.com (John Klensin)
Date: Thu, 14 Oct 2010 11:09:02 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <2d3f01ca98dd8b4c71953183a0b432d9.squirrel@webmail.cs.fiu.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<a06240834c8dcb842b520@10.0.1.21>
	<2d3f01ca98dd8b4c71953183a0b432d9.squirrel@webmail.cs.fiu.edu>
Message-ID: <AANLkTins4zmFAQRqBbbMzF+o7oib=oZvbhftA4Lv0smU@mail.gmail.com>

Well, at some level, of course.  The problem here is the hierarchy
which, wrt "ownership" bears more relationship to ownership/possession
delegation in a feudal system than to some notion of absolute
ownership.   Qwest certainly has the "ownership" rights it claims wrt
addresses delegated (or assigned, if one prefers) to its customers.
But its own "ownership" depends on its relationship with ARIN and
ARIN's "ownership" depends on its relationship with IANA.  I'm
disinclined to push back further than that for the reasons given in a
previous note, but others might.

Can Qwest retrieve address space that had been used by a previous
customer? Certainly and it is done all the time.  Can they retrieve or
reassign address space that is being used by a current customer and
force renumbering?   Certainly yes in theory.  In practice, I would
assume that the answer might have more to do with how arbitrary the
decision was and with the pain and costs of renumbering than with
abstract theories of ownership.  And I've seen more than a few ISPs
hesitate to reassign addresses to customers who would find renumbering
sufficiently painful to start hiring lawyers, precisely to avoid
testing the theory of just how far their "ownership" rights extended.

People closer to the current RIR system that I am may want to comment
on this, but I assume the situation is much the same.  I understand
that their current, non-legacy, agreements about assignment/delegation
of address space try to make it clear that such assignments don't
transfer any ownership rights.  In principle, they could take
addresses back from an active user of that address space to whom the
address space had been delegated.  In practice, trying it would
probably be really ill-advised.  And, again, that is less because of
theories of ownership than because renumbering can be expensive and
because the RIRs are consensus organizations and any perception of
significant arbitrary behavior would probably cause a revolution from
their members -- either on the basis of concerns about fairness or out
of concern about who they would come after next.

A different way of looking at all of this is that the entire
addressing allocation/ delegation/ registration model ultimately rests
on a combination of collective consent and an understanding that
having non-unique addresses floating around causes problems in routing
packets.  ISPs have lots of incentives to announce and route only
registered addresses; those who might fail to do so would be
performing a disservice to their customers (unpredictable delivery),
would antagonize their peers and probably be cut off, with the
long-term effect of putting them out of business.   Could someone
appropriate a range of addresses and start using them locally and with
their closest friends?  Sure and before RFC 1918, it used to happen
all the time.  But, as Spencer and others have pointed out, such
networks need to be kept completely isolated from the public network
and its address space; if it is not, we have non-unique addresses and
A Big Mess whether there is a debate about ownership or not.


From jmamodio at gmail.com  Thu Oct 14 08:32:19 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Thu, 14 Oct 2010 10:32:19 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB70ECA.8070202@lisse.NA>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<4CB70ECA.8070202@lisse.NA>
Message-ID: <AANLkTi=HadnLZhc=0XwsL+LZHGkOoLKxYbtVDToeGASS@mail.gmail.com>

> But then, I am not a laywer, only an elderly Obstetrician (though, to
> quote Eric "Otter" Stratton: "What's the difference?" (Yes, I know,
> wishful thinking :-)-O)), but

The issue is that for years attorneys have been trying (some of them
with certain degree of success) to get orange juice out of rocks, take
for example the DNS that was designed as a replacement as a more
effective resolution mechanism than the old HOSTS.TXT file, which
obviously didn't have a chance to scale at all.

In the early years there was not such concept of "ownership", the
WHOIS database was very consistent and a good resource to find contact
information if there was any issue related to a particular domain.
With the dramatic growth in number of eyeballs driven by "the web" and
the good senses of IP (Intellectual Property) and trademark lawyers
sniffing huge amounts of business, plus the seasonal opportunist get
rich fast spirit, a new ecosystem was created and now the DNS became a
giant billboard where everybody is fighting to get a premier space and
a global struggle is going on  to determine who is the "owner" and has
rights to the billboard.

This ecosystem actually is today a several billions worth industry
that has its own lobbyists and that if we stop paying attention soon
will derail all efforts to sustain that a bottom-up multistakeholder
process based on cooperation and collaboration is the right model for
Internet Governance.

As the IPv4 address space gets depleted, it will not take too long for
some people to start squeezing the rocks again.

I'm not against of people making money, but we need to do a better job
to remove the "ownership" claim on Internet numbers and name
resources.

My .02
Jorge


From jack at 3kitty.org  Thu Oct 14 08:49:34 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 14 Oct 2010 08:49:34 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
Message-ID: <1287071374.2914.13.camel@localhost>

On Thu, 2010-10-14 at 10:30 -0400, Ernesto Rubi wrote:
> Someone, somewhere really ought to spend a whole semester locked in a
> room
> researching this.  If only I had graduate assistants...hehehe
> 
> 

I think this will happen when somebody decides to sue somebody else
about ownership rights.  That could trigger a whole new industry around
Internet Law, not only in the US but all the other countries on the
planet.  I'm surprised it hasn't happened already.

It's not just IP addresses.  AFAIK, issues like "Email Address
Portability" haven't come up yet.  Who "owns" ernesto at cs.fiu.edu  ...?
The person?  The University?  The "owner" of the .edu domain?

Maybe we'll all become expert witnesses, and well-paid consultants to
the lawyers, legislators, FCC et al, to tell them what the words all
mean.

Let's see, what should the retainer fee be, and how much per hour is
reasonable...?

Of course, we'll all wear flip-flops and grow long beards, to establish
credibility as Internet Experts.  I think Jon would enjoy that...

/Jack




From david at post.tau.ac.il  Thu Oct 14 09:27:16 2010
From: david at post.tau.ac.il (David Sitman)
Date: Thu, 14 Oct 2010 18:27:16 +0200 (Jerusalem Standard Time)
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1287071374.2914.13.camel@localhost>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
Message-ID: <alpine.WNT.2.00.1010141816050.3520@ccfloor209>

On Thu, 14 Oct 2010, Jack Haverty wrote:

[...]
> It's not just IP addresses.  AFAIK, issues like "Email Address
> Portability" haven't come up yet.  Who "owns" ernesto at cs.fiu.edu  ...?
> The person?  The University?  The "owner" of the .edu domain?

This summer, the Ministry of Communications in Israel began considering a 
change in ISP licensing which would require ISP's to support email address 
portability, so if I were given the address david at ISP1.co.il, I could keep 
this address when I leave ISP1 and move to a different ISP. Since the 
Israeli universities have an ISP license, this has caused us quite a bit 
of consternation.

David Sitman
Tel Aviv University


From jeanjour at comcast.net  Thu Oct 14 09:49:08 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 12:49:08 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <2d3f01ca98dd8b4c71953183a0b432d9.squirrel@webmail.cs.fiu.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>   
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>   
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>   
	<a06240834c8dcb842b520@[10.0.1.21]>
	<2d3f01ca98dd8b4c71953183a0b432d9.squirrel@webmail.cs.fiu.edu>
Message-ID: <a0624083bc8dce3c5b1ef@[10.0.1.21]>

No, and the next thing you want to look at is the language between 
Quest and ARIN or whereever they got their addresses.  They are given 
a piece of address space to manage.  Theoretically, it could be 
revoked at any time.  Of course, there would be severe disruption to 
everyone if that were done.

To continue my previous rural analogy. it might be US States that 
followed the NorthWest Ordinance, where the section grid creates an 
addressing scheme for the whole state, but counties are given 
sub-authority to assign addresses in the unincorporated areas of 
their counties.

The thing is in this case we don't think about taking our address 
with us or assigning to another plant we own 500 miles away!

At 10:04 -0400 2010/10/14, Ernesto Rubi wrote:
>I understand, it's just that when 'own' is used; it implies the same
>bundle of rights that you'd have as a real or intellectual property
>'owner.'
>
>So, routing policy aside, Qwest can't 'own' in the same fee-simple
>absolute way you 'own' your real property.
>
>For example, there are severe limitations as to what Qwest can and cannot
>do with regards to alienation of the 'address' property it purpotedly
>'owns.'
>
>
>  If anyone "owns" the addresses,
>>  it is the entity assigning them to facilitate
>>  routing in its domain.
>>
>--
>Ernesto Rubi
>Network Engineer
>FIU / AMPATH
>Email:  ernesto at cs.fiu.edu



From jeanjour at comcast.net  Thu Oct 14 10:09:05 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 13:09:05 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <alpine.WNT.2.00.1010141816050.3520@ccfloor209>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
Message-ID: <a0624083cc8dce5f835c9@[10.0.1.21]>

O, good grief.  Bad idea for basically all the same reasons we have 
been discussing, just up a level!

As I said, virtually all (there may be a few exceptions but I haven't 
found it) identifiers in CS are locators in some sense.  There is a 
reason that file names in Multics (and UNIX) were called *path*names. 
So you could find them!  Files with similar names are "near" each 
other for some concept of "near."  Here in the same i-node.

Hierarchical application names are in this sense location-dependent. 
Just a very different "location" space that used for network 
addresses.

Take care,
John

At 18:27 +0200 2010/10/14, David Sitman wrote:
>On Thu, 14 Oct 2010, Jack Haverty wrote:
>
>[...]
>>It's not just IP addresses.  AFAIK, issues like "Email Address
>>Portability" haven't come up yet.  Who "owns" ernesto at cs.fiu.edu  ...?
>>The person?  The University?  The "owner" of the .edu domain?
>
>This summer, the Ministry of Communications in Israel began 
>considering a change in ISP licensing which would require ISP's to 
>support email address portability, so if I were given the address 
>david at ISP1.co.il, I could keep this address when I leave ISP1 and 
>move to a different ISP. Since the Israeli universities have an ISP 
>license, this has caused us quite a bit of consternation.
>
>David Sitman
>Tel Aviv University



From jack at 3kitty.org  Thu Oct 14 10:52:29 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 14 Oct 2010 10:52:29 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <alpine.WNT.2.00.1010141816050.3520@ccfloor209>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
Message-ID: <1287078749.2914.48.camel@localhost>

David - thanks!  It will be interesting to see how it works out.

Actually, this topic has come up before - almost 40 years ago, when
email was in its infancy. We had lots of conversations then about email,
what an address really was, etc.  It was annoying for people as they
left with their brand new degree to discover that they also lost their
identity - their email address on the ARPANet.  I used to be
jfh at mit-dms, but I lost that when I left MIT.

If I remember those ancient conversations, the situation you describe -
david at ISP1.co.il - would be characterized as a "in-care-of" mailbox.
I.E., you don't own the mailbox, that email address means "send to
david, in care of ISP1.co.il"  This is, from a legal/regulatory
perspective, a different kind of email address than a true personal
address which you own, and which is not permanently tied to an ISP.  An
in-care-of address is only valid for a specific period of time.  The key
design point is that there are different kinds of email addresses, with
different technical and legal characteristics.

Personally, I've solved the problem (for me at least) by simply getting
my own domain - "3kitty.org", which (I think) belongs to me.  At the
time we had three cats.  So my email address is inherently portable, and
I've taken it through a series of ISPs.

So, if a university simply facilitated its users to each get their own
domain name, e.g., david.sitman.il, you could move
"david at david.sitman.il" around at will, at least within IL.  Lots of the
tech-savvy crowd pretty much do this now.  If you're fast, maybe you can
snag sitman.il instead.

Before all of you other engineers flame at me, the above was my feeble
attempt at lawyer-speak.  Of course there are a few scaling problems if
suddenly a billion new domain names are registered.  Current mechanisms
might need a little tweaking...

Back in the 90s, I was involved in meetings with the US Postal Service,
who were trying to figure out their role in our brave new Internet
world.  We tried to convince them that they should step up and manage
email addresses, just as they do for physical addresses.  But the
addresses could be people, not just buildings or roles like "Customer
Service", and technologies such as digital signatures would verify
authenticity.  You would get your email address from them, and it would
be portable.   All email would in effect be "General Delivery", and the
postal system would keep track of where you wanted it delivered as you
moved around (this is the hard part, engineering-wise, but it's no worse
than cell phones).

Never happened though.  Perhaps the IL Ministry....?

/Jack


On Thu, 2010-10-14 at 18:27 +0200, David Sitman wrote:
> On Thu, 14 Oct 2010, Jack Haverty wrote:
> 
> [...]
> > It's not just IP addresses.  AFAIK, issues like "Email Address
> > Portability" haven't come up yet.  Who "owns" ernesto at cs.fiu.edu  ...?
> > The person?  The University?  The "owner" of the .edu domain?
> 
> This summer, the Ministry of Communications in Israel began considering a 
> change in ISP licensing which would require ISP's to support email address 
> portability, so if I were given the address david at ISP1.co.il, I could keep 
> this address when I leave ISP1 and move to a different ISP. Since the 
> Israeli universities have an ISP license, this has caused us quite a bit 
> of consternation.
> 
> David Sitman
> Tel Aviv University




From jmamodio at gmail.com  Thu Oct 14 10:36:55 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Thu, 14 Oct 2010 12:36:55 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <alpine.WNT.2.00.1010141816050.3520@ccfloor209>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
Message-ID: <AANLkTinXWNbHS1RxOPfRgBsEYgSxp6X8Xebu-KLLp9-=@mail.gmail.com>

> This summer, the Ministry of Communications in Israel began considering a
> change in ISP licensing which would require ISP's to support email address
> portability, so if I were given the address david at ISP1.co.il, I could keep
> this address when I leave ISP1 and move to a different ISP. Since the
> Israeli universities have an ISP license, this has caused us quite a bit of
> consternation.

That makes perfect sense for a politician, the Minister will keep his
email address as Minister even if he is no longer the Minister ...

As John said, very bad idea.

-J


From craig at aland.bbn.com  Thu Oct 14 11:10:27 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Thu, 14 Oct 2010 14:10:27 -0400
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101014181027.59FB928E137@aland.bbn.com>

> David - thanks!  It will be interesting to see how it works out.
> 
> Actually, this topic has come up before - almost 40 years ago, when
> email was in its infancy. We had lots of conversations then about email,
> what an address really was, etc.  It was annoying for people as they
> left with their brand new degree to discover that they also lost their
> identity - their email address on the ARPANet.  I used to be
> jfh at mit-dms, but I lost that when I left MIT.

It came up again briefly in the transition to SMTP, cf. RFC 773.

> Back in the 90s, I was involved in meetings with the US Postal Service,
> who were trying to figure out their role in our brave new Internet
> world.  We tried to convince them that they should step up and manage
> email addresses, just as they do for physical addresses.  But the
> addresses could be people, not just buildings or roles like "Customer
> Service", and technologies such as digital signatures would verify
> authenticity.  You would get your email address from them, and it would
> be portable.   All email would in effect be "General Delivery", and the
> postal system would keep track of where you wanted it delivered as you
> moved around (this is the hard part, engineering-wise, but it's no worse
> than cell phones).

This is actually a service that many alumni associations and professional
societies offer -- they relay mail through to your current address.

Thanks!

Craig


From jcurran at arin.net  Thu Oct 14 11:55:18 2010
From: jcurran at arin.net (John Curran)
Date: Thu, 14 Oct 2010 14:55:18 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1287071374.2914.13.camel@localhost>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
Message-ID: <6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>

On Oct 14, 2010, at 11:49 AM, Jack Haverty wrote:
> On Thu, 2010-10-14 at 10:30 -0400, Ernesto Rubi wrote:
>> Someone, somewhere really ought to spend a whole semester locked in a
>> room researching this.  If only I had graduate assistants...hehehe
> 
> I think this will happen when somebody decides to sue somebody else
> about ownership rights.  That could trigger a whole new industry around
> Internet Law, not only in the US but all the other countries on the
> planet.  I'm surprised it hasn't happened already.

It's already occurred, with the result being a District Judge reconsidered their
prior order regarding to IP addresses and instead affirming that IP addresses are 
"administered in a public trust".   Relevant law article here: 
http://www.chtlj.org/sites/default/files/media/articles/v024/v024.i2.Ryan.pdf

I provided this information to Ernesto, but apparently it made more sense 
to continue with the specious property thread rather than anchor it in any 
actual jurisprudence.

/John

John Curran
President and CEO
ARIN




From AMaitland at Commerco.Com  Thu Oct 14 13:10:26 2010
From: AMaitland at Commerco.Com (Alan J Maitland)
Date: Thu, 14 Oct 2010 14:10:26 -0600
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
Message-ID: <6.2.3.4.2.20101014134318.03fb0090@MS1.MailSys.Net>

Mr. Curran,

Thank you for the PDF link you listed in your attached, it was a good 
read (and was also wonderfully cited).  It seems clear that the IPV4 
address space is hugely challenged as regards its availability for 
allocations (and I gather also for reallocation), but for IPV6, is 
there a chance to create the markets to which your document 
refers?  In doing so, perhaps one might solve a number of issues the 
document raises.

For example, one might coax some of the holdouts on what appears to 
be holders of grandfathered IPV4 space into trading that IPV4 space 
off for some IPV6 space.

Though the point made regarding IPV6 as being finite is clearly true, 
IPV6 still represents a huge increase in the available address space 
to be allocated for and to the Internet community.  Given its size, 
perhaps it could be divided up into grouped allocations as a 
compromise to all parties involved.  For example, something like 
"reserved space", "allocated space" and "free market space", made 
available to allocation as groups such as ARIN, the RIRs and 
governing political bodies see fit or as otherwise makes sense.

Echoing a point expressed earlier in this thread, perhaps a strong 
focus and push toward moving IPV6 migration and implementation 
forward would seemingly diminish the pain associated with the ever 
shrinking IPV4 space.

Alan Maitland
The Commerce Company

At 12:55 PM 10/14/2010, John Curran wrote:
>On Oct 14, 2010, at 11:49 AM, Jack Haverty wrote:
> > On Thu, 2010-10-14 at 10:30 -0400, Ernesto Rubi wrote:
> >> Someone, somewhere really ought to spend a whole semester locked in a
> >> room researching this.  If only I had graduate assistants...hehehe
> >
> > I think this will happen when somebody decides to sue somebody else
> > about ownership rights.  That could trigger a whole new industry around
> > Internet Law, not only in the US but all the other countries on the
> > planet.  I'm surprised it hasn't happened already.
>
>It's already occurred, with the result being a District Judge 
>reconsidered their
>prior order regarding to IP addresses and instead affirming that IP 
>addresses are
>"administered in a public trust".   Relevant law article here:
>http://www.chtlj.org/sites/default/files/media/articles/v024/v024.i2.Ryan.pdf
>
>I provided this information to Ernesto, but apparently it made more sense
>to continue with the specious property thread rather than anchor it in any
>actual jurisprudence.
>
>/John
>
>John Curran
>President and CEO
>ARIN




From jcurran at arin.net  Thu Oct 14 13:31:25 2010
From: jcurran at arin.net (John Curran)
Date: Thu, 14 Oct 2010 16:31:25 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTins4zmFAQRqBbbMzF+o7oib=oZvbhftA4Lv0smU@mail.gmail.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<a06240834c8dcb842b520@10.0.1.21>
	<2d3f01ca98dd8b4c71953183a0b432d9.squirrel@webmail.cs.fiu.edu>
	<AANLkTins4zmFAQRqBbbMzF+o7oib=oZvbhftA4Lv0smU@mail.gmail.com>
Message-ID: <80F9FB4C-4AC7-4BE2-944D-BE0C4D9B047B@arin.net>

On Oct 14, 2010, at 11:09 AM, John Klensin wrote:

> A different way of looking at all of this is that the entire
> addressing allocation/ delegation/ registration model ultimately rests
> on a combination of collective consent and an understanding that
> having non-unique addresses floating around causes problems in routing
> packets. 

Collective consent is very important, and one could argue more so than any 
particular legal theory or contract clause.  This is why ARIN administers 
the Whois database according to the community developed address policy; it 
is what the ISPs want and expect in keeping their networks running; it is 
what is necessary for ARIN to be true to the mission for which we were 
established.

At present, ARIN encourages parties to return unused address space, and has a
community-developed specified transfer policy which lets one party which has 
qualified for more space get it from second party (potentially by providing 
financial incentive for the second to renumber if necessary)  This encourages 
return of little used space into the system and higher overall utilization, 
while still providing stewardship per the guidelines in RFC2008 and RFC2050.
There are some folks out there who object to this approach on principle (i.e.
it's not a wide-open free market), but it is what the community adopted.

Some likely object because may prevent them from maximizing their IP address 
monetization game and pursuing the ip-addresses-are-property angle might 
give higher yields.  The actual community direction and possible operational 
impacts aren't even a consideration for that crowd; they simply have to argue
that property rights should apply to address assignments because it is good
business to do so.  This has been expected for years and will be handled in 
due course.

/John

John Curran
President and CEO
ARIN





From jnc at mercury.lcs.mit.edu  Thu Oct 14 13:55:55 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Thu, 14 Oct 2010 16:55:55 -0400 (EDT)
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101014205555.9C0276BE5AA@mercury.lcs.mit.edu>

    > From: Alan J Maitland <AMaitland at Commerco.Com>

    > for IPV6, is there a chance to create the markets to which your
    > document refers?

There's simply no need for one - there's supposed to be more than enough for
decades (centuries) to come. (And if not, WTF are switching to it anyway?)

    > one might coax some of the holdouts on what appears to be holders of
    > grandfathered IPV4 space into trading that IPV4 space off for some IPV6
    > space.

Let's see, trade something very valuable (and useful) for something of
basically no value (supply >>>> demand -> extremely low value)? No, I don't
think so.

	Noel


From jeanjour at comcast.net  Thu Oct 14 14:14:17 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 17:14:17 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <6.2.3.4.2.20101014134318.03fb0090@MS1.MailSys.Net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<6.2.3.4.2.20101014134318.03fb0090@MS1.MailSys.Net>
Message-ID: <a06240840c8dd222b2950@[10.0.1.21]>

>
>
>Though the point made regarding IPV6 as being finite is clearly 
>true, IPV6 still represents a huge increase in the available address 
>space to be allocated for and to the Internet community.  Given its 
>size, perhaps it could be divided up into grouped allocations as a 
>compromise to all parties involved.  For example, something like 
>"reserved space", "allocated space" and "free market space", made 
>available to allocation as groups such as ARIN, the RIRs and 
>governing political bodies see fit or as otherwise makes sense.

How many ways do we have to say it?  There is no such thing as a free 
market in IP addresses.  IP addresses are assigned to facilitate 
routing.  Do anything else will severely break the Internet 
regardless of how big the address is. Actually the bigger it is 
faster it gets broken.

Or perhaps I should say it the other way around.  Please create a 
free market in IPv6 addresses as soon as possible so that we get to 
see how many ways the Internet can suffer a meltdown.  This will be 
interesting to watch.

Take care,
John


From jcurran at arin.net  Thu Oct 14 14:23:57 2010
From: jcurran at arin.net (John Curran)
Date: Thu, 14 Oct 2010 17:23:57 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <6.2.3.4.2.20101014134318.03fb0090@MS1.MailSys.Net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<6.2.3.4.2.20101014134318.03fb0090@MS1.MailSys.Net>
Message-ID: <E3C56E48-925C-4DB2-9DB4-E17FD9CF449E@arin.net>

On Oct 14, 2010, at 4:10 PM, Alan J Maitland wrote:

> Mr. Curran,
> 
> Thank you for the PDF link you listed in your attached, it was a good 
> read (and was also wonderfully cited).  It seems clear that the IPV4 
> address space is hugely challenged as regards its availability for 
> allocations (and I gather also for reallocation), but for IPV6, is 
> there a chance to create the markets to which your document 
> refers?  In doing so, perhaps one might solve a number of issues the 
> document raises.
> 
> For example, one might coax some of the holdouts on what appears to 
> be holders of grandfathered IPV4 space into trading that IPV4 space 
> off for some IPV6 space.

We currently have very wide open IPv6 policies right now, to the extent
that nearly any holder of IPv4 space can get an IPv6 allocation asap.
Also, as described, we have the specified transfer policy, which does 
result in parties being incentivized to free up little-used address 
space and put back into the system. (Reference the ARIN website
at https://www.arin.net/policy/nrpm.html to see these policies.)
I'm not saying the additional policies aren't needed or welcome;
just bringing the existing ones to your attention.  You can find
information on how to propose new policies in that same section
of the website.

> Though the point made regarding IPV6 as being finite is clearly true, 
> IPV6 still represents a huge increase in the available address space 
> to be allocated for and to the Internet community.  Given its size, 
> perhaps it could be divided up into grouped allocations as a 
> compromise to all parties involved.  For example, something like 
> "reserved space", "allocated space" and "free market space", made 
> available to allocation as groups such as ARIN, the RIRs and 
> governing political bodies see fit or as otherwise makes sense.

Yes, and no - see below.

> Echoing a point expressed earlier in this thread, perhaps a strong 
> focus and push toward moving IPV6 migration and implementation 
> forward would seemingly diminish the pain associated with the ever 
> shrinking IPV4 space.

The move to IPv6 will definitely help reduce this pain, and is 
something that all of the RIRs, ICANN, and ISOC are aggressively
trying to encourage.  The real issue that we'll run into with IPv6 
is the opposite problem: an abundance of addresses so large that 
having an assignment becomes nearly meaningless compared to the 
value of getting it routed.  The real value of the Regional Internet
Registries in this environment is not purely in issuing addresses
but in having the community forums which allow open discussion 
of what is acceptable for new allocations (and therefore entitled
to be uniquely routed globally).  Wrote a paper about that as well, 
when the idea that registries being somehow valuable independent 
of their policy framework came up with the ITU promoted theory of 
Country Internet Registries (CIRs) -
<http://www.scribd.com/doc/29487289/ARIN-Contribution-to-ITU-T-IPv6-Study-Group>

For those who really love market approaches, all you really need is a market 
for obtaining global routing table slots for IPv6 prefixes and then we quite
literally could have hundreds of registries of all shapes and sizes (maybe
even including vending machines, as a wise man once suggested). The cost of 
obtain unique numbers would be appropriately low given it's relative abundance 
with IPv6 (now everyone can be a registry! :-)  

Folks who used addresses under their ISP's prefix would incur no additional 
cost, and those that wanted "portability" would pay routing their unique prefix 
in the global routing table ($$)  Establishing such a global market for routing 
table slots is left as an exercise to the reader, and we have now fully left 
behind any and all discussion of "Internet History" (and should perhaps move 
the discussion elsewhere... :-)

/John

John Curran
President and CEO
ARIN




From ernesto at cs.fiu.edu  Thu Oct 14 16:53:49 2010
From: ernesto at cs.fiu.edu (Ernie Rubi)
Date: Thu, 14 Oct 2010 19:53:49 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
Message-ID: <4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>

I'm not sure I read that decision as broadly as you do; in fact, judicial districts are separate entities.

Even if it was to be read broadly the issue could be decided totally differently in other jurisdictions.

The law review article itself recognizes that case law is scarce.

I'm not looking for any particular angle, I'd just like to explore all sides of the debate.  My paper would likely be 3 pages long if I took that USDC decision as binding precedent not only here in the US, but internationally.


On Oct 14, 2010, at 2:55 PM, John Curran <jcurran at arin.net> wrote:

> On Oct 14, 2010, at 11:49 AM, Jack Haverty wrote:
>> On Thu, 2010-10-14 at 10:30 -0400, Ernesto Rubi wrote:
>>> Someone, somewhere really ought to spend a whole semester locked in a
>>> room researching this.  If only I had graduate assistants...hehehe
>> 
>> I think this will happen when somebody decides to sue somebody else
>> about ownership rights.  That could trigger a whole new industry around
>> Internet Law, not only in the US but all the other countries on the
>> planet.  I'm surprised it hasn't happened already.
> 
> It's already occurred, with the result being a District Judge reconsidered their
> prior order regarding to IP addresses and instead affirming that IP addresses are 
> "administered in a public trust".   Relevant law article here: 
> http://www.chtlj.org/sites/default/files/media/articles/v024/v024.i2.Ryan.pdf
> 
> I provided this information to Ernesto, but apparently it made more sense 
> to continue with the specious property thread rather than anchor it in any 
> actual jurisprudence.
> 
> /John
> 
> John Curran
> President and CEO
> ARIN



From jeanjour at comcast.net  Thu Oct 14 17:33:52 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 20:33:52 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
Message-ID: <a06240846c8dd4ff6e4d0@[10.0.1.21]>

As people have tried to explain.  There is no room for debate or any 
legal input on this.  IP addresses must be assigned to facilitate 
routing.  If they are not, there is no Internet.

Yes, there are poor misguided souls who do not understand how 
networks including the Internet work who misconstrue the nature of 
addresses and think they can apply concepts from the law.  They are 
simply foolish.  Some of these misguided souls can be very loud and 
confuse the issues in some very public forums.  This doesn't change 
anything.

These forays are in the same category as a state legislature passing 
a law to make pi equal 22/7.  No matter how irrationally they behave, 
pi will remain irrational.

Perhaps, you should expend your efforts explaining to the legal 
profession why these concepts do not apply to addresses.  That could 
be quite helpful.

Take care,
John

At 19:53 -0400 2010/10/14, Ernie Rubi wrote:
>I'm not sure I read that decision as broadly as you do; in fact, 
>judicial districts are separate entities.
>
>Even if it was to be read broadly the issue could be decided totally 
>differently in other jurisdictions.
>
>The law review article itself recognizes that case law is scarce.
>
>I'm not looking for any particular angle, I'd just like to explore 
>all sides of the debate.  My paper would likely be 3 pages long if I 
>took that USDC decision as binding precedent not only here in the 
>US, but internationally.
>
>
>On Oct 14, 2010, at 2:55 PM, John Curran <jcurran at arin.net> wrote:
>
>>  On Oct 14, 2010, at 11:49 AM, Jack Haverty wrote:
>>>  On Thu, 2010-10-14 at 10:30 -0400, Ernesto Rubi wrote:
>>>>  Someone, somewhere really ought to spend a whole semester locked in a
>>>>  room researching this.  If only I had graduate assistants...hehehe
>>>
>>>  I think this will happen when somebody decides to sue somebody else
>>>  about ownership rights.  That could trigger a whole new industry around
>>>  Internet Law, not only in the US but all the other countries on the
>>>  planet.  I'm surprised it hasn't happened already.
>>
>>  It's already occurred, with the result being a District Judge 
>>reconsidered their
>>  prior order regarding to IP addresses and instead affirming that 
>>IP addresses are
>>  "administered in a public trust".   Relevant law article here:
>> 
>>http://www.chtlj.org/sites/default/files/media/articles/v024/v024.i2.Ryan.pdf
>>
>>  I provided this information to Ernesto, but apparently it made more sense
>>  to continue with the specious property thread rather than anchor it in any
>>  actual jurisprudence.
>>
>>  /John
>>
>>  John Curran
>>  President and CEO
>>  ARIN



From richard at bennett.com  Thu Oct 14 18:12:57 2010
From: richard at bennett.com (Richard Bennett)
Date: Thu, 14 Oct 2010 18:12:57 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240846c8dd4ff6e4d0@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
Message-ID: <4CB7AA99.7000505@bennett.com>

  The trouble is, John, that any system that's widely used, as the 
Internet is, requires a legal framework for resolving disputes over 
terms of use, contracts, and ownership, so simply running off the 
lawyers isn't really an option. The technology defines some boundaries 
around the work of the lawyers, but it doesn't make them irrelevant. The 
Internet's global system of addressing requires a central authority to 
ensure address and route uniqueness; that's something we have to live 
with. Law is therefore necessary to ensure that the behavior of the 
authority and its designees is consistent with public policy.

RB

On 10/14/2010 5:33 PM, John Day wrote:
> As people have tried to explain.  There is no room for debate or any 
> legal input on this.  IP addresses must be assigned to facilitate 
> routing.  If they are not, there is no Internet.
>
> Yes, there are poor misguided souls who do not understand how networks 
> including the Internet work who misconstrue the nature of addresses 
> and think they can apply concepts from the law.  They are simply 
> foolish.  Some of these misguided souls can be very loud and confuse 
> the issues in some very public forums.  This doesn't change anything.
>
> These forays are in the same category as a state legislature passing a 
> law to make pi equal 22/7.  No matter how irrationally they behave, pi 
> will remain irrational.
>
> Perhaps, you should expend your efforts explaining to the legal 
> profession why these concepts do not apply to addresses.  That could 
> be quite helpful.
>
> Take care,
> John
>
> At 19:53 -0400 2010/10/14, Ernie Rubi wrote:
>> I'm not sure I read that decision as broadly as you do; in fact, 
>> judicial districts are separate entities.
>>
>> Even if it was to be read broadly the issue could be decided totally 
>> differently in other jurisdictions.
>>
>> The law review article itself recognizes that case law is scarce.
>>
>> I'm not looking for any particular angle, I'd just like to explore 
>> all sides of the debate.  My paper would likely be 3 pages long if I 
>> took that USDC decision as binding precedent not only here in the US, 
>> but internationally.
>>
>>
>> On Oct 14, 2010, at 2:55 PM, John Curran <jcurran at arin.net> wrote:
>>
>>>  On Oct 14, 2010, at 11:49 AM, Jack Haverty wrote:
>>>>  On Thu, 2010-10-14 at 10:30 -0400, Ernesto Rubi wrote:
>>>>>  Someone, somewhere really ought to spend a whole semester locked 
>>>>> in a
>>>>>  room researching this.  If only I had graduate assistants...hehehe
>>>>
>>>>  I think this will happen when somebody decides to sue somebody else
>>>>  about ownership rights.  That could trigger a whole new industry 
>>>> around
>>>>  Internet Law, not only in the US but all the other countries on the
>>>>  planet.  I'm surprised it hasn't happened already.
>>>
>>>  It's already occurred, with the result being a District Judge 
>>> reconsidered their
>>>  prior order regarding to IP addresses and instead affirming that IP 
>>> addresses are
>>>  "administered in a public trust".   Relevant law article here:
>>>
>>> http://www.chtlj.org/sites/default/files/media/articles/v024/v024.i2.Ryan.pdf 
>>>
>>>
>>>  I provided this information to Ernesto, but apparently it made more 
>>> sense
>>>  to continue with the specious property thread rather than anchor it 
>>> in any
>>>  actual jurisprudence.
>>>
>>>  /John
>>>
>>>  John Curran
>>>  President and CEO
>>>  ARIN
>

-- 
Richard Bennett



From jack at 3kitty.org  Thu Oct 14 18:53:19 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 14 Oct 2010 18:53:19 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240846c8dd4ff6e4d0@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
Message-ID: <1287107599.2914.115.camel@localhost>

On Thu, 2010-10-14 at 20:33 -0400, John Day wrote:
> IP addresses must be assigned to facilitate 
> routing.  If they are not, there is no Internet. 

True, with today's implementation.

Getting back to history...   IIRC, the Telephone Companies used the same
argument about telephone numbers - they had a structure which was
meaningful to the switching system, and was needed to establish the
circuit through the various COs to end at a particular wire inside your
house.

Somewhere between then and now, Number Portability happened despite
their objections, cell phones were attached to everyone's ear, and, at
least for some numbers, I can take my number and phone with me around
the planet and from company to company.  Telephone Companies had to
reinvent the innards of their networks, so that numbers were no longer
tied to addresses and/or routes.  And the Telephone System still works.
There's some pretty strong precedents in that evolution.

Seems like The Internet could make a similar evolutionary step if it was
needed....or mandated by government or customer pressures, as happened
to the Telephone networks.

/Jack






From jack at 3kitty.org  Thu Oct 14 18:53:19 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 14 Oct 2010 18:53:19 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240846c8dd4ff6e4d0@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
Message-ID: <1287107599.2914.115.camel@localhost>

On Thu, 2010-10-14 at 20:33 -0400, John Day wrote:
> IP addresses must be assigned to facilitate 
> routing.  If they are not, there is no Internet. 

True, with today's implementation.

Getting back to history...   IIRC, the Telephone Companies used the same
argument about telephone numbers - they had a structure which was
meaningful to the switching system, and was needed to establish the
circuit through the various COs to end at a particular wire inside your
house.

Somewhere between then and now, Number Portability happened despite
their objections, cell phones were attached to everyone's ear, and, at
least for some numbers, I can take my number and phone with me around
the planet and from company to company.  Telephone Companies had to
reinvent the innards of their networks, so that numbers were no longer
tied to addresses and/or routes.  And the Telephone System still works.
There's some pretty strong precedents in that evolution.

Seems like The Internet could make a similar evolutionary step if it was
needed....or mandated by government or customer pressures, as happened
to the Telephone networks.

/Jack






From jack at 3kitty.org  Thu Oct 14 18:53:19 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 14 Oct 2010 18:53:19 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240846c8dd4ff6e4d0@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
Message-ID: <1287107599.2914.115.camel@localhost>

On Thu, 2010-10-14 at 20:33 -0400, John Day wrote:
> IP addresses must be assigned to facilitate 
> routing.  If they are not, there is no Internet. 

True, with today's implementation.

Getting back to history...   IIRC, the Telephone Companies used the same
argument about telephone numbers - they had a structure which was
meaningful to the switching system, and was needed to establish the
circuit through the various COs to end at a particular wire inside your
house.

Somewhere between then and now, Number Portability happened despite
their objections, cell phones were attached to everyone's ear, and, at
least for some numbers, I can take my number and phone with me around
the planet and from company to company.  Telephone Companies had to
reinvent the innards of their networks, so that numbers were no longer
tied to addresses and/or routes.  And the Telephone System still works.
There's some pretty strong precedents in that evolution.

Seems like The Internet could make a similar evolutionary step if it was
needed....or mandated by government or customer pressures, as happened
to the Telephone networks.

/Jack






From jack at 3kitty.org  Thu Oct 14 18:53:19 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 14 Oct 2010 18:53:19 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240846c8dd4ff6e4d0@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
Message-ID: <1287107599.2914.115.camel@localhost>

On Thu, 2010-10-14 at 20:33 -0400, John Day wrote:
> IP addresses must be assigned to facilitate 
> routing.  If they are not, there is no Internet. 

True, with today's implementation.

Getting back to history...   IIRC, the Telephone Companies used the same
argument about telephone numbers - they had a structure which was
meaningful to the switching system, and was needed to establish the
circuit through the various COs to end at a particular wire inside your
house.

Somewhere between then and now, Number Portability happened despite
their objections, cell phones were attached to everyone's ear, and, at
least for some numbers, I can take my number and phone with me around
the planet and from company to company.  Telephone Companies had to
reinvent the innards of their networks, so that numbers were no longer
tied to addresses and/or routes.  And the Telephone System still works.
There's some pretty strong precedents in that evolution.

Seems like The Internet could make a similar evolutionary step if it was
needed....or mandated by government or customer pressures, as happened
to the Telephone networks.

/Jack






From jack at 3kitty.org  Thu Oct 14 18:53:19 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 14 Oct 2010 18:53:19 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240846c8dd4ff6e4d0@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
Message-ID: <1287107599.2914.115.camel@localhost>

On Thu, 2010-10-14 at 20:33 -0400, John Day wrote:
> IP addresses must be assigned to facilitate 
> routing.  If they are not, there is no Internet. 

True, with today's implementation.

Getting back to history...   IIRC, the Telephone Companies used the same
argument about telephone numbers - they had a structure which was
meaningful to the switching system, and was needed to establish the
circuit through the various COs to end at a particular wire inside your
house.

Somewhere between then and now, Number Portability happened despite
their objections, cell phones were attached to everyone's ear, and, at
least for some numbers, I can take my number and phone with me around
the planet and from company to company.  Telephone Companies had to
reinvent the innards of their networks, so that numbers were no longer
tied to addresses and/or routes.  And the Telephone System still works.
There's some pretty strong precedents in that evolution.

Seems like The Internet could make a similar evolutionary step if it was
needed....or mandated by government or customer pressures, as happened
to the Telephone networks.

/Jack






From jeanjour at comcast.net  Thu Oct 14 18:48:54 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 21:48:54 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7AA99.7000505@bennett.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]> <4CB7AA99.7000505@bennett.com>
Message-ID: <a0624084bc8dd629f445b@[10.0.1.21]>

Correct. Totally agree.  There needs to be a legal framework between 
the central authority and those it delegates to administer 
sub-domains of the address space,  In this case where we have 
provider based addresses, the providers.  (This is like my earlier 
analogy of the state delegating to the counties to address their 
roads.)  However, the customers of the providers do not enter into 
this.

This is what Curran has been pointing to as I understand it.

At 18:12 -0700 2010/10/14, Richard Bennett wrote:
>  The trouble is, John, that any system that's widely used, as the 
>Internet is, requires a legal framework for resolving disputes over 
>terms of use, contracts, and ownership, so simply running off the 
>lawyers isn't really an option. The technology defines some 
>boundaries around the work of the lawyers, but it doesn't make them 
>irrelevant. The Internet's global system of addressing requires a 
>central authority to ensure address and route uniqueness; that's 
>something we have to live with. Law is therefore necessary to ensure 
>that the behavior of the authority and its designees is consistent 
>with public policy.
>
>RB
>
>On 10/14/2010 5:33 PM, John Day wrote:
>>As people have tried to explain.  There is no room for debate or 
>>any legal input on this.  IP addresses must be assigned to 
>>facilitate routing.  If they are not, there is no Internet.
>>
>>Yes, there are poor misguided souls who do not understand how 
>>networks including the Internet work who misconstrue the nature of 
>>addresses and think they can apply concepts from the law.  They are 
>>simply foolish.  Some of these misguided souls can be very loud and 
>>confuse the issues in some very public forums.  This doesn't change 
>>anything.
>>
>>These forays are in the same category as a state legislature 
>>passing a law to make pi equal 22/7.  No matter how irrationally 
>>they behave, pi will remain irrational.
>>
>>Perhaps, you should expend your efforts explaining to the legal 
>>profession why these concepts do not apply to addresses.  That 
>>could be quite helpful.
>>
>>Take care,
>>John
>>
>>At 19:53 -0400 2010/10/14, Ernie Rubi wrote:
>>>I'm not sure I read that decision as broadly as you do; in fact, 
>>>judicial districts are separate entities.
>>>
>>>Even if it was to be read broadly the issue could be decided 
>>>totally differently in other jurisdictions.
>>>
>>>The law review article itself recognizes that case law is scarce.
>>>
>>>I'm not looking for any particular angle, I'd just like to explore 
>>>all sides of the debate.  My paper would likely be 3 pages long if 
>>>I took that USDC decision as binding precedent not only here in 
>>>the US, but internationally.
>>>
>>>
>>>On Oct 14, 2010, at 2:55 PM, John Curran <jcurran at arin.net> wrote:
>>>
>>>>  On Oct 14, 2010, at 11:49 AM, Jack Haverty wrote:
>>>>>  On Thu, 2010-10-14 at 10:30 -0400, Ernesto Rubi wrote:
>>>>>>  Someone, somewhere really ought to spend a whole semester locked in a
>>>>>>  room researching this.  If only I had graduate assistants...hehehe
>>>>>
>>>>>  I think this will happen when somebody decides to sue somebody else
>>>>>  about ownership rights.  That could trigger a whole new industry around
>>>>>  Internet Law, not only in the US but all the other countries on the
>>>>>  planet.  I'm surprised it hasn't happened already.
>>>>
>>>>  It's already occurred, with the result being a District Judge 
>>>>reconsidered their
>>>>  prior order regarding to IP addresses and instead affirming that 
>>>>IP addresses are
>>>>  "administered in a public trust".   Relevant law article here:
>>>>
>>>>http://www.chtlj.org/sites/default/files/media/articles/v024/v024.i2.Ryan.pdf
>>>>
>>>>  I provided this information to Ernesto, but apparently it made more sense
>>>>  to continue with the specious property thread rather than anchor it in any
>>>>  actual jurisprudence.
>>>>
>>>>  /John
>>>>
>>>>  John Curran
>>>>  President and CEO
>>>>  ARIN
>>
>
>--
>Richard Bennett



From jeanjour at comcast.net  Thu Oct 14 18:59:20 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 21:59:20 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7AF93.2010708@cox.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]> <4CB7AC90.2030600@cox.net>
	<a06240848c8dd5dd023e6@[10.0.1.21]> <4CB7AF93.2010708@cox.net>
Message-ID: <a0624084cc8dd638a7b5a@[10.0.1.21]>

>On 10/14/2010 8:24 PM, John Day wrote:
>>Sorry, what was true of telephone numbers?

Snip

>>>We used to think that was true of telephone numbers.
>
>("these concepts do not apply to addresses")

We already went though this. Although perhaps too indirectly.

The cell phone system changed phone numbers from being 
physical-connection-endpoint-addresses into application names and put 
other addressing schemes underneath the phone numbers so tehy could 
route to them and didn't tell anyone.  It became so useful, that it 
was decided to make all telephone numbers application names.

So what would you like to do?  Make IPv6 addresses application names? 
Then what would be put under them to be addresses so we could route 
to them?  MAC addresses?   That would be really neat.  Then we could 
route based on the manufacturer who built the interface rather than 
the provider where the interface was.  That ought to work really well!

Take care,
John


From jmamodio at gmail.com  Thu Oct 14 19:10:54 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Thu, 14 Oct 2010 21:10:54 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7AA99.7000505@bennett.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@10.0.1.21> <4CB7AA99.7000505@bennett.com>
Message-ID: <AANLkTikoU8SiPTfm1dbaYzPAd8xoWnTwSytDF1hwBDMP@mail.gmail.com>

> lawyers, but it doesn't make them irrelevant. The Internet's global system
> of addressing requires a central authority to ensure address and route
> uniqueness; that's something we have to live with.

Addressing yes, routing no. With unique routes we'll all be screwed
and back to X.25

Cheers
- J


From larrysheldon at cox.net  Thu Oct 14 19:28:19 2010
From: larrysheldon at cox.net (Larry Sheldon)
Date: Thu, 14 Oct 2010 21:28:19 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a0624084cc8dd638a7b5a@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]> <4CB7AC90.2030600@cox.net>
	<a06240848c8dd5dd023e6@[10.0.1.21]> <4CB7AF93.2010708@cox.net>
	<a0624084cc8dd638a7b5a@[10.0.1.21]>
Message-ID: <4CB7BC43.1030109@cox.net>

[Copied to the list]

On 10/14/2010 8:59 PM, John Day wrote:
>> On 10/14/2010 8:24 PM, John Day wrote:
>>> Sorry, what was true of telephone numbers?
>
> Snip
>
>>>> We used to think that was true of telephone numbers.
>>
>> ("these concepts do not apply to addresses")
>
> We already went though this. Although perhaps too indirectly.
>
> The cell phone system changed phone numbers from being
> physical-connection-endpoint-addresses into application names and put
> other addressing schemes underneath the phone numbers so tehy could
> route to them and didn't tell anyone. It became so useful, that it was
> decided to make all telephone numbers application names.
>
> So what would you like to do? Make IPv6 addresses application names?
> Then what would be put under them to be addresses so we could route to
> them? MAC addresses? That would be really neat. Then we could route
> based on the manufacturer who built the interface rather than the
> provider where the interface was. That ought to work really well!

I don't know how to do it.

When I was a Pacific Telephone employee long ago, we didn't know how to 
do it with telephone numbers either.

Once upon a time we said iot was impossible to fly.  Then to fly across 
oceans.  Then across space.

Now we are just down to we don't want to.

But if we did, we could.  Somehow.


From jeanjour at comcast.net  Thu Oct 14 19:44:41 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 22:44:41 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7BBE7.2030502@cox.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]> <4CB7AC90.2030600@cox.net>
	<a06240848c8dd5dd023e6@[10.0.1.21]> <4CB7AF93.2010708@cox.net>
	<a0624084cc8dd638a7b5a@[10.0.1.21]> <4CB7BBE7.2030502@cox.net>
Message-ID: <a0624084fc8dd6e5f057d@[10.0.1.21]>

We explained how to do it.

You jack up the phone number space and make them application names. 
But you let the users think nothing has changed.

Then you slide another address space under it that looks just like 
phone numbers but are used for routing.  You maintain a mapping from 
the first assignments to these numbers, so that *what* you want to 
get to is distinct from "where" it is.

In the case of cell phones there is one more set of mappings to the 
addresses of the cell towers.

The point is IP addresses aren't really what you want for application 
names.  If you start using them as such which is what "number 
portability" implies then you need some thing else and you are into 
an  infinite regress!  Something has to tell you where the whats are! 
;-)

As far as your contention that at one time flying seemed impossible. 
This isn't in that category.

This is in the category that once it appeared that perpetual motion 
was impossible. Then they discovered the Second Law of Thermodynamics 
and knew that it really was impossible.

Take care,
John

At 21:26 -0500 2010/10/14, Larry Sheldon wrote:
>On 10/14/2010 8:59 PM, John Day wrote:
>>>On 10/14/2010 8:24 PM, John Day wrote:
>>>>Sorry, what was true of telephone numbers?
>>
>>Snip
>>
>>>>>We used to think that was true of telephone numbers.
>>>
>>>("these concepts do not apply to addresses")
>>
>>We already went though this. Although perhaps too indirectly.
>>
>>The cell phone system changed phone numbers from being
>>physical-connection-endpoint-addresses into application names and put
>>other addressing schemes underneath the phone numbers so tehy could
>>route to them and didn't tell anyone. It became so useful, that it was
>>decided to make all telephone numbers application names.
>>
>>So what would you like to do? Make IPv6 addresses application names?
>>Then what would be put under them to be addresses so we could route to
>>them? MAC addresses? That would be really neat. Then we could route
>>based on the manufacturer who built the interface rather than the
>>provider where the interface was. That ought to work really well!
>
>I don't know how to do it.
>
>When I was a Pacific Telephone employee long ago, we didn't know how 
>to do it with telephone numbers either.
>
>Once upon a time we said iot was impossible to fly.  Then to fly 
>across oceans.  Then across space.
>
>Now we are just down to we don't want to.
>
>But if we did, we could.  Somehow.



From richard at bennett.com  Thu Oct 14 19:51:31 2010
From: richard at bennett.com (Richard Bennett)
Date: Thu, 14 Oct 2010 19:51:31 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTikoU8SiPTfm1dbaYzPAd8xoWnTwSytDF1hwBDMP@mail.gmail.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>	<a06240846c8dd4ff6e4d0@10.0.1.21>	<4CB7AA99.7000505@bennett.com>
	<AANLkTikoU8SiPTfm1dbaYzPAd8xoWnTwSytDF1hwBDMP@mail.gmail.com>
Message-ID: <4CB7C1B3.9010906@bennett.com>

  Well, yeah, X.25 or something like MPLS. My point was that I don't 
want the route from Comcast.net to Bennett.com to go through China. Not 
so much unique routes as user-controlled ones.

On 10/14/2010 7:10 PM, Jorge Amodio wrote:
>> lawyers, but it doesn't make them irrelevant. The Internet's global system
>> of addressing requires a central authority to ensure address and route
>> uniqueness; that's something we have to live with.
> Addressing yes, routing no. With unique routes we'll all be screwed
> and back to X.25
>
> Cheers
> - J

-- 
Richard Bennett



From richard at bennett.com  Thu Oct 14 19:55:29 2010
From: richard at bennett.com (Richard Bennett)
Date: Thu, 14 Oct 2010 19:55:29 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a0624084cc8dd638a7b5a@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
	<4CB7AC90.2030600@cox.net>	<a06240848c8dd5dd023e6@[10.0.1.21]>
	<4CB7AF93.2010708@cox.net> <a0624084cc8dd638a7b5a@[10.0.1.21]>
Message-ID: <4CB7C2A1.80109@bennett.com>

  Right, the phone number analogy doesn't wash because it's an 
identifier and not a device address. The phone company maps it to the 
functional equivalent of a MAC address and then locates it. I don't even 
know what the current IP address is for Bennett.com, and nobody needs to 
know it but DNS.

Embedding MAC addresses inside IPv6 addresses was one of the dumbest 
ideas in the history of networking.

RB

On 10/14/2010 6:59 PM, John Day wrote:
>> On 10/14/2010 8:24 PM, John Day wrote:
>>> Sorry, what was true of telephone numbers?
>
> Snip
>
>>>> We used to think that was true of telephone numbers.
>>
>> ("these concepts do not apply to addresses")
>
> We already went though this. Although perhaps too indirectly.
>
> The cell phone system changed phone numbers from being 
> physical-connection-endpoint-addresses into application names and put 
> other addressing schemes underneath the phone numbers so tehy could 
> route to them and didn't tell anyone.  It became so useful, that it 
> was decided to make all telephone numbers application names.
>
> So what would you like to do?  Make IPv6 addresses application names? 
> Then what would be put under them to be addresses so we could route to 
> them?  MAC addresses?   That would be really neat.  Then we could 
> route based on the manufacturer who built the interface rather than 
> the provider where the interface was.  That ought to work really well!
>
> Take care,
> John

-- 
Richard Bennett



From dhc2 at dcrocker.net  Thu Oct 14 20:01:41 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Thu, 14 Oct 2010 23:01:41 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <alpine.WNT.2.00.1010141816050.3520@ccfloor209>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
Message-ID: <4CB7C415.6030004@dcrocker.net>



On 10/14/2010 12:27 PM, David Sitman wrote:
> This summer, the Ministry of Communications in Israel began considering a change
> in ISP licensing which would require ISP's to support email address portability,
...
>      this has caused us quite a bit of consternation.


It should.

Like many appealing ideas, it suffers upon careful consideration of the changes 
needed to make it happen.

Email addressing, registration and routing each have significant design and 
operations differences from the original telephone system.  Jack's example of a 
forwarding mailbox hints at the difference:  The address is tied to a mailbox. 
If you go elsewhere, the message still has to route through the old place.  With 
telephone number portability, the actual conversation does not "go through" the 
original provider.  (There is a routing layer that is separate from the 
conversation layer, which is not true for email.)

In addition, note that the domain name portion of the email address is a "name" 
of the provider.  That carries massive semantics, in contrast with the 
neutrality of a telephone number.  One would think that portability should not 
forever tie you to the name of your original provider.

Still, it's worth asking whether it is at all practical to create email 
portability?

The answer is not only yes, but... it's been done repeatedly and without 
mandating anything:

    Create an independent service that offers "portable" addresses.  Namely, it 
just is a forwarding service.(*)  (There are elaborations of this design that 
might get clever with per-user domain names and MX records, but I'll keep it 
simple. In reality, making the lookup handling be helpfully different from the 
message transfer handling -- that is, allowing the message communications to be 
"direct" -- is actually quite difficult, at a per-user granularity, relative to 
the current system.)

    This is a value-add overlay to the existing service... with no change to the 
existing service.  As long as the forwarding service stays in business you can 
have your actual mailbox anywhere you want.

    Note that going out of business is another point of difference between the 
telephone number management system versus the email portability idea.  The 
former doesn't have to worry about continuity of service in the face of 
bankruptcy while the email one does.


d/

-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From larrysheldon at cox.net  Thu Oct 14 20:03:22 2010
From: larrysheldon at cox.net (Larry Sheldon)
Date: Thu, 14 Oct 2010 22:03:22 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a0624084fc8dd6e5f057d@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]> <4CB7AC90.2030600@cox.net>
	<a06240848c8dd5dd023e6@[10.0.1.21]> <4CB7AF93.2010708@cox.net>
	<a0624084cc8dd638a7b5a@[10.0.1.21]> <4CB7BBE7.2030502@cox.net>
	<a0624084fc8dd6e5f057d@[10.0.1.21]>
Message-ID: <4CB7C47A.4030409@cox.net>

On 10/14/2010 9:44 PM, John Day wrote:

> This is in the category that once it appeared that perpetual motion was
> impossible. Then they discovered the Second Law of Thermodynamics and
> knew that it really was impossible.

Is it, really?

One explanation of the Way Things Are says the universe is expanding and 
will forever.  I recognize that there is an unresolved frame of 
reference issue, but doesn't that suggest that stuff will be moveing in 
perpetuity?

But that all is beyond my point.

Once upon a time somebody (NANPA?) said telephone numbers had 
characterisitics similar to what IP addresses have today.

Given enough wantto, I can see several systems already in my experiences 
(No additional inventing required).) that could be used to achieve that 
(including one already in use in most IP using shops).


That was my whole point--"can't be did" is probably wrong.


From jeanjour at comcast.net  Thu Oct 14 20:17:39 2010
From: jeanjour at comcast.net (John Day)
Date: Thu, 14 Oct 2010 23:17:39 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7C1B3.9010906@bennett.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@10.0.1.21>	<4CB7AA99.7000505@bennett.com>
	<AANLkTikoU8SiPTfm1dbaYzPAd8xoWnTwSytDF1hwBDMP@mail.gmail.com>
	<4CB7C1B3.9010906@bennett.com>
Message-ID: <a06240851c8dd77cc3aec@[10.0.1.21]>

You are focusing too much on the how and not the what.  It should be 
in terms of the service.  You don't want to control where it is 
routed, you want to ensure where it is not routed.

So if China is off limits, can I divert it through North Korea?  What 
about Congo?
  ;-)

At 19:51 -0700 2010/10/14, Richard Bennett wrote:
>  Well, yeah, X.25 or something like MPLS. My point was that I don't 
>want the route from Comcast.net to Bennett.com to go through China. 
>Not so much unique routes as user-controlled ones.
>
>On 10/14/2010 7:10 PM, Jorge Amodio wrote:
>>>lawyers, but it doesn't make them irrelevant. The Internet's global system
>>>of addressing requires a central authority to ensure address and route
>>>uniqueness; that's something we have to live with.
>>Addressing yes, routing no. With unique routes we'll all be screwed
>>and back to X.25
>>
>>Cheers
>>- J
>
>--
>Richard Bennett



From richard at bennett.com  Thu Oct 14 20:26:49 2010
From: richard at bennett.com (Richard Bennett)
Date: Thu, 14 Oct 2010 20:26:49 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7C47A.4030409@cox.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
	<4CB7AC90.2030600@cox.net>	<a06240848c8dd5dd023e6@[10.0.1.21]>
	<4CB7AF93.2010708@cox.net>	<a0624084cc8dd638a7b5a@[10.0.1.21]>
	<4CB7BBE7.2030502@cox.net>	<a0624084fc8dd6e5f057d@[10.0.1.21]>
	<4CB7C47A.4030409@cox.net>
Message-ID: <4CB7C9F9.4070105@bennett.com>

  All it takes is motivation. The phone company was motivated to 
separate phone numbers from switch ports by mobility. Is there a similar 
motivation in the IP world to make a 128 bit hex number permanent 
personal property?

On 10/14/2010 8:03 PM, Larry Sheldon wrote:
> On 10/14/2010 9:44 PM, John Day wrote:
>
>> This is in the category that once it appeared that perpetual motion was
>> impossible. Then they discovered the Second Law of Thermodynamics and
>> knew that it really was impossible.
>
> Is it, really?
>
> One explanation of the Way Things Are says the universe is expanding 
> and will forever.  I recognize that there is an unresolved frame of 
> reference issue, but doesn't that suggest that stuff will be moveing 
> in perpetuity?
>
> But that all is beyond my point.
>
> Once upon a time somebody (NANPA?) said telephone numbers had 
> characterisitics similar to what IP addresses have today.
>
> Given enough wantto, I can see several systems already in my 
> experiences (No additional inventing required).) that could be used to 
> achieve that (including one already in use in most IP using shops).
>
>
> That was my whole point--"can't be did" is probably wrong.

-- 
Richard Bennett



From cos at aaaaa.org  Thu Oct 14 20:35:51 2010
From: cos at aaaaa.org (Ofer Inbar)
Date: Thu, 14 Oct 2010 23:35:51 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1287107599.2914.115.camel@localhost>
References: <m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
	<1287107599.2914.115.camel@localhost>
Message-ID: <20101015033551.GI18254@mip.aaaaa.org>

On Thu, Oct 14, 2010 at 06:53:19PM -0700,
Jack Haverty <jack at 3kitty.org> wrote:
> Getting back to history...   IIRC, the Telephone Companies used the same
> argument about telephone numbers - they had a structure which was
> meaningful to the switching system, and was needed to establish the
> circuit through the various COs to end at a particular wire inside your
> house.
> 
> Somewhere between then and now, Number Portability happened despite
> their objections, cell phones were attached to everyone's ear, and, at
[...]
> Seems like The Internet could make a similar evolutionary step if it was
> needed....or mandated by government or customer pressures, as happened
> to the Telephone networks.

Perhaps if people routinely gave out business cards with their IP
addresses on them, and companies spent money advertising their IP
addresses, and people scribbled friends' and business partners' IP
addresses on notes and put them in their contact files, it might
make sense to consider that ever happening.

In our world, it makes no sense.

P.S. This thread makes me realize that I can't even remember
nic.near.net's IP address anymore :)
  -- Cos


From richard at bennett.com  Thu Oct 14 20:55:49 2010
From: richard at bennett.com (Richard Bennett)
Date: Thu, 14 Oct 2010 20:55:49 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7C415.6030004@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net>
Message-ID: <4CB7D0C5.40604@bennett.com>

  I like the idea of an email forwarding service. Let's call it gmail.com.

Seriously, though, if some political body wants to provide its voters 
with portable email addresses, all it has to do is fund a forwarding 
service. They don't approach it that way because they'd rather pass the 
costs off to an evil capitalist.

RB

On 10/14/2010 8:01 PM, Dave CROCKER wrote:
>
>
> On 10/14/2010 12:27 PM, David Sitman wrote:
>> This summer, the Ministry of Communications in Israel began 
>> considering a change
>> in ISP licensing which would require ISP's to support email address 
>> portability,
> ...
>>      this has caused us quite a bit of consternation.
>
>
> It should.
>
> Like many appealing ideas, it suffers upon careful consideration of 
> the changes needed to make it happen.
>
> Email addressing, registration and routing each have significant 
> design and operations differences from the original telephone system.  
> Jack's example of a forwarding mailbox hints at the difference:  The 
> address is tied to a mailbox. If you go elsewhere, the message still 
> has to route through the old place.  With telephone number 
> portability, the actual conversation does not "go through" the 
> original provider.  (There is a routing layer that is separate from 
> the conversation layer, which is not true for email.)
>
> In addition, note that the domain name portion of the email address is 
> a "name" of the provider.  That carries massive semantics, in contrast 
> with the neutrality of a telephone number.  One would think that 
> portability should not forever tie you to the name of your original 
> provider.
>
> Still, it's worth asking whether it is at all practical to create 
> email portability?
>
> The answer is not only yes, but... it's been done repeatedly and 
> without mandating anything:
>
>    Create an independent service that offers "portable" addresses.  
> Namely, it just is a forwarding service.(*)  (There are elaborations 
> of this design that might get clever with per-user domain names and MX 
> records, but I'll keep it simple. In reality, making the lookup 
> handling be helpfully different from the message transfer handling -- 
> that is, allowing the message communications to be "direct" -- is 
> actually quite difficult, at a per-user granularity, relative to the 
> current system.)
>
>    This is a value-add overlay to the existing service... with no 
> change to the existing service.  As long as the forwarding service 
> stays in business you can have your actual mailbox anywhere you want.
>
>    Note that going out of business is another point of difference 
> between the telephone number management system versus the email 
> portability idea.  The former doesn't have to worry about continuity 
> of service in the face of bankruptcy while the email one does.
>
>
> d/
>

-- 
Richard Bennett



From jack at 3kitty.org  Thu Oct 14 21:12:03 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 14 Oct 2010 21:12:03 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7C415.6030004@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net>
Message-ID: <1287115923.2914.133.camel@localhost>

Dave,

If "jack at 3kitty.org" is the example of forwarding you have in mind,
that's not how it works.  When I change providers, I move my 3kitty.org
service from one provider to another, and the appropriate DNS records
reflect the change.  Mail goes directly to my current server as
reflected by the DNS data, not through the old place.  

I do have a bunch of xxx at 3kitty.org mailboxes, and they all must move
together.  If my current provider disappears, I'll just put 3kitty.org
on a new one.  Essentially I'm using a domain name to create a set of
mailboxes that are portable.

The important feature when I move is the (lack of) effect on people
trying to send me mail.  I never have to send out "please change my
email address in your contacts". I use the DNS to get "Email Address
Portability".  But, as I said earlier, it only works if not too many of
us do it.

3kitty.org is not an ISP, it's just me.  Getting back to the original
question - I have a better case for ownership of jack at 3kitty.org than I
have on the IP address I'm using right now.

/Jack

On Thu, 2010-10-14 at 23:01 -0400, Dave CROCKER wrote:
> 
> On 10/14/2010 12:27 PM, David Sitman wrote:
> > This summer, the Ministry of Communications in Israel began considering a change
> > in ISP licensing which would require ISP's to support email address portability,
> ...
> >      this has caused us quite a bit of consternation.
> 
> 
> It should.
> 
> Like many appealing ideas, it suffers upon careful consideration of the changes 
> needed to make it happen.
> 
> Email addressing, registration and routing each have significant design and 
> operations differences from the original telephone system.  Jack's example of a 
> forwarding mailbox hints at the difference:  The address is tied to a mailbox. 
> If you go elsewhere, the message still has to route through the old place.  With 
> telephone number portability, the actual conversation does not "go through" the 
> original provider.  (There is a routing layer that is separate from the 
> conversation layer, which is not true for email.)
> 
> In addition, note that the domain name portion of the email address is a "name" 
> of the provider.  That carries massive semantics, in contrast with the 
> neutrality of a telephone number.  One would think that portability should not 
> forever tie you to the name of your original provider.
> 
> Still, it's worth asking whether it is at all practical to create email 
> portability?
> 
> The answer is not only yes, but... it's been done repeatedly and without 
> mandating anything:
> 
>     Create an independent service that offers "portable" addresses.  Namely, it 
> just is a forwarding service.(*)  (There are elaborations of this design that 
> might get clever with per-user domain names and MX records, but I'll keep it 
> simple. In reality, making the lookup handling be helpfully different from the 
> message transfer handling -- that is, allowing the message communications to be 
> "direct" -- is actually quite difficult, at a per-user granularity, relative to 
> the current system.)
> 
>     This is a value-add overlay to the existing service... with no change to the 
> existing service.  As long as the forwarding service stays in business you can 
> have your actual mailbox anywhere you want.
> 
>     Note that going out of business is another point of difference between the 
> telephone number management system versus the email portability idea.  The 
> former doesn't have to worry about continuity of service in the face of 
> bankruptcy while the email one does.
> 
> 
> d/
> 




From richard at bennett.com  Thu Oct 14 21:29:03 2010
From: richard at bennett.com (Richard Bennett)
Date: Thu, 14 Oct 2010 21:29:03 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1287115923.2914.133.camel@localhost>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>	<4CB7C415.6030004@dcrocker.net>
	<1287115923.2914.133.camel@localhost>
Message-ID: <4CB7D88F.6020002@bennett.com>

  It's about time somebody explained email to Dave Crocker.

RB

On 10/14/2010 9:12 PM, Jack Haverty wrote:
> Dave,
>
> If "jack at 3kitty.org" is the example of forwarding you have in mind,
> that's not how it works.  When I change providers, I move my 3kitty.org
> service from one provider to another, and the appropriate DNS records
> reflect the change.  Mail goes directly to my current server as
> reflected by the DNS data, not through the old place.
>
> I do have a bunch of xxx at 3kitty.org mailboxes, and they all must move
> together.  If my current provider disappears, I'll just put 3kitty.org
> on a new one.  Essentially I'm using a domain name to create a set of
> mailboxes that are portable.
>
> The important feature when I move is the (lack of) effect on people
> trying to send me mail.  I never have to send out "please change my
> email address in your contacts". I use the DNS to get "Email Address
> Portability".  But, as I said earlier, it only works if not too many of
> us do it.
>
> 3kitty.org is not an ISP, it's just me.  Getting back to the original
> question - I have a better case for ownership of jack at 3kitty.org than I
> have on the IP address I'm using right now.
>
> /Jack
>
> On Thu, 2010-10-14 at 23:01 -0400, Dave CROCKER wrote:
>> On 10/14/2010 12:27 PM, David Sitman wrote:
>>> This summer, the Ministry of Communications in Israel began considering a change
>>> in ISP licensing which would require ISP's to support email address portability,
>> ...
>>>       this has caused us quite a bit of consternation.
>>
>> It should.
>>
>> Like many appealing ideas, it suffers upon careful consideration of the changes
>> needed to make it happen.
>>
>> Email addressing, registration and routing each have significant design and
>> operations differences from the original telephone system.  Jack's example of a
>> forwarding mailbox hints at the difference:  The address is tied to a mailbox.
>> If you go elsewhere, the message still has to route through the old place.  With
>> telephone number portability, the actual conversation does not "go through" the
>> original provider.  (There is a routing layer that is separate from the
>> conversation layer, which is not true for email.)
>>
>> In addition, note that the domain name portion of the email address is a "name"
>> of the provider.  That carries massive semantics, in contrast with the
>> neutrality of a telephone number.  One would think that portability should not
>> forever tie you to the name of your original provider.
>>
>> Still, it's worth asking whether it is at all practical to create email
>> portability?
>>
>> The answer is not only yes, but... it's been done repeatedly and without
>> mandating anything:
>>
>>      Create an independent service that offers "portable" addresses.  Namely, it
>> just is a forwarding service.(*)  (There are elaborations of this design that
>> might get clever with per-user domain names and MX records, but I'll keep it
>> simple. In reality, making the lookup handling be helpfully different from the
>> message transfer handling -- that is, allowing the message communications to be
>> "direct" -- is actually quite difficult, at a per-user granularity, relative to
>> the current system.)
>>
>>      This is a value-add overlay to the existing service... with no change to the
>> existing service.  As long as the forwarding service stays in business you can
>> have your actual mailbox anywhere you want.
>>
>>      Note that going out of business is another point of difference between the
>> telephone number management system versus the email portability idea.  The
>> former doesn't have to worry about continuity of service in the face of
>> bankruptcy while the email one does.
>>
>>
>> d/
>>
>

-- 
Richard Bennett



From tony.li at tony.li  Thu Oct 14 22:57:32 2010
From: tony.li at tony.li (Tony Li)
Date: Thu, 14 Oct 2010 22:57:32 -0700
Subject: [ih] In the interests of historical accuracy...
In-Reply-To: <1287071374.2914.13.camel@localhost>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
Message-ID: <7303AD87-6B21-465C-90F5-CE393824EDEF@tony.li>


> Of course, we'll all wear flip-flops and grow long beards, to establish
> credibility as Internet Experts.  I think Jon would enjoy that...

Jon wore first generation Teva sandals.  With black socks.

;-)

Tony



From tony.li at tony.li  Thu Oct 14 22:59:32 2010
From: tony.li at tony.li (Tony Li)
Date: Thu, 14 Oct 2010 22:59:32 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240840c8dd222b2950@[10.0.1.21]>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<6.2.3.4.2.20101014134318.03fb0090@MS1.MailSys.Net>
	<a06240840c8dd222b2950@[10.0.1.21]>
Message-ID: <FAE31608-B2F9-4582-A8A7-F223DCBCE40C@tony.li>


> Or perhaps I should say it the other way around.  Please create a free market in IPv6 addresses as soon as possible so that we get to see how many ways the Internet can suffer a meltdown.  This will be interesting to watch.


Done and done.  RIRs are now handing out PI prefixes freely.  The routing table is now growing at 60% y/y.

We live in interesting times,
Tony




From adrian at creative.net.au  Thu Oct 14 23:43:42 2010
From: adrian at creative.net.au (Adrian Chadd)
Date: Fri, 15 Oct 2010 14:43:42 +0800
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <FAE31608-B2F9-4582-A8A7-F223DCBCE40C@tony.li>
References: <m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<6.2.3.4.2.20101014134318.03fb0090@MS1.MailSys.Net>
	<a06240840c8dd222b2950@[10.0.1.21]>
	<FAE31608-B2F9-4582-A8A7-F223DCBCE40C@tony.li>
Message-ID: <20101015064342.GA32602@skywalker.creative.net.au>

On Thu, Oct 14, 2010, Tony Li wrote:

> Done and done.  RIRs are now handing out PI prefixes freely.  The routing table is now growing at 60% y/y.
> 
> We live in interesting times,

IPv6 will be even interesting-er times. Orgs right now want to be able to route
their space, PI or not, and carry it between providers/announce it to >1 upstream.
This will continue with IPv6, even if IP purists believe/want otherwise.

Yay. :)


Adrian



From nigel at channelisles.net  Fri Oct 15 00:56:30 2010
From: nigel at channelisles.net (Nigel Roberts)
Date: Fri, 15 Oct 2010 08:56:30 +0100
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101015033551.GI18254@mip.aaaaa.org>
References: <m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>	<a06240846c8dd4ff6e4d0@[10.0.1.21]>	<1287107599.2914.115.camel@localhost>
	<20101015033551.GI18254@mip.aaaaa.org>
Message-ID: <4CB8092E.60306@channelisles.net>


> P.S. This thread makes me realize that I can't even remember
> nic.near.net's IP address anymore :)
>    -- Cos

This thread makes me realise, quite bizarrely that I CAN remember the 
physical host addresses for MIT-AI (where I had my first ever email 
address as a tourist) and MIT-DM (where we MUD hackers were inspired by 
ZORK).

134 and 70, respectively, wasn't it?

  I think it's because I used the physical addresses (@O on the London 
TIP) to get there.


Nigel


From jcurran at arin.net  Fri Oct 15 03:33:25 2010
From: jcurran at arin.net (John Curran)
Date: Fri, 15 Oct 2010 06:33:25 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101015033551.GI18254@mip.aaaaa.org>
References: <m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
	<1287107599.2914.115.camel@localhost>
	<20101015033551.GI18254@mip.aaaaa.org>
Message-ID: <6C80C797-3E85-49A7-AFC2-2CE6E0B2402A@arin.net>

On Oct 14, 2010, at 11:35 PM, Ofer Inbar wrote:

> On Thu, Oct 14, 2010 at 06:53:19PM -0700,
> Jack Haverty <jack at 3kitty.org> wrote:
>> Getting back to history...   IIRC, the Telephone Companies used the same
>> argument about telephone numbers - they had a structure which was
>> meaningful to the switching system, and was needed to establish the
>> circuit through the various COs to end at a particular wire inside your
>> house.
>> 
>> Somewhere between then and now, Number Portability happened despite
>> their objections, cell phones were attached to everyone's ear, and, at
> [...]
>> Seems like The Internet could make a similar evolutionary step if it was
>> needed....or mandated by government or customer pressures, as happened
>> to the Telephone networks.
> 
> Perhaps if people routinely gave out business cards with their IP
> addresses on them, and companies spent money advertising their IP
> addresses, and people scribbled friends' and business partners' IP
> addresses on notes and put them in their contact files, it might
> make sense to consider that ever happening.
> 
> In our world, it makes no sense.
> 
> P.S. This thread makes me realize that I can't even remember
> nic.near.net's IP address anymore :)

192.68.71.4

/John




From jeanjour at comcast.net  Fri Oct 15 03:50:13 2010
From: jeanjour at comcast.net (John Day)
Date: Fri, 15 Oct 2010 06:50:13 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB7C2A1.80109@bennett.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>	<4CB7AC90.2030600@cox.net>
	<a06240848c8dd5dd023e6@[10.0.1.21]>	<4CB7AF93.2010708@cox.net>
	<a0624084cc8dd638a7b5a@[10.0.1.21]> <4CB7C2A1.80109@bennett.com>
Message-ID: <a06240854c8dde0d0bc69@[10.0.1.21]>

>
>Embedding MAC addresses inside IPv6 addresses was one of the dumbest 
>ideas in the history of networking.

Embedding an (N-1)-address in an (N)-address is one of those things 
that sounds right on first blush,but upon closer inspection turns out 
to be a very bad idea.

There was language to that effect in the early versions of the OSI 
Reference Model, e.g. as far back as  N227.  Once we started the work 
on the Naming and Addressing addendum to the Model about 1982 and 
looked at it carefully, we realized our error.  At that point the 
model in the approval process, politics required some slight of hand 
to fix it.

But we have known it is a bad idea for over a quarter century.  It is 
mildly amusing how many professors still propose it.  Tells you how 
carefully they have thought about it.  ;-)  Of cousre, the IPv6 group 
falls in the same category, but then you kind of expect that.

Take care,
John

>
>RB
>
>On 10/14/2010 6:59 PM, John Day wrote:
>>>On 10/14/2010 8:24 PM, John Day wrote:
>>>>Sorry, what was true of telephone numbers?
>>
>>Snip
>>
>>>>>We used to think that was true of telephone numbers.
>>>
>>>("these concepts do not apply to addresses")
>>
>>We already went though this. Although perhaps too indirectly.
>>
>>The cell phone system changed phone numbers from being 
>>physical-connection-endpoint-addresses into application names and 
>>put other addressing schemes underneath the phone numbers so tehy 
>>could route to them and didn't tell anyone.  It became so useful, 
>>that it was decided to make all telephone numbers application names.
>>
>>So what would you like to do?  Make IPv6 addresses application 
>>names? Then what would be put under them to be addresses so we 
>>could route to them?  MAC addresses?   That would be really neat. 
>>Then we could route based on the manufacturer who built the 
>>interface rather than the provider where the interface was.  That 
>>ought to work really well!
>>
>>Take care,
>>John
>
>--
>Richard Bennett



From jcurran at arin.net  Fri Oct 15 03:54:07 2010
From: jcurran at arin.net (John Curran)
Date: Fri, 15 Oct 2010 06:54:07 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <6C80C797-3E85-49A7-AFC2-2CE6E0B2402A@arin.net>
References: <m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@[10.0.1.21]>
	<1287107599.2914.115.camel@localhost>
	<20101015033551.GI18254@mip.aaaaa.org>
	<6C80C797-3E85-49A7-AFC2-2CE6E0B2402A@arin.net>
Message-ID: <D45255A5-5F27-4F8E-AE75-B5168AAFBBBF@arin.net>

On Oct 15, 2010, at 6:33 AM, John Curran wrote:
>> P.S. This thread makes me realize that I can't even remember
>> nic.near.net's IP address anymore :)
> 
> 192.68.71.4

Oops; actually 192.52.71.4!  Apparently some gray matter has
moved from my brain to my beard over the years...

It proves the point that we're much better off using names in
any case.

/John




From dhc2 at dcrocker.net  Fri Oct 15 04:40:54 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Fri, 15 Oct 2010 07:40:54 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1287115923.2914.133.camel@localhost>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>	<4CB7C415.6030004@dcrocker.net>
	<1287115923.2914.133.camel@localhost>
Message-ID: <4CB83DC6.3000203@dcrocker.net>



On 10/15/2010 12:12 AM, Jack Haverty wrote:
> If "jack at 3kitty.org" is the example of forwarding you have in mind,
> that's not how it works.  When I change providers, I move my 3kitty.org
> service from one provider to another,
...
> I do have a bunch of xxx at 3kitty.org mailboxes, and they all must move
> together.


You are emulating a version of exactly the service I described towards the end 
of my note, modulo the extra forwarding hop. It's key feature is that it is 
independent of ISPs and it does not require their cooperation.

However, as you note, all of the mailboxes must move together:  granularity is 
at the domain name level, not the mailbox level.  To get per-user granularity, 
you have to encode it in the domain name, given the way email routing works.

The reason you can have "direct" routing, without having to go through a 
forwarder is that you control the DNS MX record.  In effect that means an MX 
record per "customer", if not per "mailbox".  Again, that's doable today and it 
is done today.  The challenge is scaling that model up to a mass market.  In 
effect, it means an MX per user (or maybe per family).  Yuch.

d/

ps.  I figure you're still unhappy we didn't adopt your encoding scheme when we 
did RFC 733...



-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From jmamodio at gmail.com  Fri Oct 15 05:08:05 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Fri, 15 Oct 2010 07:08:05 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB83DC6.3000203@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net>
	<1287115923.2914.133.camel@localhost>
	<4CB83DC6.3000203@dcrocker.net>
Message-ID: <AANLkTinaTkGBdOktSDzHFJTSuGmMyQFOBnAwrQwrDhy2@mail.gmail.com>

> The reason you can have "direct" routing, without having to go through a
> forwarder is that you control the DNS MX record. ?In effect that means an MX
> record per "customer", if not per "mailbox". ?Again, that's doable today and
> it is done today. ?The challenge is scaling that model up to a mass market.
> ?In effect, it means an MX per user (or maybe per family). ?Yuch.

Did anybody experimented with the "experimental" MB, MR RR's (RFC1035) ?

-J



From jeanjour at comcast.net  Fri Oct 15 05:26:57 2010
From: jeanjour at comcast.net (John Day)
Date: Fri, 15 Oct 2010 08:26:57 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB83DC6.3000203@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net>	<1287115923.2914.133.camel@localhost>
	<4CB83DC6.3000203@dcrocker.net>
Message-ID: <a06240859c8ddf8a3516f@[10.0.1.21]>

But the real point here is that routing email, is fundamentally no 
different than routing packets, except there is no TTL with email. 
All of the same principles hold.


At 7:40 -0400 2010/10/15, Dave CROCKER wrote:
>On 10/15/2010 12:12 AM, Jack Haverty wrote:
>>If "jack at 3kitty.org" is the example of forwarding you have in mind,
>>that's not how it works.  When I change providers, I move my 3kitty.org
>>service from one provider to another,
>...
>>I do have a bunch of xxx at 3kitty.org mailboxes, and they all must move
>>together.
>
>
>You are emulating a version of exactly the service I described 
>towards the end of my note, modulo the extra forwarding hop. It's 
>key feature is that it is independent of ISPs and it does not 
>require their cooperation.
>
>However, as you note, all of the mailboxes must move together: 
>granularity is at the domain name level, not the mailbox level.  To 
>get per-user granularity, you have to encode it in the domain name, 
>given the way email routing works.
>
>The reason you can have "direct" routing, without having to go 
>through a forwarder is that you control the DNS MX record.  In 
>effect that means an MX record per "customer", if not per "mailbox". 
>Again, that's doable today and it is done today.  The challenge is 
>scaling that model up to a mass market.  In effect, it means an MX 
>per user (or maybe per family).  Yuch.
>
>d/
>
>ps.  I figure you're still unhappy we didn't adopt your encoding 
>scheme when we did RFC 733...
>
>
>
>--
>
>   Dave Crocker
>   Brandenburg InternetWorking
>   bbiw.net



From spencer at mcsr-labs.org  Fri Oct 15 06:19:08 2010
From: spencer at mcsr-labs.org (Spencer Dawkins)
Date: Fri, 15 Oct 2010 08:19:08 -0500
Subject: [ih] Some Questions over IPv4 Ownership
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu><m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com><2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu><AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com><2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu><1287071374.2914.13.camel@localhost><alpine.WNT.2.00.1010141816050.3520@ccfloor209><4CB7C415.6030004@dcrocker.net>	<1287115923.2914.133.camel@localhost><4CB83DC6.3000203@dcrocker.net>
	<a06240859c8ddf8a3516f@[10.0.1.21]>
Message-ID: <BC0FDDB358B94349B6F21816CAE46A40@china.huawei.com>

I do wonder what year the Internet first carried more *packets per second* 
than the current Internet carries *emails per second* - if that's clearly 
stated, I'm thinking Jack's point is that the only real difference is 
scaling.

And if that was the point, I agree!

Spencer


> But the real point here is that routing email, is fundamentally no 
> different than routing packets, except there is no TTL with email. All of 
> the same principles hold.
>
>
> At 7:40 -0400 2010/10/15, Dave CROCKER wrote:
>>On 10/15/2010 12:12 AM, Jack Haverty wrote:
>>>If "jack at 3kitty.org" is the example of forwarding you have in mind,
>>>that's not how it works.  When I change providers, I move my 3kitty.org
>>>service from one provider to another,
>>...
>>>I do have a bunch of xxx at 3kitty.org mailboxes, and they all must move
>>>together.
>>
>>
>>You are emulating a version of exactly the service I described towards the 
>>end of my note, modulo the extra forwarding hop. It's key feature is that 
>>it is independent of ISPs and it does not require their cooperation.
>>
>>However, as you note, all of the mailboxes must move together: granularity 
>>is at the domain name level, not the mailbox level.  To get per-user 
>>granularity, you have to encode it in the domain name, given the way email 
>>routing works.
>>
>>The reason you can have "direct" routing, without having to go through a 
>>forwarder is that you control the DNS MX record.  In effect that means an 
>>MX record per "customer", if not per "mailbox". Again, that's doable today 
>>and it is done today.  The challenge is scaling that model up to a mass 
>>market.  In effect, it means an MX per user (or maybe per family).  Yuch.
>>
>>d/
>>
>>ps.  I figure you're still unhappy we didn't adopt your encoding scheme 
>>when we did RFC 733...
>>
>>
>>
>>--
>>
>>   Dave Crocker
>>   Brandenburg InternetWorking
>>   bbiw.net
> 



From craig at aland.bbn.com  Fri Oct 15 06:29:45 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Fri, 15 Oct 2010 09:29:45 -0400
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101015132945.89F4328E137@aland.bbn.com>

> > The reason you can have "direct" routing, without having to go through a
> > forwarder is that you control the DNS MX record. ?In effect that means an MX
> > record per "customer", if not per "mailbox". ?Again, that's doable today and
> > it is done today. ?The challenge is scaling that model up to a mass market.
> > ?In effect, it means an MX per user (or maybe per family). ?Yuch.
> 
> Did anybody experimented with the "experimental" MB, MR RR's (RFC1035) ?

Yes, I did as part of the experimentation on MD and MF that lead to MX RRs.

I no longer remember what I did and my notes/emails have long since gone
away -- all I remember was wallking away thinking it was a bad idea.

Thanks!

Craig


From dhc2 at dcrocker.net  Fri Oct 15 07:06:09 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Fri, 15 Oct 2010 10:06:09 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <BC0FDDB358B94349B6F21816CAE46A40@china.huawei.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu><m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com><2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu><AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com><2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu><1287071374.2914.13.camel@localhost><alpine.WNT.2.00.1010141816050.3520@ccfloor209><4CB7C415.6030004@dcrocker.net>	<1287115923.2914.133.camel@localhost><4CB83DC6.3000203@dcrocker.net>	<a06240859c8ddf8a3516f@[10.0.1.21]>
	<BC0FDDB358B94349B6F21816CAE46A40@china.huawei.com>
Message-ID: <4CB85FD1.4080802@dcrocker.net>



On 10/15/2010 9:19 AM, Spencer Dawkins wrote:
> I do wonder what year the Internet first carried more *packets per second* than
> the current Internet carries *emails per second* - if that's clearly stated, I'm

It's an interesting question no matter how it's stated.

It suggests some sort of meta-analysis with different layers of 
chunk/packet/segment/message services and making historical quantum transitions 
of activity from one to the next.  Downward, given the way you asked your 
question.  (Is this like moving things into hardware?)


> thinking Jack's point is that the only real difference is scaling.

Alas, some differences in degree become differences in kind.  The latency and 
loss characteristics of messaging like email are typically so different from 
what is usual for IP datagrams as to permit/require all sorts of different 
design decisions.

d/

-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From jmamodio at gmail.com  Fri Oct 15 06:16:45 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Fri, 15 Oct 2010 08:16:45 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <a06240859c8ddf8a3516f@10.0.1.21>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net>
	<1287115923.2914.133.camel@localhost>
	<4CB83DC6.3000203@dcrocker.net> <a06240859c8ddf8a3516f@10.0.1.21>
Message-ID: <AANLkTik7j-8RdsUx3mw2vO3xSQUMg_imuxfrpmeczPpD@mail.gmail.com>

> But the real point here is that routing email, is fundamentally no different
> than routing packets, except there is no TTL with email. All of the same
> principles hold.

How ? I don't see email routing comparable to IP routing.

-J


From sbrim at cisco.com  Fri Oct 15 08:36:59 2010
From: sbrim at cisco.com (Scott Brim)
Date: Fri, 15 Oct 2010 11:36:59 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTik7j-8RdsUx3mw2vO3xSQUMg_imuxfrpmeczPpD@mail.gmail.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>	<4CB7C415.6030004@dcrocker.net>	<1287115923.2914.133.camel@localhost>	<4CB83DC6.3000203@dcrocker.net>
	<a06240859c8ddf8a3516f@10.0.1.21>
	<AANLkTik7j-8RdsUx3mw2vO3xSQUMg_imuxfrpmeczPpD@mail.gmail.com>
Message-ID: <4CB8751B.5060505@cisco.com>

On 10/15/2010 09:16 EDT, Jorge Amodio wrote:
>> But the real point here is that routing email, is fundamentally no different
>> than routing packets, except there is no TTL with email. All of the same
>> principles hold.
> 
> How ? I don't see email routing comparable to IP routing.
> 
> -J

It's all store-and-forward, and the forwarding step uses whatever
external information the forwarder can glean.  A difference is the
routing mechanism.  (Typical) Internet routing gathers information in
advance because it has to forward its packets fast (and suffers when
reality doesn't match the information it has gathered), while mailers
have a nice mapping system built for them called DNS that they use on
demand.


From jack at 3kitty.org  Fri Oct 15 08:38:57 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Fri, 15 Oct 2010 08:38:57 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB83DC6.3000203@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net> <1287115923.2914.133.camel@localhost>
	<4CB83DC6.3000203@dcrocker.net>
Message-ID: <1287157138.2914.148.camel@localhost>

Yep, we're on the same page.  Email portability exists now, if you just
get a domain name per recipient.  I suspect the DNS today couldn't
handle that if too many people did it.  Yes, yuch.  But a similar
mechanism must exist today inside the telephone world so it's proven
possible.

I think eventually we'll all have portable email addresses.  Not too
long either.  Our addresses might just be our cell phone numbers.  Why
not...email and voicemail aren't all that different except at the very
endpoints of the path.

Naah, I'm not mad about the encoding in 733.  But I'm happy that
structured data encoding has finally caught on with XML.  If those now
ancient mail protocols got a facelift today, I bet they'd be based on
XML.

/Jack


On Fri, 2010-10-15 at 07:40 -0400, Dave CROCKER wrote:
> 
> On 10/15/2010 12:12 AM, Jack Haverty wrote:
> > If "jack at 3kitty.org" is the example of forwarding you have in mind,
> > that's not how it works.  When I change providers, I move my 3kitty.org
> > service from one provider to another,
> ...
> > I do have a bunch of xxx at 3kitty.org mailboxes, and they all must move
> > together.
> 
> 
> You are emulating a version of exactly the service I described towards the end 
> of my note, modulo the extra forwarding hop. It's key feature is that it is 
> independent of ISPs and it does not require their cooperation.
> 
> However, as you note, all of the mailboxes must move together:  granularity is 
> at the domain name level, not the mailbox level.  To get per-user granularity, 
> you have to encode it in the domain name, given the way email routing works.
> 
> The reason you can have "direct" routing, without having to go through a 
> forwarder is that you control the DNS MX record.  In effect that means an MX 
> record per "customer", if not per "mailbox".  Again, that's doable today and it 
> is done today.  The challenge is scaling that model up to a mass market.  In 
> effect, it means an MX per user (or maybe per family).  Yuch.
> 
> d/
> 
> ps.  I figure you're still unhappy we didn't adopt your encoding scheme when we 
> did RFC 733...
> 
> 
> 




From jmamodio at gmail.com  Fri Oct 15 09:08:55 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Fri, 15 Oct 2010 11:08:55 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB8751B.5060505@cisco.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net>
	<1287115923.2914.133.camel@localhost>
	<4CB83DC6.3000203@dcrocker.net> <a06240859c8ddf8a3516f@10.0.1.21>
	<AANLkTik7j-8RdsUx3mw2vO3xSQUMg_imuxfrpmeczPpD@mail.gmail.com>
	<4CB8751B.5060505@cisco.com>
Message-ID: <AANLkTinyfO3Jqd+THz66J+LgS=9cC3CUs0RSk7dkkziV@mail.gmail.com>

> It's all store-and-forward, and the forwarding step uses whatever
> external information the forwarder can glean. ?A difference is the
> routing mechanism. ?(Typical) Internet routing gathers information in
> advance because it has to forward its packets fast (and suffers when
> reality doesn't match the information it has gathered), while mailers
> have a nice mapping system built for them called DNS that they use on
> demand.

I always thought email routing as end-to-end, if you take a message as
a single data unit that message will be delivered (in theory) to
whatever host has the MX record for the destination domain, instead IP
packets are routed on a hop by hop basis with a chance of routing
dynamically changing when the packet left the source and is in
transit.

I don't see further forwarding by the receiving MX host as "routing".

-J



From vint at google.com  Fri Oct 15 09:18:21 2010
From: vint at google.com (Vint Cerf)
Date: Fri, 15 Oct 2010 12:18:21 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1287157138.2914.148.camel@localhost>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net>
	<1287115923.2914.133.camel@localhost>
	<4CB83DC6.3000203@dcrocker.net>
	<1287157138.2914.148.camel@localhost>
Message-ID: <AANLkTikWeov2twuYsJsBJ87hnW33fw7LbeToV3EH1cdw@mail.gmail.com>

i hope HTML5 instead :-)

v


On Fri, Oct 15, 2010 at 11:38 AM, Jack Haverty <jack at 3kitty.org> wrote:
> Yep, we're on the same page. ?Email portability exists now, if you just
> get a domain name per recipient. ?I suspect the DNS today couldn't
> handle that if too many people did it. ?Yes, yuch. ?But a similar
> mechanism must exist today inside the telephone world so it's proven
> possible.
>
> I think eventually we'll all have portable email addresses. ?Not too
> long either. ?Our addresses might just be our cell phone numbers. ?Why
> not...email and voicemail aren't all that different except at the very
> endpoints of the path.
>
> Naah, I'm not mad about the encoding in 733. ?But I'm happy that
> structured data encoding has finally caught on with XML. ?If those now
> ancient mail protocols got a facelift today, I bet they'd be based on
> XML.
>
> /Jack
>
>
> On Fri, 2010-10-15 at 07:40 -0400, Dave CROCKER wrote:
>>
>> On 10/15/2010 12:12 AM, Jack Haverty wrote:
>> > If "jack at 3kitty.org" is the example of forwarding you have in mind,
>> > that's not how it works. ?When I change providers, I move my 3kitty.org
>> > service from one provider to another,
>> ...
>> > I do have a bunch of xxx at 3kitty.org mailboxes, and they all must move
>> > together.
>>
>>
>> You are emulating a version of exactly the service I described towards the end
>> of my note, modulo the extra forwarding hop. It's key feature is that it is
>> independent of ISPs and it does not require their cooperation.
>>
>> However, as you note, all of the mailboxes must move together: ?granularity is
>> at the domain name level, not the mailbox level. ?To get per-user granularity,
>> you have to encode it in the domain name, given the way email routing works.
>>
>> The reason you can have "direct" routing, without having to go through a
>> forwarder is that you control the DNS MX record. ?In effect that means an MX
>> record per "customer", if not per "mailbox". ?Again, that's doable today and it
>> is done today. ?The challenge is scaling that model up to a mass market. ?In
>> effect, it means an MX per user (or maybe per family). ?Yuch.
>>
>> d/
>>
>> ps. ?I figure you're still unhappy we didn't adopt your encoding scheme when we
>> did RFC 733...
>>
>>
>>
>
>
>



From craig at aland.bbn.com  Fri Oct 15 09:24:00 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Fri, 15 Oct 2010 12:24:00 -0400
Subject: [ih] Some Questions over IPv4 Ownership
Message-ID: <20101015162401.0017E28E137@aland.bbn.com>

> > It's all store-and-forward, and the forwarding step uses whatever
> > external information the forwarder can glean. ?A difference is the
> > routing mechanism. ?(Typical) Internet routing gathers information in
> > advance because it has to forward its packets fast (and suffers when
> > reality doesn't match the information it has gathered), while mailers
> > have a nice mapping system built for them called DNS that they use on
> > demand.
> 
> I always thought email routing as end-to-end, if you take a message as
> a single data unit that message will be delivered (in theory) to
> whatever host has the MX record for the destination domain, instead IP
> packets are routed on a hop by hop basis with a chance of routing
> dynamically changing when the packet left the source and is in
> transit.

Go re-read RFC 974 -- you can route directly, you can route through
intermediate nodes.  All carefully designed to allow for failures and
failovers.  Also, your MX destination is not necessary the point of
delivery.  E.g. Email to bbn.com may be redirected internally to one
of several mailbox servers from which folks retrieve their email.
(Think of it as dynamic routing [MX] coupled with static routing
within an organization).

Thanks!

Craig


From bob.hinden at gmail.com  Fri Oct 15 10:27:36 2010
From: bob.hinden at gmail.com (Bob Hinden)
Date: Fri, 15 Oct 2010 10:27:36 -0700
Subject: [ih] In the interests of historical accuracy...
In-Reply-To: <7303AD87-6B21-465C-90F5-CE393824EDEF@tony.li>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<7303AD87-6B21-465C-90F5-CE393824EDEF@tony.li>
Message-ID: <87714AC3-45E5-43D8-B516-F18803942748@gmail.com>

Tony,

On Oct 14, 2010, at 10:57 PM, Tony Li wrote:

> 
>> Of course, we'll all wear flip-flops and grow long beards, to establish
>> credibility as Internet Experts.  I think Jon would enjoy that...
> 
> Jon wore first generation Teva sandals.  With black socks.

I remember him wearing Birkenstock sandals.

Bob




From sbrim at cisco.com  Fri Oct 15 10:33:14 2010
From: sbrim at cisco.com (Scott Brim)
Date: Fri, 15 Oct 2010 13:33:14 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTinyfO3Jqd+THz66J+LgS=9cC3CUs0RSk7dkkziV@mail.gmail.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>	<4CB7C415.6030004@dcrocker.net>	<1287115923.2914.133.camel@localhost>	<4CB83DC6.3000203@dcrocker.net>	<a06240859c8ddf8a3516f@10.0.1.21>	<AANLkTik7j-8RdsUx3mw2vO3xSQUMg_imuxfrpmeczPpD@mail.gmail.com>	<4CB8751B.5060505@cisco.com>
	<AANLkTinyfO3Jqd+THz66J+LgS=9cC3CUs0RSk7dkkziV@mail.gmail.com>
Message-ID: <4CB8905A.3010508@cisco.com>

On 10/15/2010 12:08 EDT, Jorge Amodio wrote:
>> It's all store-and-forward, and the forwarding step uses whatever
>> external information the forwarder can glean.  A difference is the
>> routing mechanism.  (Typical) Internet routing gathers information in
>> advance because it has to forward its packets fast (and suffers when
>> reality doesn't match the information it has gathered), while mailers
>> have a nice mapping system built for them called DNS that they use on
>> demand.
> 
> I always thought email routing as end-to-end, if you take a message as
> a single data unit that message will be delivered (in theory) to
> whatever host has the MX record for the destination domain, instead IP
> packets are routed on a hop by hop basis with a chance of routing
> dynamically changing when the packet left the source and is in
> transit.
> 
> I don't see further forwarding by the receiving MX host as "routing".
> 
> -J

Check out "layer networks" in ITU-T G.805 and following.  Email routing
is hop-by-hop within its layer network context.


From jmamodio at gmail.com  Fri Oct 15 11:10:01 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Fri, 15 Oct 2010 13:10:01 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB8905A.3010508@cisco.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net>
	<1287115923.2914.133.camel@localhost>
	<4CB83DC6.3000203@dcrocker.net> <a06240859c8ddf8a3516f@10.0.1.21>
	<AANLkTik7j-8RdsUx3mw2vO3xSQUMg_imuxfrpmeczPpD@mail.gmail.com>
	<4CB8751B.5060505@cisco.com>
	<AANLkTinyfO3Jqd+THz66J+LgS=9cC3CUs0RSk7dkkziV@mail.gmail.com>
	<4CB8905A.3010508@cisco.com>
Message-ID: <AANLkTing6SiYgFpEcDMtYo+kxDBgBAYrkscX1+rK-5oe@mail.gmail.com>

> Check out "layer networks" in ITU-T G.805 and following. ?Email routing
> is hop-by-hop within its layer network context.

I guess you mean from an abstract/architectural point of view right ?

-J

PS. hate those ITU ex CCITT docs ...



From dhc2 at dcrocker.net  Fri Oct 15 11:23:11 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Fri, 15 Oct 2010 14:23:11 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTinyfO3Jqd+THz66J+LgS=9cC3CUs0RSk7dkkziV@mail.gmail.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>	<4CB7C415.6030004@dcrocker.net>	<1287115923.2914.133.camel@localhost>	<4CB83DC6.3000203@dcrocker.net>
	<a06240859c8ddf8a3516f@10.0.1.21>	<AANLkTik7j-8RdsUx3mw2vO3xSQUMg_imuxfrpmeczPpD@mail.gmail.com>	<4CB8751B.5060505@cisco.com>
	<AANLkTinyfO3Jqd+THz66J+LgS=9cC3CUs0RSk7dkkziV@mail.gmail.com>
Message-ID: <4CB89C0F.1050007@dcrocker.net>



On 10/15/2010 12:08 PM, Jorge Amodio wrote:
> I always thought email routing as end-to-end,

but it ain't.

the 'routing' that we are used to thinking about is actually only between border 
MTA's.  In a sense, the standard use of MX is like the role of BGP, which leaves 
open the question of interior routing.  (That's most often also done with 
internal MXs.)


> I don't see further forwarding by the receiving MX host as "routing".

but alas, it is exactly that.


Because quite a lot of email is between very large ISPs, there is a tendency to 
believe that modern email is only one-hop.  This misses all sorts of legitimate 
and very present scenarios, especially between enterprises.

d/

-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From dhc2 at dcrocker.net  Fri Oct 15 11:25:34 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Fri, 15 Oct 2010 14:25:34 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <20101015162401.0017E28E137@aland.bbn.com>
References: <20101015162401.0017E28E137@aland.bbn.com>
Message-ID: <4CB89C9E.4090305@dcrocker.net>



On 10/15/2010 12:24 PM, Craig Partridge wrote:
> Go re-read RFC 974

what the heck.  if we're going to do product pitches, also read:

    RFC 5598 <http://bbiw.net/recent.html#emailarch>

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From dhc2 at dcrocker.net  Fri Oct 15 11:27:53 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Fri, 15 Oct 2010 14:27:53 -0400
Subject: [ih] In the interests of historical accuracy...
In-Reply-To: <87714AC3-45E5-43D8-B516-F18803942748@gmail.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<7303AD87-6B21-465C-90F5-CE393824EDEF@tony.li>
	<87714AC3-45E5-43D8-B516-F18803942748@gmail.com>
Message-ID: <4CB89D29.2050808@dcrocker.net>



On 10/15/2010 1:27 PM, Bob Hinden wrote:
>> Jon wore first generation Teva sandals.  With black socks.
>
> I remember him wearing Birkenstock sandals.


The only time I saw Jon wear socks was with the hiking boots he wore to Montreal 
during a very cold winter.

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From vint at google.com  Fri Oct 15 11:57:22 2010
From: vint at google.com (Vint Cerf)
Date: Fri, 15 Oct 2010 14:57:22 -0400
Subject: [ih] In the interests of historical accuracy...
In-Reply-To: <4CB89D29.2050808@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<7303AD87-6B21-465C-90F5-CE393824EDEF@tony.li>
	<87714AC3-45E5-43D8-B516-F18803942748@gmail.com>
	<4CB89D29.2050808@dcrocker.net>
Message-ID: <AANLkTinkfi9ABMXNDCA+USBY1B115pvKddP4_EH0w0Dp@mail.gmail.com>

concur - open toed sandals and bare feet most of the time.

vint


On 10/15/10, Dave CROCKER <dhc2 at dcrocker.net> wrote:
>
>
> On 10/15/2010 1:27 PM, Bob Hinden wrote:
>>> Jon wore first generation Teva sandals.  With black socks.
>>
>> I remember him wearing Birkenstock sandals.
>
>
> The only time I saw Jon wear socks was with the hiking boots he wore to
> Montreal
> during a very cold winter.
>
> d/
> --
>
>    Dave Crocker
>    Brandenburg InternetWorking
>    bbiw.net
>


From sbrim at cisco.com  Fri Oct 15 12:05:41 2010
From: sbrim at cisco.com (Scott Brim)
Date: Fri, 15 Oct 2010 15:05:41 -0400
Subject: [ih] In the interests of historical accuracy...
In-Reply-To: <4CB89D29.2050808@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<7303AD87-6B21-465C-90F5-CE393824EDEF@tony.li>	<87714AC3-45E5-43D8-B516-F18803942748@gmail.com>
	<4CB89D29.2050808@dcrocker.net>
Message-ID: <4CB8A605.8040500@cisco.com>

On 10/15/2010 14:27 EDT, Dave CROCKER wrote:
> 
> 
> On 10/15/2010 1:27 PM, Bob Hinden wrote:
>>> Jon wore first generation Teva sandals.  With black socks.
>>
>> I remember him wearing Birkenstock sandals.
> 
> 
> The only time I saw Jon wear socks was with the hiking boots he wore to
> Montreal during a very cold winter.

There was the time (I heard from Ari Ollikainen, I wasn't there) where
he put on socks to get on an airplane because they wouldn't let him on
barefoot.


From jmamodio at gmail.com  Fri Oct 15 12:18:52 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Fri, 15 Oct 2010 14:18:52 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB89C9E.4090305@dcrocker.net>
References: <20101015162401.0017E28E137@aland.bbn.com>
	<4CB89C9E.4090305@dcrocker.net>
Message-ID: <AANLkTiksprqYsZ=XJu1-_Ciw-eJYZVAxO0XJREWd+O7h@mail.gmail.com>

> what the heck. ?if we're going to do product pitches, also read:

Food fight !!!

> ? RFC 5598 <http://bbiw.net/recent.html#emailarch>

That's an interesting page, thanks Dave.

-J



From vint at google.com  Fri Oct 15 13:02:41 2010
From: vint at google.com (Vint Cerf)
Date: Fri, 15 Oct 2010 16:02:41 -0400
Subject: [ih] In the interests of historical accuracy...
In-Reply-To: <4CB8A605.8040500@cisco.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<7303AD87-6B21-465C-90F5-CE393824EDEF@tony.li>
	<87714AC3-45E5-43D8-B516-F18803942748@gmail.com>
	<4CB89D29.2050808@dcrocker.net> <4CB8A605.8040500@cisco.com>
Message-ID: <AANLkTikMgkWv=rRue8nbGBno75m_qg+kPDRnKrr3nFmT@mail.gmail.com>

i think that is a true story.


On 10/15/10, Scott Brim <sbrim at cisco.com> wrote:
> On 10/15/2010 14:27 EDT, Dave CROCKER wrote:
>>
>>
>> On 10/15/2010 1:27 PM, Bob Hinden wrote:
>>>> Jon wore first generation Teva sandals.  With black socks.
>>>
>>> I remember him wearing Birkenstock sandals.
>>
>>
>> The only time I saw Jon wear socks was with the hiking boots he wore to
>> Montreal during a very cold winter.
>
> There was the time (I heard from Ari Ollikainen, I wasn't there) where
> he put on socks to get on an airplane because they wouldn't let him on
> barefoot.
>


From richard at bennett.com  Fri Oct 15 13:05:22 2010
From: richard at bennett.com (Richard Bennett)
Date: Fri, 15 Oct 2010 13:05:22 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB85FD1.4080802@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu><m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com><2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu><AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com><2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu><1287071374.2914.13.camel@localhost><alpine.WNT.2.00.1010141816050.3520@ccfloor209><4CB7C415.6030004@dcrocker.net>	<1287115923.2914.133.camel@localhost><4CB83DC6.3000203@dcrocker.net>	<a06240859c8ddf8a3516f@[10.0.1.21]>	<BC0FDDB358B94349B6F21816CAE46A40@china.huawei.com>
	<4CB85FD1.4080802@dcrocker.net>
Message-ID: <4CB8B402.2020609@bennett.com>

  Another interesting question is when the number of emails rejected by 
RBL exceeded the number that were actually carried to their ultimate 
delivery point.

On 10/15/2010 7:06 AM, Dave CROCKER wrote:
>
>
> On 10/15/2010 9:19 AM, Spencer Dawkins wrote:
>> I do wonder what year the Internet first carried more *packets per 
>> second* than
>> the current Internet carries *emails per second* - if that's clearly 
>> stated, I'm
>
> It's an interesting question no matter how it's stated.
>
> It suggests some sort of meta-analysis with different layers of 
> chunk/packet/segment/message services and making historical quantum 
> transitions of activity from one to the next.  Downward, given the way 
> you asked your question.  (Is this like moving things into hardware?)
>
>
>> thinking Jack's point is that the only real difference is scaling.
>
> Alas, some differences in degree become differences in kind.  The 
> latency and loss characteristics of messaging like email are typically 
> so different from what is usual for IP datagrams as to permit/require 
> all sorts of different design decisions.
>
> d/
>

-- 
Richard Bennett



From larrysheldon at cox.net  Fri Oct 15 13:15:18 2010
From: larrysheldon at cox.net (Larry Sheldon)
Date: Fri, 15 Oct 2010 15:15:18 -0500
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB89C0F.1050007@dcrocker.net>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>	<m2k4lle73w.wl%randy@psg.com>	<4CB6EEC6.2050107@cisco.com>	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>	<1287071374.2914.13.camel@localhost>	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>	<4CB7C415.6030004@dcrocker.net>	<1287115923.2914.133.camel@localhost>	<4CB83DC6.3000203@dcrocker.net>	<a06240859c8ddf8a3516f@10.0.1.21>	<AANLkTik7j-8RdsUx3mw2vO3xSQUMg_imuxfrpmeczPpD@mail.gmail.com>	<4CB8751B.5060505@cisco.com>	<AANLkTinyfO3Jqd+THz66J+LgS=9cC3CUs0RSk7dkkziV@mail.gmail.com>
	<4CB89C0F.1050007@dcrocker.net>
Message-ID: <4CB8B656.4020409@cox.net>

On 10/15/2010 1:23 PM, Dave CROCKER wrote:
> On 10/15/2010 12:08 PM, Jorge Amodio wrote:
>> I always thought email routing as end-to-end,
>
> but it ain't.
>
> the 'routing' that we are used to thinking about is actually only
> between border MTA's. In a sense, the standard use of MX is like the
> role of BGP, which leaves open the question of interior routing. (That's
> most often also done with internal MXs.)

We used aliases files to tell the boundary MTA where to route the 
incoming email traffic.

I believe some of the departments had further distributions.

And within the organization, we had email portability, if people 
followed the example at example.com address structure because we could 
update the control aliases file pretty quickly when people moved from 
one department to another.

And even mail you perceive as "one hop" very likely made several hops 
within ${DepartureISP}, maybe a couple through ${spamFilkterHouse} and 
more than one within ${terminatingISP} to your mbox.

If your client will show them to you, look at (and count) the Received: 
headers on messages you read.


From jklensin at gmail.com  Fri Oct 15 13:29:12 2010
From: jklensin at gmail.com (John Klensin)
Date: Fri, 15 Oct 2010 16:29:12 -0400
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <1287107599.2914.115.camel@localhost>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<6066055B-511F-4C9B-ACA8-4707E1C9175F@arin.net>
	<4CA54010-0B7A-4D05-8CEE-0CB579766C3C@cs.fiu.edu>
	<a06240846c8dd4ff6e4d0@10.0.1.21>
	<1287107599.2914.115.camel@localhost>
Message-ID: <AANLkTimqg=-yMyig2RuRHnihj=3mpvdhVO6sQCEhqnD=@mail.gmail.com>

On 10/14/10, Jack Haverty <jack at 3kitty.org> wrote:

>...
> Seems like The Internet could make a similar evolutionary step if it was
> needed....or mandated by government or customer pressures, as happened
> to the Telephone networks.

We did make such a step.  It is called the DNS, which provides
identifiers that are very portable wrt the routing structure and IP
addresses.

The mistake here is not in asserting that IP addresses should be
portable but in assuming that the telephone number concept and the IP
address concept should map to each other because they are both
user-visible and mostly numeric.  The Internet analogy to a telephone
number is a domain name.

And, if you want someone to sell you a very portable address with
either your phone number of the set of digits you would like to think
of as any other sort of portable identifier, I'm sure you would have
no problem finding a domain administrator willing to do that.

   john


From jack at 3kitty.org  Fri Oct 15 18:25:00 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Fri, 15 Oct 2010 18:25:00 -0700
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTikWeov2twuYsJsBJ87hnW33fw7LbeToV3EH1cdw@mail.gmail.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu>
	<m2k4lle73w.wl%randy@psg.com> <4CB6EEC6.2050107@cisco.com>
	<2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu>
	<AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com>
	<2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu>
	<1287071374.2914.13.camel@localhost>
	<alpine.WNT.2.00.1010141816050.3520@ccfloor209>
	<4CB7C415.6030004@dcrocker.net> <1287115923.2914.133.camel@localhost>
	<4CB83DC6.3000203@dcrocker.net> <1287157138.2914.148.camel@localhost>
	<AANLkTikWeov2twuYsJsBJ87hnW33fw7LbeToV3EH1cdw@mail.gmail.com>
Message-ID: <1287192300.2914.187.camel@localhost>

Hi Vint,

HTML5 was born somewhat after my tour of duty in W3C, so I'm not
familiar with the details.  But I think it's intent is primarily to
encode rich multi-media content - "documents" - so that they can be
passed across a wide range of user interfaces (PC screens, phones, etc.)
It's very good for adding structure to documents.

I was referring to XML as a way of conveying email *header* information,
among other things.  Conveying the email metadata in a structured way
isn't a new idea - it was embodied in RFC713 and RFC722 back in 1976,
from the sometimes pounding-head-against-a-wall experience I had in
building one of the MIT email systems. 

Having structured data enables all sorts of new user functionality -
e.g., archiving email and being able to later do very powerful searches.
The email program I wrote did this kind of stuff within its own world,
but the interactions across the net to other email systems and to the
DataComputer were too unwieldy and error-prone to do any fancy stuff in
a distributed way.  Hence the motivation for structured data in the
protocols.

So, yes, HTML5 would be the way to transport email content.   But XML or
something with similar characteristics would handle the metadata of the
headers and events (like delivery, reply, forward, addressees,
distribution lists, etc.).  Couple that with today's hardware and
database technology, and very cool things are possible.

On the other hand, most of the planet now seems enamored with typing
one-liner messages with their thumbs...

/Jack

PS We seem to have wandered pretty far away from "IPV4 Ownership", so
I'll be quiet now...

On Fri, 2010-10-15 at 12:18 -0400, Vint Cerf wrote:
> i hope HTML5 instead :-)
> 
> v
> 
> 
> On Fri, Oct 15, 2010 at 11:38 AM, Jack Haverty <jack at 3kitty.org> wrote:
> > Yep, we're on the same page.  Email portability exists now, if you just
> > get a domain name per recipient.  I suspect the DNS today couldn't
> > handle that if too many people did it.  Yes, yuch.  But a similar
> > mechanism must exist today inside the telephone world so it's proven
> > possible.
> >
> > I think eventually we'll all have portable email addresses.  Not too
> > long either.  Our addresses might just be our cell phone numbers.  Why
> > not...email and voicemail aren't all that different except at the very
> > endpoints of the path.
> >
> > Naah, I'm not mad about the encoding in 733.  But I'm happy that
> > structured data encoding has finally caught on with XML.  If those now
> > ancient mail protocols got a facelift today, I bet they'd be based on
> > XML.
> >
> > /Jack
> >
> >
> > On Fri, 2010-10-15 at 07:40 -0400, Dave CROCKER wrote:
> >>
> >> On 10/15/2010 12:12 AM, Jack Haverty wrote:
> >> > If "jack at 3kitty.org" is the example of forwarding you have in mind,
> >> > that's not how it works.  When I change providers, I move my 3kitty.org
> >> > service from one provider to another,
> >> ...
> >> > I do have a bunch of xxx at 3kitty.org mailboxes, and they all must move
> >> > together.
> >>
> >>
> >> You are emulating a version of exactly the service I described towards the end
> >> of my note, modulo the extra forwarding hop. It's key feature is that it is
> >> independent of ISPs and it does not require their cooperation.
> >>
> >> However, as you note, all of the mailboxes must move together:  granularity is
> >> at the domain name level, not the mailbox level.  To get per-user granularity,
> >> you have to encode it in the domain name, given the way email routing works.
> >>
> >> The reason you can have "direct" routing, without having to go through a
> >> forwarder is that you control the DNS MX record.  In effect that means an MX
> >> record per "customer", if not per "mailbox".  Again, that's doable today and it
> >> is done today.  The challenge is scaling that model up to a mass market.  In
> >> effect, it means an MX per user (or maybe per family).  Yuch.
> >>
> >> d/
> >>
> >> ps.  I figure you're still unhappy we didn't adopt your encoding scheme when we
> >> did RFC 733...
> >>
> >>
> >>
> >
> >
> >




From johnl at iecc.com  Sat Oct 16 21:51:39 2010
From: johnl at iecc.com (John Levine)
Date: 17 Oct 2010 04:51:39 -0000
Subject: [ih] IP addresses are not phone numbers,
	was Some Questions over IPv4 Ownership
In-Reply-To: <4CB7C9F9.4070105@bennett.com>
Message-ID: <20101017045139.33536.qmail@joyce.lan>

>  All it takes is motivation. The phone company was motivated to
>separate phone numbers from switch ports by mobility. Is there a
>similar motivation in the IP world to make a 128 bit hex number
>permanent personal property?

I suppose that if the motivation comes along with trillions of
dollars, maybe.  In the real world, no.

RFC 3482 has a well-written overview of the way that telephone number
portability works.  Anyone interested in the topic should read it
before speculating about how it works or making analogies to other
services.  Go read it now, I'll wait.

There's a couple of rather essential differences between routing phone
calls and routing IP packets.  One is that phone numbers, unlike IP
addresses still have a great deal of geographic locality, since number
allocations are tied to geography, and number portability is limited
to within local areas.  Regardless of portability, all numbers that
start with +33 are in France, and all numbers that start +1212 are in
New York City.

This means that the phone system doesn't need anything analogous to
the Internet's backbone routers that know all 300,000 randomly
allocated IP address prefixes.  All they need is a few hundred
international country codes, or a few thousand prefixes in a
portability area.

Another rather important difference is that the phone system only
needs to do one portability lookup per phone call or per SMS.  Even
the most enthusiastic teen texter sends only a few messages per
minute, as opposed to a computer which can easily send hundreds or
thousands of IP packets per second, each of which has an IP address
that needs to be routed.

Mobile phone roaming doesn't work like number portability; it's
basically call forwarding.  No matter where in the world you go, your
phone and phone number are still tied to your home carrier.  If you
and I are standing next to each other in the US, and you call my
mobile phone which has a UK phone number, the call goes from the US to
my carrier in the UK, and then back to the US.  The home carrier needs
only to remember which network a roaming user is on, and perhaps a
routing number on that network.  Roaming databases update very slowly;
even the most peripatetic traveller is unlikely to register on more
than a few networks a day.

So while it is sort of plausible that people could have portable email
addresses using DNS lookups, e.g. whatever at user.isp.com rather than
user at isp.com, there's no way that's going to happen for IP addresses.
Fortunately, since we have the DNS, it doesn't much matter what your
IP addresses are.

R's,
John


From AMaitland at Commerco.Com  Sat Oct 16 23:55:26 2010
From: AMaitland at Commerco.Com (Alan J Maitland)
Date: Sun, 17 Oct 2010 00:55:26 -0600
Subject: [ih] IP addresses are not phone numbers,
 was Some  Questions over IPv4 Ownership
In-Reply-To: <20101017045139.33536.qmail@joyce.lan>
References: <4CB7C9F9.4070105@bennett.com>
	<20101017045139.33536.qmail@joyce.lan>
Message-ID: <6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>

John,

Good points and I really enjoyed your use of the term "peripatetic traveler".

Does anyone on the list know if there is a plan to more 
geographically allocate IP addresses in V6?  It seems like doing that 
might make it easier to control routing table sizes.  In other words, 
I'm going back to a section of one of John Curran's contributions to 
this discussion earlier talking about CIRs, which seem to flow down 
from RIRs (see 
http://www.scribd.com/doc/29487289/ARIN-Contribution-to-ITU-T-IPv6-Study-Group 
).

What I gleaned from that link was that using the RIR to CIR structure 
looks a little like the hierarchical structure of DNS.  If portions 
of the V6 address space could be allocated in a way that was 
geographically aligned to the RIRs and CIRs and maybe even regional 
routes within a country by major carrier, would that not take some 
load off the primary routing resources?

Best,

Alan

At 10:51 PM 10/16/2010, you wrote:
> >  All it takes is motivation. The phone company was motivated to
> >separate phone numbers from switch ports by mobility. Is there a
> >similar motivation in the IP world to make a 128 bit hex number
> >permanent personal property?
>
>I suppose that if the motivation comes along with trillions of
>dollars, maybe.  In the real world, no.
>
>RFC 3482 has a well-written overview of the way that telephone number
>portability works.  Anyone interested in the topic should read it
>before speculating about how it works or making analogies to other
>services.  Go read it now, I'll wait.
>
>There's a couple of rather essential differences between routing phone
>calls and routing IP packets.  One is that phone numbers, unlike IP
>addresses still have a great deal of geographic locality, since number
>allocations are tied to geography, and number portability is limited
>to within local areas.  Regardless of portability, all numbers that
>start with +33 are in France, and all numbers that start +1212 are in
>New York City.
>
>This means that the phone system doesn't need anything analogous to
>the Internet's backbone routers that know all 300,000 randomly
>allocated IP address prefixes.  All they need is a few hundred
>international country codes, or a few thousand prefixes in a
>portability area.
>
>Another rather important difference is that the phone system only
>needs to do one portability lookup per phone call or per SMS.  Even
>the most enthusiastic teen texter sends only a few messages per
>minute, as opposed to a computer which can easily send hundreds or
>thousands of IP packets per second, each of which has an IP address
>that needs to be routed.
>
>Mobile phone roaming doesn't work like number portability; it's
>basically call forwarding.  No matter where in the world you go, your
>phone and phone number are still tied to your home carrier.  If you
>and I are standing next to each other in the US, and you call my
>mobile phone which has a UK phone number, the call goes from the US to
>my carrier in the UK, and then back to the US.  The home carrier needs
>only to remember which network a roaming user is on, and perhaps a
>routing number on that network.  Roaming databases update very slowly;
>even the most peripatetic traveller is unlikely to register on more
>than a few networks a day.
>
>So while it is sort of plausible that people could have portable email
>addresses using DNS lookups, e.g. whatever at user.isp.com rather than
>user at isp.com, there's no way that's going to happen for IP addresses.
>Fortunately, since we have the DNS, it doesn't much matter what your
>IP addresses are.
>
>R's,
>John




From vint at google.com  Sun Oct 17 04:24:44 2010
From: vint at google.com (Vint Cerf)
Date: Sun, 17 Oct 2010 07:24:44 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
Message-ID: <481c6802282a4977985b7f30ec614b87@mail.gmail.com>

I hope not. Allocation needs to be based on consumption rate and topology
not geography. V

----- Original Message -----
From: internet-history-bounces at postel.org
<internet-history-bounces at postel.org>
To: John Levine <johnl at iecc.com>
Cc: internet-history at postel.org <internet-history at postel.org>
Sent: Sun Oct 17 02:55:26 2010
Subject: Re: [ih] IP addresses are not phone numbers, was Some  Questions
over IPv4 Ownership

John,

Good points and I really enjoyed your use of the term "peripatetic
traveler".

Does anyone on the list know if there is a plan to more
geographically allocate IP addresses in V6?  It seems like doing that
might make it easier to control routing table sizes.  In other words,
I'm going back to a section of one of John Curran's contributions to
this discussion earlier talking about CIRs, which seem to flow down
from RIRs (see
http://www.scribd.com/doc/29487289/ARIN-Contribution-to-ITU-T-IPv6-Study-Group
).

What I gleaned from that link was that using the RIR to CIR structure
looks a little like the hierarchical structure of DNS.  If portions
of the V6 address space could be allocated in a way that was
geographically aligned to the RIRs and CIRs and maybe even regional
routes within a country by major carrier, would that not take some
load off the primary routing resources?

Best,

Alan

At 10:51 PM 10/16/2010, you wrote:
> >  All it takes is motivation. The phone company was motivated to
> >separate phone numbers from switch ports by mobility. Is there a
> >similar motivation in the IP world to make a 128 bit hex number
> >permanent personal property?
>
>I suppose that if the motivation comes along with trillions of
>dollars, maybe.  In the real world, no.
>
>RFC 3482 has a well-written overview of the way that telephone number
>portability works.  Anyone interested in the topic should read it
>before speculating about how it works or making analogies to other
>services.  Go read it now, I'll wait.
>
>There's a couple of rather essential differences between routing phone
>calls and routing IP packets.  One is that phone numbers, unlike IP
>addresses still have a great deal of geographic locality, since number
>allocations are tied to geography, and number portability is limited
>to within local areas.  Regardless of portability, all numbers that
>start with +33 are in France, and all numbers that start +1212 are in
>New York City.
>
>This means that the phone system doesn't need anything analogous to
>the Internet's backbone routers that know all 300,000 randomly
>allocated IP address prefixes.  All they need is a few hundred
>international country codes, or a few thousand prefixes in a
>portability area.
>
>Another rather important difference is that the phone system only
>needs to do one portability lookup per phone call or per SMS.  Even
>the most enthusiastic teen texter sends only a few messages per
>minute, as opposed to a computer which can easily send hundreds or
>thousands of IP packets per second, each of which has an IP address
>that needs to be routed.
>
>Mobile phone roaming doesn't work like number portability; it's
>basically call forwarding.  No matter where in the world you go, your
>phone and phone number are still tied to your home carrier.  If you
>and I are standing next to each other in the US, and you call my
>mobile phone which has a UK phone number, the call goes from the US to
>my carrier in the UK, and then back to the US.  The home carrier needs
>only to remember which network a roaming user is on, and perhaps a
>routing number on that network.  Roaming databases update very slowly;
>even the most peripatetic traveller is unlikely to register on more
>than a few networks a day.
>
>So while it is sort of plausible that people could have portable email
>addresses using DNS lookups, e.g. whatever at user.isp.com rather than
>user at isp.com, there's no way that's going to happen for IP addresses.
>Fortunately, since we have the DNS, it doesn't much matter what your
>IP addresses are.
>
>R's,
>John


From jcurran at arin.net  Sun Oct 17 04:34:20 2010
From: jcurran at arin.net (John Curran)
Date: Sun, 17 Oct 2010 07:34:20 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some  Questions over IPv4 Ownership
In-Reply-To: <6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
References: <4CB7C9F9.4070105@bennett.com>
	<20101017045139.33536.qmail@joyce.lan>
	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
Message-ID: <84911698-0FA4-4B2A-A931-4A36BF3BBBE2@arin.net>

On Oct 17, 2010, at 2:55 AM, Alan J Maitland wrote:

> John,
> 
> Good points and I really enjoyed your use of the term "peripatetic traveler".
> 
> Does anyone on the list know if there is a plan to more geographically allocate IP addresses in V6?  It seems like doing that might make it easier to control routing table sizes.  In other words, I'm going back to a section of one of John Curran's contributions to this discussion earlier talking about CIRs, which seem to flow down from RIRs (see http://www.scribd.com/doc/29487289/ARIN-Contribution-to-ITU-T-IPv6-Study-Group ).
> 
> What I gleaned from that link was that using the RIR to CIR structure looks a little like the hierarchical structure of DNS.  If portions of the V6 address space could be allocated in a way that was geographically aligned to the RIRs and CIRs and maybe even regional routes within a country by major carrier, would that not take some load off the primary routing resources?


Alan - 
 
  There have been discussions for geographic addressing in IPv6, but
  it doesn't improve routing (actually, the converse) unless the path
  of connectivity actually follows the geography.  In order words, it
  can have useful returns where the entire country has one mandatory
  ISP which everyone much utilize for their transit to the rest of the
  Internet.  In that circumstance, you can switch "local" ISPs and 
  have no impact on the global routing table since the global table
  has only a single aggregate for the entire country.

  The problem with this model is that it's completely divorced from
  typical reality, whereby numerous distinct international ISPs all 
  are vying for business in a country.  If you attempt to overlay a
  geographic addressing model, each ISP needs to announce slices of
  those geographic prefixes to the greater Internet.  This is indeed
  like what happens presently with local number portability, but one
  must remember that phone destinations are looked up via SS7 once at
  the beginning of the call, and the typical phone call lasts tens or 
  hundreds of seconds.  This yields an acceptable ratio of lookup 
  overhead in the circuit switched world. Attempting to do the same 
  in the packet network results in that lookup being performed for 
  every packet, as it arrives as every backbone router, globally.
  The imputed cost of a *single* non-hierarchical route in the global
  Internet has been estimated to be $15000 to $25000 worth of overall
  capital costs, so trying to avoid needless routes (or accurately 
  recover the associated costs) is a high priority.

/John




From dhc2 at dcrocker.net  Sun Oct 17 06:12:28 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Sun, 17 Oct 2010 09:12:28 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some  Questions over IPv4 Ownership
In-Reply-To: <84911698-0FA4-4B2A-A931-4A36BF3BBBE2@arin.net>
References: <4CB7C9F9.4070105@bennett.com>	<20101017045139.33536.qmail@joyce.lan>	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
	<84911698-0FA4-4B2A-A931-4A36BF3BBBE2@arin.net>
Message-ID: <4CBAF63C.70806@dcrocker.net>



On 10/17/2010 7:34 AM, John Curran wrote:
> There have been discussions for geographic addressing in IPv6, but
>    it doesn't improve routing (actually, the converse) unless the path
>    of connectivity actually follows the geography.

Deering conducted discussions as part of his original proposal effort (the first 
SIP) that fed into the current IPv6.  As I understand it, there was followon 
research in the area (via Lixia?).  Note that none of this surfaced all that 
publicly, which says a lot about how it progressed.

In order to have it work somewhat in the style of a locator, there would need to 
be topological constraints, with "local" internet exchanges.  This would permit 
routing to the area and then dispatching to the correct local carrier.

The existing scheme does not impose topological constraints, so this would 
constitute a massive paradigm change.

The alternative approach is to have the address operate strictly as an 
identifier, with a mapping layer down to an actual locator.  We'd get to invent 
another 'address', along with the mapping mechanism.  Whoopee...

My own view is that IP Addresses need to be treated strictly as locators and 
that the infrastructure needs to be tailored solely for that.  This is a variant 
of saying that the current Internet IP model is fine and we should not mess with 
it, except to make it more efficient at doing what it has been doing so well.


>    The problem with this model is that it's completely divorced from
>    typical reality, whereby numerous distinct international ISPs all
>    are vying for business in a country.

When Deering started his discussions, I felt that geographic addressing was 
essential, in order to buy customers freedom from ISP lock-in.

Per the above view, I now believe that the most powerful path is to get 
"identifier" functions entirely out of the IP Address, and allow IP Addresses to 
have rather late binding, to permit easier re-binding.  Freedom comes from not 
embedding lower-layer addressing information into hosts.  That puts me into the 
"move it out of the core infrastructure camp" which is not where most sympathies 
or efforts seem to lie.

That is, make host-level rendez-vous activities rely on an identifier and not a 
locator.  (Except for highly dynamic transitions, I remain a fan of having 
domain names fill this role.)  The work, then, is on dynamic registration and 
binding, and name-based applications.

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From mfidelman at meetinghouse.net  Sun Oct 17 06:38:11 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Sun, 17 Oct 2010 09:38:11 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <20101017045139.33536.qmail@joyce.lan>
References: <20101017045139.33536.qmail@joyce.lan>
Message-ID: <4CBAFC43.20002@meetinghouse.net>

John Levine wrote:
> There's a couple of rather essential differences between routing phone
> calls and routing IP packets.  One is that phone numbers, unlike IP
> addresses still have a great deal of geographic locality, since number
> allocations are tied to geography, and number portability is limited
> to within local areas.  Regardless of portability, all numbers that
> start with +33 are in France, and all numbers that start +1212 are in
> New York City.
>    
Of course things change a bit as we worry about mobile IP.  I sort of 
wonder when about how soon we'll get to the point where mobile devices 
outnumber fixed devices.  Come to think about it, it's already the case 
for both my household and business.

Miles Fidelman

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From larrysheldon at cox.net  Sun Oct 17 08:18:20 2010
From: larrysheldon at cox.net (Larry Sheldon)
Date: Sun, 17 Oct 2010 10:18:20 -0500
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <4CBAFC43.20002@meetinghouse.net>
References: <20101017045139.33536.qmail@joyce.lan>
	<4CBAFC43.20002@meetinghouse.net>
Message-ID: <4CBB13BC.1020004@cox.net>

On 10/17/2010 8:38 AM, Miles Fidelman wrote:
> John Levine wrote:
>> There's a couple of rather essential differences between routing phone
>> calls and routing IP packets. One is that phone numbers, unlike IP
>> addresses still have a great deal of geographic locality, since number
>> allocations are tied to geography, and number portability is limited
>> to within local areas. Regardless of portability, all numbers that
>> start with +33 are in France, and all numbers that start +1212 are in
>> New York City.
> Of course things change a bit as we worry about mobile IP. I sort of
> wonder when about how soon we'll get to the point where mobile devices
> outnumber fixed devices. Come to think about it, it's already the case
> for both my household and business.

Which is exactly the point I have been trying to raise--and had given up on.

We don't _know_ what will be driven by lawyers, marketers, and actual 
demand.

And we certainly don't know what technology will enable.  (Once upon a 
time telephone numbers were driven by switchboard multiples and then 
Stroweger switches with the concept of portability simply undefined).

We have several mobile devices (calling them "telephones" is as silly as 
talking about "dialing") that have one or more IP addresses.  So far we 
are surviving because they all use DHCP (and that might be the answer if 
coupled with a much faster DNS--I think the route will be via an 
advanced ARP).

I think (and will continue to think even if I am silenced) that saying 
anything (including the second law issues) is impossible guantees thqt 
you will be found to be in error someday.

I will be silent on the matter now, if I have to drop off the list.


From tony.li at tony.li  Sun Oct 17 09:43:31 2010
From: tony.li at tony.li (Tony Li)
Date: Sun, 17 Oct 2010 09:43:31 -0700
Subject: [ih] IP addresses are not phone numbers,
	was Some  Questions over IPv4 Ownership
In-Reply-To: <4CBAF63C.70806@dcrocker.net>
References: <4CB7C9F9.4070105@bennett.com>	<20101017045139.33536.qmail@joyce.lan>	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
	<84911698-0FA4-4B2A-A931-4A36BF3BBBE2@arin.net>
	<4CBAF63C.70806@dcrocker.net>
Message-ID: <A4B3CD6E-9F08-454D-9633-4BFBCBF7EAC4@tony.li>


> Per the above view, I now believe that the most powerful path is to get "identifier" functions entirely out of the IP Address, and allow IP Addresses to have rather late binding, to permit easier re-binding.  Freedom comes from not embedding lower-layer addressing information into hosts.  That puts me into the "move it out of the core infrastructure camp" which is not where most sympathies or efforts seem to lie.


This implies adding an explicit identifier to the semantics somehow.  Most contributors seem to prefer a map-and-encap architecture, where the outer 'address' only has global locator semantics, while the inner 'address' retains a local locator and identifier approach.  The strong preference for this is largely due to the opinion that we can only reasonably change the network side of the equation, and that changing host semantics at this point is intractable.

Not everyone agrees with this, but it is the popular sentiment.

Shameless plug: This discussion has been going on for about three years now over on the routing research group mailing list: rrg at irtf.org.  While we're officially nearly done with our work, there is still plenty of opportunity to influence the broader audience, and experienced network architects have been in short supply.

Regards,
Tony




From johnl at iecc.com  Sun Oct 17 10:52:44 2010
From: johnl at iecc.com (John R. Levine)
Date: 17 Oct 2010 13:52:44 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
References: <4CB7C9F9.4070105@bennett.com>
	<20101017045139.33536.qmail@joyce.lan>
	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
Message-ID: <alpine.BSF.2.00.1010171317050.52060@joyce.lan>

> Does anyone on the list know if there is a plan to more geographically 
> allocate IP addresses in V6?  It seems like doing that might make it 
> easier to control routing table sizes.

No, and it wouldn't help since in the Internet unlike the phone system 
there is little relationship between routes and national boundaries.

Route tables are a different issue.  IPv6 will make them smaller, since 
people are more likely to be able to get a single allocation that's big 
enough for the long time, but it's also easier for people who aren't 
network operators to get their own PI space.  But none of this has to do 
with whether they're geographic or not.  The most effective way to limit 
route table growth is to figure out some way to charge a small amount for 
publishing an entry, since fully half of the current table is redundant.

R's,
John


From mfidelman at meetinghouse.net  Sun Oct 17 11:04:48 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Sun, 17 Oct 2010 14:04:48 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <4CBB13BC.1020004@cox.net>
References: <20101017045139.33536.qmail@joyce.lan>	<4CBAFC43.20002@meetinghouse.net>
	<4CBB13BC.1020004@cox.net>
Message-ID: <4CBB3AC0.6000400@meetinghouse.net>

Larry Sheldon wrote:
> On 10/17/2010 8:38 AM, Miles Fidelman wrote:
>> Of course things change a bit as we worry about mobile IP. I sort of
>> wonder when about how soon we'll get to the point where mobile devices
>> outnumber fixed devices. Come to think about it, it's already the case
>> for both my household and business.
>
> Which is exactly the point I have been trying to raise--and had given 
> up on.
>
> We have several mobile devices (calling them "telephones" is as silly 
> as talking about "dialing") that have one or more IP addresses.  So 
> far we are surviving because they all use DHCP (and that might be the 
> answer if coupled with a much faster DNS--I think the route will be 
> via an advanced ARP).
DHCP isn't necessarily a bad solution - for devices that stand still 
when they're being used (e.g., laptops, under most conditions).  The 
notion of binding an IP address to physical infrastructure, and identity 
to a domain name, and using DNS to map things, seems to work pretty well.

Where it seems to be just a bit trickier is for devices that are moving 
while being used - smartphones, systems mounted in vehicles, and so 
forth.  And where things get particularly interesting is in the world of 
mesh networks, and particularly MANETs, where you don't have any carrier 
to "own" IP addresses.

Associating IP addresses with physical network assets, and "ownership" 
with the carriers who own those physical assets makes both logical and 
practical sense.  But when computers and network devices are merged, and 
there is no carrier, the question of who "owns" and administers IP 
addresses takes on a whole new flavor.

Miles Fidelman

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From mfidelman at meetinghouse.net  Sun Oct 17 12:43:39 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Sun, 17 Oct 2010 15:43:39 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <4CBB3AC0.6000400@meetinghouse.net>
References: <20101017045139.33536.qmail@joyce.lan>	<4CBAFC43.20002@meetinghouse.net>	<4CBB13BC.1020004@cox.net>
	<4CBB3AC0.6000400@meetinghouse.net>
Message-ID: <4CBB51EB.1080500@meetinghouse.net>

Miles Fidelman wrote:
> DHCP isn't necessarily a bad solution - for devices that stand still 
> when they're being used (e.g., laptops, under most conditions).  The 
> notion of binding an IP address to physical infrastructure, and 
> identity to a domain name, and using DNS to map things, seems to work 
> pretty well.
>
> Where it seems to be just a bit trickier is for devices that are 
> moving while being used - smartphones, systems mounted in vehicles, 
> and so forth.  And where things get particularly interesting is in the 
> world of mesh networks, and particularly MANETs, where you don't have 
> any carrier to "own" IP addresses.
>
> Associating IP addresses with physical network assets, and "ownership" 
> with the carriers who own those physical assets makes both logical and 
> practical sense.  But when computers and network devices are merged, 
> and there is no carrier, the question of who "owns" and administers IP 
> addresses takes on a whole new flavor.
I guess I should also add the case of roaming, where one wants to 
maintain TCP sessions while handing off link-level connectivity across 
assets owned by multiple carriers.

Miles

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From tony.li at tony.li  Sun Oct 17 13:40:36 2010
From: tony.li at tony.li (Tony Li)
Date: Sun, 17 Oct 2010 13:40:36 -0700
Subject: [ih] IP addresses are not phone numbers,
	was Some Questions over IPv4 Ownership
In-Reply-To: <alpine.BSF.2.00.1010171317050.52060@joyce.lan>
References: <4CB7C9F9.4070105@bennett.com>
	<20101017045139.33536.qmail@joyce.lan>
	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
	<alpine.BSF.2.00.1010171317050.52060@joyce.lan>
Message-ID: <F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>


On Oct 17, 2010, at 10:52 AM, John R. Levine wrote:

>> Does anyone on the list know if there is a plan to more geographically allocate IP addresses in V6?  It seems like doing that might make it easier to control routing table sizes.
> 
> No, and it wouldn't help since in the Internet unlike the phone system there is little relationship between routes and national boundaries.


There is a theory that as we allocate major blocks from the RIRs, that we will be able to do some proxy aggregation at the continental level and would recover a significant amount of routing table space.  However, as with any proxy aggregation approach, the interaction with traffic engineering and the mechanism for effective deployment is an unresolved issue.

One thought that has been in the hallways is establishing a routing arbiter function, analogous to IANA that would help coordinate where proxy aggregation would occur.


> Route tables are a different issue.  IPv6 will make them smaller, since people are more likely to be able to get a single allocation that's big enough for the long time, but it's also easier for people who aren't network operators to get their own PI space.  


To date, the number of stub ASes showing up in the routing table suggests that the cost of the pervasive PI allocation far exceeds the benefits of a single prefix for growing ISPs.

Tony




From ernesto at cs.fiu.edu  Sun Oct 17 14:21:52 2010
From: ernesto at cs.fiu.edu (Ernie Rubi)
Date: Sun, 17 Oct 2010 17:21:52 -0400
Subject: [ih] IP addresses are not phone numbers,
	was Some Questions over IPv4 Ownership
In-Reply-To: <F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>
References: <4CB7C9F9.4070105@bennett.com>
	<20101017045139.33536.qmail@joyce.lan>
	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
	<alpine.BSF.2.00.1010171317050.52060@joyce.lan>
	<F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>
Message-ID: <DBA46DE4-E9BC-4E59-AA05-DC429FC5C81E@cs.fiu.edu>

Thats an interesting thought, but under what legal authority? Or just a community effort with no real enforcement bite?

On Oct 17, 2010, at 4:40 PM, Tony Li <tony.li at tony.li> wrote:

> One thought that has been in the hallways is establishing a routing arbiter function, analogous to IANA that would help coordinate where proxy aggregation would occur.



From tony.li at tony.li  Sun Oct 17 15:33:39 2010
From: tony.li at tony.li (Tony Li)
Date: Sun, 17 Oct 2010 15:33:39 -0700
Subject: [ih] IP addresses are not phone numbers,
	was Some Questions over IPv4 Ownership
In-Reply-To: <DBA46DE4-E9BC-4E59-AA05-DC429FC5C81E@cs.fiu.edu>
References: <4CB7C9F9.4070105@bennett.com>
	<20101017045139.33536.qmail@joyce.lan>
	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
	<alpine.BSF.2.00.1010171317050.52060@joyce.lan>
	<F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>
	<DBA46DE4-E9BC-4E59-AA05-DC429FC5C81E@cs.fiu.edu>
Message-ID: <6F3CC2DF-9F13-432A-8F86-CA493B4E4CE8@tony.li>


On Oct 17, 2010, at 2:21 PM, Ernie Rubi wrote:

> Thats an interesting thought, but under what legal authority? 


I'm no lawyer, but my understanding is that the routing architecture is under the purview of ISOC, just as the namespace is.  So it would be essentially analogous to ICANN.

Tony




From sbrim at cisco.com  Sun Oct 17 16:15:22 2010
From: sbrim at cisco.com (Scott Brim)
Date: Sun, 17 Oct 2010 16:15:22 -0700
Subject: [ih] IP addresses are not phone numbers,
 was Some  Questions over IPv4 Ownership
In-Reply-To: <4CBAF63C.70806@dcrocker.net>
References: <4CB7C9F9.4070105@bennett.com>	<20101017045139.33536.qmail@joyce.lan>	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>	<84911698-0FA4-4B2A-A931-4A36BF3BBBE2@arin.net>
	<4CBAF63C.70806@dcrocker.net>
Message-ID: <4CBB838A.60805@cisco.com>

On 10/17/2010 06:12 PDT, Dave CROCKER wrote:
> 
> 
> On 10/17/2010 7:34 AM, John Curran wrote:
>> There have been discussions for geographic addressing in IPv6, but
>>    it doesn't improve routing (actually, the converse) unless the path
>>    of connectivity actually follows the geography.
> 
> Deering conducted discussions as part of his original proposal effort
> (the first SIP) that fed into the current IPv6.  As I understand it,
> there was followon research in the area (via Lixia?).  Note that none of
> this surfaced all that publicly, which says a lot about how it progressed.

Tony Hain probably carried it the furthest and most recently.

> My own view is that IP Addresses need to be treated strictly as locators
> and that the infrastructure needs to be tailored solely for that.  This
> is a variant of saying that the current Internet IP model is fine and we
> should not mess with it, except to make it more efficient at doing what
> it has been doing so well.

Well, yes, the current IP model is fine but that doesn't mean that all
routing/forwarding must be restricted to it.  There should at least be a
topological space with aggregatable locators that are globally unique
and globally routed like the current Internet.  That doesn't mean that
there can't be other topological spaces with locators that are not
globally routed (e.g. LISP classic).

> When Deering started his discussions, I felt that geographic addressing
> was essential, in order to buy customers freedom from ISP lock-in.

Joel Halpern and I once had a nice long walk-and-talk on the beach in
Waikiki where I became convinced that while metro-based addressing is
better when everything is simple, it becomes "grubbier" very fast as you
add policy, operational exceptions, etc. to complicate things.

> Per the above view, I now believe that the most powerful path is to get
> "identifier" functions entirely out of the IP Address

Absolutely.  The fundamental problem in locator/identifier separation
has nothing to do with location.  Rather, it's to wean identification
functions away from depending on topology-dependent inputs.  Regardless
of where or at what layer those identification functions are placed.

Scott


From sbrim at cisco.com  Sun Oct 17 16:26:31 2010
From: sbrim at cisco.com (Scott Brim)
Date: Sun, 17 Oct 2010 16:26:31 -0700
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <4CBAFC43.20002@meetinghouse.net>
References: <20101017045139.33536.qmail@joyce.lan>
	<4CBAFC43.20002@meetinghouse.net>
Message-ID: <4CBB8627.1080202@cisco.com>

On 10/17/2010 06:38 PDT, Miles Fidelman wrote:
> Of course things change a bit as we worry about mobile IP.  I sort of
> wonder when about how soon we'll get to the point where mobile devices
> outnumber fixed devices.  Come to think about it, it's already the case
> for both my household and business.

I've heard a forecast of 2013.  I think it will be sooner.  Please
include "M2M" devices such as sensors and vending machines that are
connected via the cellular network because it's easier to do mass
deployments that way.

On 10/17/2010 08:18 PDT, Larry Sheldon wrote:
> And we certainly don't know what technology will enable.  

Not only is the future stranger than you imagine, it is stranger than
you _can_ imagine :-)

> We have several mobile devices (calling them "telephones" is as silly as
> talking about "dialing") 

I like "mobiles" as in "Internet mobiles".

On 10/17/2010 11:04 PDT, Miles Fidelman wrote:
> DHCP isn't necessarily a bad solution - for devices that stand still
> when they're being used (e.g., laptops, under most conditions).  The
> notion of binding an IP address to physical infrastructure, and identity
> to a domain name, and using DNS to map things, seems to work pretty well.

See several drafts and RFCs, including
<http://tools.ietf.org/html/draft-brim-mobility-and-privacy>, for
concerns about using public mapping systems like DNS to allow mobility.

On 10/17/2010 12:43 PDT, Miles Fidelman wrote:
> I guess I should also add the case of roaming, where one wants to
> maintain TCP sessions while handing off link-level connectivity across
> assets owned by multiple carriers.

When multiple carriers are involved, why does the hand-off have to be at
(so-called) link-level?

Thanks ... Scott


From mfidelman at meetinghouse.net  Sun Oct 17 17:58:39 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Sun, 17 Oct 2010 20:58:39 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <4CBB8627.1080202@cisco.com>
References: <20101017045139.33536.qmail@joyce.lan>
	<4CBAFC43.20002@meetinghouse.net> <4CBB8627.1080202@cisco.com>
Message-ID: <4CBB9BBF.7090804@meetinghouse.net>

Scott Brim wrote:
> On 10/17/2010 12:43 PDT, Miles Fidelman wrote:
>    
>> I guess I should also add the case of roaming, where one wants to
>> maintain TCP sessions while handing off link-level connectivity across
>> assets owned by multiple carriers.
>>      
> When multiple carriers are involved, why does the hand-off have to be at
> (so-called) link-level?
>    
Well... for current software to work right, maintaining a TCP connection 
requires that the underlying IP addresses remain the same (at least I 
think it does).  So somehow the IP numbers have to stay the same, while 
the underlying network connection changes.

It's not necessarily a link-layer function to do the handoff, but what 
is actually changing is happening at the link layer - one physical link 
is dropping, and a new one is coming up.

The hand-off is really a routing function - changing the mapping between 
a specific IP address and a specific physical link.  That shouldn't be 
too hard to do if one carrier is handing things from one asset to 
another, seems a bit more difficult when two carriers are involved - and 
brings us back to the original question of who "owns" the IP address 
when more than one carrier is involved. :-)

Miles



-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From sbrim at cisco.com  Sun Oct 17 18:30:51 2010
From: sbrim at cisco.com (Scott Brim)
Date: Sun, 17 Oct 2010 18:30:51 -0700
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <4CBB9BBF.7090804@meetinghouse.net>
References: <20101017045139.33536.qmail@joyce.lan>	<4CBAFC43.20002@meetinghouse.net>
	<4CBB8627.1080202@cisco.com> <4CBB9BBF.7090804@meetinghouse.net>
Message-ID: <4CBBA34B.90301@cisco.com>

On 10/17/2010 17:58 PDT, Miles Fidelman wrote:
> Scott Brim wrote:
>> On 10/17/2010 12:43 PDT, Miles Fidelman wrote:
>>   
>>> I guess I should also add the case of roaming, where one wants to
>>> maintain TCP sessions while handing off link-level connectivity across
>>> assets owned by multiple carriers.
>>>      
>> When multiple carriers are involved, why does the hand-off have to be at
>> (so-called) link-level?
>>    
> Well... for current software to work right, maintaining a TCP connection
> requires that the underlying IP addresses remain the same (at least I
> think it does).  So somehow the IP numbers have to stay the same, while
> the underlying network connection changes.
> 
> It's not necessarily a link-layer function to do the handoff, but what
> is actually changing is happening at the link layer - one physical link
> is dropping, and a new one is coming up.
> 
> The hand-off is really a routing function - changing the mapping between
> a specific IP address and a specific physical link.  That shouldn't be
> too hard to do if one carrier is handing things from one asset to
> another, seems a bit more difficult when two carriers are involved - and
> brings us back to the original question of who "owns" the IP address
> when more than one carrier is involved. :-)

Got it.  You're trying to work with current software.  That can be done
below the L3/L4 monolith, or above it (in the app layer) and of course
we're trying to break the monolith apart.  Every mechanism has areas
where it's useful.


From mfidelman at meetinghouse.net  Sun Oct 17 18:45:08 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Sun, 17 Oct 2010 21:45:08 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <4CBBA34B.90301@cisco.com>
References: <20101017045139.33536.qmail@joyce.lan>	<4CBAFC43.20002@meetinghouse.net>
	<4CBB8627.1080202@cisco.com>
	<4CBB9BBF.7090804@meetinghouse.net> <4CBBA34B.90301@cisco.com>
Message-ID: <4CBBA6A4.1000506@meetinghouse.net>

Scott Brim wrote:
> On 10/17/2010 17:58 PDT, Miles Fidelman wrote:
>    
>> Well... for current software to work right, maintaining a TCP connection
>> requires that the underlying IP addresses remain the same (at least I
>> think it does).  So somehow the IP numbers have to stay the same, while
>> the underlying network connection changes.
>>
>> It's not necessarily a link-layer function to do the handoff, but what
>> is actually changing is happening at the link layer - one physical link
>> is dropping, and a new one is coming up.
>>
>> The hand-off is really a routing function - changing the mapping between
>> a specific IP address and a specific physical link.  That shouldn't be
>> too hard to do if one carrier is handing things from one asset to
>> another, seems a bit more difficult when two carriers are involved - and
>> brings us back to the original question of who "owns" the IP address
>> when more than one carrier is involved. :-)
>>      
> Got it.  You're trying to work with current software.  That can be done
> below the L3/L4 monolith, or above it (in the app layer) and of course
> we're trying to break the monolith apart.  Every mechanism has areas
> where it's useful.
>    
I've always been a big fan of layering - it's been instrumental in 
enabling the rather rapid evolution we've seen at all layers of the 
Internet.  I expect that we'll make more progress by localizing changes 
in ways that don't break the layers that work, and don't need to 
change.  I'd really hate to see anything that resembles the (IMHO) 
abortion we see in W3C-style web services, where all kinds of lower 
layer functions have been pushed to somewhere above HTTP.

Cheers,

Miles

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From adrian at creative.net.au  Sun Oct 17 19:50:44 2010
From: adrian at creative.net.au (Adrian Chadd)
Date: Mon, 18 Oct 2010 10:50:44 +0800
Subject: [ih] IP addresses are not phone numbers,
	was Some Questions over IPv4 Ownership
In-Reply-To: <4CBB9BBF.7090804@meetinghouse.net>
References: <20101017045139.33536.qmail@joyce.lan>
	<4CBAFC43.20002@meetinghouse.net> <4CBB8627.1080202@cisco.com>
	<4CBB9BBF.7090804@meetinghouse.net>
Message-ID: <20101018025044.GA3573@skywalker.creative.net.au>

On Sun, Oct 17, 2010, Miles Fidelman wrote:

> It's not necessarily a link-layer function to do the handoff, but what 
> is actually changing is happening at the link layer - one physical link 
> is dropping, and a new one is coming up.
> 
> The hand-off is really a routing function - changing the mapping between 
> a specific IP address and a specific physical link.  That shouldn't be 
> too hard to do if one carrier is handing things from one asset to 
> another, seems a bit more difficult when two carriers are involved - and 
> brings us back to the original question of who "owns" the IP address 
> when more than one carrier is involved. :-)

I've had people claim "websockets" is the future, where all that complicated
stuff is hidden behind a HTTP gateway. Your address can change but as your
sockets are all over HTTP, none of your code needs to know.

Of course, websockets has its fair share of problems (incl. but not limited
to a previous draft wanting to speak HTTP enough to fool a proxy, but not
speak HTTP - making proxying it through a hierarchy of caches rather
unreliable. Sigh. When will people learn. :-)



Adrian



From dhc2 at dcrocker.net  Sun Oct 17 22:16:07 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Mon, 18 Oct 2010 01:16:07 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <6F3CC2DF-9F13-432A-8F86-CA493B4E4CE8@tony.li>
References: <4CB7C9F9.4070105@bennett.com>	<20101017045139.33536.qmail@joyce.lan>	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>	<alpine.BSF.2.00.1010171317050.52060@joyce.lan>	<F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>	<DBA46DE4-E9BC-4E59-AA05-DC429FC5C81E@cs.fiu.edu>
	<6F3CC2DF-9F13-432A-8F86-CA493B4E4CE8@tony.li>
Message-ID: <4CBBD817.7010603@dcrocker.net>



On 10/17/2010 6:33 PM, Tony Li wrote:
> routing architecture is under the purview of ISOC


ISOC ???

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From tony.li at tony.li  Sun Oct 17 22:31:11 2010
From: tony.li at tony.li (Tony Li)
Date: Sun, 17 Oct 2010 22:31:11 -0700
Subject: [ih] IP addresses are not phone numbers,
	was Some Questions over IPv4 Ownership
In-Reply-To: <4CBBD817.7010603@dcrocker.net>
References: <4CB7C9F9.4070105@bennett.com>	<20101017045139.33536.qmail@joyce.lan>	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>	<alpine.BSF.2.00.1010171317050.52060@joyce.lan>	<F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>	<DBA46DE4-E9BC-4E59-AA05-DC429FC5C81E@cs.fiu.edu>
	<6F3CC2DF-9F13-432A-8F86-CA493B4E4CE8@tony.li>
	<4CBBD817.7010603@dcrocker.net>
Message-ID: <8A6C2979-CF69-4477-A3E3-A9B7B9D81590@tony.li>


On Oct 17, 2010, at 10:16 PM, Dave CROCKER wrote:

> On 10/17/2010 6:33 PM, Tony Li wrote:
>> routing architecture is under the purview of ISOC
> 
> ISOC ???


Am I incorrect?  The documents that I can find claim that routing and addressing are both architecturally under the IETF.  As that's not a legal entity, the relevant parent is ISOC.

>From there, the addressing portion is delegated to ICANN (specifically IANA), in turn to NRO, and in turn to RIRs.

Please see draft-azinger-scalable-addressing-00 (Sections 2.1 and 2.2) for my understanding and references.

If there's something factually incorrect, please educate me.

Tony




From dhc2 at dcrocker.net  Sun Oct 17 22:43:32 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Mon, 18 Oct 2010 01:43:32 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <8A6C2979-CF69-4477-A3E3-A9B7B9D81590@tony.li>
References: <4CB7C9F9.4070105@bennett.com>	<20101017045139.33536.qmail@joyce.lan>	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>	<alpine.BSF.2.00.1010171317050.52060@joyce.lan>	<F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>	<DBA46DE4-E9BC-4E59-AA05-DC429FC5C81E@cs.fiu.edu>
	<6F3CC2DF-9F13-432A-8F86-CA493B4E4CE8@tony.li>
	<4CBBD817.7010603@dcrocker.net>
	<8A6C2979-CF69-4477-A3E3-A9B7B9D81590@tony.li>
Message-ID: <4CBBDE84.3070101@dcrocker.net>



On 10/18/2010 1:31 AM, Tony Li wrote:
> Am I incorrect?  The documents that I can find claim that routing and
> addressing are both architecturally under the IETF.  As that's not a legal
> entity, the relevant parent is ISOC.


Ah, ok.  The 'isoc has authority over the ietf' model.

I think you meant that the IETF is not incorporated.  As I understand it, there 
are a number of unincorporated legal entities (with the most familiar being 
Doing Business As -- DBA.)  I'm told that they do have legal standing.  (One of 
the discussions on this topic included an attorney telling us all what the legal 
term was for an original, IETF-styled arrangement but I forget what it was.

Of course, now that there is the IETF Trust, even the IETF has an incorporated 
component (albeit only designed for handling IPR.)

ISOC's oversight is extremely constrained and pertains strictly to process -- 
although their approval of an IAB Nomcom slate does sometimes get rather more 
substantive.

I believe administrative delegations come from IETF standards.  I had not heard 
that ISOC got into the middle of such processes.

d/


-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From vint at google.com  Mon Oct 18 00:09:21 2010
From: vint at google.com (Vint Cerf)
Date: Mon, 18 Oct 2010 03:09:21 -0400
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <4CBBDE84.3070101@dcrocker.net>
References: <4CB7C9F9.4070105@bennett.com>
	<20101017045139.33536.qmail@joyce.lan>
	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
	<alpine.BSF.2.00.1010171317050.52060@joyce.lan>
	<F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>
	<DBA46DE4-E9BC-4E59-AA05-DC429FC5C81E@cs.fiu.edu>
	<6F3CC2DF-9F13-432A-8F86-CA493B4E4CE8@tony.li>
	<4CBBD817.7010603@dcrocker.net>
	<8A6C2979-CF69-4477-A3E3-A9B7B9D81590@tony.li>
	<4CBBDE84.3070101@dcrocker.net>
Message-ID: <AANLkTinBVr5vRtYJVAMjCFZzk-4S8s0_u=PaXkjRVk1n@mail.gmail.com>

ISOC does not get into the middle of allocations and assignments -
that's officially the role of IETF and sometimes IAB, both of which
function under the auspices of but not the direction of ISOC.

vint



On Mon, Oct 18, 2010 at 1:43 AM, Dave CROCKER <dhc2 at dcrocker.net> wrote:
>
>
> On 10/18/2010 1:31 AM, Tony Li wrote:
>>
>> Am I incorrect? ?The documents that I can find claim that routing and
>> addressing are both architecturally under the IETF. ?As that's not a legal
>> entity, the relevant parent is ISOC.
>
>
> Ah, ok. ?The 'isoc has authority over the ietf' model.
>
> I think you meant that the IETF is not incorporated. ?As I understand it,
> there are a number of unincorporated legal entities (with the most familiar
> being Doing Business As -- DBA.) ?I'm told that they do have legal standing.
> ?(One of the discussions on this topic included an attorney telling us all
> what the legal term was for an original, IETF-styled arrangement but I
> forget what it was.
>
> Of course, now that there is the IETF Trust, even the IETF has an
> incorporated component (albeit only designed for handling IPR.)
>
> ISOC's oversight is extremely constrained and pertains strictly to process
> -- although their approval of an IAB Nomcom slate does sometimes get rather
> more substantive.
>
> I believe administrative delegations come from IETF standards. ?I had not
> heard that ISOC got into the middle of such processes.
>
> d/
>
>
> --
>
> ?Dave Crocker
> ?Brandenburg InternetWorking
> ?bbiw.net
>



From dot at dotat.at  Mon Oct 18 06:17:01 2010
From: dot at dotat.at (Tony Finch)
Date: Mon, 18 Oct 2010 14:17:01 +0100
Subject: [ih] Some Questions over IPv4 Ownership
In-Reply-To: <4CB8B402.2020609@bennett.com>
References: <20101013223948.C983B6BE57A@mercury.lcs.mit.edu><m2k4lle73w.wl%randy@psg.com>
	<4CB6EEC6.2050107@cisco.com><2a9f1fa91891ee11a7cf49be4c2eff09.squirrel@webmail.cs.fiu.edu><AANLkTi=5nZwv4E_9Wt1L137bbxymRaGjMjSHsOSiSfh6@mail.gmail.com><2eba43e02e1bae29115f7d684ae2c7d3.squirrel@webmail.cs.fiu.edu><1287071374.2914.13.camel@localhost><alpine.WNT.2.00.1010141816050.3520@ccfloor209><4CB7C415.6030004@dcrocker.net>
	<1287115923.2914.133.camel@localhost><4CB83DC6.3000203@dcrocker.net>
	<a06240859c8ddf8a3516f@[10.0.1.21]>
	<BC0FDDB358B94349B6F21816CAE46A40@china.huawei.com>
	<4CB85FD1.4080802@dcrocker.net> <4CB8B402.2020609@bennett.com>
Message-ID: <alpine.LSU.2.00.1010181416300.29333@hermes-2.csi.cam.ac.uk>

On Fri, 15 Oct 2010, Richard Bennett wrote:

> Another interesting question is when the number of emails rejected by RBL
> exceeded the number that were actually carried to their ultimate delivery
> point.

That's been about 10:1 rejected:delivered for years now.

Tony.
-- 
f.anthony.n.finch  <dot at dotat.at>  http://dotat.at/
HUMBER THAMES DOVER WIGHT PORTLAND: NORTH BACKING WEST OR NORTHWEST, 5 TO 7,
DECREASING 4 OR 5, OCCASIONALLY 6 LATER IN HUMBER AND THAMES. MODERATE OR
ROUGH. RAIN THEN FAIR. GOOD.


From jmamodio at gmail.com  Tue Oct 19 05:49:00 2010
From: jmamodio at gmail.com (Jorge Amodio)
Date: Tue, 19 Oct 2010 07:49:00 -0500
Subject: [ih] IP addresses are not phone numbers,
 was Some Questions over IPv4 Ownership
In-Reply-To: <AANLkTinBVr5vRtYJVAMjCFZzk-4S8s0_u=PaXkjRVk1n@mail.gmail.com>
References: <4CB7C9F9.4070105@bennett.com>
	<20101017045139.33536.qmail@joyce.lan>
	<6.2.3.4.2.20101017002516.0400bdf8@MS1.MailSys.Net>
	<alpine.BSF.2.00.1010171317050.52060@joyce.lan>
	<F212ADFA-A72F-47AE-85FF-7B5F7A6E1503@tony.li>
	<DBA46DE4-E9BC-4E59-AA05-DC429FC5C81E@cs.fiu.edu>
	<6F3CC2DF-9F13-432A-8F86-CA493B4E4CE8@tony.li>
	<4CBBD817.7010603@dcrocker.net>
	<8A6C2979-CF69-4477-A3E3-A9B7B9D81590@tony.li>
	<4CBBDE84.3070101@dcrocker.net>
	<AANLkTinBVr5vRtYJVAMjCFZzk-4S8s0_u=PaXkjRVk1n@mail.gmail.com>
Message-ID: <AANLkTi=hUK7H4N0Yp6YM0F0wHGqXsfdcuWkVpR-gwPB=@mail.gmail.com>

It seems that after all somebody decided to confess ...

http://gcn.com/articles/2010/10/18/cybereye-box-vint-cerf-blame-for-ipv6.aspx

Hey Vint, since when you became "Godfather" ? it makes me feel that we
are sort of members of the TCP/IP MOB :-).

William it is "Vint" not "Vinnie" ...

Cheers
Jorge


From braden at isi.edu  Wed Oct 20 11:15:55 2010
From: braden at isi.edu (Bob Braden)
Date: Wed, 20 Oct 2010 11:15:55 -0700
Subject: [ih] internet-history Digest, Vol 47, Issue 26
In-Reply-To: <mailman.9943.1287366342.753.internet-history@postel.org>
References: <mailman.9943.1287366342.753.internet-history@postel.org>
Message-ID: <4CBF31DB.6060802@isi.edu>


This discussion, although it is not without interest, seems to have 
little to do with the primary purpose of this list, Internet history.

Bob Braden

>
>
> Today's Topics:
>
>     1. Re: IP addresses are not phone numbers, was Some Questions
>        over IPv4 Ownership (Miles Fidelman)
>     2. Re: IP addresses are not phone numbers,	was Some Questions
>        over IPv4 Ownership (Tony Li)
>     3. Re: IP addresses are not phone numbers,	was Some Questions
>        over IPv4 Ownership (Ernie Rubi)
>     4. Re: IP addresses are not phone numbers,	was Some Questions
>        over IPv4 Ownership (Tony Li)
>     5. Re: IP addresses are not phone numbers, was Some  Questions
>        over IPv4 Ownership (Scott Brim)
>     6. Re: IP addresses are not phone numbers, was Some Questions
>        over IPv4 Ownership (Scott Brim)
>     7. Re: IP addresses are not phone numbers, was Some Questions
>        over IPv4 Ownership (Miles Fidelman)
>     8. Re: IP addresses are not phone numbers, was Some Questions
>        over IPv4 Ownership (Scott Brim)
>     9. Re: IP addresses are not phone numbers, was Some Questions
>        over IPv4 Ownership (Miles Fidelman)
>
>
> ----------------------------------------------------------------------
>


From mfidelman at meetinghouse.net  Wed Oct 20 11:55:47 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Wed, 20 Oct 2010 14:55:47 -0400
Subject: [ih] internet-history Digest, Vol 47, Issue 26
In-Reply-To: <4CBF31DB.6060802@isi.edu>
References: <mailman.9943.1287366342.753.internet-history@postel.org>
	<4CBF31DB.6060802@isi.edu>
Message-ID: <4CBF3B33.8050305@meetinghouse.net>

Future History?

Bob Braden wrote:
>
> This discussion, although it is not without interest, seems to have 
> little to do with the primary purpose of this list, Internet history.
>
> Bob Braden
>
>>
>>
>> Today's Topics:
>>
>>     1. Re: IP addresses are not phone numbers, was Some Questions
>>        over IPv4 Ownership (Miles Fidelman)
>>     2. Re: IP addresses are not phone numbers,    was Some Questions
>>        over IPv4 Ownership (Tony Li)
>>     3. Re: IP addresses are not phone numbers,    was Some Questions
>>        over IPv4 Ownership (Ernie Rubi)
>>     4. Re: IP addresses are not phone numbers,    was Some Questions
>>        over IPv4 Ownership (Tony Li)
>>     5. Re: IP addresses are not phone numbers, was Some  Questions
>>        over IPv4 Ownership (Scott Brim)
>>     6. Re: IP addresses are not phone numbers, was Some Questions
>>        over IPv4 Ownership (Scott Brim)
>>     7. Re: IP addresses are not phone numbers, was Some Questions
>>        over IPv4 Ownership (Miles Fidelman)
>>     8. Re: IP addresses are not phone numbers, was Some Questions
>>        over IPv4 Ownership (Scott Brim)
>>     9. Re: IP addresses are not phone numbers, was Some Questions
>>        over IPv4 Ownership (Miles Fidelman)
>>
>>
>> ----------------------------------------------------------------------
>>


-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From mfidelman at meetinghouse.net  Thu Oct 28 05:19:26 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 08:19:26 -0400
Subject: [ih] birth of the Internet?
Message-ID: <4CC96A4E.2070704@meetinghouse.net>

Hi Folks,

I've been part of an interesting discussion, on another list, that's led 
to the question of when the Internet was actually "born."

Was it:
- when the ARPANET moved from NCP to TCP/IP?
- when the ARPANET was split into two, linked backbones (ARPANET/MILNET)?
- was there an earlier point at which IP was linking LANs across the 
ARPANET?
- some other event?

Miles Fidelman

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From vint at google.com  Thu Oct 28 05:44:23 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 08:44:23 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC96A4E.2070704@meetinghouse.net>
References: <4CC96A4E.2070704@meetinghouse.net>
Message-ID: <AANLkTimXc5_H1jetdWuNft4pJsC-xu78wLe9P6Pv-U-r@mail.gmail.com>

1. initial design: Spring-Fall 1973
2. first published paper: May 1974
3. first full spec of TCP: December 1974 (RFC 675)
4. More complete specs after splitting off of IP: RFC791, 792, 793 - 1981
5. official launch of Internet: January 1, 1983 (which also coincided
with the creation of MILNET split from former research ARPANET) and
the incorporation of the Packet Satellite and Packet Radio Net(s) into
the DARPA Internet. this was also the date of replacement of NCP with
TCP/IP
6. As I recall, there were tests well before 1983 in which the Xerox
PARC ethernet was linked to Internet by way of a packet radio at PARC
connecting via the SRI International operated Packet Radio network.
Larry Masinter's name comes to mind as having had something to do with
this but others on the list may have clearer recollections. The first
Packet Radio/ARPANET tests of TCP/IP occurred in the 1975 time frame
using a BBN developed gateway. The first three-network tests were done
on November 22, 1977 linking ARPANET, the SF Bay Area Packet Radio net
and the Packet Satellite Network.

vint





On Thu, Oct 28, 2010 at 8:19 AM, Miles Fidelman
<mfidelman at meetinghouse.net> wrote:
> Hi Folks,
>
> I've been part of an interesting discussion, on another list, that's led to
> the question of when the Internet was actually "born."
>
> Was it:
> - when the ARPANET moved from NCP to TCP/IP?
> - when the ARPANET was split into two, linked backbones (ARPANET/MILNET)?
> - was there an earlier point at which IP was linking LANs across the
> ARPANET?
> - some other event?
>
> Miles Fidelman
>
> --
> In theory, there is no difference between theory and practice.
> In<fnord> ?practice, there is. ? .... Yogi Berra
>
>
>



From mfidelman at meetinghouse.net  Thu Oct 28 05:48:54 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 08:48:54 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <AANLkTimXc5_H1jetdWuNft4pJsC-xu78wLe9P6Pv-U-r@mail.gmail.com>
References: <4CC96A4E.2070704@meetinghouse.net>
	<AANLkTimXc5_H1jetdWuNft4pJsC-xu78wLe9P6Pv-U-r@mail.gmail.com>
Message-ID: <4CC97136.8060909@meetinghouse.net>

Thanks Vint,

Always helpful to have the official word from the man who was paying the 
bills at the time.

Miles

Vint Cerf wrote:
> 1. initial design: Spring-Fall 1973
> 2. first published paper: May 1974
> 3. first full spec of TCP: December 1974 (RFC 675)
> 4. More complete specs after splitting off of IP: RFC791, 792, 793 - 1981
> 5. official launch of Internet: January 1, 1983 (which also coincided
> with the creation of MILNET split from former research ARPANET) and
> the incorporation of the Packet Satellite and Packet Radio Net(s) into
> the DARPA Internet. this was also the date of replacement of NCP with
> TCP/IP
> 6. As I recall, there were tests well before 1983 in which the Xerox
> PARC ethernet was linked to Internet by way of a packet radio at PARC
> connecting via the SRI International operated Packet Radio network.
> Larry Masinter's name comes to mind as having had something to do with
> this but others on the list may have clearer recollections. The first
> Packet Radio/ARPANET tests of TCP/IP occurred in the 1975 time frame
> using a BBN developed gateway. The first three-network tests were done
> on November 22, 1977 linking ARPANET, the SF Bay Area Packet Radio net
> and the Packet Satellite Network.
>
> vint
>
>    

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From jcurran at istaff.org  Thu Oct 28 05:53:32 2010
From: jcurran at istaff.org (John Curran)
Date: Thu, 28 Oct 2010 08:53:32 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC96A4E.2070704@meetinghouse.net>
References: <4CC96A4E.2070704@meetinghouse.net>
Message-ID: <A2151671-FB23-415C-B1B1-31B7CA8DC0E7@istaff.org>

On Oct 28, 2010, at 8:19 AM, Miles Fidelman wrote:

> Hi Folks,
> 
> I've been part of an interesting discussion, on another list, that's led to the question of when the Internet was actually "born."
> 
> Was it:
> - when the ARPANET moved from NCP to TCP/IP?
> - when the ARPANET was split into two, linked backbones (ARPANET/MILNET)?
> - was there an earlier point at which IP was linking LANs across the ARPANET?
> - some other event?

Best to determine the criteria in the abstract, and then apply the 
criteria to the above situations to determine the earliest match...

If one considers the Internet to be "a global system of interconnected 
computer networks that use the standard Internet Protocol Suite (TCP/IP)",
then one would expect to have: 1) multiple interconnected networks, and 
2) TCP/IP protocol.

/John


From mbaer at cs.tu-berlin.de  Thu Oct 28 07:37:50 2010
From: mbaer at cs.tu-berlin.de (=?ISO-8859-1?Q?Matthias_B=E4rwolff?=)
Date: Thu, 28 Oct 2010 16:37:50 +0200
Subject: [ih] birth of the Internet?
In-Reply-To: <AANLkTimXc5_H1jetdWuNft4pJsC-xu78wLe9P6Pv-U-r@mail.gmail.com>
References: <4CC96A4E.2070704@meetinghouse.net>
	<AANLkTimXc5_H1jetdWuNft4pJsC-xu78wLe9P6Pv-U-r@mail.gmail.com>
Message-ID: <4CC98ABE.3050608@cs.tu-berlin.de>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

It is probably fair to add that the TCP/IP split, which was really at
the heart of getting the Internet going, took place in 1977/1978, a
process nicely reflected in the IEN notes from the time. UDP was drafted
in 1979; by then it was probably fairly clear that there would be no
going back to the initial monolithic TCP.

I have recently written up a pretty detailed account of those
developments, drop me a note off list in case you want a PDF copy of that.

Going back to the original question -- when was the Internet born? -- I
am surely not the first to note that it is probably a futile exercise to
try and find any precise date that may adequately be taken as the birth
date of the Internet; it would probably not even be far-fetched to go
back to the Greek and some beacon communication scheme (which already
had a fair amount of "control plane elaboration" to it).

Matthias


On 10/28/2010 02:44 PM, Vint Cerf wrote:
> 1. initial design: Spring-Fall 1973
> 2. first published paper: May 1974
> 3. first full spec of TCP: December 1974 (RFC 675)
> 4. More complete specs after splitting off of IP: RFC791, 792, 793 - 1981
> 5. official launch of Internet: January 1, 1983 (which also coincided
> with the creation of MILNET split from former research ARPANET) and
> the incorporation of the Packet Satellite and Packet Radio Net(s) into
> the DARPA Internet. this was also the date of replacement of NCP with
> TCP/IP
> 6. As I recall, there were tests well before 1983 in which the Xerox
> PARC ethernet was linked to Internet by way of a packet radio at PARC
> connecting via the SRI International operated Packet Radio network.
> Larry Masinter's name comes to mind as having had something to do with
> this but others on the list may have clearer recollections. The first
> Packet Radio/ARPANET tests of TCP/IP occurred in the 1975 time frame
> using a BBN developed gateway. The first three-network tests were done
> on November 22, 1977 linking ARPANET, the SF Bay Area Packet Radio net
> and the Packet Satellite Network.
> 
> vint
> 
> 
> 
> 
> 
> On Thu, Oct 28, 2010 at 8:19 AM, Miles Fidelman
> <mfidelman at meetinghouse.net> wrote:
>> Hi Folks,
>>
>> I've been part of an interesting discussion, on another list, that's led to
>> the question of when the Internet was actually "born."
>>
>> Was it:
>> - when the ARPANET moved from NCP to TCP/IP?
>> - when the ARPANET was split into two, linked backbones (ARPANET/MILNET)?
>> - was there an earlier point at which IP was linking LANs across the
>> ARPANET?
>> - some other event?
>>
>> Miles Fidelman
>>
>> --
>> In theory, there is no difference between theory and practice.
>> In<fnord>  practice, there is.   .... Yogi Berra
>>
>>
>>
> 

- -- 
Matthias B?rwolff
www.b?rwolff.de
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iQEcBAEBAgAGBQJMyYq+AAoJEFEeq050tDy5dZwH/25HxbwrFcVp3s9wVMziS8S7
wfIt+oN48URYk5UyrdgX2yzqBs/QHXwEI5BLapT540vs8PCbt5LYH1UKwbPYkoNL
KAfKi2/7Uuwzj2I+ABroRb5zqj5as6duM1wEWem9XCfQOv1cl55PYcpQRPOt978V
5Qug+FbTkTOpVZ+/URWyLY+x7QlhwKSW9U6CeIIL++sYyphEXKpEvE1WcsUsCFAP
VKi3D7dxGpoXANLwjDcdPmRNbJz1ls+OxOEiQBghx7lGUnkyniEvh4jg+iwK/jiH
W5mnOmhDK2jAnQIZPhq3BZ1NewGQPbldMvRhad7SDb1U0DRPZ4kC3mPLBEmiwXg=
=l1Ed
-----END PGP SIGNATURE-----


From dave.walden.family at gmail.com  Thu Oct 28 08:05:27 2010
From: dave.walden.family at gmail.com (Dave Walden)
Date: Thu, 28 Oct 2010 11:05:27 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC96A4E.2070704@meetinghouse.net>
References: <4CC96A4E.2070704@meetinghouse.net>
Message-ID: <4cc99149.d2a1df0a.3e06.24f7@mx.google.com>

A couple of years ago, in response to a query from Bob Kahn, I went 
to the BBN library to look at old Quarterly Technical Reports to ARPA 
to see when the first connection of multiple networks happened.  My 
memory without looking closely again is that BBN did two network 
experiments over SatNet before the first 3-net work, and that Bob's 
memory was that the first three network connection was done in late 1977.

At 08:19 AM 10/28/2010, Miles Fidelman wrote:
>Hi Folks,
>
>I've been part of an interesting discussion, on another list, that's 
>led to the question of when the Internet was actually "born."
>
>Was it:
>- when the ARPANET moved from NCP to TCP/IP?
>- when the ARPANET was split into two, linked backbones (ARPANET/MILNET)?
>- was there an earlier point at which IP was linking LANs across the ARPANET?
>- some other event?
>
>Miles Fidelman
>
>--
>In theory, there is no difference between theory and practice.
>In<fnord>  practice, there is.   .... Yogi Berra
>


--
home address: 12 Linden Rd., E. Sandwich, MA 02537; ph/fax=508-888-7655/4168
email address:  dave at walden-family.com; 
website(s):  http://www.walden-family.com/ 



From jnc at mercury.lcs.mit.edu  Thu Oct 28 09:06:07 2010
From: jnc at mercury.lcs.mit.edu (Noel Chiappa)
Date: Thu, 28 Oct 2010 12:06:07 -0400 (EDT)
Subject: [ih] birth of the Internet?
Message-ID: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>

    > From: Miles Fidelman <mfidelman at meetinghouse.net>

    > I've been part of an interesting discussion, on another list, that's
    > led to the question of when the Internet was actually "born."

To me, one of two events would make sense as the official birth (and
credit to others on the list who've said something close to this, I'm
tweaking their take):

i) the first time data was successfully moved from a computer on one
network to a computer on another over TCP (no IP in the early days)

or

ii) the first time an actual user used the internetwork (as above) to
actually do sometime; i.e. no just simply a test - e.g. to move a file
around, or print a document, or something.


Actually, now that I think about it, the credit for the one of these
milestones probably belongs to the PARC guys, using PUP. (I don't know if
Vint's early TCP work at Stanford met i), or the PARC guys might have snagged
that one too.)

	Noel



From vint at google.com  Thu Oct 28 09:39:18 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 12:39:18 -0400
Subject: [ih] birth of the Internet?
Message-ID: <b0123c4964f70450ba13ddd3246fe733@mail.gmail.com>

The pup system was in operation before tcp. Assuming this was a system of
interconnected ethernets, their system would have been rather more
homogeneous at that time compared to the internet. V

----- Original Message -----
From: internet-history-bounces at postel.org
<internet-history-bounces at postel.org>
To: internet-history at postel.org <internet-history at postel.org>
Cc: jnc at mercury.lcs.mit.edu <jnc at mercury.lcs.mit.edu>
Sent: Thu Oct 28 12:06:07 2010
Subject: Re: [ih] birth of the Internet?

    > From: Miles Fidelman <mfidelman at meetinghouse.net>

    > I've been part of an interesting discussion, on another list, that's
    > led to the question of when the Internet was actually "born."

To me, one of two events would make sense as the official birth (and
credit to others on the list who've said something close to this, I'm
tweaking their take):

i) the first time data was successfully moved from a computer on one
network to a computer on another over TCP (no IP in the early days)

or

ii) the first time an actual user used the internetwork (as above) to
actually do sometime; i.e. no just simply a test - e.g. to move a file
around, or print a document, or something.


Actually, now that I think about it, the credit for the one of these
milestones probably belongs to the PARC guys, using PUP. (I don't know if
Vint's early TCP work at Stanford met i), or the PARC guys might have
snagged
that one too.)

	Noel


From cos at aaaaa.org  Thu Oct 28 09:44:01 2010
From: cos at aaaaa.org (Ofer Inbar)
Date: Thu, 28 Oct 2010 12:44:01 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC96A4E.2070704@meetinghouse.net>
References: <4CC96A4E.2070704@meetinghouse.net>
Message-ID: <20101028164401.GN18254@mip.aaaaa.org>

Miles Fidelman <mfidelman at meetinghouse.net> wrote:
> I've been part of an interesting discussion, on another list, that's led 
> to the question of when the Internet was actually "born."
> 
> Was it:
> - when the ARPANET moved from NCP to TCP/IP?
> - when the ARPANET was split into two, linked backbones (ARPANET/MILNET)?
> - was there an earlier point at which IP was linking LANs across the 
> ARPANET?
> - some other event?

I think there are two separate kinds of ways of looking at it:

1. Pick some criteria for what makes the Internet special, and look
   for the first time something was done / existed that met them?

2. Trace the direct lineage of today's Internet to its organizational
   or physical predecessors and see where it started.  That it, what
   was the birth of the thing that eventually turned into the Internet.

If you go with #1, you may pick something that was a one-shot, or that
developed into some separate entity that either got discontinued or
absorbed into the Internet long after the Internet had separately
evolved to meet your criteria.  That is, the thing that you pick may
not be part of the Internet's primary direct lineage.

If you go with #2, you may pick something that did *not* develop its
unique Internet-qualifying aspects until after they had all been done
by someone else somewhere else.

I think #2 is a better fit for the question, but that implies that
picking some key criteria of technical achievement blindly and then
looking for the first occurrence of that in history is the wrong
approach.

That aside, though, if I have to pick what makes the Internet the
Internet, I'd go with: allowing multiple computer networks under
*separate, independent administration* to connect and support the same
set of common protocols and apps across them.  Where "multiple" isn't
just two, who work out their own unique bilateral; I mean that it lets
more than two accomplish this goal in similar ways without having to
work out new technical means of doing so for each pair.
  -- Cos


From lpress at csudh.edu  Thu Oct 28 09:53:07 2010
From: lpress at csudh.edu (Larry Press)
Date: Thu, 28 Oct 2010 09:53:07 -0700
Subject: [ih] birth of the Internet?
In-Reply-To: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
Message-ID: <4CC9AA73.7020801@csudh.edu>

Folks,

Does NSFNet figure into this discussion at all?

Would it be fair to say that NSFNet grew up to be The Internet, where 
some of the other things being mentioned were experiments or 
developments in internetworking?

Larry


From dhc2 at dcrocker.net  Thu Oct 28 09:59:09 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Thu, 28 Oct 2010 09:59:09 -0700
Subject: [ih] birth of the Internet?
In-Reply-To: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
Message-ID: <4CC9ABDD.3030404@dcrocker.net>


> Actually, now that I think about it, the credit for the one of these
> milestones probably belongs to the PARC guys, using PUP. (I don't know if
> Vint's early TCP work at Stanford met i), or the PARC guys might have snagged
> that one too.)


There is often an interesting challenging in distinguishing between credit for 
winning concepts, credit for important research demonstration, versus credit for 
delivering a 'successful' solution.  Each step is essential, but different.

The original web was the last of these.  Essentially all of its ideas has been 
demonstrated far earlier and one could reasonably argue that there had been 
meaningful "integration" demonstrations earlier also.  What the Web did was a 
winning systems integration that became a production network.

My understanding is that PARC had meaningful internetworking running in the 
earlier 70s, so that they probably win on the first two milestones.  (A careful 
comparison of the technical differences between their design choices and the 
Internet's could be useful for teaching.)

On the other hand, I heard that a session at the first ICCC (where the Arpanet 
was demonstrated in the fall of 1972) had the first discussions on 
Internetworking.  So that might be where the credit for the 'idea' goes.

If this sounds like it serves to reduce the credit due specifically to the 
Internet work, it very much does not.

We tend to vastly underrate the task of "systems engineering" that produces 
successful services out of established technology components (ideas).  Given how 
often systems designs fail, this lack of due credit always mystifies me.

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From mfidelman at meetinghouse.net  Thu Oct 28 10:33:47 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 13:33:47 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9AA73.7020801@csudh.edu>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
	<4CC9AA73.7020801@csudh.edu>
Message-ID: <4CC9B3FB.80707@meetinghouse.net>

Larry Press wrote:
> Would it be fair to say that NSFNet grew up to be The Internet, where 
> some of the other things being mentioned were experiments or 
> developments in internetworking?
I think you have to review your timing.  The Internet predates the 
NSFnet by several years.

I think Vint nailed it (and he should know):

"5. official launch of Internet: January 1, 1983 (which also coincided 
with the creation of MILNET split from former research ARPANET) and the 
incorporation of the Packet Satellite and Packet Radio Net(s) into
the DARPA Internet. this was also the date of replacement of NCP with 
TCP/IP"

with some earlier experiments:

"6. ..The first Packet Radio/ARPANET tests of TCP/IP occurred in the 
1975 time frame
using a BBN developed gateway. The first three-network tests were done
on November 22, 1977 linking ARPANET, the SF Bay Area Packet Radio net
and the Packet Satellite Network."



-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From braden at isi.edu  Thu Oct 28 10:40:02 2010
From: braden at isi.edu (Bob Braden)
Date: Thu, 28 Oct 2010 10:40:02 -0700
Subject: [ih] Origination date for the Internet
In-Reply-To: <mailman.10036.1288283998.753.internet-history@postel.org>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
Message-ID: <4CC9B572.9080300@isi.edu>

It is perhaps fun to identify all the technical milestones during
R&D on internetworking, on IP, on TCP, on PUP, on packet radios in bread 
trucks, in packet satellites, ... , perhaps even on packet switching, 
none of these seem very germane to the original question -- when did 
*THE Internet* enter service.

Speaking as one of the soldiers in the trenches at the time,
I have always counted the beginning of THE Internet as Jan 1, 1983. 
That was (to some us, memorable and for Dan Lynch, traumatic)
the crunch time when we freed ourselves from the good old reliable
NCP and leapt into the TCP/IP world.  (Was the new thing REALLY
going to work at some realistic scale?) And we took with us a
significant population of real network users, which justifies the
capital I on Internet.

Someone had a T shirt bearing the inscription "I survived Jan 1, 1983" 
or something like that. I wish I had one.

Bob Braden



On 10/28/2010 9:39 AM, internet-history-request at postel.org wrote:
> Send internet-history mailing list submissions to
> 	internet-history at postel.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
> 	http://mailman.postel.org/mailman/listinfo/internet-history
> or, via email, send a message with subject or body 'help' to
> 	internet-history-request at postel.org
>
> You can reach the person managing the list at
> 	internet-history-owner at postel.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of internet-history digest..."
>
>
> Today's Topics:
>
>     1. birth of the Internet? (Miles Fidelman)
>     2. Re: birth of the Internet? (Vint Cerf)
>     3. Re: birth of the Internet? (Miles Fidelman)
>     4. Re: birth of the Internet? (John Curran)
>     5. Re: birth of the Internet? (Matthias B?rwolff)
>     6. Re: birth of the Internet? (Dave Walden)
>     7. Re: birth of the Internet? (Noel Chiappa)
>     8. Re: birth of the Internet? (Vint Cerf)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 28 Oct 2010 08:19:26 -0400
> From: Miles Fidelman<mfidelman at meetinghouse.net>
> Subject: [ih] birth of the Internet?
> To: internet-history at postel.org
> Message-ID:<4CC96A4E.2070704 at meetinghouse.net>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Hi Folks,
>
> I've been part of an interesting discussion, on another list, that's led
> to the question of when the Internet was actually "born."
>
> Was it:
> - when the ARPANET moved from NCP to TCP/IP?
> - when the ARPANET was split into two, linked backbones (ARPANET/MILNET)?
> - was there an earlier point at which IP was linking LANs across the
> ARPANET?
> - some other event?
>
> Miles Fidelman
>


From sbrim at cisco.com  Thu Oct 28 10:47:54 2010
From: sbrim at cisco.com (Scott Brim)
Date: Thu, 28 Oct 2010 13:47:54 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9AA73.7020801@csudh.edu>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
	<4CC9AA73.7020801@csudh.edu>
Message-ID: <4CC9B74A.5070402@cisco.com>

On 10/28/2010 12:53 EDT, Larry Press wrote:
> Folks,
> 
> Does NSFNet figure into this discussion at all?
> 
> Would it be fair to say that NSFNet grew up to be The Internet, where
> some of the other things being mentioned were experiments or
> developments in internetworking?

I wouldn't say that.  Yes NSFNet spawned or inspired children that ended
up dominating the development of the Internet but IMHO "the Internet"
existed long before.


From dhc2 at dcrocker.net  Thu Oct 28 11:00:57 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Thu, 28 Oct 2010 11:00:57 -0700
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9B3FB.80707@meetinghouse.net>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>	<4CC9AA73.7020801@csudh.edu>
	<4CC9B3FB.80707@meetinghouse.net>
Message-ID: <4CC9BA59.8050508@dcrocker.net>



On 10/28/2010 10:33 AM, Miles Fidelman wrote:
> Larry Press wrote:
>> Would it be fair to say that NSFNet grew up to be The Internet, where some of
>> the other things being mentioned were experiments or developments in
>> internetworking?
> I think you have to review your timing. The Internet predates the NSFnet by
> several years.


Right.

NSFNet really qualifies as the /final/ stage of development of the Internet, 
before fully commercial adoption.

It's introduction of an additional backbone forced core changes to the routing 
technology, but otherwise it had to do with expanding the Internet 
operationally, rather than in 'creating' the Internet.

Besides the forcing function on BGP, it's 'innovation' was to seed organizations 
that created a commercial core to the public Internet.

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From lpress at csudh.edu  Thu Oct 28 11:13:47 2010
From: lpress at csudh.edu (Larry Press)
Date: Thu, 28 Oct 2010 11:13:47 -0700
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9B3FB.80707@meetinghouse.net>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>	<4CC9AA73.7020801@csudh.edu>
	<4CC9B3FB.80707@meetinghouse.net>
Message-ID: <4CC9BD5B.6070300@csudh.edu>

On 10/28/2010 10:33 AM, Miles Fidelman wrote:

> Larry Press wrote:

> I think you have to review your timing.  The Internet predates the
> NSFnet by several years.


For sure -- I was trying to differentiate between "the Internet" (a 
proper noun), which is what the initial query asked about, and 
"inter-networking."

Asking what the first example of inter-networking was or what the first 
inter-networking experiment or protocol was is not, unless we define it 
to be so, the same as asking about the birth of the Internet.

Larry



From mfidelman at meetinghouse.net  Thu Oct 28 11:26:35 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 14:26:35 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9BD5B.6070300@csudh.edu>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>	<4CC9AA73.7020801@csudh.edu>
	<4CC9B3FB.80707@meetinghouse.net> <4CC9BD5B.6070300@csudh.edu>
Message-ID: <4CC9C05B.1070700@meetinghouse.net>

Larry Press wrote:
> For sure -- I was trying to differentiate between "the Internet" (a 
> proper noun), which is what the initial query asked about, and 
> "inter-networking."
>
> Asking what the first example of inter-networking was or what the 
> first inter-networking experiment or protocol was is not, unless we 
> define it to be so, the same as asking about the birth of the Internet.
Sure.  Is there anybody here who doesn't think that the (capital I) 
Internet started with the ARPANET as its backbone, well before the 
NSFnet was a gleam in anybody's eye?



-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From vint at google.com  Thu Oct 28 11:26:46 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 14:26:46 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9BA59.8050508@dcrocker.net>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
	<4CC9AA73.7020801@csudh.edu> <4CC9B3FB.80707@meetinghouse.net>
	<4CC9BA59.8050508@dcrocker.net>
Message-ID: <AANLkTimS6oM32Afo6Yt1bBRLC-maaZxCSN_oEjUxCM8P@mail.gmail.com>

NSF program managers also espoused and supported the Network Access
Points for interconnection in place of the NSFNET as it was retired in
1995.

v


On Thu, Oct 28, 2010 at 2:00 PM, Dave CROCKER <dhc2 at dcrocker.net> wrote:
>
>
> On 10/28/2010 10:33 AM, Miles Fidelman wrote:
>>
>> Larry Press wrote:
>>>
>>> Would it be fair to say that NSFNet grew up to be The Internet, where
>>> some of
>>> the other things being mentioned were experiments or developments in
>>> internetworking?
>>
>> I think you have to review your timing. The Internet predates the NSFnet
>> by
>> several years.
>
>
> Right.
>
> NSFNet really qualifies as the /final/ stage of development of the Internet,
> before fully commercial adoption.
>
> It's introduction of an additional backbone forced core changes to the
> routing technology, but otherwise it had to do with expanding the Internet
> operationally, rather than in 'creating' the Internet.
>
> Besides the forcing function on BGP, it's 'innovation' was to seed
> organizations that created a commercial core to the public Internet.
>
> d/
> --
>
> ?Dave Crocker
> ?Brandenburg InternetWorking
> ?bbiw.net
>



From vint at google.com  Thu Oct 28 11:30:53 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 14:30:53 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CC9B572.9080300@isi.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
Message-ID: <AANLkTikrcficRyocf54-SU4VbXqfZV9OM0+J7E2NZVGV@mail.gmail.com>

bob: red button "I survived the TCP/IP transition" issued by Interop.

v


On Thu, Oct 28, 2010 at 1:40 PM, Bob Braden <braden at isi.edu> wrote:
> It is perhaps fun to identify all the technical milestones during
> R&D on internetworking, on IP, on TCP, on PUP, on packet radios in bread
> trucks, in packet satellites, ... , perhaps even on packet switching, none
> of these seem very germane to the original question -- when did *THE
> Internet* enter service.
>
> Speaking as one of the soldiers in the trenches at the time,
> I have always counted the beginning of THE Internet as Jan 1, 1983. That was
> (to some us, memorable and for Dan Lynch, traumatic)
> the crunch time when we freed ourselves from the good old reliable
> NCP and leapt into the TCP/IP world. ?(Was the new thing REALLY
> going to work at some realistic scale?) And we took with us a
> significant population of real network users, which justifies the
> capital I on Internet.
>
> Someone had a T shirt bearing the inscription "I survived Jan 1, 1983" or
> something like that. I wish I had one.
>
> Bob Braden
>
>
>
> On 10/28/2010 9:39 AM, internet-history-request at postel.org wrote:
>>
>> Send internet-history mailing list submissions to
>> ? ? ? ?internet-history at postel.org
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>> ? ? ? ?http://mailman.postel.org/mailman/listinfo/internet-history
>> or, via email, send a message with subject or body 'help' to
>> ? ? ? ?internet-history-request at postel.org
>>
>> You can reach the person managing the list at
>> ? ? ? ?internet-history-owner at postel.org
>>
>> When replying, please edit your Subject line so it is more specific
>> than "Re: Contents of internet-history digest..."
>>
>>
>> Today's Topics:
>>
>> ? ?1. birth of the Internet? (Miles Fidelman)
>> ? ?2. Re: birth of the Internet? (Vint Cerf)
>> ? ?3. Re: birth of the Internet? (Miles Fidelman)
>> ? ?4. Re: birth of the Internet? (John Curran)
>> ? ?5. Re: birth of the Internet? (Matthias B?rwolff)
>> ? ?6. Re: birth of the Internet? (Dave Walden)
>> ? ?7. Re: birth of the Internet? (Noel Chiappa)
>> ? ?8. Re: birth of the Internet? (Vint Cerf)
>>
>>
>> ----------------------------------------------------------------------
>>
>> Message: 1
>> Date: Thu, 28 Oct 2010 08:19:26 -0400
>> From: Miles Fidelman<mfidelman at meetinghouse.net>
>> Subject: [ih] birth of the Internet?
>> To: internet-history at postel.org
>> Message-ID:<4CC96A4E.2070704 at meetinghouse.net>
>> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>>
>> Hi Folks,
>>
>> I've been part of an interesting discussion, on another list, that's led
>> to the question of when the Internet was actually "born."
>>
>> Was it:
>> - when the ARPANET moved from NCP to TCP/IP?
>> - when the ARPANET was split into two, linked backbones (ARPANET/MILNET)?
>> - was there an earlier point at which IP was linking LANs across the
>> ARPANET?
>> - some other event?
>>
>> Miles Fidelman
>>
>



From eric.gade at gmail.com  Thu Oct 28 11:34:42 2010
From: eric.gade at gmail.com (Eric Gade)
Date: Thu, 28 Oct 2010 19:34:42 +0100
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9BA59.8050508@dcrocker.net>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
	<4CC9AA73.7020801@csudh.edu> <4CC9B3FB.80707@meetinghouse.net>
	<4CC9BA59.8050508@dcrocker.net>
Message-ID: <AANLkTim+Dxi+JYNjyTiP6LV3YsGSnHFED3yqG7fbqN74@mail.gmail.com>

>
> NSFNet really qualifies as the /final/ stage of development of the
> Internet, before fully commercial adoption.
>
Shoudl the question be recharacterized as a difference between "When was The
Internet?" and "When was the Worldnet?"

On Thu, Oct 28, 2010 at 7:00 PM, Dave CROCKER <dhc2 at dcrocker.net> wrote:

>
>
> On 10/28/2010 10:33 AM, Miles Fidelman wrote:
>
>> Larry Press wrote:
>>
>>> Would it be fair to say that NSFNet grew up to be The Internet, where
>>> some of
>>> the other things being mentioned were experiments or developments in
>>> internetworking?
>>>
>> I think you have to review your timing. The Internet predates the NSFnet
>> by
>> several years.
>>
>
>
> Right.
>
> NSFNet really qualifies as the /final/ stage of development of the
> Internet, before fully commercial adoption.
>
> It's introduction of an additional backbone forced core changes to the
> routing technology, but otherwise it had to do with expanding the Internet
> operationally, rather than in 'creating' the Internet.
>
> Besides the forcing function on BGP, it's 'innovation' was to seed
> organizations that created a commercial core to the public Internet.
>
>
> d/
> --
>
>  Dave Crocker
>  Brandenburg InternetWorking
>  bbiw.net
>



-- 
Eric
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20101028/944847d7/attachment.htm>

From mfidelman at meetinghouse.net  Thu Oct 28 12:02:42 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 15:02:42 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9BA59.8050508@dcrocker.net>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>	<4CC9AA73.7020801@csudh.edu>	<4CC9B3FB.80707@meetinghouse.net>
	<4CC9BA59.8050508@dcrocker.net>
Message-ID: <4CC9C8D2.1060805@meetinghouse.net>

Dave CROCKER wrote:
> NSFNet really qualifies as the /final/ stage of development of the 
> Internet, before fully commercial adoption.
As I recall, Roger Fradenburgh had the job of turning off the ARPANET, 
at its end-of-life.  I was in the next office at BBN when net 10 went 
down, and bits just kept humming across the Internet.   One of the 
sadder moments in Internet history.

Miles

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From richard at bennett.com  Thu Oct 28 12:08:42 2010
From: richard at bennett.com (Richard Bennett)
Date: Thu, 28 Oct 2010 12:08:42 -0700
Subject: [ih] birth of the Internet?
In-Reply-To: <AANLkTimS6oM32Afo6Yt1bBRLC-maaZxCSN_oEjUxCM8P@mail.gmail.com>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>	<4CC9AA73.7020801@csudh.edu>
	<4CC9B3FB.80707@meetinghouse.net>	<4CC9BA59.8050508@dcrocker.net>
	<AANLkTimS6oM32Afo6Yt1bBRLC-maaZxCSN_oEjUxCM8P@mail.gmail.com>
Message-ID: <4CC9CA3A.5030208@bennett.com>

They actually funded the NAPs, as I recall.  It didn't take long for the 
NAPs to be replaced by the carrier-neutral IXs.

RB

On 10/28/2010 11:26 AM, Vint Cerf wrote:
> NSF program managers also espoused and supported the Network Access
> Points for interconnection in place of the NSFNET as it was retired in
> 1995.
>
> v
>
>
> On Thu, Oct 28, 2010 at 2:00 PM, Dave CROCKER<dhc2 at dcrocker.net>  wrote:
>>
>> On 10/28/2010 10:33 AM, Miles Fidelman wrote:
>>> Larry Press wrote:
>>>> Would it be fair to say that NSFNet grew up to be The Internet, where
>>>> some of
>>>> the other things being mentioned were experiments or developments in
>>>> internetworking?
>>> I think you have to review your timing. The Internet predates the NSFnet
>>> by
>>> several years.
>>
>> Right.
>>
>> NSFNet really qualifies as the /final/ stage of development of the Internet,
>> before fully commercial adoption.
>>
>> It's introduction of an additional backbone forced core changes to the
>> routing technology, but otherwise it had to do with expanding the Internet
>> operationally, rather than in 'creating' the Internet.
>>
>> Besides the forcing function on BGP, it's 'innovation' was to seed
>> organizations that created a commercial core to the public Internet.
>>
>> d/
>> --
>>
>>   Dave Crocker
>>   Brandenburg InternetWorking
>>   bbiw.net
>>

-- 
--
Richard Bennett



From bob.hinden at gmail.com  Thu Oct 28 12:14:47 2010
From: bob.hinden at gmail.com (Bob Hinden)
Date: Thu, 28 Oct 2010 12:14:47 -0700
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CC9B572.9080300@isi.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
Message-ID: <CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>

Bob,

On Oct 28, 2010, at 10:40 AM, Bob Braden wrote:

> It is perhaps fun to identify all the technical milestones during
> R&D on internetworking, on IP, on TCP, on PUP, on packet radios in bread trucks, in packet satellites, ... , perhaps even on packet switching, none of these seem very germane to the original question -- when did *THE Internet* enter service.
> 
> Speaking as one of the soldiers in the trenches at the time,
> I have always counted the beginning of THE Internet as Jan 1, 1983. That was (to some us, memorable and for Dan Lynch, traumatic)
> the crunch time when we freed ourselves from the good old reliable
> NCP and leapt into the TCP/IP world.  (Was the new thing REALLY
> going to work at some realistic scale?) And we took with us a
> significant population of real network users, which justifies the
> capital I on Internet.
> 
> Someone had a T shirt bearing the inscription "I survived Jan 1, 1983" or something like that. I wish I had one.
> 

I still have my "I Survived the TCP Transition 1/1/83" red button.

In my view this was the time when the Internet became operational as people starting using it for their day to day work, instead of a set of researchers.  Conception and birth occurred earlier :-)

Bob





From sbrim at cisco.com  Thu Oct 28 12:15:29 2010
From: sbrim at cisco.com (Scott Brim)
Date: Thu, 28 Oct 2010 15:15:29 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <AANLkTim+Dxi+JYNjyTiP6LV3YsGSnHFED3yqG7fbqN74@mail.gmail.com>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>	<4CC9AA73.7020801@csudh.edu>
	<4CC9B3FB.80707@meetinghouse.net>	<4CC9BA59.8050508@dcrocker.net>
	<AANLkTim+Dxi+JYNjyTiP6LV3YsGSnHFED3yqG7fbqN74@mail.gmail.com>
Message-ID: <4CC9CBD1.7090703@cisco.com>

On 10/28/2010 14:34 EDT, Eric Gade wrote:
> Shoudl the question be recharacterized as a difference between "When was
> The Internet?" and "When was the Worldnet?"

If so, then the answer to the second half would have to be "not yet".


From jjd at jjd.com  Thu Oct 28 12:29:17 2010
From: jjd at jjd.com (James J Dempsey)
Date: Thu, 28 Oct 2010 15:29:17 -0400
Subject: [ih] birth of the Internet?
Message-ID: <30606.1288294157@vegas.jjd.com>

Miles asks:

> Sure.  Is there anybody here who doesn't think that the (capital I) 
> Internet started with the ARPANET as its backbone, well before the 
> NSFnet was a gleam in anybody's eye?

With help from someone else's memory, I could argue that The Internet
started *before* the ARPANET was connected to it.

I recall being in the BBN NOC the day the ARPANET transitioned from NCP to
TCP.  That NOC was on the 5th floor of BBN's building 6.  Sometime pretty
soon after that, the NOC (and our offices) moved to BBN's building 1.  After
the move, the NOC was much larger, managed more networks and had windows to
the hallway that allowed visitors to look in at what was going on.

At about the time of that move, I remember people talking about the "ARPANET
NOC" as being the one in the lab, while the "Internet NOC" was (if I
remember correctly) in Eric Rosen's office.  Soon after that, the terminals
and displays monitoring the Internet gateways moved into a corner of the NOC
proper along with the ARPANET monitoring.

I'm not sure if this "Internet NOC" existed before 1 January 1983, but it
very well might have.  I'll bet Eric Rosen or Bob Hinden could say so more
definitively.  If there was an "Internet NOC" monitoring the Internet
gateways before January 1983, I would argue that at that point The Internet
did exist, even if the ARPANET was not yet connected to it via TCP/IP.

                       --Jim Dempsey--
                         jjd at jjd.com


From mfidelman at meetinghouse.net  Thu Oct 28 12:40:59 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 15:40:59 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <30606.1288294157@vegas.jjd.com>
References: <30606.1288294157@vegas.jjd.com>
Message-ID: <4CC9D1CB.7070504@meetinghouse.net>

James J Dempsey wrote:
> I'm not sure if this "Internet NOC" existed before 1 January 1983, but it
> very well might have.  I'll bet Eric Rosen or Bob Hinden could say so more
> definitively.  If there was an "Internet NOC" monitoring the Internet
> gateways before January 1983, I would argue that at that point The Internet
> did exist, even if the ARPANET was not yet connected to it via TCP/IP.
>
>    
Hmmm.... does that suggest that the mailbridge gateways were the 
embryonic Internet? :-)

Miles

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From benc at hawaga.org.uk  Thu Oct 28 12:44:44 2010
From: benc at hawaga.org.uk (Ben Clifford)
Date: Thu, 28 Oct 2010 19:44:44 +0000 (GMT)
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9C8D2.1060805@meetinghouse.net>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
	<4CC9AA73.7020801@csudh.edu>
	<4CC9B3FB.80707@meetinghouse.net> <4CC9BA59.8050508@dcrocker.net>
	<4CC9C8D2.1060805@meetinghouse.net>
Message-ID: <Pine.LNX.4.64.1010281943010.18042@dildano.hawaga.org.uk>


> As I recall, Roger Fradenburgh had the job of turning off the ARPANET, 
> at its end-of-life.  I was in the next office at BBN when net 10 went 
> down

What was 'turning off the ARPANET' in terms of actual switches to switch? 
Turning off some specific single piece of hardware that made it all go off 
or an IP-level router to the rest of the internet or something else?

-- 



From mfidelman at meetinghouse.net  Thu Oct 28 12:52:01 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 15:52:01 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <Pine.LNX.4.64.1010281943010.18042@dildano.hawaga.org.uk>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
	<4CC9AA73.7020801@csudh.edu> <4CC9B3FB.80707@meetinghouse.net>
	<4CC9BA59.8050508@dcrocker.net> <4CC9C8D2.1060805@meetinghouse.net>
	<Pine.LNX.4.64.1010281943010.18042@dildano.hawaga.org.uk>
Message-ID: <4CC9D461.90304@meetinghouse.net>

Ben Clifford wrote:
>> As I recall, Roger Fradenburgh had the job of turning off the ARPANET,
>> at its end-of-life.  I was in the next office at BBN when net 10 went
>> down
>>      
> What was 'turning off the ARPANET' in terms of actual switches to switch?
> Turning off some specific single piece of hardware that made it all go off
> or an IP-level router to the rest of the internet or something else?
>    
Probably something like:
- log into NU (the management software used to monitor and control the 
ARPANET IMPs)
- send a shutdown command to each node
- physically pull the plugs, disconnect cables, haul away the IMPs
- I'm really not sure if anything needed to be done to the routing 
tables, or if that all happened automagically at that point

I'll ask Roger and report back.

Miles

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From louie at transsys.com  Thu Oct 28 13:00:02 2010
From: louie at transsys.com (Louis Mamakos)
Date: Thu, 28 Oct 2010 16:00:02 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9CA3A.5030208@bennett.com>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>	<4CC9AA73.7020801@csudh.edu>
	<4CC9B3FB.80707@meetinghouse.net>	<4CC9BA59.8050508@dcrocker.net>
	<AANLkTimS6oM32Afo6Yt1bBRLC-maaZxCSN_oEjUxCM8P@mail.gmail.com>
	<4CC9CA3A.5030208@bennett.com>
Message-ID: <DC3F4A5A-72E3-482B-9F47-16CCF9D7072F@transsys.com>

The reverse actually happened.  Carrier neutral, commercial
interconnects were already well in operation by the time the
NAP concept came along.

There were carrier neutral exchange points in existence which at
least UUNET, Sprint, PSI and other had in use; these were operated
as a service by MFS Datanet and instigated by some of those initially
connected.  Initial interconnects were 10Mb/s Ethernet and various
other things followed.  For example, MAE-EAST, MAE-WEST.

Later, the NSF instigated these NAP interconnects, which were of
moderate success, depending on who you asked, and the particular
NAP exchange point in question.  MFS expanded the scope of their
interconnection solution and that mostly worked OK.  There was
the NY NAP in New Jersey operated by Sprint and I believe a Chicago
NAP operated by Ameritech at the time.  The latter was ATM based
and I suspect many lessons were learned from that experience.

To some extent, the NAPs were a bit redundant given some public
exchanges that were already in operation.  One of the other drivers
was as a funding source for some of the NSF regional networks for
connections.

There was also another federally funded exchange point or two,
the "FIX" that existed at around this time.  Various agency
networks (DoE, NASA, etc.) were interconnected here and at least
the FIX interconnect on the east coast in Washington was generally
aligned with the MAE-EAST interconnect fabric.

Louis Mamakos

On Oct 28, 2010, at 3:08 PM, Richard Bennett wrote:

> They actually funded the NAPs, as I recall.  It didn't take long for the NAPs to be replaced by the carrier-neutral IXs.
> 
> RB
> 
> On 10/28/2010 11:26 AM, Vint Cerf wrote:
>> NSF program managers also espoused and supported the Network Access
>> Points for interconnection in place of the NSFNET as it was retired in
>> 1995.
>> 
>> v
>> 
>> 
>> On Thu, Oct 28, 2010 at 2:00 PM, Dave CROCKER<dhc2 at dcrocker.net>  wrote:
>>> 
>>> On 10/28/2010 10:33 AM, Miles Fidelman wrote:
>>>> Larry Press wrote:
>>>>> Would it be fair to say that NSFNet grew up to be The Internet, where
>>>>> some of
>>>>> the other things being mentioned were experiments or developments in
>>>>> internetworking?
>>>> I think you have to review your timing. The Internet predates the NSFnet
>>>> by
>>>> several years.
>>> 
>>> Right.
>>> 
>>> NSFNet really qualifies as the /final/ stage of development of the Internet,
>>> before fully commercial adoption.
>>> 
>>> It's introduction of an additional backbone forced core changes to the
>>> routing technology, but otherwise it had to do with expanding the Internet
>>> operationally, rather than in 'creating' the Internet.
>>> 
>>> Besides the forcing function on BGP, it's 'innovation' was to seed
>>> organizations that created a commercial core to the public Internet.
>>> 
>>> d/
>>> --
>>> 
>>>  Dave Crocker
>>>  Brandenburg InternetWorking
>>>  bbiw.net
>>> 
> 
> -- 
> --
> Richard Bennett
> 




From vint at google.com  Thu Oct 28 13:05:42 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 16:05:42 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <30606.1288294157@vegas.jjd.com>
References: <30606.1288294157@vegas.jjd.com>
Message-ID: <AANLkTikWdWPzoqQSfa2EXpm6jRuvtF4faK6CT2xCS85F@mail.gmail.com>

- there was clearly an experimental activity up until 1/1/1983. The
folks in Europe were forced to use the Packet Satellite network and
X.25 networks to access the US during 1982 at least, so it could be
that Eric Rosen was monitoring that as a operational if somewhat
experimental system? Peter Kirstein would know since he managed at
least the UK part of the operational SATNET access.

v

On Thu, Oct 28, 2010 at 3:29 PM, James J Dempsey <jjd at jjd.com> wrote:
> Miles asks:
>
>> Sure. ?Is there anybody here who doesn't think that the (capital I)
>> Internet started with the ARPANET as its backbone, well before the
>> NSFnet was a gleam in anybody's eye?
>
> With help from someone else's memory, I could argue that The Internet
> started *before* the ARPANET was connected to it.
>
> I recall being in the BBN NOC the day the ARPANET transitioned from NCP to
> TCP. ?That NOC was on the 5th floor of BBN's building 6. ?Sometime pretty
> soon after that, the NOC (and our offices) moved to BBN's building 1. ?After
> the move, the NOC was much larger, managed more networks and had windows to
> the hallway that allowed visitors to look in at what was going on.
>
> At about the time of that move, I remember people talking about the "ARPANET
> NOC" as being the one in the lab, while the "Internet NOC" was (if I
> remember correctly) in Eric Rosen's office. ?Soon after that, the terminals
> and displays monitoring the Internet gateways moved into a corner of the NOC
> proper along with the ARPANET monitoring.
>
> I'm not sure if this "Internet NOC" existed before 1 January 1983, but it
> very well might have. ?I'll bet Eric Rosen or Bob Hinden could say so more
> definitively. ?If there was an "Internet NOC" monitoring the Internet
> gateways before January 1983, I would argue that at that point The Internet
> did exist, even if the ARPANET was not yet connected to it via TCP/IP.
>
> ? ? ? ? ? ? ? ? ? ? ? --Jim Dempsey--
> ? ? ? ? ? ? ? ? ? ? ? ? jjd at jjd.com
>



From vint at google.com  Thu Oct 28 13:06:15 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 16:06:15 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <Pine.LNX.4.64.1010281943010.18042@dildano.hawaga.org.uk>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
	<4CC9AA73.7020801@csudh.edu> <4CC9B3FB.80707@meetinghouse.net>
	<4CC9BA59.8050508@dcrocker.net> <4CC9C8D2.1060805@meetinghouse.net>
	<Pine.LNX.4.64.1010281943010.18042@dildano.hawaga.org.uk>
Message-ID: <AANLkTikjwE570x5aMDobP0dfQ2jJ66pS_cuOqRU5gpog@mail.gmail.com>

shutting off the IMPs.


On Thu, Oct 28, 2010 at 3:44 PM, Ben Clifford <benc at hawaga.org.uk> wrote:
>
>> As I recall, Roger Fradenburgh had the job of turning off the ARPANET,
>> at its end-of-life. ?I was in the next office at BBN when net 10 went
>> down
>
> What was 'turning off the ARPANET' in terms of actual switches to switch?
> Turning off some specific single piece of hardware that made it all go off
> or an IP-level router to the rest of the internet or something else?
>
> --
>
>



From mfidelman at meetinghouse.net  Thu Oct 28 13:08:34 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 16:08:34 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
Message-ID: <4CC9D842.4010906@meetinghouse.net>

Bob,

Bob Hinden wrote:
> I still have my "I Survived the TCP Transition 1/1/83" red button.
>
> In my view this was the time when the Internet became operational as people starting using it for their day to day work, instead of a set of researchers.  Conception and birth occurred earlier :-)
>    
Actually, that raises another interesting question: At what point, prior 
to 1/1/83, if any, was there a minimal set of networks, gateways, and 
end systems that were passing IP packets on an ongoing basis - as 
opposed to being cobbled together to run some experiment or other, and 
then brought back down?  Can we isolate a date when IP packets started 
flowing and never stopped?

Miles



-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From sbrim at cisco.com  Thu Oct 28 13:24:33 2010
From: sbrim at cisco.com (Scott Brim)
Date: Thu, 28 Oct 2010 16:24:33 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9D1CB.7070504@meetinghouse.net>
References: <30606.1288294157@vegas.jjd.com>
	<4CC9D1CB.7070504@meetinghouse.net>
Message-ID: <4CC9DC01.20601@cisco.com>

On 10/28/2010 15:40 EDT, Miles Fidelman wrote:
> Hmmm.... does that suggest that the mailbridge gateways were the
> embryonic Internet? :-)

They weren't fully transparent IP forwarders.


From mfidelman at meetinghouse.net  Thu Oct 28 13:29:18 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 16:29:18 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9DC01.20601@cisco.com>
References: <30606.1288294157@vegas.jjd.com>
	<4CC9D1CB.7070504@meetinghouse.net> <4CC9DC01.20601@cisco.com>
Message-ID: <4CC9DD1E.1030300@meetinghouse.net>

>
> On 10/28/2010 15:40 EDT, Miles Fidelman wrote:
>    
>> Hmmm.... does that suggest that the mailbridge gateways were the
>> embryonic Internet? :-)
>>      
> They weren't fully transparent IP forwarders.
>    
As I recall, they weren't supposed to be transparent forwarders, but in 
practice nobody every configured the filters.  Bob Hinden - care to comment?

Miles

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From dhc2 at dcrocker.net  Thu Oct 28 13:40:30 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Thu, 28 Oct 2010 13:40:30 -0700
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9D1CB.7070504@meetinghouse.net>
References: <30606.1288294157@vegas.jjd.com>
	<4CC9D1CB.7070504@meetinghouse.net>
Message-ID: <4CC9DFBE.3070801@dcrocker.net>



On 10/28/2010 12:40 PM, Miles Fidelman wrote:
> Hmmm.... does that suggest that the mailbridge gateways were the embryonic
> Internet?


There's a rational argument for that, in terms of little i, but not big I. [1]

The big-I term pretty clearly relies on having a core (default-free zone) 
backbone based on TCP/IP that is used as transit between end systems.


[1] <http://www.rfc-editor.org/rfc/rfc1775.txt>
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From vint at google.com  Thu Oct 28 14:23:28 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 17:23:28 -0400
Subject: [ih] birth of the Internet?
In-Reply-To: <DC3F4A5A-72E3-482B-9F47-16CCF9D7072F@transsys.com>
References: <20101028160607.BA55B6BE54F@mercury.lcs.mit.edu>
	<4CC9AA73.7020801@csudh.edu> <4CC9B3FB.80707@meetinghouse.net>
	<4CC9BA59.8050508@dcrocker.net>
	<AANLkTimS6oM32Afo6Yt1bBRLC-maaZxCSN_oEjUxCM8P@mail.gmail.com>
	<4CC9CA3A.5030208@bennett.com>
	<DC3F4A5A-72E3-482B-9F47-16CCF9D7072F@transsys.com>
Message-ID: <AANLkTimmz3zVR6zM3Bw1pFFxK2Afp5FKyROPT3_2So8c@mail.gmail.com>

Louis is correct. the CIX (commercial Internet eXchange) came into
being about 1989.

MAE-EAST/MAE-WEST arrived a bit later - perhaps somewhere in the 1992
time frame????

NAPS were "slope-funded" by NSF in the 1995-1998 (later?) period to
assure connectivity that had been provided by NSFNET that was retired
in 1995.

vint


On Thu, Oct 28, 2010 at 4:00 PM, Louis Mamakos <louie at transsys.com> wrote:
> The reverse actually happened. ?Carrier neutral, commercial
> interconnects were already well in operation by the time the
> NAP concept came along.
>
> There were carrier neutral exchange points in existence which at
> least UUNET, Sprint, PSI and other had in use; these were operated
> as a service by MFS Datanet and instigated by some of those initially
> connected. ?Initial interconnects were 10Mb/s Ethernet and various
> other things followed. ?For example, MAE-EAST, MAE-WEST.
>
> Later, the NSF instigated these NAP interconnects, which were of
> moderate success, depending on who you asked, and the particular
> NAP exchange point in question. ?MFS expanded the scope of their
> interconnection solution and that mostly worked OK. ?There was
> the NY NAP in New Jersey operated by Sprint and I believe a Chicago
> NAP operated by Ameritech at the time. ?The latter was ATM based
> and I suspect many lessons were learned from that experience.
>
> To some extent, the NAPs were a bit redundant given some public
> exchanges that were already in operation. ?One of the other drivers
> was as a funding source for some of the NSF regional networks for
> connections.
>
> There was also another federally funded exchange point or two,
> the "FIX" that existed at around this time. ?Various agency
> networks (DoE, NASA, etc.) were interconnected here and at least
> the FIX interconnect on the east coast in Washington was generally
> aligned with the MAE-EAST interconnect fabric.
>
> Louis Mamakos
>
> On Oct 28, 2010, at 3:08 PM, Richard Bennett wrote:
>
>> They actually funded the NAPs, as I recall. ?It didn't take long for the NAPs to be replaced by the carrier-neutral IXs.
>>
>> RB
>>
>> On 10/28/2010 11:26 AM, Vint Cerf wrote:
>>> NSF program managers also espoused and supported the Network Access
>>> Points for interconnection in place of the NSFNET as it was retired in
>>> 1995.
>>>
>>> v
>>>
>>>
>>> On Thu, Oct 28, 2010 at 2:00 PM, Dave CROCKER<dhc2 at dcrocker.net> ?wrote:
>>>>
>>>> On 10/28/2010 10:33 AM, Miles Fidelman wrote:
>>>>> Larry Press wrote:
>>>>>> Would it be fair to say that NSFNet grew up to be The Internet, where
>>>>>> some of
>>>>>> the other things being mentioned were experiments or developments in
>>>>>> internetworking?
>>>>> I think you have to review your timing. The Internet predates the NSFnet
>>>>> by
>>>>> several years.
>>>>
>>>> Right.
>>>>
>>>> NSFNet really qualifies as the /final/ stage of development of the Internet,
>>>> before fully commercial adoption.
>>>>
>>>> It's introduction of an additional backbone forced core changes to the
>>>> routing technology, but otherwise it had to do with expanding the Internet
>>>> operationally, rather than in 'creating' the Internet.
>>>>
>>>> Besides the forcing function on BGP, it's 'innovation' was to seed
>>>> organizations that created a commercial core to the public Internet.
>>>>
>>>> d/
>>>> --
>>>>
>>>> ?Dave Crocker
>>>> ?Brandenburg InternetWorking
>>>> ?bbiw.net
>>>>
>>
>> --
>> --
>> Richard Bennett
>>
>
>
>



From vint at google.com  Thu Oct 28 14:44:54 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 17:44:54 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CC9D842.4010906@meetinghouse.net>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
Message-ID: <AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>

actually ISI tracked TCP/IP capability during 1982; the primary
regular use was from Europe, especially the UK, prior to january 1983;
by then there LANS connecting to the ARPANET by way of gateways
(Proteon was around with its rings - Noel Chiappa is that correct?).
Then came Cisco but i guess after 1984.

Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
with NCP.

v


On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
<mfidelman at meetinghouse.net> wrote:
> Bob,
>
> Bob Hinden wrote:
>>
>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>
>> In my view this was the time when the Internet became operational as
>> people starting using it for their day to day work, instead of a set of
>> researchers. ?Conception and birth occurred earlier :-)
>>
>
> Actually, that raises another interesting question: At what point, prior to
> 1/1/83, if any, was there a minimal set of networks, gateways, and end
> systems that were passing IP packets on an ongoing basis - as opposed to
> being cobbled together to run some experiment or other, and then brought
> back down? ?Can we isolate a date when IP packets started flowing and never
> stopped?
>
> Miles
>
>
>
> --
> In theory, there is no difference between theory and practice.
> In<fnord> ?practice, there is. ? .... Yogi Berra
>
>
>



From bob.hinden at gmail.com  Thu Oct 28 14:55:53 2010
From: bob.hinden at gmail.com (Bob Hinden)
Date: Thu, 28 Oct 2010 14:55:53 -0700
Subject: [ih] birth of the Internet?
In-Reply-To: <4CC9DD1E.1030300@meetinghouse.net>
References: <30606.1288294157@vegas.jjd.com>
	<4CC9D1CB.7070504@meetinghouse.net> <4CC9DC01.20601@cisco.com>
	<4CC9DD1E.1030300@meetinghouse.net>
Message-ID: <F171DDDA-7F33-4D87-9B23-EDB05F551E5B@gmail.com>


On Oct 28, 2010, at 1:29 PM, Miles Fidelman wrote:

>> 
>> On 10/28/2010 15:40 EDT, Miles Fidelman wrote:
>>   
>>> Hmmm.... does that suggest that the mailbridge gateways were the
>>> embryonic Internet? :-)
>>>     
>> They weren't fully transparent IP forwarders.
>>   
> As I recall, they weren't supposed to be transparent forwarders, but in practice nobody every configured the filters.  Bob Hinden - care to comment?

We built filtering code that was deployed in the mailbridge gateways.  As I remember, except for the time when the Morris worm struck, it was not turned on.  These may have been the (or as least one of) first firewalls.

Bob







From bob.hinden at gmail.com  Thu Oct 28 15:04:03 2010
From: bob.hinden at gmail.com (Bob Hinden)
Date: Thu, 28 Oct 2010 15:04:03 -0700
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CC9D842.4010906@meetinghouse.net>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
Message-ID: <74199E59-46ED-4C0E-8634-73037F9FE87D@gmail.com>


On Oct 28, 2010, at 1:08 PM, Miles Fidelman wrote:

> Bob,
> 
> Bob Hinden wrote:
>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>> 
>> In my view this was the time when the Internet became operational as people starting using it for their day to day work, instead of a set of researchers.  Conception and birth occurred earlier :-)
>>   
> Actually, that raises another interesting question: At what point, prior to 1/1/83, if any, was there a minimal set of networks, gateways, and end systems that were passing IP packets on an ongoing basis - as opposed to being cobbled together to run some experiment or other, and then brought back down?  Can we isolate a date when IP packets started flowing and never stopped?

There was a set of operational gateways from the Arpanet to Satnet to gateways at UCL, RSRE, and NDRE, MIT, SRI, SAC, Ft. Bragg, etc.   There is a picture dated 25 Feb 1982 at:

  http://en.wikipedia.org/wiki/File:Internet_map_in_February_82.jpg

This was started earlier, but I don't have an exact date.

Bob





From galmes at tamu.edu  Thu Oct 28 15:19:04 2010
From: galmes at tamu.edu (Guy Almes)
Date: Thu, 28 Oct 2010 17:19:04 -0500
Subject: [ih] Origination date for the Internet
In-Reply-To: <AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
Message-ID: <4CC9F6D8.1040708@tamu.edu>

Vint et al.,
   I wonder about how many (mostly departmental) LANs were running 
TCP/IP and connected to the ARPAnet by 1-Jan-83?

   	-- Guy

On 10/28/10 4:44 PM, Vint Cerf wrote:
> actually ISI tracked TCP/IP capability during 1982; the primary
> regular use was from Europe, especially the UK, prior to january 1983;
> by then there LANS connecting to the ARPANET by way of gateways
> (Proteon was around with its rings - Noel Chiappa is that correct?).
> Then came Cisco but i guess after 1984.
>
> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
> with NCP.
>
> v
>
>
> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
> <mfidelman at meetinghouse.net>  wrote:
>> Bob,
>>
>> Bob Hinden wrote:
>>>
>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>
>>> In my view this was the time when the Internet became operational as
>>> people starting using it for their day to day work, instead of a set of
>>> researchers.  Conception and birth occurred earlier :-)
>>>
>>
>> Actually, that raises another interesting question: At what point, prior to
>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>> systems that were passing IP packets on an ongoing basis - as opposed to
>> being cobbled together to run some experiment or other, and then brought
>> back down?  Can we isolate a date when IP packets started flowing and never
>> stopped?
>>
>> Miles
>>
>>
>>
>> --
>> In theory, there is no difference between theory and practice.
>> In<fnord>    practice, there is.   .... Yogi Berra
>>
>>
>>
>
>


From vint at google.com  Thu Oct 28 15:25:31 2010
From: vint at google.com (Vint Cerf)
Date: Thu, 28 Oct 2010 18:25:31 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CC9F6D8.1040708@tamu.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
Message-ID: <AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>

beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
been released, right?

v


On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes <galmes at tamu.edu> wrote:
> Vint et al.,
> ?I wonder about how many (mostly departmental) LANs were running TCP/IP and
> connected to the ARPAnet by 1-Jan-83?
>
> ? ? ? ?-- Guy
>
> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>
>> actually ISI tracked TCP/IP capability during 1982; the primary
>> regular use was from Europe, especially the UK, prior to january 1983;
>> by then there LANS connecting to the ARPANET by way of gateways
>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>> Then came Cisco but i guess after 1984.
>>
>> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
>> with NCP.
>>
>> v
>>
>>
>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>> <mfidelman at meetinghouse.net> ?wrote:
>>>
>>> Bob,
>>>
>>> Bob Hinden wrote:
>>>>
>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>
>>>> In my view this was the time when the Internet became operational as
>>>> people starting using it for their day to day work, instead of a set of
>>>> researchers. ?Conception and birth occurred earlier :-)
>>>>
>>>
>>> Actually, that raises another interesting question: At what point, prior
>>> to
>>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>>> systems that were passing IP packets on an ongoing basis - as opposed to
>>> being cobbled together to run some experiment or other, and then brought
>>> back down? ?Can we isolate a date when IP packets started flowing and
>>> never
>>> stopped?
>>>
>>> Miles
>>>
>>>
>>>
>>> --
>>> In theory, there is no difference between theory and practice.
>>> In<fnord> ? ?practice, there is. ? .... Yogi Berra
>>>
>>>
>>>
>>
>>
>



From galmes at tamu.edu  Thu Oct 28 15:31:48 2010
From: galmes at tamu.edu (Guy Almes)
Date: Thu, 28 Oct 2010 17:31:48 -0500
Subject: [ih] Origination date for the Internet
In-Reply-To: <AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
Message-ID: <4CC9F9D4.8010003@tamu.edu>

Vint,
   I recall that, at the University of Washington, we were waiting for 
Berkeley to mail us a tape with the release (I think 4.1) that we 
needed.  The timeliness of those Berkeley releases in this story is 
probably important.
	-- Guy

On 10/28/10 5:25 PM, Vint Cerf wrote:
> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
> been released, right?
>
> v
>
>
> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu>  wrote:
>> Vint et al.,
>>   I wonder about how many (mostly departmental) LANs were running TCP/IP and
>> connected to the ARPAnet by 1-Jan-83?
>>
>>         -- Guy
>>
>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>
>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>> regular use was from Europe, especially the UK, prior to january 1983;
>>> by then there LANS connecting to the ARPANET by way of gateways
>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>> Then came Cisco but i guess after 1984.
>>>
>>> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
>>> with NCP.
>>>
>>> v
>>>
>>>
>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>> <mfidelman at meetinghouse.net>    wrote:
>>>>
>>>> Bob,
>>>>
>>>> Bob Hinden wrote:
>>>>>
>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>
>>>>> In my view this was the time when the Internet became operational as
>>>>> people starting using it for their day to day work, instead of a set of
>>>>> researchers.  Conception and birth occurred earlier :-)
>>>>>
>>>>
>>>> Actually, that raises another interesting question: At what point, prior
>>>> to
>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>>>> systems that were passing IP packets on an ongoing basis - as opposed to
>>>> being cobbled together to run some experiment or other, and then brought
>>>> back down?  Can we isolate a date when IP packets started flowing and
>>>> never
>>>> stopped?
>>>>
>>>> Miles
>>>>
>>>>
>>>>
>>>> --
>>>> In theory, there is no difference between theory and practice.
>>>> In<fnord>      practice, there is.   .... Yogi Berra
>>>>
>>>>
>>>>
>>>
>>>
>>
>


From galmes at tamu.edu  Thu Oct 28 15:35:52 2010
From: galmes at tamu.edu (Guy Almes)
Date: Thu, 28 Oct 2010 17:35:52 -0500
Subject: [ih] Origination date for the Internet
In-Reply-To: <74199E59-46ED-4C0E-8634-73037F9FE87D@gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>
	<74199E59-46ED-4C0E-8634-73037F9FE87D@gmail.com>
Message-ID: <4CC9FAC8.6080907@tamu.edu>

Bob,
   In this Feb-82 map, it looks like the ARPAnet and SATnet are the two 
key wide-area networks.  What was COMSAT?  Are all the rest LANs?
	-- Guy

On 10/28/10 5:04 PM, Bob Hinden wrote:
>
> On Oct 28, 2010, at 1:08 PM, Miles Fidelman wrote:
>
>> Bob,
>>
>> Bob Hinden wrote:
>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>
>>> In my view this was the time when the Internet became operational as people starting using it for their day to day work, instead of a set of researchers.  Conception and birth occurred earlier :-)
>>>
>> Actually, that raises another interesting question: At what point, prior to 1/1/83, if any, was there a minimal set of networks, gateways, and end systems that were passing IP packets on an ongoing basis - as opposed to being cobbled together to run some experiment or other, and then brought back down?  Can we isolate a date when IP packets started flowing and never stopped?
>
> There was a set of operational gateways from the Arpanet to Satnet to gateways at UCL, RSRE, and NDRE, MIT, SRI, SAC, Ft. Bragg, etc.   There is a picture dated 25 Feb 1982 at:
>
>    http://en.wikipedia.org/wiki/File:Internet_map_in_February_82.jpg
>
> This was started earlier, but I don't have an exact date.
>
> Bob
>
>
>
>


From jack at 3kitty.org  Thu Oct 28 15:44:44 2010
From: jack at 3kitty.org (Jack Haverty)
Date: Thu, 28 Oct 2010 15:44:44 -0700
Subject: [ih] birth of the Internet?
In-Reply-To: <AANLkTikWdWPzoqQSfa2EXpm6jRuvtF4faK6CT2xCS85F@mail.gmail.com>
References: <30606.1288294157@vegas.jjd.com>
	<AANLkTikWdWPzoqQSfa2EXpm6jRuvtF4faK6CT2xCS85F@mail.gmail.com>
Message-ID: <1288305884.2493.137.camel@localhost>

Eric might have been looking at the Internet to try to learn about it or
debug a problem, but I don't think he "monitored" it.  That was a job
for operators in the NOC.  Earlier it was probably the console on Mike
Brescia's or Bob Hinden's desk.  Maybe mine too for a day or two.

/Jack

Vint - FYI, when I send mail to you I get:

<vint at google.com>:
Sorry, no mailbox here by that name. (#5.1.1)



On Thu, 2010-10-28 at 16:05 -0400, Vint Cerf wrote:
> - there was clearly an experimental activity up until 1/1/1983. The
> folks in Europe were forced to use the Packet Satellite network and
> X.25 networks to access the US during 1982 at least, so it could be
> that Eric Rosen was monitoring that as a operational if somewhat
> experimental system? Peter Kirstein would know since he managed at
> least the UK part of the operational SATNET access.
> 
> v
> 
> On Thu, Oct 28, 2010 at 3:29 PM, James J Dempsey <jjd at jjd.com> wrote:
> > Miles asks:
> >
> >> Sure.  Is there anybody here who doesn't think that the (capital I)
> >> Internet started with the ARPANET as its backbone, well before the
> >> NSFnet was a gleam in anybody's eye?
> >
> > With help from someone else's memory, I could argue that The Internet
> > started *before* the ARPANET was connected to it.
> >
> > I recall being in the BBN NOC the day the ARPANET transitioned from NCP to
> > TCP.  That NOC was on the 5th floor of BBN's building 6.  Sometime pretty
> > soon after that, the NOC (and our offices) moved to BBN's building 1.  After
> > the move, the NOC was much larger, managed more networks and had windows to
> > the hallway that allowed visitors to look in at what was going on.
> >
> > At about the time of that move, I remember people talking about the "ARPANET
> > NOC" as being the one in the lab, while the "Internet NOC" was (if I
> > remember correctly) in Eric Rosen's office.  Soon after that, the terminals
> > and displays monitoring the Internet gateways moved into a corner of the NOC
> > proper along with the ARPANET monitoring.
> >
> > I'm not sure if this "Internet NOC" existed before 1 January 1983, but it
> > very well might have.  I'll bet Eric Rosen or Bob Hinden could say so more
> > definitively.  If there was an "Internet NOC" monitoring the Internet
> > gateways before January 1983, I would argue that at that point The Internet
> > did exist, even if the ARPANET was not yet connected to it via TCP/IP.
> >
> >                       --Jim Dempsey--
> >                         jjd at jjd.com
> >
> 




From louie at transsys.com  Thu Oct 28 16:08:16 2010
From: louie at transsys.com (Louis Mamakos)
Date: Thu, 28 Oct 2010 19:08:16 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CC9FAC8.6080907@tamu.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>
	<74199E59-46ED-4C0E-8634-73037F9FE87D@gmail.com>
	<4CC9FAC8.6080907@tamu.edu>
Message-ID: <9F611925-E95B-423E-B05D-703C00908B05@transsys.com>

Comsat was Comsat Labs, and that network was instigated by Dave Mills.  Though not on the picture, at around that time we had a had a low speed connection to one of Dave's fuzzballs[1] in his swamp at Comsat, at the University of Maryland with a few fuzzballs of our own on the other end.

Louis Mamakos

[1] http://www.eecis.udel.edu/~mills/gallery/gallery10.html


On Oct 28, 2010, at 6:35 PM, Guy Almes wrote:

> Bob,
>  In this Feb-82 map, it looks like the ARPAnet and SATnet are the two key wide-area networks.  What was COMSAT?  Are all the rest LANs?
> 	-- Guy
> 
> On 10/28/10 5:04 PM, Bob Hinden wrote:
>> 
>> On Oct 28, 2010, at 1:08 PM, Miles Fidelman wrote:
>> 
>>> Bob,
>>> 
>>> Bob Hinden wrote:
>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>> 
>>>> In my view this was the time when the Internet became operational as people starting using it for their day to day work, instead of a set of researchers.  Conception and birth occurred earlier :-)
>>>> 
>>> Actually, that raises another interesting question: At what point, prior to 1/1/83, if any, was there a minimal set of networks, gateways, and end systems that were passing IP packets on an ongoing basis - as opposed to being cobbled together to run some experiment or other, and then brought back down?  Can we isolate a date when IP packets started flowing and never stopped?
>> 
>> There was a set of operational gateways from the Arpanet to Satnet to gateways at UCL, RSRE, and NDRE, MIT, SRI, SAC, Ft. Bragg, etc.   There is a picture dated 25 Feb 1982 at:
>> 
>>   http://en.wikipedia.org/wiki/File:Internet_map_in_February_82.jpg
>> 
>> This was started earlier, but I don't have an exact date.
>> 
>> Bob
>> 
>> 
>> 
>> 
> 




From bob.hinden at gmail.com  Thu Oct 28 16:10:11 2010
From: bob.hinden at gmail.com (Bob Hinden)
Date: Thu, 28 Oct 2010 16:10:11 -0700
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CC9F6D8.1040708@tamu.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
Message-ID: <A71CC11E-43C5-40A1-ACD0-82FF6E17CF3F@gmail.com>

Guy,

On Oct 28, 2010, at 3:19 PM, Guy Almes wrote:

> Vint et al.,
>  I wonder about how many (mostly departmental) LANs were running TCP/IP and connected to the ARPAnet by 1-Jan-83?

I am not sure.  LANs were still pretty novel.  I think Ethernet was still the big yellow cable that you had to drill into to install a transceiver.  

Bob


> 
>  	-- Guy
> 
> On 10/28/10 4:44 PM, Vint Cerf wrote:
>> actually ISI tracked TCP/IP capability during 1982; the primary
>> regular use was from Europe, especially the UK, prior to january 1983;
>> by then there LANS connecting to the ARPANET by way of gateways
>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>> Then came Cisco but i guess after 1984.
>> 
>> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
>> with NCP.
>> 
>> v
>> 
>> 
>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>> <mfidelman at meetinghouse.net>  wrote:
>>> Bob,
>>> 
>>> Bob Hinden wrote:
>>>> 
>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>> 
>>>> In my view this was the time when the Internet became operational as
>>>> people starting using it for their day to day work, instead of a set of
>>>> researchers.  Conception and birth occurred earlier :-)
>>>> 
>>> 
>>> Actually, that raises another interesting question: At what point, prior to
>>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>>> systems that were passing IP packets on an ongoing basis - as opposed to
>>> being cobbled together to run some experiment or other, and then brought
>>> back down?  Can we isolate a date when IP packets started flowing and never
>>> stopped?
>>> 
>>> Miles
>>> 
>>> 
>>> 
>>> --
>>> In theory, there is no difference between theory and practice.
>>> In<fnord>    practice, there is.   .... Yogi Berra
>>> 
>>> 
>>> 
>> 
>> 




From galmes at tamu.edu  Thu Oct 28 17:18:51 2010
From: galmes at tamu.edu (Guy Almes)
Date: Thu, 28 Oct 2010 19:18:51 -0500
Subject: [ih] Origination date for the Internet
In-Reply-To: <A71CC11E-43C5-40A1-ACD0-82FF6E17CF3F@gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<A71CC11E-43C5-40A1-ACD0-82FF6E17CF3F@gmail.com>
Message-ID: <4CCA12EB.6030200@tamu.edu>

Bob,
   You're right about the big yellow cable.  And, at some sites, the one 
person who could consistently drill into it without shorting it out was 
the one person in the department who could not be fired.
   But this 1983 period saw the confluence of several things:
<> the ARPAnet going to IP
<> Berkeley 4.1 and later 4.2 coming out
<> the VAX-11/750 making it attractive to have multiple small 
time-sharing systems in a department
<> the beginnings of Sun and the emerging "scientific workstation" market
<> the DIX work on 10Mb/s Ethernet becoming mature
   So, the point is just that the rapid growth in LANs was important to 
the Internet becoming important, even apart from multiple wide-area 
networks.
	-- Guy

On 10/28/10 6:10 PM, Bob Hinden wrote:
> Guy,
>
> On Oct 28, 2010, at 3:19 PM, Guy Almes wrote:
>
>> Vint et al.,
>>   I wonder about how many (mostly departmental) LANs were running TCP/IP and connected to the ARPAnet by 1-Jan-83?
>
> I am not sure.  LANs were still pretty novel.  I think Ethernet was still the big yellow cable that you had to drill into to install a transceiver.
>
> Bob
>
>
>>
>>   	-- Guy
>>
>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>> regular use was from Europe, especially the UK, prior to january 1983;
>>> by then there LANS connecting to the ARPANET by way of gateways
>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>> Then came Cisco but i guess after 1984.
>>>
>>> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
>>> with NCP.
>>>
>>> v
>>>
>>>
>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>> <mfidelman at meetinghouse.net>   wrote:
>>>> Bob,
>>>>
>>>> Bob Hinden wrote:
>>>>>
>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>
>>>>> In my view this was the time when the Internet became operational as
>>>>> people starting using it for their day to day work, instead of a set of
>>>>> researchers.  Conception and birth occurred earlier :-)
>>>>>
>>>>
>>>> Actually, that raises another interesting question: At what point, prior to
>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>>>> systems that were passing IP packets on an ongoing basis - as opposed to
>>>> being cobbled together to run some experiment or other, and then brought
>>>> back down?  Can we isolate a date when IP packets started flowing and never
>>>> stopped?
>>>>
>>>> Miles
>>>>
>>>>
>>>>
>>>> --
>>>> In theory, there is no difference between theory and practice.
>>>> In<fnord>     practice, there is.   .... Yogi Berra
>>>>
>>>>
>>>>
>>>
>>>
>
>


From richard at bennett.com  Thu Oct 28 17:37:46 2010
From: richard at bennett.com (Richard Bennett)
Date: Thu, 28 Oct 2010 17:37:46 -0700
Subject: [ih] Origination date for the Internet
In-Reply-To: <AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
Message-ID: <4CCA175A.4090708@bennett.com>

3Com was founded way back in '79, but it took Ron Crane a while to 
figure out the black brick; I think their Multibus adapter was about 
'81. The first single-chip Ethernet controller, the SEEQ 8001, didn't go 
into production until '83.

BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected 
PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was 
moved to Palo Alto shortly afterward.

PAIX came long in 1996 as a carrier-neutral NAP alternative.

RB

On 10/28/2010 3:25 PM, Vint Cerf wrote:
> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
> been released, right?
>
> v
>
>
> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu>  wrote:
>> Vint et al.,
>>   I wonder about how many (mostly departmental) LANs were running TCP/IP and
>> connected to the ARPAnet by 1-Jan-83?
>>
>>         -- Guy
>>
>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>> regular use was from Europe, especially the UK, prior to january 1983;
>>> by then there LANS connecting to the ARPANET by way of gateways
>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>> Then came Cisco but i guess after 1984.
>>>
>>> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
>>> with NCP.
>>>
>>> v
>>>
>>>
>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>> <mfidelman at meetinghouse.net>    wrote:
>>>> Bob,
>>>>
>>>> Bob Hinden wrote:
>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>
>>>>> In my view this was the time when the Internet became operational as
>>>>> people starting using it for their day to day work, instead of a set of
>>>>> researchers.  Conception and birth occurred earlier :-)
>>>>>
>>>> Actually, that raises another interesting question: At what point, prior
>>>> to
>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>>>> systems that were passing IP packets on an ongoing basis - as opposed to
>>>> being cobbled together to run some experiment or other, and then brought
>>>> back down?  Can we isolate a date when IP packets started flowing and
>>>> never
>>>> stopped?
>>>>
>>>> Miles
>>>>
>>>>
>>>>
>>>> --
>>>> In theory, there is no difference between theory and practice.
>>>> In<fnord>      practice, there is.   .... Yogi Berra
>>>>
>>>>
>>>>
>>>

-- 
--
Richard Bennett



From galmes at tamu.edu  Thu Oct 28 17:59:25 2010
From: galmes at tamu.edu (Guy Almes)
Date: Thu, 28 Oct 2010 19:59:25 -0500
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CCA175A.4090708@bennett.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>	<4CC9F6D8.1040708@tamu.edu>	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com>
Message-ID: <4CCA1C6D.1030505@tamu.edu>

Richard,
   Right.
   The original CIX was in 1991 and, interestingly, done as a router 
rather than a switch.  And with a T1 circuit coming from each 
participant.  I always wondered how PSI, UUnet, and CERFnet agreed on 
routing policies.
	-- Guy

On 10/28/10 7:37 PM, Richard Bennett wrote:
> 3Com was founded way back in '79, but it took Ron Crane a while to
> figure out the black brick; I think their Multibus adapter was about
> '81. The first single-chip Ethernet controller, the SEEQ 8001, didn't go
> into production until '83.
>
> BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected
> PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was
> moved to Palo Alto shortly afterward.
>
> PAIX came long in 1996 as a carrier-neutral NAP alternative.
>
> RB
>
> On 10/28/2010 3:25 PM, Vint Cerf wrote:
>> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
>> been released, right?
>>
>> v
>>
>>
>> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu> wrote:
>>> Vint et al.,
>>> I wonder about how many (mostly departmental) LANs were running
>>> TCP/IP and
>>> connected to the ARPAnet by 1-Jan-83?
>>>
>>> -- Guy
>>>
>>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>>> regular use was from Europe, especially the UK, prior to january 1983;
>>>> by then there LANS connecting to the ARPANET by way of gateways
>>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>>> Then came Cisco but i guess after 1984.
>>>>
>>>> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
>>>> with NCP.
>>>>
>>>> v
>>>>
>>>>
>>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>>> <mfidelman at meetinghouse.net> wrote:
>>>>> Bob,
>>>>>
>>>>> Bob Hinden wrote:
>>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>>
>>>>>> In my view this was the time when the Internet became operational as
>>>>>> people starting using it for their day to day work, instead of a
>>>>>> set of
>>>>>> researchers. Conception and birth occurred earlier :-)
>>>>>>
>>>>> Actually, that raises another interesting question: At what point,
>>>>> prior
>>>>> to
>>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>>>>> systems that were passing IP packets on an ongoing basis - as
>>>>> opposed to
>>>>> being cobbled together to run some experiment or other, and then
>>>>> brought
>>>>> back down? Can we isolate a date when IP packets started flowing and
>>>>> never
>>>>> stopped?
>>>>>
>>>>> Miles
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> In theory, there is no difference between theory and practice.
>>>>> In<fnord> practice, there is. .... Yogi Berra
>>>>>
>>>>>
>>>>>
>>>>
>


From tony.li at tony.li  Thu Oct 28 17:59:51 2010
From: tony.li at tony.li (Tony Li)
Date: Thu, 28 Oct 2010 17:59:51 -0700
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CCA175A.4090708@bennett.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com>
Message-ID: <3C2B857C-7F7C-47AB-9878-78974EC711F2@tony.li>


On Oct 28, 2010, at 5:37 PM, Richard Bennett wrote:

> BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected PSI,
                                                                                  ^^^^
                                                                                  7000

Tony





From mfidelman at meetinghouse.net  Thu Oct 28 18:37:44 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Thu, 28 Oct 2010 21:37:44 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <9F611925-E95B-423E-B05D-703C00908B05@transsys.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<74199E59-46ED-4C0E-8634-73037F9FE87D@gmail.com>	<4CC9FAC8.6080907@tamu.edu>
	<9F611925-E95B-423E-B05D-703C00908B05@transsys.com>
Message-ID: <4CCA2568.6090805@meetinghouse.net>

>
> On Oct 28, 2010, at 6:35 PM, Guy Almes wrote:
>
>    
>> Bob,
>>   In this Feb-82 map, it looks like the ARPAnet and SATnet are the two key wide-area networks.  What was COMSAT?  Are all the rest LANs?
>> 	-- Guy
>>
>>      
Come to think of it, when did CSnet come online, and where did it 
connect to everything else?

Miles


-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra




From jcurran at istaff.org  Thu Oct 28 18:45:21 2010
From: jcurran at istaff.org (John Curran)
Date: Thu, 28 Oct 2010 21:45:21 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CCA1C6D.1030505@tamu.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com> <4CCA1C6D.1030505@tamu.edu>
Message-ID: <02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>

Guy - 
 
  Routing policies between peers over the CIX were agreed to on 
  a bilateral basis, just as most peering is done today. The only 
  exception I know of was due to the CIX/ANS interconnection and 
  combits quandary, which resulted in ANS being a transit network 
  for select NSF regional network who joined the CIX.  Paul Vixie 
  was the network engineer configuring the CIX router over most of 
  its operational life, and could supply the specific details...

/John

On Oct 28, 2010, at 8:59 PM, Guy Almes wrote:

> Richard,
>  Right.
>  The original CIX was in 1991 and, interestingly, done as a router rather than a switch.  And with a T1 circuit coming from each participant.  I always wondered how PSI, UUnet, and CERFnet agreed on routing policies.
>   -- Guy
> 
> On 10/28/10 7:37 PM, Richard Bennett wrote:
>> 3Com was founded way back in '79, but it took Ron Crane a while to
>> figure out the black brick; I think their Multibus adapter was about
>> '81. The first single-chip Ethernet controller, the SEEQ 8001, didn't go
>> into production until '83.
>> 
>> BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected
>> PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was
>> moved to Palo Alto shortly afterward.
>> 
>> PAIX came long in 1996 as a carrier-neutral NAP alternative.
>> 
>> RB
>> 
>> On 10/28/2010 3:25 PM, Vint Cerf wrote:
>>> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
>>> been released, right?
>>> 
>>> v
>>> 
>>> 
>>> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu> wrote:
>>>> Vint et al.,
>>>> I wonder about how many (mostly departmental) LANs were running
>>>> TCP/IP and
>>>> connected to the ARPAnet by 1-Jan-83?
>>>> 
>>>> -- Guy
>>>> 
>>>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>>>> regular use was from Europe, especially the UK, prior to january 1983;
>>>>> by then there LANS connecting to the ARPANET by way of gateways
>>>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>>>> Then came Cisco but i guess after 1984.
>>>>> 
>>>>> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
>>>>> with NCP.
>>>>> 
>>>>> v
>>>>> 
>>>>> 
>>>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>>>> <mfidelman at meetinghouse.net> wrote:
>>>>>> Bob,
>>>>>> 
>>>>>> Bob Hinden wrote:
>>>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>>> 
>>>>>>> In my view this was the time when the Internet became operational as
>>>>>>> people starting using it for their day to day work, instead of a
>>>>>>> set of
>>>>>>> researchers. Conception and birth occurred earlier :-)
>>>>>>> 
>>>>>> Actually, that raises another interesting question: At what point,
>>>>>> prior
>>>>>> to
>>>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>>>>>> systems that were passing IP packets on an ongoing basis - as
>>>>>> opposed to
>>>>>> being cobbled together to run some experiment or other, and then
>>>>>> brought
>>>>>> back down? Can we isolate a date when IP packets started flowing and
>>>>>> never
>>>>>> stopped?
>>>>>> 
>>>>>> Miles
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> --
>>>>>> In theory, there is no difference between theory and practice.
>>>>>> In<fnord> practice, there is. .... Yogi Berra
>>>>>> 
>>>>>> 
>>>>>> 
>>>>> 
>> 




From galmes at tamu.edu  Thu Oct 28 19:21:05 2010
From: galmes at tamu.edu (Guy Almes)
Date: Thu, 28 Oct 2010 21:21:05 -0500
Subject: [ih] Origination date for the Internet
In-Reply-To: <02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com> <4CCA1C6D.1030505@tamu.edu>
	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>
Message-ID: <4CCA2F91.20203@tamu.edu>

John,
   My question was really about the original PSI/ UUnet/ CERFnet design 
and, I guess to make it interesting, you'd have to posit a 4th member, 
just say FOOnet.
   Suppose PSI and UUnet each have routes to a given destination D that 
does not involve going through the CIX router.
   Suppose CERFnet and FOOnet connect to the CIX router and need to get 
to D.  Suppose, further, that CERFnet would prefer to get to D via PSI 
and that FOOnet would prefer to get to D via UUnet.
   What would be the routing table entry in the CIX router for D?
   How would "bilateral" agreements help?
   This has puzzled me for almost 20 years,
	-- Guy

On 10/28/10 8:45 PM, John Curran wrote:
> Guy -
>
>    Routing policies between peers over the CIX were agreed to on
>    a bilateral basis, just as most peering is done today. The only
>    exception I know of was due to the CIX/ANS interconnection and
>    combits quandary, which resulted in ANS being a transit network
>    for select NSF regional network who joined the CIX.  Paul Vixie
>    was the network engineer configuring the CIX router over most of
>    its operational life, and could supply the specific details...
>
> /John
>
> On Oct 28, 2010, at 8:59 PM, Guy Almes wrote:
>
>> Richard,
>>   Right.
>>   The original CIX was in 1991 and, interestingly, done as a router rather than a switch.  And with a T1 circuit coming from each participant.  I always wondered how PSI, UUnet, and CERFnet agreed on routing policies.
>>    -- Guy
>>
>> On 10/28/10 7:37 PM, Richard Bennett wrote:
>>> 3Com was founded way back in '79, but it took Ron Crane a while to
>>> figure out the black brick; I think their Multibus adapter was about
>>> '81. The first single-chip Ethernet controller, the SEEQ 8001, didn't go
>>> into production until '83.
>>>
>>> BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected
>>> PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was
>>> moved to Palo Alto shortly afterward.
>>>
>>> PAIX came long in 1996 as a carrier-neutral NAP alternative.
>>>
>>> RB
>>>
>>> On 10/28/2010 3:25 PM, Vint Cerf wrote:
>>>> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
>>>> been released, right?
>>>>
>>>> v
>>>>
>>>>
>>>> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu>  wrote:
>>>>> Vint et al.,
>>>>> I wonder about how many (mostly departmental) LANs were running
>>>>> TCP/IP and
>>>>> connected to the ARPAnet by 1-Jan-83?
>>>>>
>>>>> -- Guy
>>>>>
>>>>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>>>>> regular use was from Europe, especially the UK, prior to january 1983;
>>>>>> by then there LANS connecting to the ARPANET by way of gateways
>>>>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>>>>> Then came Cisco but i guess after 1984.
>>>>>>
>>>>>> Of course during 1982 many ARPANET sites came up on TCP/IP in parallel
>>>>>> with NCP.
>>>>>>
>>>>>> v
>>>>>>
>>>>>>
>>>>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>>>>> <mfidelman at meetinghouse.net>  wrote:
>>>>>>> Bob,
>>>>>>>
>>>>>>> Bob Hinden wrote:
>>>>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>>>>
>>>>>>>> In my view this was the time when the Internet became operational as
>>>>>>>> people starting using it for their day to day work, instead of a
>>>>>>>> set of
>>>>>>>> researchers. Conception and birth occurred earlier :-)
>>>>>>>>
>>>>>>> Actually, that raises another interesting question: At what point,
>>>>>>> prior
>>>>>>> to
>>>>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and end
>>>>>>> systems that were passing IP packets on an ongoing basis - as
>>>>>>> opposed to
>>>>>>> being cobbled together to run some experiment or other, and then
>>>>>>> brought
>>>>>>> back down? Can we isolate a date when IP packets started flowing and
>>>>>>> never
>>>>>>> stopped?
>>>>>>>
>>>>>>> Miles
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> In theory, there is no difference between theory and practice.
>>>>>>> In<fnord>  practice, there is. .... Yogi Berra
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>
>
>


From dhc2 at dcrocker.net  Thu Oct 28 22:33:15 2010
From: dhc2 at dcrocker.net (Dave CROCKER)
Date: Thu, 28 Oct 2010 22:33:15 -0700
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CCA2568.6090805@meetinghouse.net>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<74199E59-46ED-4C0E-8634-73037F9FE87D@gmail.com>	<4CC9FAC8.6080907@tamu.edu>	<9F611925-E95B-423E-B05D-703C00908B05@transsys.com>
	<4CCA2568.6090805@meetinghouse.net>
Message-ID: <4CCA5C9B.1030609@dcrocker.net>



On 10/28/2010 6:37 PM, Miles Fidelman wrote:
>>
>> On Oct 28, 2010, at 6:35 PM, Guy Almes wrote:
>>
>>> Bob,
>>> In this Feb-82 map, it looks like the ARPAnet and SATnet are the two key
>>> wide-area networks. What was COMSAT? Are all the rest LANs?
>>> -- Guy
>>>
> Come to think of it, when did CSnet come online, and where did it connect to
> everything else?


This probably requires a multi-staged answer, since CSNet went through several 
stages of connectivity.  I only worked on the first part.

Initial connectivity was email-only, relayed through telephone calls to or from 
remote sites and thence to the Arpanet.  The initial gateway was at the 
University of Delaware, with  second gateway added quickly at The Rand Corp.

Some of the 'telephone' access was via x.29/x.25 connections.  These, 
essentially, were telephone calls.  The only interesting difference was that 
matching the email protocol data unit to the x.26 packet size was important for 
efficiency and cost.  If the former was larger than the latter, it resulted in 
double the packet count...

The UDel system had a dial-up packet protocol developed by one of the other grad 
students, Ed Szurkowski.  I've always quite liked how it was designed. 
Adaptable but easy enough to use.

d/
-- 

   Dave Crocker
   Brandenburg InternetWorking
   bbiw.net


From vint at google.com  Fri Oct 29 00:20:37 2010
From: vint at google.com (Vint Cerf)
Date: Fri, 29 Oct 2010 03:20:37 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CCA2F91.20203@tamu.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com> <4CCA1C6D.1030505@tamu.edu>
	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>
	<4CCA2F91.20203@tamu.edu>
Message-ID: <AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>

Guy,

You are making an interesting assumption I think. The way I understood
CIX is that it was an ethernet with three routers, one each operated
by PSI, UUnet and CERFnet. There isn't a unique "CIX router". There
are three of them. Does that help?

v


On Thu, Oct 28, 2010 at 10:21 PM, Guy Almes <galmes at tamu.edu> wrote:
> John,
> ?My question was really about the original PSI/ UUnet/ CERFnet design and, I
> guess to make it interesting, you'd have to posit a 4th member, just say
> FOOnet.
> ?Suppose PSI and UUnet each have routes to a given destination D that does
> not involve going through the CIX router.
> ?Suppose CERFnet and FOOnet connect to the CIX router and need to get to D.
> ?Suppose, further, that CERFnet would prefer to get to D via PSI and that
> FOOnet would prefer to get to D via UUnet.
> ?What would be the routing table entry in the CIX router for D?
> ?How would "bilateral" agreements help?
> ?This has puzzled me for almost 20 years,
> ? ? ? ?-- Guy
>
> On 10/28/10 8:45 PM, John Curran wrote:
>>
>> Guy -
>>
>> ? Routing policies between peers over the CIX were agreed to on
>> ? a bilateral basis, just as most peering is done today. The only
>> ? exception I know of was due to the CIX/ANS interconnection and
>> ? combits quandary, which resulted in ANS being a transit network
>> ? for select NSF regional network who joined the CIX. ?Paul Vixie
>> ? was the network engineer configuring the CIX router over most of
>> ? its operational life, and could supply the specific details...
>>
>> /John
>>
>> On Oct 28, 2010, at 8:59 PM, Guy Almes wrote:
>>
>>> Richard,
>>> ?Right.
>>> ?The original CIX was in 1991 and, interestingly, done as a router rather
>>> than a switch. ?And with a T1 circuit coming from each participant. ?I
>>> always wondered how PSI, UUnet, and CERFnet agreed on routing policies.
>>> ? -- Guy
>>>
>>> On 10/28/10 7:37 PM, Richard Bennett wrote:
>>>>
>>>> 3Com was founded way back in '79, but it took Ron Crane a while to
>>>> figure out the black brick; I think their Multibus adapter was about
>>>> '81. The first single-chip Ethernet controller, the SEEQ 8001, didn't go
>>>> into production until '83.
>>>>
>>>> BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected
>>>> PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was
>>>> moved to Palo Alto shortly afterward.
>>>>
>>>> PAIX came long in 1996 as a carrier-neutral NAP alternative.
>>>>
>>>> RB
>>>>
>>>> On 10/28/2010 3:25 PM, Vint Cerf wrote:
>>>>>
>>>>> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
>>>>> been released, right?
>>>>>
>>>>> v
>>>>>
>>>>>
>>>>> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu> ?wrote:
>>>>>>
>>>>>> Vint et al.,
>>>>>> I wonder about how many (mostly departmental) LANs were running
>>>>>> TCP/IP and
>>>>>> connected to the ARPAnet by 1-Jan-83?
>>>>>>
>>>>>> -- Guy
>>>>>>
>>>>>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>>>>>
>>>>>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>>>>>> regular use was from Europe, especially the UK, prior to january
>>>>>>> 1983;
>>>>>>> by then there LANS connecting to the ARPANET by way of gateways
>>>>>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>>>>>> Then came Cisco but i guess after 1984.
>>>>>>>
>>>>>>> Of course during 1982 many ARPANET sites came up on TCP/IP in
>>>>>>> parallel
>>>>>>> with NCP.
>>>>>>>
>>>>>>> v
>>>>>>>
>>>>>>>
>>>>>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>>>>>> <mfidelman at meetinghouse.net> ?wrote:
>>>>>>>>
>>>>>>>> Bob,
>>>>>>>>
>>>>>>>> Bob Hinden wrote:
>>>>>>>>>
>>>>>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>>>>>
>>>>>>>>> In my view this was the time when the Internet became operational
>>>>>>>>> as
>>>>>>>>> people starting using it for their day to day work, instead of a
>>>>>>>>> set of
>>>>>>>>> researchers. Conception and birth occurred earlier :-)
>>>>>>>>>
>>>>>>>> Actually, that raises another interesting question: At what point,
>>>>>>>> prior
>>>>>>>> to
>>>>>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and
>>>>>>>> end
>>>>>>>> systems that were passing IP packets on an ongoing basis - as
>>>>>>>> opposed to
>>>>>>>> being cobbled together to run some experiment or other, and then
>>>>>>>> brought
>>>>>>>> back down? Can we isolate a date when IP packets started flowing and
>>>>>>>> never
>>>>>>>> stopped?
>>>>>>>>
>>>>>>>> Miles
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>>> In theory, there is no difference between theory and practice.
>>>>>>>> In<fnord> ?practice, there is. .... Yogi Berra
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>
>>
>>
>



From richard at bennett.com  Fri Oct 29 03:19:10 2010
From: richard at bennett.com (Richard Bennett)
Date: Fri, 29 Oct 2010 03:19:10 -0700
Subject: [ih] Origination date for the Internet
In-Reply-To: <AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>	<4CC9F6D8.1040708@tamu.edu>	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>	<4CCA175A.4090708@bennett.com>
	<4CCA1C6D.1030505@tamu.edu>	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>	<4CCA2F91.20203@tamu.edu>
	<AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
Message-ID: <4CCA9F9E.5030105@bennett.com>

Per Farooq Hussein's account, the CIX initially used a Cisco AGS 
initially. See page 6 of 
http://www.farooqhussain.org/projects/CIX%20Router%20Description%2010-03-03-final.pdf

He says it's currently at the Equinix office in Reston. The 7500 that 
replaced the AGS after it was moved to Santa Clara is now in the 
Smithsonian. It would be a fun field trip to go see them.

RB

On 10/29/2010 12:20 AM, Vint Cerf wrote:
> Guy,
>
> You are making an interesting assumption I think. The way I understood
> CIX is that it was an ethernet with three routers, one each operated
> by PSI, UUnet and CERFnet. There isn't a unique "CIX router". There
> are three of them. Does that help?
>
> v
>
>
> On Thu, Oct 28, 2010 at 10:21 PM, Guy Almes<galmes at tamu.edu>  wrote:
>> John,
>>   My question was really about the original PSI/ UUnet/ CERFnet design and, I
>> guess to make it interesting, you'd have to posit a 4th member, just say
>> FOOnet.
>>   Suppose PSI and UUnet each have routes to a given destination D that does
>> not involve going through the CIX router.
>>   Suppose CERFnet and FOOnet connect to the CIX router and need to get to D.
>>   Suppose, further, that CERFnet would prefer to get to D via PSI and that
>> FOOnet would prefer to get to D via UUnet.
>>   What would be the routing table entry in the CIX router for D?
>>   How would "bilateral" agreements help?
>>   This has puzzled me for almost 20 years,
>>         -- Guy
>>
>> On 10/28/10 8:45 PM, John Curran wrote:
>>> Guy -
>>>
>>>    Routing policies between peers over the CIX were agreed to on
>>>    a bilateral basis, just as most peering is done today. The only
>>>    exception I know of was due to the CIX/ANS interconnection and
>>>    combits quandary, which resulted in ANS being a transit network
>>>    for select NSF regional network who joined the CIX.  Paul Vixie
>>>    was the network engineer configuring the CIX router over most of
>>>    its operational life, and could supply the specific details...
>>>
>>> /John
>>>
>>> On Oct 28, 2010, at 8:59 PM, Guy Almes wrote:
>>>
>>>> Richard,
>>>>   Right.
>>>>   The original CIX was in 1991 and, interestingly, done as a router rather
>>>> than a switch.  And with a T1 circuit coming from each participant.  I
>>>> always wondered how PSI, UUnet, and CERFnet agreed on routing policies.
>>>>    -- Guy
>>>>
>>>> On 10/28/10 7:37 PM, Richard Bennett wrote:
>>>>> 3Com was founded way back in '79, but it took Ron Crane a while to
>>>>> figure out the black brick; I think their Multibus adapter was about
>>>>> '81. The first single-chip Ethernet controller, the SEEQ 8001, didn't go
>>>>> into production until '83.
>>>>>
>>>>> BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected
>>>>> PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was
>>>>> moved to Palo Alto shortly afterward.
>>>>>
>>>>> PAIX came long in 1996 as a carrier-neutral NAP alternative.
>>>>>
>>>>> RB
>>>>>
>>>>> On 10/28/2010 3:25 PM, Vint Cerf wrote:
>>>>>> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
>>>>>> been released, right?
>>>>>>
>>>>>> v
>>>>>>
>>>>>>
>>>>>> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu>    wrote:
>>>>>>> Vint et al.,
>>>>>>> I wonder about how many (mostly departmental) LANs were running
>>>>>>> TCP/IP and
>>>>>>> connected to the ARPAnet by 1-Jan-83?
>>>>>>>
>>>>>>> -- Guy
>>>>>>>
>>>>>>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>>>>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>>>>>>> regular use was from Europe, especially the UK, prior to january
>>>>>>>> 1983;
>>>>>>>> by then there LANS connecting to the ARPANET by way of gateways
>>>>>>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>>>>>>> Then came Cisco but i guess after 1984.
>>>>>>>>
>>>>>>>> Of course during 1982 many ARPANET sites came up on TCP/IP in
>>>>>>>> parallel
>>>>>>>> with NCP.
>>>>>>>>
>>>>>>>> v
>>>>>>>>
>>>>>>>>
>>>>>>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>>>>>>> <mfidelman at meetinghouse.net>    wrote:
>>>>>>>>> Bob,
>>>>>>>>>
>>>>>>>>> Bob Hinden wrote:
>>>>>>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>>>>>>
>>>>>>>>>> In my view this was the time when the Internet became operational
>>>>>>>>>> as
>>>>>>>>>> people starting using it for their day to day work, instead of a
>>>>>>>>>> set of
>>>>>>>>>> researchers. Conception and birth occurred earlier :-)
>>>>>>>>>>
>>>>>>>>> Actually, that raises another interesting question: At what point,
>>>>>>>>> prior
>>>>>>>>> to
>>>>>>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and
>>>>>>>>> end
>>>>>>>>> systems that were passing IP packets on an ongoing basis - as
>>>>>>>>> opposed to
>>>>>>>>> being cobbled together to run some experiment or other, and then
>>>>>>>>> brought
>>>>>>>>> back down? Can we isolate a date when IP packets started flowing and
>>>>>>>>> never
>>>>>>>>> stopped?
>>>>>>>>>
>>>>>>>>> Miles
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> In theory, there is no difference between theory and practice.
>>>>>>>>> In<fnord>    practice, there is. .... Yogi Berra
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>

-- 
Richard Bennett



From craig at aland.bbn.com  Fri Oct 29 04:04:42 2010
From: craig at aland.bbn.com (Craig Partridge)
Date: Fri, 29 Oct 2010 07:04:42 -0400
Subject: [ih] Origination date for the Internet
Message-ID: <20101029110442.E02A828E139@aland.bbn.com>

> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
> been released, right?

4.1 as I recall.

I arrived at BBN in October '83 at which point 4.2 was due shortly and I was
soon given an early Sun Workstation (entirely metal case, serial # around 200 as
I recall) running the first SunOS (BSD 4.1c) on the BBN corporate Ethernet
(a yellow cable strung in the ceiling).

Also of note, CSNET had been in operation for some time.  There were
stories of many folks in January, during the transition, having trouble
with transitioning to SMTP and deciding it was simplest to relay all their
non-local email to the CSNET relay, as CSNET was sorting out the biggest
messes anyway...

Thanks!

Craig


From galmes at tamu.edu  Fri Oct 29 05:50:40 2010
From: galmes at tamu.edu (Guy Almes)
Date: Fri, 29 Oct 2010 07:50:40 -0500
Subject: [ih] Origination date for the Internet
In-Reply-To: <AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>	<4CC9F6D8.1040708@tamu.edu>	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>	<4CCA175A.4090708@bennett.com>	<4CCA1C6D.1030505@tamu.edu>	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>	<4CCA2F91.20203@tamu.edu>
	<AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
Message-ID: <4CCAC320.5010903@tamu.edu>

Vint,
   That's the odd thing: it *was* a single (Cisco 7xxx) router.  Hence 
the obvious problem I noted.
   With Paul Vixie configuring the router, I suspect that what could be 
done was done.  But I am note aware of any early-90s router technology 
that would fix the problem I noted.
   I'd be interested to hear from Paul or anyone else on what the CIX 
founders planned to do about this problem.  Several of them were 
technically very savvy and would have anticipated the situation/problem.

   This contrasts, of course, with the 1995-era Chicago NAP -- 
implemented with an Ameritech ATM switch.  In that case, using ATM PVCs, 
they were able to do layer-2 switching with both local and long-distance 
circuits.

   By the way, your note shows that you are now having a reaction 
similar to the one I had back in '91.
  	-- Guy

On 10/29/10 2:20 AM, Vint Cerf wrote:
> Guy,
>
> You are making an interesting assumption I think. The way I understood
> CIX is that it was an ethernet with three routers, one each operated
> by PSI, UUnet and CERFnet. There isn't a unique "CIX router". There
> are three of them. Does that help?
>
> v
>
>
> On Thu, Oct 28, 2010 at 10:21 PM, Guy Almes<galmes at tamu.edu>  wrote:
>> John,
>>   My question was really about the original PSI/ UUnet/ CERFnet design and, I
>> guess to make it interesting, you'd have to posit a 4th member, just say
>> FOOnet.
>>   Suppose PSI and UUnet each have routes to a given destination D that does
>> not involve going through the CIX router.
>>   Suppose CERFnet and FOOnet connect to the CIX router and need to get to D.
>>   Suppose, further, that CERFnet would prefer to get to D via PSI and that
>> FOOnet would prefer to get to D via UUnet.
>>   What would be the routing table entry in the CIX router for D?
>>   How would "bilateral" agreements help?
>>   This has puzzled me for almost 20 years,
>>         -- Guy
>>
>> On 10/28/10 8:45 PM, John Curran wrote:
>>>
>>> Guy -
>>>
>>>    Routing policies between peers over the CIX were agreed to on
>>>    a bilateral basis, just as most peering is done today. The only
>>>    exception I know of was due to the CIX/ANS interconnection and
>>>    combits quandary, which resulted in ANS being a transit network
>>>    for select NSF regional network who joined the CIX.  Paul Vixie
>>>    was the network engineer configuring the CIX router over most of
>>>    its operational life, and could supply the specific details...
>>>
>>> /John
>>>
>>> On Oct 28, 2010, at 8:59 PM, Guy Almes wrote:
>>>
>>>> Richard,
>>>>   Right.
>>>>   The original CIX was in 1991 and, interestingly, done as a router rather
>>>> than a switch.  And with a T1 circuit coming from each participant.  I
>>>> always wondered how PSI, UUnet, and CERFnet agreed on routing policies.
>>>>    -- Guy
>>>>
>>>> On 10/28/10 7:37 PM, Richard Bennett wrote:
>>>>>
>>>>> 3Com was founded way back in '79, but it took Ron Crane a while to
>>>>> figure out the black brick; I think their Multibus adapter was about
>>>>> '81. The first single-chip Ethernet controller, the SEEQ 8001, didn't go
>>>>> into production until '83.
>>>>>
>>>>> BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected
>>>>> PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was
>>>>> moved to Palo Alto shortly afterward.
>>>>>
>>>>> PAIX came long in 1996 as a carrier-neutral NAP alternative.
>>>>>
>>>>> RB
>>>>>
>>>>> On 10/28/2010 3:25 PM, Vint Cerf wrote:
>>>>>>
>>>>>> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
>>>>>> been released, right?
>>>>>>
>>>>>> v
>>>>>>
>>>>>>
>>>>>> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu>    wrote:
>>>>>>>
>>>>>>> Vint et al.,
>>>>>>> I wonder about how many (mostly departmental) LANs were running
>>>>>>> TCP/IP and
>>>>>>> connected to the ARPAnet by 1-Jan-83?
>>>>>>>
>>>>>>> -- Guy
>>>>>>>
>>>>>>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>>>>>>
>>>>>>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>>>>>>> regular use was from Europe, especially the UK, prior to january
>>>>>>>> 1983;
>>>>>>>> by then there LANS connecting to the ARPANET by way of gateways
>>>>>>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>>>>>>> Then came Cisco but i guess after 1984.
>>>>>>>>
>>>>>>>> Of course during 1982 many ARPANET sites came up on TCP/IP in
>>>>>>>> parallel
>>>>>>>> with NCP.
>>>>>>>>
>>>>>>>> v
>>>>>>>>
>>>>>>>>
>>>>>>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>>>>>>> <mfidelman at meetinghouse.net>    wrote:
>>>>>>>>>
>>>>>>>>> Bob,
>>>>>>>>>
>>>>>>>>> Bob Hinden wrote:
>>>>>>>>>>
>>>>>>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>>>>>>
>>>>>>>>>> In my view this was the time when the Internet became operational
>>>>>>>>>> as
>>>>>>>>>> people starting using it for their day to day work, instead of a
>>>>>>>>>> set of
>>>>>>>>>> researchers. Conception and birth occurred earlier :-)
>>>>>>>>>>
>>>>>>>>> Actually, that raises another interesting question: At what point,
>>>>>>>>> prior
>>>>>>>>> to
>>>>>>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and
>>>>>>>>> end
>>>>>>>>> systems that were passing IP packets on an ongoing basis - as
>>>>>>>>> opposed to
>>>>>>>>> being cobbled together to run some experiment or other, and then
>>>>>>>>> brought
>>>>>>>>> back down? Can we isolate a date when IP packets started flowing and
>>>>>>>>> never
>>>>>>>>> stopped?
>>>>>>>>>
>>>>>>>>> Miles
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> In theory, there is no difference between theory and practice.
>>>>>>>>> In<fnord>    practice, there is. .... Yogi Berra
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>
>>>
>>>
>>
>


From vint at google.com  Fri Oct 29 05:54:19 2010
From: vint at google.com (Vint Cerf)
Date: Fri, 29 Oct 2010 08:54:19 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CCAC320.5010903@tamu.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com> <4CCA1C6D.1030505@tamu.edu>
	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>
	<4CCA2F91.20203@tamu.edu>
	<AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
	<4CCAC320.5010903@tamu.edu>
Message-ID: <AANLkTi=LgypJ5hZdxiCLGMQZDN9Mf=ZGAT1WvkEu1AWc@mail.gmail.com>

Guy,

so I learned from Randy Bush about the Cisco 7500 (and maybe AGS before).

now I gotta go draw some diagrams!

v


On Fri, Oct 29, 2010 at 8:50 AM, Guy Almes <galmes at tamu.edu> wrote:
> Vint,
> ?That's the odd thing: it *was* a single (Cisco 7xxx) router. ?Hence the
> obvious problem I noted.
> ?With Paul Vixie configuring the router, I suspect that what could be done
> was done. ?But I am note aware of any early-90s router technology that would
> fix the problem I noted.
> ?I'd be interested to hear from Paul or anyone else on what the CIX founders
> planned to do about this problem. ?Several of them were technically very
> savvy and would have anticipated the situation/problem.
>
> ?This contrasts, of course, with the 1995-era Chicago NAP -- implemented
> with an Ameritech ATM switch. ?In that case, using ATM PVCs, they were able
> to do layer-2 switching with both local and long-distance circuits.
>
> ?By the way, your note shows that you are now having a reaction similar to
> the one I had back in '91.
> ? ? ? ?-- Guy
>
> On 10/29/10 2:20 AM, Vint Cerf wrote:
>>
>> Guy,
>>
>> You are making an interesting assumption I think. The way I understood
>> CIX is that it was an ethernet with three routers, one each operated
>> by PSI, UUnet and CERFnet. There isn't a unique "CIX router". There
>> are three of them. Does that help?
>>
>> v
>>
>>
>> On Thu, Oct 28, 2010 at 10:21 PM, Guy Almes<galmes at tamu.edu> ?wrote:
>>>
>>> John,
>>> ?My question was really about the original PSI/ UUnet/ CERFnet design
>>> and, I
>>> guess to make it interesting, you'd have to posit a 4th member, just say
>>> FOOnet.
>>> ?Suppose PSI and UUnet each have routes to a given destination D that
>>> does
>>> not involve going through the CIX router.
>>> ?Suppose CERFnet and FOOnet connect to the CIX router and need to get to
>>> D.
>>> ?Suppose, further, that CERFnet would prefer to get to D via PSI and that
>>> FOOnet would prefer to get to D via UUnet.
>>> ?What would be the routing table entry in the CIX router for D?
>>> ?How would "bilateral" agreements help?
>>> ?This has puzzled me for almost 20 years,
>>> ? ? ? ?-- Guy
>>>
>>> On 10/28/10 8:45 PM, John Curran wrote:
>>>>
>>>> Guy -
>>>>
>>>> ? Routing policies between peers over the CIX were agreed to on
>>>> ? a bilateral basis, just as most peering is done today. The only
>>>> ? exception I know of was due to the CIX/ANS interconnection and
>>>> ? combits quandary, which resulted in ANS being a transit network
>>>> ? for select NSF regional network who joined the CIX. ?Paul Vixie
>>>> ? was the network engineer configuring the CIX router over most of
>>>> ? its operational life, and could supply the specific details...
>>>>
>>>> /John
>>>>
>>>> On Oct 28, 2010, at 8:59 PM, Guy Almes wrote:
>>>>
>>>>> Richard,
>>>>> ?Right.
>>>>> ?The original CIX was in 1991 and, interestingly, done as a router
>>>>> rather
>>>>> than a switch. ?And with a T1 circuit coming from each participant. ?I
>>>>> always wondered how PSI, UUnet, and CERFnet agreed on routing policies.
>>>>> ? -- Guy
>>>>>
>>>>> On 10/28/10 7:37 PM, Richard Bennett wrote:
>>>>>>
>>>>>> 3Com was founded way back in '79, but it took Ron Crane a while to
>>>>>> figure out the black brick; I think their Multibus adapter was about
>>>>>> '81. The first single-chip Ethernet controller, the SEEQ 8001, didn't
>>>>>> go
>>>>>> into production until '83.
>>>>>>
>>>>>> BTW, as far as I can tell, the first CIX was a Cisco 7500 that
>>>>>> connected
>>>>>> PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was
>>>>>> moved to Palo Alto shortly afterward.
>>>>>>
>>>>>> PAIX came long in 1996 as a carrier-neutral NAP alternative.
>>>>>>
>>>>>> RB
>>>>>>
>>>>>> On 10/28/2010 3:25 PM, Vint Cerf wrote:
>>>>>>>
>>>>>>> beats me - 3COM was in operation by then and Berkeley BSD 4.x had
>>>>>>> also
>>>>>>> been released, right?
>>>>>>>
>>>>>>> v
>>>>>>>
>>>>>>>
>>>>>>> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu> ? ?wrote:
>>>>>>>>
>>>>>>>> Vint et al.,
>>>>>>>> I wonder about how many (mostly departmental) LANs were running
>>>>>>>> TCP/IP and
>>>>>>>> connected to the ARPAnet by 1-Jan-83?
>>>>>>>>
>>>>>>>> -- Guy
>>>>>>>>
>>>>>>>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>>>>>>>
>>>>>>>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>>>>>>>> regular use was from Europe, especially the UK, prior to january
>>>>>>>>> 1983;
>>>>>>>>> by then there LANS connecting to the ARPANET by way of gateways
>>>>>>>>> (Proteon was around with its rings - Noel Chiappa is that
>>>>>>>>> correct?).
>>>>>>>>> Then came Cisco but i guess after 1984.
>>>>>>>>>
>>>>>>>>> Of course during 1982 many ARPANET sites came up on TCP/IP in
>>>>>>>>> parallel
>>>>>>>>> with NCP.
>>>>>>>>>
>>>>>>>>> v
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>>>>>>>> <mfidelman at meetinghouse.net> ? ?wrote:
>>>>>>>>>>
>>>>>>>>>> Bob,
>>>>>>>>>>
>>>>>>>>>> Bob Hinden wrote:
>>>>>>>>>>>
>>>>>>>>>>> I still have my "I Survived the TCP Transition 1/1/83" red
>>>>>>>>>>> button.
>>>>>>>>>>>
>>>>>>>>>>> In my view this was the time when the Internet became operational
>>>>>>>>>>> as
>>>>>>>>>>> people starting using it for their day to day work, instead of a
>>>>>>>>>>> set of
>>>>>>>>>>> researchers. Conception and birth occurred earlier :-)
>>>>>>>>>>>
>>>>>>>>>> Actually, that raises another interesting question: At what point,
>>>>>>>>>> prior
>>>>>>>>>> to
>>>>>>>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and
>>>>>>>>>> end
>>>>>>>>>> systems that were passing IP packets on an ongoing basis - as
>>>>>>>>>> opposed to
>>>>>>>>>> being cobbled together to run some experiment or other, and then
>>>>>>>>>> brought
>>>>>>>>>> back down? Can we isolate a date when IP packets started flowing
>>>>>>>>>> and
>>>>>>>>>> never
>>>>>>>>>> stopped?
>>>>>>>>>>
>>>>>>>>>> Miles
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> --
>>>>>>>>>> In theory, there is no difference between theory and practice.
>>>>>>>>>> In<fnord> ? ?practice, there is. .... Yogi Berra
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>
>>>>>>
>>>>
>>>>
>>>
>>
>



From louie at transsys.com  Fri Oct 29 06:15:11 2010
From: louie at transsys.com (Louis Mamakos)
Date: Fri, 29 Oct 2010 09:15:11 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com> <4CCA1C6D.1030505@tamu.edu>
	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>
	<4CCA2F91.20203@tamu.edu>
	<AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
Message-ID: <D942477A-BA4A-453A-85DB-EE992A9C6E7B@transsys.com>

I think the CIX was always a router interconnect.  It was MAE-EAST that manifested initially as an (10Mb/s) Ethernet fabric, and quickly used an early metro-ethernet interconnection via DS3 circuits to move out of one physical location.  It was clearly much easier to manage policy for bilateral peering arrangements using the L2 model as compared to the CIX (at least for larger, "tier 1" operators.)  

Louis Mamakos

On Oct 29, 2010, at 3:20 AM, Vint Cerf wrote:

> Guy,
> 
> You are making an interesting assumption I think. The way I understood
> CIX is that it was an ethernet with three routers, one each operated
> by PSI, UUnet and CERFnet. There isn't a unique "CIX router". There
> are three of them. Does that help?
> 
> v
> 
> 
> On Thu, Oct 28, 2010 at 10:21 PM, Guy Almes <galmes at tamu.edu> wrote:
>> John,
>>  My question was really about the original PSI/ UUnet/ CERFnet design and, I
>> guess to make it interesting, you'd have to posit a 4th member, just say
>> FOOnet.
>>  Suppose PSI and UUnet each have routes to a given destination D that does
>> not involve going through the CIX router.
>>  Suppose CERFnet and FOOnet connect to the CIX router and need to get to D.
>>  Suppose, further, that CERFnet would prefer to get to D via PSI and that
>> FOOnet would prefer to get to D via UUnet.
>>  What would be the routing table entry in the CIX router for D?
>>  How would "bilateral" agreements help?
>>  This has puzzled me for almost 20 years,
>>        -- Guy
>> 
>> On 10/28/10 8:45 PM, John Curran wrote:
>>> 
>>> Guy -
>>> 
>>>   Routing policies between peers over the CIX were agreed to on
>>>   a bilateral basis, just as most peering is done today. The only
>>>   exception I know of was due to the CIX/ANS interconnection and
>>>   combits quandary, which resulted in ANS being a transit network
>>>   for select NSF regional network who joined the CIX.  Paul Vixie
>>>   was the network engineer configuring the CIX router over most of
>>>   its operational life, and could supply the specific details...
>>> 
>>> /John
>>> 
>>> On Oct 28, 2010, at 8:59 PM, Guy Almes wrote:
>>> 
>>>> Richard,
>>>>  Right.
>>>>  The original CIX was in 1991 and, interestingly, done as a router rather
>>>> than a switch.  And with a T1 circuit coming from each participant.  I
>>>> always wondered how PSI, UUnet, and CERFnet agreed on routing policies.
>>>>   -- Guy
>>>> 
>>>> On 10/28/10 7:37 PM, Richard Bennett wrote:
>>>>> 
>>>>> 3Com was founded way back in '79, but it took Ron Crane a while to
>>>>> figure out the black brick; I think their Multibus adapter was about
>>>>> '81. The first single-chip Ethernet controller, the SEEQ 8001, didn't go
>>>>> into production until '83.
>>>>> 
>>>>> BTW, as far as I can tell, the first CIX was a Cisco 7500 that connected
>>>>> PSI, UUNET, and Cerfnet somewhere around McLean, VA in 1991. It was
>>>>> moved to Palo Alto shortly afterward.
>>>>> 
>>>>> PAIX came long in 1996 as a carrier-neutral NAP alternative.
>>>>> 
>>>>> RB
>>>>> 
>>>>> On 10/28/2010 3:25 PM, Vint Cerf wrote:
>>>>>> 
>>>>>> beats me - 3COM was in operation by then and Berkeley BSD 4.x had also
>>>>>> been released, right?
>>>>>> 
>>>>>> v
>>>>>> 
>>>>>> 
>>>>>> On Thu, Oct 28, 2010 at 6:19 PM, Guy Almes<galmes at tamu.edu>  wrote:
>>>>>>> 
>>>>>>> Vint et al.,
>>>>>>> I wonder about how many (mostly departmental) LANs were running
>>>>>>> TCP/IP and
>>>>>>> connected to the ARPAnet by 1-Jan-83?
>>>>>>> 
>>>>>>> -- Guy
>>>>>>> 
>>>>>>> On 10/28/10 4:44 PM, Vint Cerf wrote:
>>>>>>>> 
>>>>>>>> actually ISI tracked TCP/IP capability during 1982; the primary
>>>>>>>> regular use was from Europe, especially the UK, prior to january
>>>>>>>> 1983;
>>>>>>>> by then there LANS connecting to the ARPANET by way of gateways
>>>>>>>> (Proteon was around with its rings - Noel Chiappa is that correct?).
>>>>>>>> Then came Cisco but i guess after 1984.
>>>>>>>> 
>>>>>>>> Of course during 1982 many ARPANET sites came up on TCP/IP in
>>>>>>>> parallel
>>>>>>>> with NCP.
>>>>>>>> 
>>>>>>>> v
>>>>>>>> 
>>>>>>>> 
>>>>>>>> On Thu, Oct 28, 2010 at 4:08 PM, Miles Fidelman
>>>>>>>> <mfidelman at meetinghouse.net>  wrote:
>>>>>>>>> 
>>>>>>>>> Bob,
>>>>>>>>> 
>>>>>>>>> Bob Hinden wrote:
>>>>>>>>>> 
>>>>>>>>>> I still have my "I Survived the TCP Transition 1/1/83" red button.
>>>>>>>>>> 
>>>>>>>>>> In my view this was the time when the Internet became operational
>>>>>>>>>> as
>>>>>>>>>> people starting using it for their day to day work, instead of a
>>>>>>>>>> set of
>>>>>>>>>> researchers. Conception and birth occurred earlier :-)
>>>>>>>>>> 
>>>>>>>>> Actually, that raises another interesting question: At what point,
>>>>>>>>> prior
>>>>>>>>> to
>>>>>>>>> 1/1/83, if any, was there a minimal set of networks, gateways, and
>>>>>>>>> end
>>>>>>>>> systems that were passing IP packets on an ongoing basis - as
>>>>>>>>> opposed to
>>>>>>>>> being cobbled together to run some experiment or other, and then
>>>>>>>>> brought
>>>>>>>>> back down? Can we isolate a date when IP packets started flowing and
>>>>>>>>> never
>>>>>>>>> stopped?
>>>>>>>>> 
>>>>>>>>> Miles
>>>>>>>>> 
>>>>>>>>> 
>>>>>>>>> 
>>>>>>>>> --
>>>>>>>>> In theory, there is no difference between theory and practice.
>>>>>>>>> In<fnord>  practice, there is. .... Yogi Berra
>>>>>>>>> 
>>>>>>>>> 
>>>>>>>>> 
>>>>>>>> 
>>>>> 
>>> 
>>> 
>> 
> 




From randy at psg.com  Fri Oct 29 06:42:40 2010
From: randy at psg.com (Randy Bush)
Date: Fri, 29 Oct 2010 22:42:40 +0900
Subject: [ih] Origination date for the Internet
In-Reply-To: <AANLkTi=LgypJ5hZdxiCLGMQZDN9Mf=ZGAT1WvkEu1AWc@mail.gmail.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com> <4CCA1C6D.1030505@tamu.edu>
	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>
	<4CCA2F91.20203@tamu.edu>
	<AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
	<4CCAC320.5010903@tamu.edu>
	<AANLkTi=LgypJ5hZdxiCLGMQZDN9Mf=ZGAT1WvkEu1AWc@mail.gmail.com>
Message-ID: <m2tyk515f3.wl%randy@psg.com>

[ i really had planned to remain silent, sigh ]

> so I learned from Randy Bush about the Cisco 7500 (and maybe AGS
> before).

s/7500/7000/ as someone noted earlier.  and yes, the cix as number was
in the middle of the path.  and under-provisioning was a game played by
some, uu in particular, to pretend to be open about peering while not
really being so.

uu, psi, and later cerf and more were bilats, though the social
agreement was that it was to be a multi-lat.  ans bitched and whined,
and really tried to hold on to what they thought was an nsf-granted
monopoly.  susan estrada has not spoken to me since i publicly offered
to pay ans's $10k cix fee if they would stop whining and join.

this should all be on the com-priv list, wherever that is archived.  but
one minute with google did not find it.

randy


From galmes at tamu.edu  Fri Oct 29 09:32:56 2010
From: galmes at tamu.edu (Guy Almes)
Date: Fri, 29 Oct 2010 11:32:56 -0500
Subject: [ih] Origination date for the Internet
In-Reply-To: <m2tyk515f3.wl%randy@psg.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>	<4CC9B572.9080300@isi.edu>	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>	<4CC9D842.4010906@meetinghouse.net>	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>	<4CC9F6D8.1040708@tamu.edu>	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>	<4CCA175A.4090708@bennett.com>
	<4CCA1C6D.1030505@tamu.edu>	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>	<4CCA2F91.20203@tamu.edu>	<AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>	<4CCAC320.5010903@tamu.edu>	<AANLkTi=LgypJ5hZdxiCLGMQZDN9Mf=ZGAT1WvkEu1AWc@mail.gmail.com>
	<m2tyk515f3.wl%randy@psg.com>
Message-ID: <4CCAF738.3040103@tamu.edu>

Randy,
   I was trying to emphasize the technical aspect.
   Did the CIX pioneers not realize, at the outset, that using a router 
as an interconnect would make things odd in an (ahem) dynamically 
evolving business environment?
   My impression is that all subsequent interconnects have been layer-2 
switches and such.

   Regards,
	-- Guy
p.s.: And, as for the non-technical issues of that era, thanks for being 
one of the grownups

On 10/29/10 8:42 AM, Randy Bush wrote:
> [ i really had planned to remain silent, sigh ]
>
>> so I learned from Randy Bush about the Cisco 7500 (and maybe AGS
>> before).
>
> s/7500/7000/ as someone noted earlier.  and yes, the cix as number was
> in the middle of the path.  and under-provisioning was a game played by
> some, uu in particular, to pretend to be open about peering while not
> really being so.
>
> uu, psi, and later cerf and more were bilats, though the social
> agreement was that it was to be a multi-lat.  ans bitched and whined,
> and really tried to hold on to what they thought was an nsf-granted
> monopoly.  susan estrada has not spoken to me since i publicly offered
> to pay ans's $10k cix fee if they would stop whining and join.
>
> this should all be on the com-priv list, wherever that is archived.  but
> one minute with google did not find it.
>
> randy
>


From johnl at iecc.com  Fri Oct 29 11:00:04 2010
From: johnl at iecc.com (John Levine)
Date: 29 Oct 2010 18:00:04 -0000
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CCAF738.3040103@tamu.edu>
Message-ID: <20101029180004.35621.qmail@joyce.lan>

>   Did the CIX pioneers not realize, at the outset, that using a router 
>as an interconnect would make things odd in an (ahem) dynamically 
>evolving business environment?

My recollection of CIX was that it was more of a political move than a
technical one.  It was a lot harder for networks to demand high
interconnect fees when the other network could say, never mind, we'll
go through CIX.  The performance would be lousy, but that's OK, since
now the customers of the network who wanted the money would complain
about it.

R's,
John


From randy at psg.com  Fri Oct 29 11:38:09 2010
From: randy at psg.com (Randy Bush)
Date: Sat, 30 Oct 2010 03:38:09 +0900
Subject: [ih] Origination date for the Internet
In-Reply-To: <4CCAF738.3040103@tamu.edu>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com> <4CCA1C6D.1030505@tamu.edu>
	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>
	<4CCA2F91.20203@tamu.edu>
	<AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
	<4CCAC320.5010903@tamu.edu>
	<AANLkTi=LgypJ5hZdxiCLGMQZDN9Mf=ZGAT1WvkEu1AWc@mail.gmail.com>
	<m2tyk515f3.wl%randy@psg.com> <4CCAF738.3040103@tamu.edu>
Message-ID: <m2lj5g26b2.wl%randy@psg.com>

> Did the CIX pioneers not realize, at the outset, that using a router
> as an interconnect would make things odd in an (ahem) dynamically
> evolving business environment?

there was what was characterized by some as a religious discussion.
some just thought it was router envy.  other that everything looked like
a nail.  but this was pretty early, so give 'em a break.

but the cix was as much a political/business act than a long term
scalable exchange point.  as louie says, mae east was for real (except
for those damned datakits!).

the jokes were the nsf-sponsored 'naps', like the ans monopoly none of
which survived.  of the nsf dump and run, little was actually viable,
sprintlink being the notable exception, thanks steve goldstein, dima
volodin, ...

> My impression is that all subsequent interconnects have been layer-2
> switches and such.

no, just all the sane ones.  folk like greene, woodcock, ... were still
pushing layer three connects into the 21st century.  folk still push it
today, but in the poorer countries.  they test drugs there too.

</rant>  i knew i did not want to enter this discussion.

randy


From louie at transsys.com  Fri Oct 29 13:34:18 2010
From: louie at transsys.com (Louis Mamakos)
Date: Fri, 29 Oct 2010 16:34:18 -0400
Subject: [ih] Origination date for the Internet
In-Reply-To: <m2lj5g26b2.wl%randy@psg.com>
References: <mailman.10036.1288283998.753.internet-history@postel.org>
	<4CC9B572.9080300@isi.edu>
	<CA3FF539-B9DB-424A-9DB6-53A212401A44@gmail.com>
	<4CC9D842.4010906@meetinghouse.net>
	<AANLkTi=VA-uN+7KkYN7Q_R4TMZD=ceQy6OFEdocFpbTM@mail.gmail.com>
	<4CC9F6D8.1040708@tamu.edu>
	<AANLkTikt+aP8TySL59rwjm4PQz+KLZHwMeMWrrtWw8m7@mail.gmail.com>
	<4CCA175A.4090708@bennett.com> <4CCA1C6D.1030505@tamu.edu>
	<02F34EF5-1388-4677-AA22-7557710F559C@istaff.org>
	<4CCA2F91.20203@tamu.edu>
	<AANLkTi=vc0zcxwSfTUQRjf5kp9GWO1O1iq5gd6EDmQ-6@mail.gmail.com>
	<4CCAC320.5010903@tamu.edu>
	<AANLkTi=LgypJ5hZdxiCLGMQZDN9Mf=ZGAT1WvkEu1AWc@mail.gmail.com>
	<m2tyk515f3.wl%randy@psg.com> <4CCAF738.3040103@tamu.edu>
	<m2lj5g26b2.wl%randy@psg.com>
Message-ID: <91E6B985-758C-4981-9FA9-A823E29ACA49@transsys.com>


On Oct 29, 2010, at 2:38 PM, Randy Bush wrote:

>> Did the CIX pioneers not realize, at the outset, that using a router
>> as an interconnect would make things odd in an (ahem) dynamically
>> evolving business environment?
> 
> there was what was characterized by some as a religious discussion.
> some just thought it was router envy.  other that everything looked like
> a nail.  but this was pretty early, so give 'em a break.

As a practical matter, at the time it was pretty obvious how to order
a T-1 circuit between the CIX router and your router and plug the two
router interfaces together.  It wasn't nearly as obvious how to do
that with an L2 switch at the time, with the stuff you already had
deployed in your network.  The later L2/Ethernet exchanges initially
required you to colocate your router next to the switch (be it
Ethernet, FDDI, whatever).  It was later before MFS figured out
how to extend the ethernet hose to your premise over their metro
ethernet service.
> 
> but the cix was as much a political/business act than a long term
> scalable exchange point.  as louie says, mae east was for real (except
> for those damned datakits!).

I recall these were some sort of ADC box connected in a ring configuration
with DS3 circuits, or something. 

> the jokes were the nsf-sponsored 'naps', like the ans monopoly none of
> which survived.  of the nsf dump and run, little was actually viable,
> sprintlink being the notable exception, thanks steve goldstein, dima
> volodin, ...

The "Washington NAP" mostly just was an extension of MAE-EAST, and largely
duplicative.  The Sprint NAP in NY was incrementally new in capability
and location, though I'm trying to recall if UUNET eventually connected
to it or not; private bilateral peering was starting to take off by
then..

Louis Mamakos



From mfidelman at meetinghouse.net  Sat Oct 30 11:21:36 2010
From: mfidelman at meetinghouse.net (Miles Fidelman)
Date: Sat, 30 Oct 2010 14:21:36 -0400
Subject: [ih] turning off the ARPANET
Message-ID: <4CCC6230.5060603@meetinghouse.net>

Roger Fradenburgh shared this, which I repost with his permission:

-------------
Ah yes. I can't claim to have been involved with the Arpanet's birth, 
but I was in the thick of things when the DARPA/ISTO PM decided it was 
time for euthanasia.

It wasn't really "turned off" in the manner those words tend to infer - 
It would be more appropriate to think of it as having been disassembled 
over a period of several months. Jan Whitener, who I THINK was working 
at SAIC at the time, handled the user side of things, notifying various 
entities their connections were going away, so they'd best start making 
other arrangements. BBN handled the trunks and nodes as DARPA 
iteratively identified specific ones they wanted to retire (as in "stop 
paying for"), with Yours Truly acting in something of a PM role.

I don't recall the exact date the last IMP (or IMPs) went off the air, 
but along the way I relied heavily on our net analysis group to make 
sure what was left of the network could still service what was left of 
the user community. Once an IMP slated to be decommissioned was isolated 
(all trunks disconnected), it could be powered down and hauled away.

...So I guess it kinda was "turned off," but a chunk at a time, not all 
at once.
-------------

-- 
In theory, there is no difference between theory and practice.
In<fnord>  practice, there is.   .... Yogi Berra