Ronda Hauben on Thu, 16 Sep 1999 00:24:45 +0200 (CEST)


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

<nettime> Dispute about the origins of the Internet


The following post is a response to a post that was on the 
Community memory Mailing List about the importance of ARPA's office
of Information Process (IPTO) in the development of the Internet.
And this is actually a controversy over the importance of 
the role of government and of computer science in the creation
and development of the Internet.

A reporter called me later today and was interested in this
controversy as there are celebrations planned on the 30th 
anniversary of the ARPANET (1969). It's also the 25th anniversary
of the publication of the paper describing the TCP protocol
(1974).

Following are my comments in response forwarded by Bob Bickford:


Bob Bickford <rab@WELL.COM>

BB>This was posted on another list which I read.  
BB>With the author's permission,
BB>I am forwarding it here.  I generally agree with what he ways.   {grin}

>.....rab

 --BEGIN-QUOTE---------------------------------------------------------

I wonder who Tim May is quoting in this early part of his response?
And where was the original post?


>Date: 9 Sep 1999 11:34:05 -0600
>From: Tim May <tcmay@got.net>
>Subject: DARPA Considered Unnecessary

>>>DARPA funding for ARPANET, etc. used a different approach.  Rather than
>>>mandating specification before implementation of protocols designed by
>>>properly-constituted international committees (OSI), it funded competent
>>>researchers and mostly gave them their head to determine appropriate
>>>directions by actually trying things.
>
>>>This approach was rare.  Most countries that dabbled in network funding
>>>went the other way, of mandating things that didn't work: EU, most
>>>European governments, etc.

>>ARPA/DARPA is the darling example always cited as the rationale for "good"
>>government funding. The funding _did_ work. (Caveat about my bias: my first
>>ARPANet account was around 1973.)

The point here that is missed is that ARPA's Information Processing
Techniques Office (IPTO) didn't just *fund* researchers. It was
a component and important part of the research paradigm.

That is well hidden,however, by those who talk about ARPA funding.

It makes it seem as if it is only a way of giving money to researchers.

But actually ARPA/IPTO was a component and crucial part of the
computer science research community, and the leadership of IPTO
was for an important period of time in IPTO's lifetime a crucial
part of the research community.

My recent paper on IPTO begins to document this.

See http://www.ais.org/~ronda/new.papers/arpa_ipto.txt


TM>>Would networks and user interfaces and other such things have advanced as
TM>>much without DARPA funding?

TM>>I think so. And John McCarthy even wrote an article about 15 or so years
TM>>ago explicating just this point, and even claiming that probably we would
TM>>have a _better_ network model, more akin to the point-to-point phone model
TM>>than the "spoke and hub" model championed by the ARPANet.

Not much of an argument by the writer of this post. Just that he
thinks so. This leaves out that JCR Licklider was the guiding
force at ARPA/IPTO making time sharing (which was McCarthy's idea)
a reality by the way he built a community of computer science
researchers, who advised him in turn on what he was doing
and which he called the Intergalactic network.

See http://www.ais.org/~ronda/new.papers/internet.txt

Also the ARPANET was the result of the work done by IPTO

And the Internet the next advance, *not* the same as the
packet switching advance of the ARPANET.

While the ARPANET was one network controlled as a network,
the Internet is built on the model of the intercommunication
of diverse packet switching networks.

The Internet model made it possible to scale as it welcomed
diverse networks under the open architecture principle
developed by Bob Kahn, then at ARPA/IPTO.

All were welcomed to become part of the Internet.

Also a lot of the important early work developing the Internet
was done out of the IPT office itself by the program manager
or director etc.

TM>>Technology was advancing relentlessly, and, by the way, it was not being
TM>>driven by ARPA or DARPA funding in any significant way.

It's easy to make such statements without giving any evidence
to back them up.

The batch processing mode of computers was the mode in 1962 and would
have continued to be the mode for a number of years if not for
ARPA/IPTO's research work developing time sharing and interactive
computing.

And it wasn't that any commercial entities were out to put in the
millions of dollars and years of research work and support of
researchers that was needed to develop packet switching through
the ARPANET research, nor the open architecture principle
and protocol development that made the Internet a reality.

It took 10 years from 1973 from the date that Kahn and Cerf
designed the protocol for internetworking, TCP as it was then
called, till the cutover to TCP/IP on the ARPANET in January 1983.

This wasn't the kind of research that industry was doing or would
do.

So May's comments are denying the need for both computer science
researchers and for a government role in that research.

And that would leave us now with batch processing and without

Unix
Usenet
packet switching
the Internet

TM>>(I joined Intel in 1974, working in Technology Development, and I can
TM>>assure you all that neither memory chips nor microprocessors nor the early
TM>>personal computers were affected in any noticeable way by what DARPA at
TM>>that time was doing. I sometimes see claims that VLSI was pioneered by
TM>>DARPA, or that computer-aided design was invented by Mead and Conway, and
TM>>so on. Just ain't so.)

How strange. By 1974 ARPA/IPTO had accomplished a revolution in
the form of computing

ARPA/IPTO started in 1962.

And it did its work well.

Also I have only read a bit about VLSI, but as I understand it
each chip manufacturer had its own proprietary designs and VSLI
took on that challenge and opened up the process.

And the Information Science Institute (ISI) became a center for
the work in this field.


TM>>Don't get me wrong. I think what DARPA threw money at was mostly money
TM>>well-spent. Ditto for what Xerox PARC was spending (sometimes with links to
TM>>DARPA projects).

But it wasn't that ARPA/IPTO was a question of "throwing money".
It was a crucial part of a computer science research community.


TM>>But many people have been drawing overbroad conclusions about the effects
TM>>DARPA had. It is almost certain that the pace of technological development

What are these overbroad conclusions?

ARPA/IPTO fundamentally changed the computer science paradigm.

Also companies like Sun etc wouldn't exist.

TM>>would have been about the same had DARPA never existed. The same applies,
TM>>by the way, for other players, including Intel, Microsoft, Apple, etc.

Well Microsoft and Apple in fact are the product of work done
by Kemeny creating basic as part of the Dartmouth Time Sharing
system.

And DOS by Microsoft is a beneficiary of the work done developing
time sharing and then CTSS and MULTICS and its beneficiary UNIX.

And the work on time sharing and CTSS was the result of ARPA/IPTO's
work.

The programming languages for the personal computer were the offspring
of the programming language developed for time sharing systems like
Kemeny's DTSS.


TM>>The actual history would have been different, obviously. In some ways less
TM>>progress, in some ways more progress. But there was a certain inexorability
TM>>about most of the technological developments. Other companies would have
>>introduced much the same products.

No the paradigm was fundamentally changed.

And the participatory role of users on the Internet is *not* something
that industry seems to recognize as important as they seem to
equate computer users or users on the Internet as customers.

The whole concept of Netizens is a development in fact of
the computer - human relationship that Wiener (Norbert Wiener)
recognized was the crucial relationship to be explored as
part of his work on Cybernetics.

So it isn't that the technology is rushing ahead at all, nor that
industry, at least in the computer field, is the great creator
of the future.

It is that the development of computer science and of cybernetic
science has led to the ability to make certain leaps and
the computer science research community which was developed
as part of ARPA/IPTO was able to have both the vision and the
ability to develop the science that has made these technological
developments a reality.


TM>>In summary, I'm not dissing DARPA, just questioning the notion that
TM>>DARPA-like funding is needed.


The vision and research that ARPA/IPTO carried out is needed more
than ever. And we need to learn the lessons from these developments
*not* to deny that they were made.


TM>>- --Tim May

Ronda
ronda@ais.org
-----------------
                 Netizens: On the History and Impact
                     of Usenet and the Internet
                Published by IEEE Computer Society Press
                      ISBN # 0-8186-7706-6
               http://www.columbia.edu/~hauben/netbook/

______________________________________________________________________
            Posted by David S. Bennahum (davidsol@panix.com)
                    Moderator: Community Memory
            http://memex.org/community-memory.html
         A CPSR Project -- http://www.cpsr.org -- cpsr@cpsr.org
 Materials may be reposted in their *entirety* for non-commercial use.

 Get this list in digest form:          SET CYHIST DIGEST
 Leave this list:                       SIGNOFF CYHIST
 Send these commands to:                LISTSERV@MAELSTROM.STJOHNS.EDU
______________________________________________________________________



#  distributed via <nettime>: no commercial use without permission
#  <nettime> is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo@bbs.thing.net and "info nettime-l" in the msg body
#  archive: http://www.nettime.org contact: nettime@bbs.thing.net