Some Myths about the Internet

By Patrice Lyons with input from Robert E. Kahn, Corporation for National Research Initiatives—The Internet is used in many of our daily activities, usually without us giving much thought to what it is. There is often an awareness of this ease of access to information for a wide variety of services, but there is often a misunderstanding about the Internet’s intrinsic meaning. The scholarly publishing industry has moved in recent years from a focus on printed publications to the dissemination of scholarly works represented in digital form, and the Internet has become an important resource in this context.

Many myths could be explored about what it means to be the Internet, but, for purposes of this brief note, we have purposely limited our choices to the following two:

Myth #1: The Internet is simply one example of an Internet Protocol (IP)-based packet network; and IP-based networks may exist independently of the Internet.

Reality: While the Internet means many things to many people, in essence, the Internet is a global information system that involves a set of protocols and procedures that enable interoperability thus making it possible for multiple networks, computers, devices and other resources to communicate with each other. While the architecture of the Internet has accommodated the addition of new network technology from the start, and, today, there are many networks deployed in the Internet, including more recently next generation networks (or NGNs), the Internet itself is not a packet network.

That an IP-based network or platform could exist independently of the Internet belies a basic understanding of the Internet. The notion of gateways (now routers) was introduced originally to deal with IP addresses and routing because of the myriad issues associated with changing the three existing packet networks (intended to be a part of the initial Internet) to support IP intrinsically. If some or all of the networks in the Internet today support IP intrinsically, they could be called IP-based networks or IP-enabled networks; in any event, such IP-based networks would be part of the Internet.

Myth #2: The “end-to-end” principle is an essential requirement of the Internet.

Reality: A key element in the original Internet design was the role of TCP (transmission control protocol) and IP. While IP remains the basic addressing component of the Internet, certain of the services and even applications that make use of IP, now or in the future, may be more effectively implemented with the active participation of individual networks rather than having those services relegated to what are sometimes called the “edges” or “ends” of the Internet and being “layered” on it. While the basic Internet design did not require adherence to an end-to-end concept, it was a part of the initial Internet implementation so as not to require the initial networks to change their internal operations in order to accommodate the Internet. This was simply an early implementation choice. Current advances in technology go beyond these early choices. For example, CNRI’s Handle System has been shown to facilitate high-level addressing for mobile devices and for digital objects, more generally. This addressing capability, via the use of unique identifiers, can be provided internal to a network, as well as external to it as a layered service. The Handle System is perhaps best known in the scholarly publishing community through the branded DOI System developed by the International DOI Foundation.

In 1995, the Federal Networking Council adopted a resolution defining what it meant to be the Internet. Since then, the Internet technology has continued to evolve along different dimensions. During the Working Group on Internet Governance (WGIG) in 2004, there was some discussion of what it meant to be the “end-to-end” principle with respect to the Internet. On that occasion, CNRI submitted a paper in which it proposed the introduction of words in the FNC’s “Definition of the Internet” to recognize that innovation might take place not just at the “ends,” but inside particular networks.

Ms. Patrice Lyons is Corporate Counsel to the Corporation for National Research Initiatives (CNRI) where, for over twenty years, she has been involved in a wide variety of issues relating to the Internet. Before working with CNRI, she served as an attorney at UNESCO in Paris and in the Office of General Counsel of the U.S. Copyright Office.