In all the dreck and dross of Internet books here is a brilliant
gem. This remarkably well-written work ought, at least, to clear up
some confusion about what the Internet is. The big barriers to
understanding the Internet are its cloak of technical wizardry and
an incredible amount of hype, pro and con, devoted to the subject.
Hafner and Lyon's book is exceptionally good at explaining the
technical aspects of the Internet and is a small step toward
dispelling the hype. Even without the hype or the wizard's
mumbo-jumbo, the Internet is magic, and seems more so the more one
knows about it.
First we meet Bob Taylor, who works in an office in the
Pentagon. In 1966 he has, in a room next to his office, three
computer terminals. These are nothing like today's terminals, but
are glorified typewriters and Teletype machines. Each terminal is
connected to a different computer: two computers in California and
one at MIT. Each terminal works in a different way, and the
computers each require a different command to perform essentially
the same task. Thus Hafner and Lyon set up the problem for which
the Internet is the solution: how Bob can have only one terminal
that will connect him to any of the three computers and will let
him use standard commands that each computer can translate into its
own dialect.
This history does not confine itself to the development of the
technical solution to Bob's problem. Bob's office was in the
Pentagon, but Bob, although extended the courtesies due a one-star
general, was a civilian. He worked for the Advanced Research
Projects Agency (ARPA), and the network he was concerned about
creating was ARPAnet. It would be twenty years before ARPAnet,
joined with many other, younger networks, became known as the
Internet. To understand that the Internet never was part of the
nation's nuclear defense (or retaliation) system, and that it never
was possible for a kid to crack codes on the Internet and launch a
nuclear first strike, we first have to learn a bit about ARPA.
The more-science panic that swept America after 1957, when the
Soviets became the first country to launch a satellite, gave birth
to ARPA. Although ARPA funded projects with immediate military
applications, it had a free hand to engage in pure research, and in
this regard we are unlikely to see anything like ARPA again.
Perhaps ARPA owed its independence to being created during the
administration of a military man who did not trust the military. (I
wonder how many people who worry about the military-industrial
complex, know that Eisenhower coined the term and was worried about
it, too.) In many ways ARPA's survival was as great a political
feat as the invention of computer networking was a technical one.
Surely the military cast a covetous eye on the large budget of this
mysterious agency set down in its midst, and the Republican
National Committee did have a word with Ike about funding projects
that employed so many people who lacked enthusiasm for the
Republican ticket. But ARPA survived, and well enough that its
ARPAnet contractor—BBN, whose previous claim to fame was the
failed design of Lincoln Center—could send a man (Severo
Ornstein) wearing a Resistance lapel button to brief the generals
in the Pentagon in 1969.
The technical aspects of the Internet are easier to understand
in this book than in most because developments are presented in
their evolutionary order, one part built on another. Books that
take the Internet as it appears to today's new user and attempt to
explain it by parts, I'm afraid, will never be so clear. The
analogies made for nontechnical readers are apt, if sometimes
cutesy. Naturally the book is more fun for readers who have some
experience and can cheer on a favorite aspect as it develops. A
good example of this is the "Request For Comments" (RFC).
Getting computers all to use the same internal language was
impossible. Every interaction between computers had to be
anticipated and given a common language in which to occur. That
required standards. Setting those standards is what a Request For
Comments, really little more than a memo, was all about. We are
shown Steve Crocker issuing RFC 1, in every way as tentative and
open to suggestion as the title "Request for Comments" suggests.
This is more amusing if you know that RFCs today (numbering into
four figures) are more like the law of the Medes and the Persians.
Compliance is expected now, not comment.
The technical side of the story makes a clear case for pure
research. Nothing could be more obvious than that these
guys—and it was almost all guys in those days—had
little idea what they were doing, why they were doing it, or what
they would have when they were done. Linking the country's most
powerful computers in the country's major universities seemed a
good idea, yet no one had a clear picture of what the computers
would do once they were hooked up. By the time seven or eight nodes
were connected, linking places like Stanford, UCLA, and MIT, the
use of the network was only 3 percent of capacity, and much of that
was for testing the network, not doing anything useful with it. The
solution was ahead of its problem.
Some uses of the Internet were anticipated, such as replacing
Bob's three terminals with a single device. Other uses such as
e-mail, now regarded as essential, were unanticipated and
discovered accidentally. Those who are not familiar with computers
will no doubt be surprised to discover that games played an
important role in the development of the Internet, just as games
continue to teach new users how to use both the Internet and the
computers it connects. Today's wizards learned their trade by
pressing buttons just to see what they did (and hardly anything
could redress the badly imbalanced sex ratio of Internet users more
than encouraging little girls to push buttons just to see what they
do).
Wizards leaves off in 1989, about the time that Internet
connections started to be available to the public. The World Wide
Web—all of the Internet that some new users know—did
not begin until 1990. The last chapter has not been written yet. In
many ways, the Internet is still a solution looking for a
problem—unless one is willing to believe that unsolicited
commercial e-mail, which has become a plague within just the last
year, is the Internet's best use. Neither has the last of the myths
been debunked. In truth the Internet is neither an information
superhighway nor a pedophile's lair, dripping with
pornography—it still requires some skill to dredge up either
useful information or dirty pictures. Perhaps useful information
will get on-line in a way that ordinary people can find it easily,
or perhaps the Internet will become like satellite
TV—hundreds of channels carrying little of real value.
Meanwhile, it is a magic place with wizards and, yes, some dragons,
and most of all buttons to push just for the joy of finding out
what they do.
