[ih] Essential components of Internet technology and operation

[Jack Haverty]

Not to discourage capturing history, but I think this is a really hard task.

Consider for example "packet switching", as the "operational
innovation" of breaking a message up into smaller pieces, sending the
pieces independently and/or redundantly, and putting it all together
at the destination.

The earliest instance of this technique (as opposed to specific
computer implementation) that I've encountered was ... during the
heyday of the Roman Empire.   Roman generals out in remote parts of
the frontier had to communicate with Rome reliably and securely.  So,
slaves were tasked to write a message on a scroll, possibly even
making several scrolls with the same message.  Each scroll was then
ripped into strips, each containing only an unintelligible fragment of
the message.  Each strip was given to a different courier, who would
run to the destination.  Different couriers were sent by different
means, e.g., one by land and one by sea, one by mountains, one through
the desert, and would choose their route based on local conditions
they encountered.    When enough couriers reached the destination the
message could be reassembled and read.  Spies or hostile forces along
the way might intercept some couriers, but they couldn't read the
message.  Slaves were expendable "packets".  With enough couriers, a
message would very likely get through.

Sounds like "Packet Switching" to me.  Once computers got invented
slaves were no longer the best choice, so now we employ computers to
do our packet switching.  Credit Caesar, roughly 2000 years ago.  One
of his staff probably actually defined the technique, after hearing
about it at the local pub from a traveler from a distant land.

Sadly, this was all documented in RFC MCLVII, which was never issued.

/Jack


On Thu, Jul 26, 2012 at 2:56 PM, Dave Crocker <dhc2 at dcrocker.net> wrote:
> Folks,
>
> As evidenced by several efforts, including the latest, I doubt that
> considering different criteria for defining the start of the Internet is
> ever going to lead to rough consensus about a single definition.
>
> In the face of an impasse about an entirety, it often is useful to break
> things down into components and consider them separately.
>
> So I suggest an exercise a deconstructive exercise.
>
> For today's Internet:
>
>    a.  Consider a technical, administrative or operational innovation that
> is generally viewed as important for making the Internet work.
>
>    b.  Identify when it was innovated and by whom.
>
>    c.  Rinse, repeat, developing a list of essential components and their
> origins.
>
>
> For example, a couple of components that aren't near the margins of
> importance:
>
>    1.  Packet switching
>
>        While there is some debate about fine-grain of details about
> innovation, its conceptualization was roughly the mid-60s by one or very few
> folk, and its demonstration in a network was, perhaps, 1969 in the Arpanet.
>
>    2.  Hyperlinks
>
>        Conceptualized by Nelson and Engelbart, apparently separately. (I
> don't know enough about the internals of Engelbart's project to know exactly
> how it developed there and who exactly should get credit for it.)
> Terminology from Nelson.  Demonstrated operationally by Engelbart, in a
> standalone system.  Demonstrated in a distributed system by Berners-Lee.
>
>
> To the extent that my statements are inaccurate or incomplete, I suspect the
> debate and repair effort can be more constrained for most of the items (as
> long as we keep away from the invention of email...)
>
>
> With a small amount of diligence, this ought to produce an interesting
> timeline, possibly with, ummm, seven layers...
>
> Thoughts?
>
>
> d/
>
> --
>  Dave Crocker
>  Brandenburg InternetWorking
>  bbiw.net