[chapter-delegates] How about a World Internet Day?

Robert Kahn rkahn at cnri.reston.va.us
Mon May 23 15:49:56 PDT 2005


Bob,

I'm sure that's all exactly right, but keep in mind that this was almost a 
full time job at DARPA for me as well for many weeks.

bob

At 12:31 PM 5/23/2005, Bob Braden wrote:
>   *>
>   *> Bob,
>   *>
>   *> Actually, to follow the analogy a bit further, the birth may have 
> started
>   *> on 1/1/83, but (depending on your definition of birth) it was a ten 
> year
>   *> gestation period and the birth was not really over until either a 
> few weeks
>   *> or a few months after january 1st. By mid february, 1983 we had 
> about 80%
>   *> of the ARPANET hosts converted from NCP to TCP, and a lot of new LAN 
> based
>   *> "workstation hosts" that used only tcp and not the arpanet host 
> protocol;
>   *> there were also some longer term stragglers on the ARPANET that did not
>   *> take the january date seriously. In reality, I don't think much 
> happened on
>   *> that new years day, and we decided to keep both NCP and TCP protocols
>   *> running in parallel for at least six months to insure continued
>   *> connectivity for everyone.
>
>Bob,
>
>This is all true, of course, but those of us protocol hackers who spent
>our New Year's day 1983 hard at work in the office KNEW when the
>swithover was, and what a milestone it was.  The event had been
>orchestrated by Jon; the evidence is still in the RFC series.
>As I recall, Dan Lynch, ops manager at ISI, did not get much sleep on
>Jan 1 or a few days afterwards, as production usage exposed flaws in
>the TCP/IP code on TOPS20 systems at ISI.
>
>So, even though it did not happen instantly, it was a clearly-defined
>event for its participants.  I think there was a T shirt around with "I
>survived the ARPAnet/Internet transition -- Jan 1, 1983", or something
>like that.
>
>Bob
>
>
>
>   *>
>   *> bob
>   *>
>   *> At 08:13 PM 5/20/2005, Bob Braden wrote:
>   *>
>   *>
>   *>
>   *> >>There are many milestones so I think an attempt to celebrate a 
> birthday
>   *> >>is hard.
>   *> >>For Internet one might reasonably pick 1/1/1983 as the date the 
> system
>   *> >>was deployed on all of the networks supported by DARPA.
>   *> >>[cut]
>   *> >
>   *> >For those of us involved with making the Internet happen, I expect 
> that
>   *> >Jan 1, 1983 is the true birth of the Internet as a
>   *> >network of networks using TCP/IP, and operational as  opposed to 
> experimental.
>   *> >
>   *> >Bob Braden
>   *>



More information about the Chapter-delegates mailing list