Internet, one must understand something about its original design. And
then, even more important, to understand as well that changes to this origi-
nal architecture are likely to affect the reach of innovation here.

so which code matters? Which parts of the architecture?[3-34]

The Internet is not a novel or a symphony. No one authored a beginning,
middle, and end. At any particular point in its history, it certainly has a
structure, or architecture, that is implemented through a set of protocols
and conventions. But this architecture was never fully planned; no one de-
signed it from the bottom up. It is more like the architecture of an old Eu-
ropean city, with a central section that is clear and well worn, but with
additions that are many and sometimes confused.

At various points in the history of the Net's development, there have been
efforts at restating its principles. Something called "RFC 1958," published
in 1996, is perhaps the best formal effort. The Internet was built upon "re-
quests for comments," or RFCs. Researchers -- essentially grad students --
charged with the task of developing the protocols that would eventually
build the Internet developed these protocols through these humble requests
for comments. RFC 1 was written by Steve Crocker and outlined an under-
standing about the protocols for host ("IMP") software. Some RFCs specify
particular Internet protocols; some wax philosophical. RFC 1958 is clearly
in the latter camp -- an "informational" document about the "Architectural
Principles of the Internet."[3-35]

According to RFC 1958, though "[m]any members of the Internet com-
munity would argue that there is no architecture," this document reports
that "the community" generally "believes" this about the Internet: "that the
goal is connectivity, the tool is the Internet protocol and the intelligence is
end-to-end rather than hidden in the network."[3-36] "The network's job is to
transmit datagrams as efficiently and flexibly as possible. Everything else
should be done at the fringes."[3-37]

This design has important consequences for innovation -- indeed, we can
count three:

____ * First, because applications run on computers at the edge of the net-
____ work, innovators with new applications need only connect their com-
____ puters to the network to let their applications run. No change to the
____ computers within the network is required. If you are a developer, for ex-
____ ample, who wants to use the Internet to make telephone calls, you need

[[36]]

p035 _ -chap- _ toc-1 _ p036w _ toc-2 _ +chap+ _ p037

v?