[Info-vax] Reading Gordon Bell's VAX strategy document

[Dan Cross]

In article <uf3m7s$lv$4 at news.misty.com>,
Johnny Billquist  <bqt at softjar.se> wrote:
>On 2023-09-28 08:09, Lars Brinkhoff wrote:
>> Johnny Billquist wrote:
>>> Dan Cross wrote:
>>>>> In fact, were there any [VAXen] at all on the ARPANET before the
>>>>> 1983 flag day?
>>>> Almost certainly.  The initial TCP/IP implementation work for Unix
>>>> was being done at BBN at the time, and I imagine that meant VAXen
>>>> connected to ARPANET.
>> 
>> I meant VAX machines talking the NCP protocol.
>> 
>>> Well, before flag day, ARPANET wasn't speaking TCP/IP...
>> 
>> Yet, there were experiments with TCP long before the flag day so it's
>> not a 100% either/or situation.  I get the feeling (but I have no
>> evidence handy) some subset of nodes got started using TCP before NCP
>> was shut down.  On several occasions in 1982 (and maybe earlier?) BBN
>> arranged for NCP "brown-outs" to encourage speedy development.
>
>There were of course development, and testing done between machines and 
>so on. But that was not "ARPANET". ARPANET was running NCP until flag 
>day, when it officially switched to IP.

...and for a while after that, too!

>And at some point after that, 
>all of ARPANET because just the 10.* addresses on the Internet, and then 
>ARPANET was turned off, and it was decided that 10.* should not exist on 
>the Internet anymore...

What you wrote above is certainly true, but given that the
ARPANET was the original Internet backbone and that the initial
TCP work for VAX Unix was being done at BBN, it's not
unreasonable to believe that they did an NCP implementation for
the VAX before TCP/IP proper.  Of course, that's speculation (I
have no evidence) but it's not unreasonable.

	- Dan C.