IX - Creating an Internet
The protocol suite that makes the Internet possible is known as the TCP/IP protocol suite (Transmission Control Protocol/Internet Protocol). Lundh explains the extensive effort needed to transform the design into functioning protocol specifications. He describes the years of experiments, analysis of the results, and the design of new experiments to test the theory developed from the experimental
process. Failures or surprises from the actual experience of the researchers helped them to make the needed changes in the implementation efforts. Lundh writes:
"Those protocols resulted from an extremely thorough analysis and design.
'No stone was left unturned' during the development which took several
years. Theoretical analyses were complemented by experiments. Combinations
of traffic types and requirements, network topologies and application types
were imagined, tried, failed, changed and tried again. The 'final' TCP and
IP were not easily postulated and approved. Nobody can ever reproduce in a
laboratory the chaotic traffic pattern of a lively telecom or computing
network and even less the diverse demands of information exchange. The
growing active dynamic traffic situation in the ARPANET prevailed during
onwards development of its own underlying technology. That may be one reason
for the robustness, elegance and survivability of the result." (Lundh, 12)
Lundh emphasizes the importance of a functional network with actual users and traffic as a laboratory for the researchers. He describes how theory grew out of experimental research and then was used to guide the experimental process. In this way, the theory was verified or modified.
Recalling his experience, Lundh writes, "During a period of intensively active development, methods were conceived and perfected until functioning well in an environment which was closer to reality than anyone might have dreamt up in a 'sterile' laboratory." This experimental process was closely intertwined with theoretical development. He adds:
"At the same time a profound theoretical understanding was developed.
It kept its scrutiny on experimental results and was both guiding and
following up the work in an admirable teamwork."(Lundh, 12)
Describing the political conditions that had to be accommodated to create a protocol that would function for the international community, Spilling explains the rationale of the TCP design:
"In order to allow Host computers, connected to different networks to
communicate, these networks have to be interconnected. This is not a
trivial matter, since different networks, in general, are supported by
organizations with different requirements and therefore will develop
differently. Any changes in existing networks in order to interconnect
these, will be costly and impeded by political factors. The obvious
approach therefore, would be to leave the local nets undisturbed and
to perform the interconnections outside them. This is one of the main
ideas behind the TCP." (Spilling, Proposal to Nato, 5)
The protocol requirements were such that the networks participating in the Internet would not be limited in their internal development or activities.(21) The use of gateway computers helped in this process. Gateway computers would reformat the packets of data from the form needed by one network into the form to meet the requirements of the next network on their journey to their final destination. The gateway software would also determine the best next path for the packets of data to take to get to their destination.
Spilling explains that when Host 1 (on Net 1) wants to exchange data with Host 2 (on Net 2), it forms the data into Internet packets according to the TCP format and encloses them in the format required by Net 1. This action, he says, is called "wrapping."(Spilling, Proposal to Nato, 6) Spilling attributes the term "wrapping" to an article by Louis Pouzin and H. Zimmerman. Internet packets are then transported to the gateway where they are unwrapped from the Net 1 format and rewrapped in the format for Net 2 for transmission to Host 2 (on Net 2).
| 