IP: some More on Breaking the Tech Myths (before they break us) VERY INTERESTING STUFF

[Dave Farber <farber () cis upenn edu>]

To: farber () cis upenn edu
Date: Fri, 11 Dec 98 09:23:05 PST
From: "Willis H. Ware" <willis () rand org>

--
Dave:

What prompted the prior msg about y our sustem is the following two messages
which I sent you, not knowing that you were busily involved in the DC scene.
These have apparently fallen into your backlog, but they were timely when
sent.  If you consider them still timely, how about adding a brief preface
to the effect that they concern


------- Forwarded Messages

To: farber () cis upenn edu
Cc: willis () rand org
Reply-To: willis () rand org
Subject: Re: IP: More on Breaking the Tech Myths (before they break us) 
In-reply-to: Your message of Mon, 07 Dec 98 06:56:17 EST.
             <4.1.19981207065552.03c687b0 () linc cis upenn edu> 
Date: Mon, 07 Dec 98 09:23:09 PST
From: "Willis H. Ware" <willis () conrad rand org>

- --
Folder: YES
- --
Dave:

Danny Cohen's recitation of facts and dates is accurate.  The Baran work
(done in the early 60s) was documented in a series of 12 research
memorandum all of which are now on the Rand external Web page under
"Classics" (http://www.rand.org).  Indeed this work was used by ARPA to
persuade DoD and others to invest in packet technology and all that came
along with it.  All of this history is well documented and cited in various
places with proper credits.

At one point, Larry Roberts commented to me personally: "..please come to
Washington and help us make people understand how we're going to change the
world." That is an accurate verbatim quote.

Just a little point in passing (also well documented).  Paul Baran used the
phrase "distributed communications" in his writings.  The phrase "packet
switching" was originated by Donald Davies who came along with the concept
soon after Baran's work was published.  It is not clear to my knowledge
whether Davies had knowledge of Baran's work, especially since one of the
original series of documents was classified.

                                Willis

------- Message 2

To: farber () cis upenn edu
Cc: willis () rand org
Reply-To: willis () rand org
Subject: Re: IP: Internet Myths 
In-reply-to: Your message of Sat, 05 Dec 98 15:52:05 EST.
             <4.1.19981205155125.00bb5e70 () mail earthlink net> 
Date: Mon, 07 Dec 98 09:41:11 PST
From: "Willis H. Ware" <willis () conrad rand org>

- --
Folder: YES
- --
Dave:

After replying to Danny Cohen's message, I went back and read the original
mythology discussion.  A few more observations ...

1.  The genesis of the Baran work was a concern by the USAF about survival
of command-control during a nuclear war.  The USAF leadership expressed
this concern to Frank Collbohm, president of Rand at the time, and he
relayed it to Keith Uncapher, Paul Baran and others.  I'm a little hazy
about who carried the message from the USAF but it might have been Curtis
LeMay.

Among other views, it was thought possible to exploit the AM network which
was widespread and with numerous nodes (stations) even in the 60s.  I well
remember a variety of discussions between Paul, AT&T leadership, and others
about the coming transition from an analog world to a digital world.  It
was a hard sell, and except for ARPA's infusion of leadership and funds, the
digital networks of today could have been much slower in arriving.

2. In regard to the following:

> > This myth has served to help us rationalize constant expenditures on new
> > computers and upgrades. But while a 300mhz computer costs less today than a
> > 200mhz model did a year ago, we spend more total cash if we continually
> > replace our machines for little or no added benefits. In a flourish of
> > planned obsolescence bravado, computer companies and software writers create
> > chips that require more advanced operating systems, and operating systems
> > that require newer chips and more memory. Buying more computers, even if
> > progressively cheaper, costs more.

There is nothing new about that statement.  In the days during which IBM
(and others') "big iron" (mainframes) were dominant, one of the inside jokes
and truisms was that:

        a.  With every new introduction of a machine, you would wind up
spending the same amount of money; and

        b.  The increased performance was inevitably diluted by the new and
more demanding system-level software.

And in the same vein:

> > Otherwise known as Moore's Law, the myth is true when taken in isolation.

Moore's law (although I'm not certain when the name was attached to it) was
around long before Microsoft became the powerful corporation that it is.
The author was careless about the facts of history.

What goes around comes around ...  it's hard to understand where this author
is coming from.

                                                Willis Ware

------- End of Forwarded Messages