1. The Global Internet Project: Who We Are

The GIP is an international group of senior executives committed to spurring the growth of the Internet worldwide. It is composed of 13 senior executives from companies around the world. Its founder is Dr. James Clark, the founder of Netscape Communications Corporation. A full list of members is available at http://www.witsa.org/gip/about/members.htm.

The Internet Society (ISOC) has also endorsed this GIP paper. ISOC is a non-governmental International organization for global cooperation and coordination for the Internet, its internetworking technologies and applications. The Society’s individual and organizational members, from over 150 countries, are bound by a common stake in maintaining the viability and global scaling of the Internet

The GIP strives to inform and educate governments, industry, international organizations and the public on the nature and significance of the Internet and what it can become. The GIP is committed to promoting the Internet as the global platform for computing and communications.

As an international group, GIP members believe that expansion of the Internet will depend to a great extent on the ability of companies and consumers to obtain products and services in a secure, flexible, convenient and easy-to-use manner. A fundamental precept of the GIP is that because the Internet is a global medium, it is critical to address its challenges globally. The group is working with appropriate national and international bodies to find answers to a variety of difficult issues to assure the best possible future for all members of the Internet community.

2. The Internet's Rapid Growth

In its recent paper on "Internet Foundations: Breaking Technology Bottlenecks" -- the Global Internet Project identified several growing pressures on the Internet's infrastructure. In particular, the Internet is experiencing a rapid growth in the number of Internet users and the demand for more sophisticated applications by these users. One of the major bottlenecks identified in this paper are bottlenecks that occur at the server level.

By almost all measures, the impact of the Internet has been profound . . . and continues to grow. At the close of 1998, there were roughly 140 million users on the Internet worldwide. The number of World Wide Web pages increases by an estimated 300,000 per week. Traffic over the Internet's infrastructure is doubling every year, and some ISPs see traffic growth of 1000% a year.

Moreover, the spectacular growth of the Internet is a global phenomenon. Between 1993 and 1996, the number of Internet hosts in Europe increased by about 600 percent, according to figures developed by the International Telecommunication Union. During that same period, the growth in Internet hosts in Africa and Asia was even more meteoric – more than 800 percent for each continent.

This exponential growth will enable information access and electronic commerce in innovative ways that will have a fundamental, positive impact on how people live, learn and conduct business. The ongoing growth of the Internet promises to have a far-reaching impact on overall network technology and evolution.

The enormous success of the Internet as a source of information and as a platform for e-commerce has led to some challenges. Users’ expectations for a fruitful and enjoyable experience browsing the Internet can lead to some frustration due to occasional slow performance and unreliable connections. At times, Web sites can take a long time to access and pages a while to download. However, a technology called caching plays a big role in relieving Internet congestion.

3. Caching's Role in Relieving Internet Bottlenecks

Caching is a fundamental and growing part of the way the World Wide Web works today. Caching is the automatic creation of temporary copies of information residing on computers other than a host server in order to make this information more easily available to people around the world, and to prevent traffic jams when users wish to access a Web site. Caching technology, in short, allows for the temporary storage of digitized materials. Without caching, users would have to go all the way back through the Internet to the original server of the publisher to access data. In addition, publishers would need to make significant investments and increase the capacity of their web servers in order to meet the increasing demand for their information. Caching reduces the need to re-transmit information from the source server unnecessarily, and this, in turn, reduces absolute capacity requirements in the Internet. Caching is carried out through an automated technical process, and can take place at the user’s computer or at intermediate computers in the network between the user and a web site server. The rules associated with caching have been standardized and endorsed by all major Internet companies and the governing bodies of the Internet.

Caching is playing an increasing role in relieving bottlenecks at the server level and ensuring that Internet web sites can be accessed easily anywhere in the world at a reasonable cost. It is also important to note that caching only involves making copies of material that already is accessible on the Internet. It does not give people access to information that is not available on the Internet. So caching merely improves the ability of the network to provide access to information which is already there.

Given the dual pressures of exponential user growth and increasing demand for sophisticated, bandwidth-hungry applications that involve features such as Java applets, electronic commerce, multimedia and graphics, caching's role in the future of the Internet will become even more important.

4. Types of Caching

There are three basic types of caching in use today. They are:

1. PC or Browser Caching

   This form of caching takes place on an individual's Personal Computer. When an individual clicks on the back button of his/her web browser, the software is showing a temporary reproduction of a site that the person had just visited (as opposed to going out on the Internet to find that site again). This simple technological tool allows the user to quickly see the last web site visited, and relieves a great deal of congestion on the Internet. If the user decides that he/she would like to actually visit the web site again to see if there is any updated information, the user would click on the "refresh" icon or type in the URL or address into the browser. Most browser software periodically erases all data stored on these temporary caches in order to avoid filling the user’s hard drive.
   
   

2. Proxy Server Caching

   In Proxy Server Caching, temporary copies of web pages are copied and stored on other servers to allow for easier and more rapid access to this information. This is a technological tool used by Internet Server Providers (ISPs) and other network providers, and is also used heavily by large web sites who attract many visitors and "hits." Proxy server caching is also a way to distribute content geographically around the world to allow for easier access and to reduce the telecommunications costs of access to web sites far away. Proxy server caching, for example, may be part of the solution to concerns in Asia that Internet access costs are too high, which partially result from the large number of content and web sites based in the U.S.

   ISPs and other network providers that have deployed caching systems from Inktomi Corporation, one of the leading makers of caching software, have experienced "hit rates" of up to 80% from the local cache. This means that up to 80% of the information requested by end users is served from the local cache, rather than retrieved from original servers located across the Internet. This reduces the amount of redundant traffic on the Internet, which substantially improves the efficiency of the network and the experience for the end user.

   One of the challenges associated with proxy server caching is allowing for web site owners to retrieve accurate and detailed information related to how many visitors their sites are attracting. Such information is important for advertising revenues. Therefore, service providers must comply with rules concerning updating material or transmitting material back to the originating source, but the originating source cannot abuse these rules.
   
   

3. Mirroring or Mirror Caching

   "Mirroring" refers to separate but identical sites being set up on different servers so that if one server is experiencing more traffic than it can handle or is not functioning at all or "down," then the traffic is automatically diverted to the mirror site. Mirroring also can be a useful tool if the servers are distributed geographically to allow for more rapid and inexpensive access to web sites. For instance, during the Nagano Olympics, IBM created five mirror sites in Japan, the United States and Germany, to enable web users throughout the world to get easy and rapid access to the latest results from the winter games.

   Mirror caching is usually based on contractual arrangements between the original source and the provider of the service or mirror caching.

   An approach called "replication" may also become more significant in the future. Replication refers to duplicating data on servers at several points on the network, in anticipation of user demand, rather than caching it in response to user demand.

5. GIP Positions and Recommendations

• The GIP believes that caching offers many benefits to Internet users around the world.
• It is an important part of the answer to relieving Internet bottlenecks and making information on the Internet easily available to people globally at a reasonable cost.
• If caching were no longer permitted, the Internet would experience many more "traffic jams." This would increase the frustration due to what some have called the "World Wide Wait." Banning caching would exacerbate the problem of providing sufficient capacity to meet today’s demand, let alone tomorrow’s. Without caching, the Internet could lose its important role as a vehicle to promote electronic commerce.
• If caching were banned, the cost of Internet access would increase substantially, because Internet Service Providers (ISPs) would need to make additional investments to install more network and high performance communications technology.
• Even worse, the cost of hosting web sites and making more information available on the Internet would increase substantially, thereby reducing the amount and diversity of content available on the web.
• The GIP strongly supports the goal of protecting copyrighted works in this global digital economy.
• The GIP believes that there are technological solutions available to ensure the protection of copyrighted materials, such as software preventing anyone from making temporary copies of the information on a web site, digital watermarks, authentication tools and other approaches.
• The GIP applauds the efforts of the World Wide Web Consortium and the Internet Engineering Task Force (IETF) to refine and expand the standards that make caching efficient and ensure that it is done in a way that meets the needs of ISPs, authors and publishers, software manufacturers, technology tool designers, and others. The hypertext transfer protocol (HTTP) is a major element in ensuring that data transfer related standards are robust and adhered to by network operators.
• The GIP supports the compromise struck between content owners and technology providers in the Digital Millennium Copyright Act (DMCA) passed last year by the U.S. Congress. The DMCA permits caching and was the result of extensive negotiations among interested parties, many of whom have a significant presence in Europe. Among other provisions, the DMCA:
  1. Permits an online services provider (OSP) to make "a temporary copy of popular Internet material requested by a user so that the OSP can deliver that copy to subsequent users, which can be done more quickly and efficiently than obtaining the original material for each subsequent user."
  2. Requires the OSP to "comply with any rules on updating the cached material that are specified by the Originator using a generally accepted industry standard protocol."

April 5, 1999