Network visibility

Miles Fidelman mfidelman at
Thu Oct 21 18:16:19 UTC 2021


You guys were in grade school, some of us were there at the beginning 
(well, in my case, 2 years after the beginning).  I can assure you that 
folks made a big deal about what was and wasn't the Internet, and the 
distinction between "an internet" and "the (capital I) Internet."  
Opinions varied then, and opinions vary now.

But... by and large, as I understand the general zeitgeist:

- you're either on the Internet, or you're not - the key question is 
whether you can send & receive IP packets from the public address space 
(i.e., the classified segments are internets, but not part of THE 
Internet).  There are also disagreements on where the Internet ends - at 
the demarc, or at the IP stack in your machine (I argue the latter, but 
that's debatable)

- as to when the Internet was born... that's also debatable.  The 
ARPANET started passing it's first packets in Sept. 1969 - that's a 
known point in time.  One could probably find the date when the first IP 
packet crossed transited a router between two networks. Beyond that, the 
Flag Day is about as good a date as any - before that there it all was a 
gaggle of networks, some routers (then called gateways), supporting 
various internetworking protocols, including IP.  But the Flag Day made 
it all official - except for a few special exceptions, that marks the 
date that every machine on the net was reachable by IP, and NOT by NCP.

So... how about dropping all the pontification.  It just makes you look 

Miles Fidelman

Owen DeLong via NANOG wrote:
>> On Oct 21, 2021, at 08:55 , Mel Beckman <mel at 
>> <mailto:mel at>> wrote:
>>> On Oct 21, 2021, at 8:19 AM, Owen DeLong <owen at 
>>> <mailto:owen at>> wrote:
>>> No, but you are ignoring the point of my message…
>>> The TCP/IP internet existed _BEFORE_ the flag day you mentioned. The 
>>> flag day was the end of NCP, not the beginning of TCP/IP. IIRC, at 
>>> the time,
>> Owen,
>> But we’re not talking about the birth of TCP/IP. We’re talking about 
>> the birth of the capital-I Internet, which by definition runs 
>> exclusively on TCP/IP, and that didn’t start until Jan 1, 1983. 
>> Although there was /experimentation /using IP during 1982, that was 
>> still ARPANET. It was the /guaranteed exclusive /availability of IP 
>> that made 1983 the Internet’s birth date.
> IMHO, that’s an absurd definition. It was still ARPANET after January 
> 1, 1983 too. Prior to 1982, it was ARPANET on NCP. During 1982, it was 
> ARPANET running on NCP+TCP/IP, much like the Internet runs dual stack 
> today on IPv4 and IPv6.
> In 1983, NCP was removed from most of the backbone, as I hope will 
> happen with IPv4 in the next few years.
>> And no, it’s not analogous to the eventual IPv6 transition, because 
>> both IPv5 and IPv4 are Capital-I Internet standard protocols.
> You’re picking arbitrary definitions of Capital-I Internet standards. 
> NCP was every bit as standardized as TCP/IP in 1982.
> Both were documented in the same IEN series of documents.
> IEN later (well after TCP/IP) evoked to become RFC.
> Don’t believe me? Look at the hosts.txt file from IPv4 days which 
> still referenced IEN116.
> Owen

In theory, there is no difference between theory and practice.
In practice, there is.  .... Yogi Berra

Theory is when you know everything but nothing works.
Practice is when everything works but no one knows why.
In our lab, theory and practice are combined:
nothing works and no one knows why.  ... unknown

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the NANOG mailing list