ADVANCEMENT OF THE INTERNET DURING THE 1970'S
In today's society, everyone is using the Internet. Whether you are a
student, teacher, or businessman, the Internet is vital. It is impossible
to go through a whole day without hearing the word Internet. The Internet
is also growing. It is projected that by the year 2000 there may be a
million different sites on the Internet (Brader and Mankin 3). Many
believe the Internet is the best thing that has happened this decade. And
while the Internet seems to be a recent achievement, the foundation of the
Internet was set in the 1970's. It was then that scientists and engineers
got together and created the backbone and protocols that the Internet still
uses today.
The idea of the Internet started in the 1960's. Ian Hardy describes how
the US military created the Advanced Research Projects Agency (ARPA) to
create technologies useful to the military. One project was to develop a
method of communications that would be almost impossible to destroy.
J.C.R. Licklider, head of ARPA at the time, talked of a "Galactic Network,"
in which any one person could exchange data or communications with anyone
else (Hardy). This sounds much like our Internet today.
ARPA set forth to create such a network, and they called it the ARPANet.
They contracted corporations and universities to create the technologies
needed to implement this network. Bolt Beranek and Newman (BBN) got the
contract to create the Interface Message Processors (IMP). These devices
acted like modems, and translated computer data into a format that could be
sent across high capacity lines. UCLA got a contract to measure the speed
and efficiency of the network. Network Analysis Corp. got the contract to
design the topology
of the net, in other words, to map out the ARPANet.
When the first IMP was created, it was given to UCLA, since their site was
set up to gauge the functionality of the net. Stanford Research Institute
received the next IMP. Their site was to be the Network Information Center
(NIC), which is still used today. It holds information about the data
stored on different servers and hosts. Once these two sites were
connected, the ARPANet was born. The connection worked flawlessly, and two
more sites were added, UCSB and the University of Utah. These sites did
not have network related data on them. They were to show that the ARPANet
could be used to exchange scientific and engineering data. The four sites
worked well together, demonstrating the feasibility of such a network. By
1971, there were 15 different sites on the ARPANet (Zakon).
The ARPANet grew over the years, and other networks came into being once
the concept was proven. Packet Radio Network, a system similar to the
ARPANet, was developed. It didn't use telephone or trunk lines, and
communicated only through radio waves. Another network that was
independent of land-based communications was Packet Satellite. This system
relayed data through a series of satellites and satellite dishes. These
two networks showed that the if a global network were to be created, it
could not rely on just one form of communication. Other networks were
created using the same means as the ARPANet. ALOHANet was created at the
University of Hawaii in 1970, and was connected to the ARPANet in 1972
(Zakon). TheoryNet was created by Larry Landweber at University of
Wisconsin to provide email to 100's of computer science researchers
(Zakon). Telenet, the first public network, was opened by BBN in 1974.
All these networks were connected to the ARPANet, creating a much larger
network. This network of networks is what we call the Internet (Manger
336). These networks are used today as a backbone to the Internet. They
serve as a foundation to new networks that connect to the Internet. Packet
Radio and Packet Satellite were also connected to the Internet later.
Other networks that were separate and incompatible with the Internet were
connected by using translators and gateways. USENET, a network of
discussion groups, is not part of the Internet; however, anyone on the
Internet can connect to a USENET discussion group (Delphi). Tymnet, a
commercial network provide by Tymshare Inc., allowed a person in
Philadelphia to use a computer in California. However if one used the
Internet, this connection was invisible.
How is one network part of the Internet and another one not? If a network
uses the Internet Protocol (IP), and it's connected to what is already
called the Internet, it's part of the Internet. When the ARPANet was
launched, IP did not exist at the time. However the designers of the
ARPANet set up many of the things that IP uses.
The first thing that the ARPANet did was to incorporate Packet Switching
Theory (Internet Society). Early experiments with networking computers
started by trying to send all the data at one time, in one big heap. This
is called circuit networking. However, as anyone who uses a computer
network knows, there are network errors. Therefore sometimes the heap of
data would be corrupted causing the data to be resent or worse, cause the
network to crash. The idea of Packet Switching Theory is to break the data
into little bits, or packets, and then send them to the other computer.
The other computer would reassemble the packets into a heap of data. If
one packet was lost or corrupted, it could easily be resent. Packet
networks would be much more reliable than circuit networks. Also they
would be generally faster, because they resend less data. Packets are
still used in the Internet today, and are its basis of communications. The
Internet of today could not have grown as it has if were not the Packet
Switching Theory of the 70's.
Another idea used to develop ARPANet was dynamic routing. Previously data
would follow one route through the network. If this route became unusable,
the network would just stop working. Dynamic routing was the idea that if
a route became unusable, for instance a computer crashed or a phone line
went down, data on that route could take a detour (Levine, Baroudi, Young,
21). This meant that if one computer crashed in Wisconsin your email from
New York to California wouldn't be lost. It also meant that no computer
was special. Every computer was just the same as every other computer.
When dynamic routing was applied to the Internet, it applied to whole
networks, and not just single computers. Dynamic routing is especially
important to the Internet today. With thousands of different networks all
interconnected, there is not a day that goes by without at least one server
going down. If it were not for dynamic routing, this would kill everyone's
access to the Internet.
When designing the ARPANet, there needed to be a language for the computers
to speak to each other. This is called the protocol. After the first four
computers were connected to the ARPANet a protocol called Network Control
Protocol was developed. It was simply a format for the packets that could
tell other computers where a packet was going, how large it was, etc.
However when Packet Radio was being developed, they found that NCP was not
adequate to their needs. Packet Radio has to be able to deal with
interference and blackouts, and there needs to have a more reliable
communications protocol. Robert Kahn, head of the Packet Radio Project,
began to devise a new protocol that would be more reliable. At the same
time he made sure it would be expandable to allow new networks. He began
working with Vint Cerf to spell out the details of the new protocol
(Internet Society). The Internet Society, in their paper, spell out the
key aspects to their thinking. First each network or computer must be
stand alone. This means if it were separated from the network, it should
not crash and cause problems. All communications would be a on a best
effort basis. Previously it was assumed there would be no errors. Finally
they restated that there must not be one special computer or network.
They presented their new protocol to the International Network Working
Group (INWG) (Zakon). This group was formed to regulate the growth of the
Internet. The group approved the protocol, and it was tested on several
machines. It became know as Transmission Control Protocol (TCP). TCP was
implemented on Unix machines and on home PCs as TCP showed itself to be the
future of networking. Experiments with sending voice data over the
Internet showed that TCP error handling services were not adequate to all
situations (Internet Society). Therefore, TCP was split into TCP and IP.
TCP would handle error handling and would be based on IP. IP would only
contain address information (Bradner and Mankin, 159). For services that
needed IP but not TCP, there was User Datagram Protocol (UDP). UDP was
similar to TCP but without the complicated error handling (Internet
Society). TCP/IP is the basis of all Internet activity today. Without it,
the Internet would not exist as it does.
Development of TCP/IP led to another design philosophy. This was the Open
Architecture Network. The Internet Society describes that in such a
network, each sub-network could be designed separately. However at places
where these sub-networks were connected together, they would speak TCP/IP.
If I designed a network in New York and used my own protocol, I could still
connect to Internet if I could translate my protocol to TCP/IP. This was
important to the development of the Internet because it meant the Internet
wouldn't need to be changed for each new network that was added. Brader
and Mankin note that since the Internet has seen lots of uncontrolled
growth, adaptability is especially important (3).
Closely connected with Open Architecture is Open Development. Michael
Hauben says that this meant that anyone could work on the Internet project.
Requests for Comments (RFC's) were developed as a informal way of
discussing ideas. Anyone who had an idea could quickly write an informal
message that anyone else could read. Reading a RFC could spawn responses,
experiments, or even new ideas (Internet Society). A timely message was
better than one with proper grammar. The tradition has kept on, and RFC's
are still written today.
By using RFC's to communicate, programmers wrote applications for the
Internet in the field. NCP was developed once the ARPANet was connected.
TCP/IP was designed on paper, but changed when applied to computers. Email
was designed by one person, but many people wrote all the programs that
made it possible. This tradition continues today. Both Netscape and
Microsoft defined their own extensions to HTML. Had it not been for the
democratic methods developed during the 70's, the extensions we use in HTML
such as tables and frames would be in dispute by some Internet council.
They would be useless to the user.
Many of the programs that we use today were developed during the 1970's.
Robert Zakon notes that most important of all, email, was developed by Ray
Tomlinson. He extended a local mail program to use the whole Internet.
Within a year someone else wrote a program that allowed people to have in
boxes, out boxes, and all the luxuries we use today. Engineers working on
the Internet used email to exchange data with each other, however David
Mayr notes they also used it to gossip. Email is the premier use of the
Internet today. Without email, few people would have connected to the
Internet, which would leave only a handful of universities on the Internet.
File Transfer Protocol (FTP) was developed, and as Robert Zakon notes,
allowed programmers to exchange programs they were working on. Scientists
could use FTP to transfer the results of experiment from the lab to their
home computers. FTP is used by business today to distribute software,
patches, and documention. FTP also serves as a foundation to other
programs such as Gopher and HTTP. Telnet, which allows a user to remotely
log in to another computer, was also developed during the 1970's (BBN).
This gave students the ability to use an inexpensive terminal to connect to
a powerful mainframe to work on research or experiments, at a fraction of
the cost of their own super computer (Levine, Baroudi, Young, 22). Telnet
formed the basis for dial-in connections. Today all home users connect to
the Internet by dialing up to a mainframe, and invisibly begin a telnet
connection. Without telnet, few home users could afford to use the
Internet. Network Voice Protocol was designed to allow telephone-like
connections over the Internet. Unix to Unix Copy Protocol (UUCP) was also
developed (Zakon). It serves the same purpose as FTP, however it is not
TCP/IP dependent, so it is not considered part of the Internet. It is the
basis of USENET, which is where newsgroups reside (Levine, Baroudi, Young,
169). A Multi User Dungeon (MUD), a game that users could log into and
play role playing games together was created. It showed the Internet was
not just for work but also for play (Zakon).
After the seventies, only a few things happened to the Internet. The World
Wide Web (WWW) was formed. Although it seems like a great achievement, it
is just based on the TCP/IP services developed in the seventies. Also the
Internet has seen tremendous growth, yet the methods defined in the
seventies have withstood the stresses of expansion. The power of TCP/IP
was proven during the Gulf War. Jason Manger tells how Iraq set up a
TCP/IP network for their command structure, and the US could not take this
network out (2). Without the work done in the seventies, the Internet as
we know it would not exist.
Home | Events | People | Inventions | Art
Literature | Music | Entertainment | Fashions
Movie Reviews | Research Papers