[e2e] Nature mag, DARPA, and the Internet
dhc2 at dcrocker.net
Fri Aug 8 19:59:46 PDT 2003
>> So, Darpa did the packet-related work and one or two of the core
JT> "packet-related work" like IP, UDP, TCP, DNS, SMTP, FTP, etc.?
IP/UDP/TCP/DNS and a bit more, are the packet work. SMTP and FTP are
the core applications. The "etc." is pretty much the part that is in
JT> If that's not the Internet, what exactly is?
The web and search engines were the two areas I cited.
I have already been roundly and soundly taken to task for short-changing
Darpa's role in search-engine technology. (And, no, I had not
appreciated that bit of history. Darpa's poor public relations skills
continue...) However I'll note that the engine work were for later
contributors. The early efforts, notably starting with Archie, were not
Similarly, whatever Darpa is contributing to the Web, now, it had
nothing to do with the origins of the service.
JT> I agree that there's more - email, the web, mice, windows, etc. that
Well, sigh. This becomes ironic.
In fact it *was* ARPA research that invented the mouse and, one could
argue, windows. The most significant contribution to that area of
computing was Doug Englelbart's SRI-ARC team, of course. ARPA money.
Though one could also argue that ARPA did not carried it forward to make
it generally useful, the way it did with the packet net.
JT> Would there even _be_ a web, URLs, etc. without the DNS? or TCP over
JT> which HTTP runs?
My point was not that all the arpa stuff was not -- is not -- essential.
Rather it was/is that the popular experience of the thing called the
"Internet" required/requires more than that stuff.
The danger in these sorts of discussions about attributing
responsibility is where to draw the line?
For example, I'm not inclined to credit AT&T for the Internet, even
though they invented the transistor.
Dave Crocker <mailto:dcrocker at brandenburg.com>
Brandenburg InternetWorking <http://www.brandenburg.com>
Sunnyvale, CA USA <tel:+1.408.246.8253>, <fax:+1.866.358.5301>
More information about the end2end-interest