[ih] distributed network control: Usenet

[vinton cerf]

wow, I missed the NNSC entirely - but in 1992 I was busy with starting
ISOC...

v


On Tue, Jul 20, 2021 at 10:23 PM Jack Haverty <jack at 3kitty.org> wrote:

> Thanks Vint.  Dates have always been hard (for me at least) to keep sorted.
>
> Per the authoritative voice of Wikipedia - "Support for NSFNET end-users
> was provided by the NSF Network Service Center (NNSC), located at BBN
> Technologies <https://en.wikipedia.org/wiki/BBN_Technologies>".   But I
> can't say I remember anything about that at all.
>
> Also I agree we didn't get involved with the Fuzzball network as it got
> rolled out to NSF et al.  However, I remember clearly various "events" that
> occurred several years earlier while Dave was developing his Fuzzball code
> and a few Fuzzies were part of the fledgling Internet.   At some of the
> meetings, we would jokingly rib each other.   Dave would report on
> something his Fuzzies tried to do; I would report on the ensuing
> disruptions in the Internet we were trying to make reliable.   I joked that
> Dave's tendency was to poke at the Internet in some new way and say "Hey,
> look, it turned pink, let's see what else we can do."   I would respond
> with something like "Please don't do that."   Dave was the scientist,
> exploring the unknown.  I was the engineer, trying to get the beast to keep
> running.
>
> That was all before 1982 and was a strong motivation for creating the
> notion of ASes and EGP.   EGP allowed us both to be happy and keep
> scientific experimentation and operational engineering from doing battle
> inside the Internet.
>
> /Jack
>
> On 7/20/21 6:49 PM, vinton cerf wrote:
>
> The fuzzball net arrived about 1986 at 50Kb/s - congested quickly and the
> IBM/MCI/MERIT version of NSFNET launched in 1988 at 1.5 Mb/s
>
> I am not aware of any involvement of BBN in either the fuzzball network or
> the subsequent NSFNET except that presumably Mills implemented EGP. BGP
> doesn't arrive until 1989 and as I recall, Yakov Rekhter of IBM and Keith
> Loughead at Cisco wrote RFC 1105 describing it.
>
> v
>
>
> On Tue, Jul 20, 2021 at 8:02 PM Jack Haverty via Internet-history <
> internet-history at elists.isoc.org> wrote:
>
>> It's the time frame.   My comment was about the period before EGP was
>> created in 1982.   The IETF didn't exist yet.   I don't remember which
>> if any of those networks existed before 1982.   But if they did, I think
>> they had a single manager.
>>
>> /Jack
>>
>>
>> On 7/20/21 4:45 PM, Tony Li wrote:
>> > No, all of the regionals and other networks were various independent
>> organizations. There was no centralization, just the chaos of trying to
>> keep things working through the informal network of operator’s personal
>> connections. For this, the IETF and NANOG were indispensable.
>> >
>> > Tony
>> >
>> >
>> >
>> >> On Jul 20, 2021, at 4:39 PM, Jack Haverty via Internet-history <
>> internet-history at elists.isoc.org> wrote:
>> >>
>> >> Weren't those all managed by the same organization or its contractor,
>> in the early 80s before EGP?
>> >>
>> >> I remember that at one point BBN was the contractor managing CSNET
>> (Dick Edmiston).   NSFNET started in mid-80s and IIRC was thoroughly
>> dominated by Dave Mills' Fuzzballs.  Our experiences when Dave was
>> experimenting with connecting his Fuzzies to the core Internet was a
>> primary motivator for EGP, which made it possible for Fuzzies to connect
>> and do their thing without impacting the core.   BBN had some managerial
>> role in NSFNET too IIRC.
>> >>
>> >> After EGP, and probably more importantly BGP, the world of
>> Internetworking changed.
>> >>
>> >> /Jack
>> >>
>> >>
>> >> On 7/20/21 4:03 PM, Miles Fidelman via Internet-history wrote:
>> >>>> Jack Haverty wrote:
>> >>>>> What I was referencing was a non-technical design decision -- the
>> notion
>> >>>>> that there shouldn't be any single person, corporation, or
>> organization
>> >>>>> "managing the network".   The ARPANET, and IIRC all other networks
>> of
>> >>>>> the day, were under a single organization's control.
>> >>> Really?  NASA SPAN, DOEnet, then CSnet, and then the Supercomputer
>> Center Networks, and the NSFnet regionals & Backbone?
>> >>>>
>> >>>
>> >>
>> >> --
>> >> Internet-history mailing list
>> >> Internet-history at elists.isoc.org
>> >> https://elists.isoc.org/mailman/listinfo/internet-history
>>
>>
>> --
>> Internet-history mailing list
>> Internet-history at elists.isoc.org
>> https://elists.isoc.org/mailman/listinfo/internet-history
>>
>
>