1 2 3 4 5 6

alliances with just about everybody.

Ironically, the suggestion that the Internet should adopt settlement comes at a time when, according to the Economist,[8] the telephone industry is considering scrapping settlement in favor of a sender-keeps-all system. American companies complain that they are subsidizing inefficient foreign carriers by having to fork out those billions every year. At the same time, the magazine notes, the much higher cost of leased lines outside the United States means that it's more expensive for those countries to add bandwidth than it is for the United States to do the same. Since about 60 percent of all Internet sites are in the United States, connecting to the United States is a higher priority for non-American carriers than connecting to the rest of the world is for Americans. The upshot is that the costs of upgrading international bandwidth fall disproportionately on overseas carriers, who complain that the rest of the planet is subsidizing the United States. Until the United States cares as much about connecting to the rest of the world as the rest of the world cares about connecting to the United States, this situation is likely to remain.

In a massive article for Wired on the laying of the world's longest wire--the first privately financed (by a consortium of companies including Nynex, New York-based Gulf Associates, and others from Saudi Arabic and Japan) high-capacity fiber optic cable from the United Kingdom to Japan--science fiction writer Neal Stephenson pointed out that settlement has also backfired on the telephone companies in the form of "callback" services, which take advantage of the price differentials between the United States and elsewhere.[9] Someone subscribing to such a service in London, for example, dials a computer in the United States, which immediately dials back with an open line from which the caller can dial out to any number. Even from London, which probably has some of the cheapest telephone rates outside the United States, such services can cut the cost of a call to a third of the price. Such services are illegal in many places, but just as Netheads find ways to alter spellings to bypass computerized censors, the callback services find ways to counter blocking mechanisms local telephone companies put in place. Given the international nature of the Internet and its connections, it's hard to see how the Internet equivalent of such services could be stopped if settlement became a way of life on the Net.

Another problem with settlement is how to charge. Services like CompuServe ran complex, tiered pricing schemes for years. On CompuServe, you pay for the amount of time you spend in some areas and the amount of material you download in others. Contributing files is free. But that's on a single proprietary system that was designed from the outset to handle such billing. The Internet was not so designed, and is not a single system that can be easily metered. Charging by the amount of traffic is also a complex matter on something like the Web, where a user may send a tiny amount of information one way--a URL--and get an unpredictably large amount back. Charging by the packet wouldn't lead anywhere useful, since each one is acknowledged by the return of another packet.

At least some of the bandwidth problem posed by the Internet phone and other applications such as video conferencing via products like CU-SeeMe, fancy advertising using Java, complex graphics, computer animations, and the potential resource guzzling of automated software agents and robots[10] will be solved technically. Better compression to cut down on the amount of data, local storage of popular audio and video files, and new developments such as satellite broadcasting of some sites will take some of the load off the Net. But part of the solution has to be structural: if millions of people worldwide use the Internet for real-time multimedia applications the network will have to be upgraded to cope--particularly if, as many old-time Netizens hope, there are to be two-way broadband connections to each home so that any consumer can also be a publisher. Ironically, the same advertising that may sponsor free content on the Net may add to access costs by requiring bigger bandwidth. The money to create that will have to come from someplace, and if it's going to come from commercial entities such as telephone companies they want evidence that they will be able to make the money back. Charging ISPs for access to the phone network is one way; being able to sell information services such as video on demand and home shopping (if you count those as information) is another.