[ih] IANA

Vint Cerf vint at google.com
Sun Aug 30 12:11:42 PDT 2009


Thanks jake!

----- Original Message -----
From: Elizabeth Feinler <feinler at earthlink.net>
To: Vint Cerf <vint at google.com>
Cc: Steve Crocker <steve at shinkuro.com>; johnnyryan1 at gmail.com
<johnnyryan1 at gmail.com>; internet-history at postel.org
<internet-history at postel.org>; Bob Braden <braden at ISI.EDU>
Sent: Sun Aug 30 14:32:52 2009
Subject: Re: [ih] IANA

Vint, Steve, et al,

My recollection was that Jon came to SRI about 1973 and left to join
USC-ISI when Engelbart's group was planning to spin off to TimeShare -
about 1976.  If you want the exact dates for your records, I can check
with Personnel at SRI.

Jon never maintained any host tables, even when he was at SRI.  That
was always done by the NIC.  In the early days BBN informed the NIC
when a new IMP was added to the network.  BBN provided us with any
contacts they had, and the NIC followed up to get a host name,
Technical Liaison, etc.  Peggy Karp was the first keeper of the host
table.  Jeanne North (Reddy Dively) under Dick Watson took over for a
short time, then, when I joined the NIC in 1972 it became my job.
Mike Kudlick and I formulated some rules early on to help formalize
the host table.  The NIC maintained both the "flat" table and then the
DNS host table until the NIC project left SRI in 1991.   Mary Stahl
and Sue Romano were in charge of this effort.  In the early days folks
refreshed their host tables from ours.  Soon that created a bottle
neck, so we distributed the table every night to several other sites
that could help with the "refresh".   (ISI was one of those
"refresher" sites. ). Then the NIC provided host information under the
WHOIS and NAMSER servers, and also came up with the TLD naming scheme
that is still in use today, i.e., .com,  .gov, .mil, .org, .edu.  We
coordinated the group called Host Administrators who had authority
over host matters at the various sites.  Jon (by this time at ISI) and
Joyce Reynolds maintained the Assigned Numbers list until the mid
eighties (can't remember exact date, but can look it up if anyone is
interested).  At that time the NIC took over the maintenance of
Assignd Numbers as well.  The split between Assigned Numbers and Host
Tables was always a little odd (one was DCA funded the other ARPA
funded) so we all kept in close contact with each other at all times.
For several years I  kept Jon as a consultant on my contract to advise
the NIC and keep things in synch.  This was especially useful when we
switched to the DNS.

Hope this helps.

Regards,

Jake


On Aug 29, 2009, at 10:23 AM, Vint Cerf wrote:

> Steve,
>
> thanks for trying to get this straight. It's amazing how much we've
> forgotten.
>
> Jake, when did Jon come to SRI? Did the NIC maintain the host.txt
> file initially (that sounds right to me)?
>
> vint
>
> On Aug 29, 2009, at 1:05 PM, Steve Crocker wrote:
>
>> Johnny,
>>
>> I don't think this quite right.  The addresses for machines in the
>> Arpanet were formed two numbers.  One number was the address of the
>> IMP.  The other was from the particular interface on the IMP that
>> connected to the host.  The IMPs were numbered from 1 to 63.  The
>> interfaces were numbered from 0 to 3.  The host address was the IMP
>> address plus 64 times the interface.  UCLA had IMP number 1.  The
>> first host attached to it was addressed as host 1.  The second
>> host's address was 65.  Etc.
>>
>> IMP numbers were assigned by BBN in the order they were installed.
>> Host interfaces were used pretty much in order at each institution.
>>
>> I don't remember who precisely kept track of the overall list.  It
>> wasn't at UCLA and it wasn't in the early RFCs.  I believe SRI took
>> on this job, and eventually distributed the information in the form
>> of the host table.  Pretty early in the process we started using
>> names for the hosts, e.g. "UCLA" for host number 1, "SRI" for host
>> number 2, and then longer names to distinguish between hosts at
>> each institution, e.g. "UCLA-CCN" (I think) host 65.
>>
>> My memory of this is probably not 100% perfect, so I apologize in
>> advance if some details are wrong.  Please do cross check.
>>
>> Bob Braden can probably tell you much more, and perhaps ex SRI
>> people can fill in some blanks.  Check with Jake Feinler.  I have
>> cc'd Bob and Jake as well as Vint.
>>
>> Cheers,
>>
>> Steve
>>
>>
>>
>>
>>
>> On Aug 29, 2009, at 12:49 PM, Johnny RYAN wrote:
>>
>>> Dear Steve,
>>>
>>> I'm checking the chronology regarding IANA, the governance of DNS
>>> etc., and I wonder if you could clarify when Jon Postel started the
>>> task of keeping track of the addresses of the host machines on the
>>> ARPANET? Vint Cerf suggests that it may have been when the RFCs
>>> started in April 1969 but that you would know best.
>>>
>>> Does that sound right?
>>>
>>> Johnny
>>>
>>> --
>>> My Next Book... http://johnnyryan.wordpress.com/books/net-history-2010/
>>>
>>> Profile: http://www.linkedin.com/in/johnnyryan1
>>>
>>>
>>>
>>> On Sat, Aug 29, 2009 at 3:51 PM, Vint Cerf<vint at google.com> wrote:
>>>>
>>>> On Aug 29, 2009, at 10:18 AM, Johnny RYAN wrote:
>>>>
>>>>> Dear All,
>>>>>
>>>>> I'd to make sure I have the chronology and causality right in the
>>>>> development of ICANN
>>>>>
>>>>>> From 1968 onward, Jon Postel volunteered to keep track of the
>>>>
>>>> no, more like 1969 - when the RFCs were started (April). Steve
>>>> Crocker
>>>> would know of course whether Jon volunteered sooner than that.
>>>>>>
>>>>> addresses of host machines on ARPANET and then Internet.
>>>>> In 1983, the same year as Mockapetris developed the DNS, Postel
>>>>> setup
>>>>> the IANA to manage the addresses and domain names
>>>>> 1992, NSF issues solicitation for bidders to perform network
>>>>> information service management
>>>>> 1993-1998, Network Solutions contracted to perform domain name
>>>>> registration service - working with / under (unclear here) the
>>>>> IANA
>>>>> 1995, controversy arises when Network Solutions proposes a fee for
>>>>> registering domains - the fee was apparently envisaged by NSF at
>>>>> the
>>>>> outset,
>>>>
>>>> no, it think the fee matter arose as the registration rate
>>>> increased and
>>>> NSF concluded that spending research $ on what was evidently a
>>>> commercial activity did not make sense. It authorized $50/year with
>>>> a two year minimum per domain name at first. This became very
>>>> controversial.
>>>>
>>>>> and the Dec 1994 review of Network Solutions work recommended
>>>>> the fee. Controversy however came from the hesitancy on the part
>>>>> of
>>>>> some about Network Solutions having a monopoly and the prospect of
>>>>> their having no oversight from 1998 onward
>>>>
>>>> you seem to leave out a two year period when the community tried to
>>>> self organize. An Ah Hoc committee was formed but did not converge
>>>> entirely. As I recall, Jon initiated discussion about 1996 on the
>>>> matter
>>>> of institutionalizing IANA functions in part because USC was
>>>> unwilling
>>>> to provide legal protection for potential controversies over
>>>> rights to
>>>> register particular domain names, etc. The attempt to self-organize
>>>> led to a proposed Geneva-based entity but this ignited a firestorm
>>>> among some members of Congress (who somehow thought that
>>>> the US still "controlled" the Internet) and led to Ira Magaziner's
>>>> involvement with the support of (or maybe at the request of?)
>>>> then President Bill Clinton.
>>>>>
>>>>> 1997, prompted by Ira Magaziner, Clinton administration commits to
>>>>> privitization in its 'Framework for Global Electronic Commerce'
>>>>> 1998, January, US Department of Commerce Green Paper 'A Proposal
>>>>> to
>>>>> Improve the Technical Management of Internet Names and Addresses'
>>>>> proposes a private, not for profit corporation to coordinate DNS
>>>>> 1998, US Department of Commerce publishes statement of policy on
>>>>> Internet names and addresses and announces the beginning of a
>>>>> transition of DNS functions to the new corporation, following a
>>>>> period
>>>>> of further study and consultation.
>>>>
>>>> October 1998, Jon Postel dies - he would have served as the CTO of
>>>> ICANN.
>>>>>
>>>>> November 1998, ICANN signs memorandum of understanding with the US
>>>>> Department of Commerce to undertake joint study of the future
>>>>> for DNS
>>>>> management that would be operated by the private sector.
>>>>> December 1998, ICANN signs a transition agreement with the
>>>>> University
>>>>> of Southern California in which it assumed the functions
>>>>> previously
>>>>> fulfilled by the IANA.
>>>>>
>>>>> If this chronology is accurate, then the next problem I have is
>>>>> understanding where the controversy from 2005 came from...
>>>>>
>>>>> I hope all of you are well, any input would be much appreciated.
>>>>>
>>>>> Johnny
>>>>>
>>>>>
>>>>> --
>>>>> My Next Book...
>>>>> http://johnnyryan.wordpress.com/books/net-history-2010/
>>>>>
>>>>> Profile: http://www.linkedin.com/in/johnnyryan1
>>>>
>>>>
>>
>



More information about the Internet-history mailing list