The growth of the railroads in the United States is also a good story
to study. There are many parallels with the Internet story.
Another story to study is the telephone industry...
In the "old days", telephone systems were based on a model where
the central core switching system was very expensive and the telephones
were cheap. Centralized authorities made some sense, partly because
the equipment capitalization was enormous. Common control was easier
and could be performed by a few expensive systems.
With the Internet, the capitalization is largely incremental. The core
"switching system" is low-cost compared to telephone systems. Unlike
low-cost telephones, PCs and servers are placed on the "edges" of the
system and help to contribute to the common control.
This incremental capitalization allows people to start small and to
grow at their own pace. Also, companies can select prime urban markets
and bypass rural areas. This is similar to what happened with railroads.
As companies grow (like railroads) they start to encounter other growing
companies and they best serve their customers by coordinating routes
that allow packets (passengers) to transfer from one place to another
with ease. It does not do a customer any good to be stranded in Peoria
with no route to Des Moines.
The Internet is only one component of the rapidly evolving story of
a decentralized telecommunications system. The Internet would not
be here if UNIX had not been set free, if AT&T had not been split,
and if PC technology had not been allowed to evolve without centralized
Unfortunately, the lack of centralized control allows the Internet to evolve
but also discourages coordination. Eventually, customers (passengers)
will be the ones that get to decide which ISPs (trains) they ride. ISPs
that work together to provide customers with the best service will survive.
Those that do not will end up as carnival rides without connections to