At 1:22 AM +0000 8/4/02, olympus-digest wrote:
>
>Date: Sat, 03 Aug 2002 18:44:43 -0600
>From: Garth Wood <garth@xxxxxxxxxxxxxxx>
>Subject: RE: [OM] Further on technology (R&D)
>
>At 07:17 PM 8/3/2002 -0400, Joe Gwinn wrote:
>
>[snip]
>
> >>But government - in any country - did not "invent" the internet. That is
> >>certainly just not true.
> >
> >Umm. No, it *is* true. US DoD's ARPA (Advanced Research Projects Agency)
> >provided all the original funding. Their intent was to build a
> >communications system that would survive nuclear war, which meant that the
> >communications system could not be centralized. Because the system had to
> >be able to survive the abrupt loss of any subset of nodes and links, all
> >nodes and links had to be able to constitute themselves into a working
> >network without manual intervention. (There are a number of histories of
> >the birth of the ARPANET, the predecessor of the Internet, that tell the
> >story.)
>
>And virtually all of them get the story wrong, too. One of the actual
>architects of the early ARPA initiative was interviewed for a lengthy article
>a few years back in "The Sciences" (IIRC), disabusing his audience of the
>notion that the Internet grew out of an initiative to have a network that
>could survive a nuclear war. The *real* reason the basics of the network were
>created back then was because researchers in various locations around the
>country all wanted access to the same mainframe computers at the same time,
>and back then, it was pretty much a "one computer, one terminal" affair.
>After about a half hour of lobbying his immediate manager for funds, this
>individual walked out of the meeting with $500 G's of funding (and in the
>late-1960s, to boot) and a few ideas.
$500K was a lot of money then (late 1960s), but still this is seed money only,
and is orders of magnitude too small to build even the initial arpanet. All
the hardware except computers had to be invented from scratch, and
manufactured. Universities simply didn't have that kind of money.
>The rest is history. Details in the article "Casting the Net" in the
>September/October 1996 issue of "The Sciences," published by the New York
>Academy of Sciences. (Can't find my issue at the moment, but whatever...)
I've read lots of histories too, with varying viewpoints, but ARPA funding is
well established, as are their original motives. The academic community was
interested in remote access to mainframes (which were rare and expensive then)
for sure, but that isn't why ARPA funded the effort. ARPA is supposed to show
a possible military payoff to their grants.
It's precisely the damage-tolerance designed into the internet protocols that
have allowed those protocols to displace all competing protocols, because
damage tolerance implies robustness in the extreme, precisely what's needed for
reliable operation in the real world. I read have some of the early articles
on making the arpanet secure and robust as well; none of it was classified.
It's very common for the renegade programmers of today to rebel at the thought
that the birth of the internet is associated with nuclear weapons, but it's
true nonetheless. The list of technologies that were developed by or for the
military is quite long, and many of these technologies go on to become
commonplace. Science has been the handmaiden of War for millennia.
Ultimately, the true value of government in non-government research is in the
funding of tens of blue-sky proposals for every one that succeeds, and hundreds
of proposals for every one that becomes famous.
Joe Gwinn
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|