On Friday, October 23rd, Vinton Cerf came to DePaul to talk about the Internet: the history, the present, and the future. Vinton Cerf, who came as part of the Distinguished Lecturer Series, entitled his presentation "Interplanetary Internet." This is what he had to say:
In the late 60's, Cerf was teaching at the University of California, when ARPA, a government organization, came to him to help work on designing a way of linking all the major government computers together for the Department of Defense. The method that Cerf devised was called Transmission Control Protocol (TCP), which used information sent by packets. Basic research of information packets went on from1969 to 1985. In October of 1972, ARPANET was launched. This was the basis of inter-networking (or today, Internet). The theory of ARPANET was that Defense could use computers for mobile control and battle. Of course, it would be rather hard to run wire to submarines and tanks so that radio and satellite had to be used. Information is sent from one computer to another in small packages of data or packets.
While Cerf was at Stanford in 1973, he wrote a discourse entitled "Internetting" and he brought it to formal publication in 1974. However, the Internet's true groundwork was not laid until 1983. Cerf set a sentinel date of January 1, 1983 as the date to have all systems in his academic group communicating by TCP. He had to coerce his fellow professors and administrators by shutting down all non-TCP compliant communications systems twice before the date, a day in June and two days in October of 1982, until they switched over from the old systems. On January 1st, they shut down the old communications and found that 70% of the systems worked. After a couple of months everything came up to par.
At this time, there were only 400 computers in the Internet and they were all mainframes. No one turned a profit on the Internet until 1986,when companies like Cisco Systems and 3Com started producing routers which they sold to schools and the government. However, at that time, commercial services were still not allowed on the government-ran networks. In 1989, Cerf persuaded the government to allow MCI to provide mail services for one year. Of course, many companies sprang up saying that they too needed to provide service or MCI would have an unfair market advantage. They were allowed on and haven't been off since. In 1990, UUNet became the first Internet Service Provider (i.e. for websites) but at that time, websites were still text-based. In 1992, NCSA built Mosaic, the first graphical web-browser. They gave away a million copies since they were non-profit, educationally sponsored. Cerf noted that "people said this is better than text." In 1994, the World Wide Web came to the public's attention with quick and prolific media coverage, and in 1995, the World Wide Web came to Microsoft's attention. At that same time, the government moved away from supporting the Internet's bandwidth hungry backbone. However, the government still manages the Domain Name Server entries, which matches unique domain names with corresponding numerical representation, through Network Solutions, Inc.
Cerf stated that "we're in an Internet gold rush. Of course, you know who gets rich in a gold rush. It's the ones selling picks and shovels." Current estimates state there are 13 million Domains, with 37 million hosts, 100 million users of the Internet, and between 206 and 246 countries that are on-line. Interestingly, Finland has the highest per-capita of households that have Internet service, but there are 50 million users in the United States. UUNet, the Internet backbone (where all transactions go through) supports 60 terabytes of data a week, 75% of which is in World Wide Web traffic (as opposed to mail, true ftp, etc.) Cerf stated there were at least 350 million web pages, but he believes 90% of the Internet is, as he said, "dark" (or created on the fly like CGI-based pages). He predicted 300 million Internet users by January 1, 2000 and from 100 to 200 million computers by 2001.
However, he stated that companies put more money into Intranet (web sites, mail systems, etc ran entirely enclosed inside a business's network) than Internet. However, there will be a sling-back effect when they realize they'll have to connect to the Internet to interact with other divisions and companies. (Currently, MCI is working with the US Post Office for Internet transactions and Wall Street's NASDAQ system. NASDAQ has to maintain a 99.99% reliability rating and connect with every system using it within milliseconds.) On the other side is Extranets. These connect private companies with each other to compare. (Currently, Bell Corporation is designing a closed Extranet for GM, Ford, and Chrysler to interact.)
By 2002, estimates state that Internet-based commerce (or more trendier, "e-commerce") will reach 32 billion American dollars a year or approximately 1% of the world's 30 trillion dollar world economy.
All this promise is very exciting for IT people, Software Engineers, and especially those working in the Internet-based market. However, it also brings up a lot of issues. The first issue is information security. Cryptography is a growing field that focuses on three issues:
To do this, businesses are now being authenticated by banks and the Better Business Bureau. Also, with security on a personal computing level, a business would want their employees to have security keys to their computers. Then, of course, they would need key recovery schemes in case of an emergency. Firewalling is the common security tool on today's networks, but firewalls only protect information from sources outside the network, when 80% of all information crimes are inside jobs, so there is a need for inside firewalling.
Another problem is the ever-present demand for bandwidth. Fiber optics have already been tested to move terabits of data per second, but switches and routers just aren't that fast. Routers are slowly getting better. This need for large amounts of information passing is primarily due to multimedia requests. Currently, there are 3500 radio stations that exist entirely on the Internet and do not broadcast in air waves. One-way transmissions like radio, television, and video are bandwidth-hungry. Digital Subscriber Loop (DSL) cabling would require the rewiring of 30% of all homes in America. In addition, fiber would need to replace coaxial cable for cable modems to get DSL speeds.
Still, the Internet has revolutionized marketing. However, it isn't the first time that a new technology had brought sales out of the store and into the home. 100 years ago, Sears and Roebuck sent out a catalog of products to everyone in America for them to order from home. Of course, the Sears and Roebuck catalog worked a double purpose a hundred years ago. To better understand this, think of a consumer working on his laptop in a bathroom stall and realizing he has no toilet paper. As Cerf put it, "Maybe this paperless society wasn't such a good idea after all." To put Internet sales into perspective, Dell sold 600 million dollars in the month of October last year (1997). Amazon brought in 87 million dollars in revenue in their first quarter of 1998, while Autobytel sold 1.5 billion dollars in cars.
The way the Internet works is changing how commerce works, too. Self-service, where the customer surfs a site and finds products, using a "shopping cart" to manage them, has taken the need for many employees away. Even help is automated with Frequently Asked Questions (FAQs) pages, automated response centers, and Customer Forums, where customers talk to each other in a chatroom system to discuss solutions. (Of course, when AOL tried Customer Forums, the users took the opportunity to bash AOL and AOL shut it down.)
Some day, Cerf could see people having dedicated lines (where you are constantly connected to the Internet, just like you are constantly connected to the phone system) instead of dialing-up to an Internet Service Provider. That would be the age of Internet-enabled appliances, (say for example, a bathroom scale that automatically sends output to your doctor.) One company has already attested that they could put TCP/IP connectivity on a microchip for 35 cents a chip. There could be clothes that monitor your bio-information or global tracking, or any other possibility. Some companies have already introduced micro-computers the size of pagers that have functionality to play radio, act as a pager, work as a cellular phone, and send and receive e-mail.
Cerf told us a story about how amazing technology is. When his wife was 3, she lost most of her hearing to a childhood ailment. She never regained it, but was fit with a hearing aid. The hearing aid gave her sound at a 100 decibel loss. (That's about the point where she could stand next to a jet engine and not hear it.) Well, she heard of a procedure on the Internet for people who had hearing impediments at her level. She searched around and found some people who had the procedure and were very satisfied with the results. After some discussion, she found that there was a doctor at John Hopkins who would perform the procedure. She worked to contact him with no answer. She e-mailed him and he responded the next day. Anyway, she flew to the hospital and they performed a cochlear implant, and she flew back home the same day. There was a two-week period of activation to make sure the surgery took. She flew back again and had it activated, once again in an outpatient procedure.. The device consisted of a pager-size computer that she would keep with her, a microphone to hook onto her lapel, and a dual-magnet system. 20 minutes after they finished the procedure, she called him up on the phone.
Other issues that are present in the Internet are:
At this point, Cerf segued into his current project: the Interplanetary Internet.
In 25 years, there will be manned base stations on the moon and on Mars, much like there are ones at the poles today. Current TCP/IP technology won't cut it with the three-hour delay in one-way transmission from Earth to Mars. The Jet Propulsion Laboratory is working with Cerf on designing new communications technologies. This is a truly exciting step for NASA and other space-driven organizations because until now, each space mission used its own communication designs and now there will be a set standard.
Planetary ("dirt-side" or on each planet) Internets will be on each inhibited planet or moon for communication between bases and vehicles. Then, there will be planetary gateways, transmission centers off the planet, which must compensate for the big delays as well as background noise. Communications satellites will be used. However, since signals can't be sent through the sun, there will be a ring of satellites in a non-elliptical orbit around the sun. The Mars mission will be in 2001. This mission and each subsequent mission will carry a piece to slowly build this communications system.
In the meantime, the Planetary Society helped children build remote controlled robots in large sandboxes to simulate the Pathfinder vehicle. The controls are run by computer and the robots are controlled over the Internet. Using multiple cameras per sandbox and multiple sandboxes, and the lag time between signal sending and signal receiving, these children are displaying what controlling these Martian vehicles will really be like.
So when this dream becomes a reality, everyone will want a chance to interact with Mars, taking pictures remotely and whatnot. Cerf pictures a lottery system to give people a chance to do this. He also predicts 2005 to be the year where Mars will be an accessible node and 2018 as the year for the earliest possible manned mission to Mars. However, he will be publicly reporting the long-term vision for the Interplanetary Internet.
Besides being Clinton's primary advisor on Technology affairs, Cerf is also the president of the Internet Society (ISOC). At the end of his lecture, he noted that there is a Chicago-based Chapter currently opening.