Notes from 53rd International STC Conference
Las Vegas, Nevada, May 7-10, 2006

Keynote Address: Fireside Chat with "Fathers" of the Internet

Patrick Whitney

Watson and Crick had DNA. Cerf and Kahn had the Internet. To be fair, Vinton G. Cerf and Robert E. Kahn are only two of many scientists and researchers considered "Fathers of the Internet." But like Watson and Crick, Cerf and Kahn are known for their teamwork. In the early 1970s, they designed the communication protocols that allow computer networks to communicate with each other: transmission-control protocol (TCP), which breaks data into "packets" and routes it to host computers, and internet protocol (IP), which addresses and forwards the packets. Without standards that establish how dissimilar computer networks to talk to each other, there would be no e-mail, no streaming video, no Google, no Amazon. In short, no internet. Cerf and Kahn are STC’s Honorary Fellows for 2006. The awards were conferred before their keynote address.

Session Description: The "fathers of the internet" took an interesting look back and forward on the communication medium that has become the linchpin of both business and personal communication worldwide.

• Bob Kahn came up with the original idea in 1972, based on earlier work on the ARPANET.
• Local networking (as opposed to batch processing) was originally ridiculed as an impractical idea, but the inventors of the ARPANET persevered nonetheless due to the need to share valuable computer resources among researchers.
• Kahn and Cerf saw the need to "federate" all the diverse networks together into a seamless online environment (the internet).
• Early DARPA funding of the internet was not easy to get, because initially the users didn't even have computers, let alone networks. The idea was somewhat ahead of its time. Once computers proliferated and networks emerged, the architecture was ready to create what amounted to an intranet (ARPANET).
• Originally, the system was contained to the defense department, but the TCP/IP protocols were never patented; from the outset, they were free to the world. This decision was a key factor in the evolution of this new idea into the biggest breakthrough in communication since the invention of the printing press.
• What, exactly, IS the internet: Is it a "network of networks"? To this day, there is no one definitive description.
• Present estimate of internet users is about 1 billion.
• It took 20 years from the creation of the original architecture (1973) for the internet to hit the marketplace (1993), when the "worldwide web" emerged. E-mail spread the popularity of the new tool, and commercialization has followed.
• The opening up of the web resulted in an avalanche of material being "dumped" onto the internet, ranging from useless to profound.
• Search engines added usefulness to the internet. They continue to be refined. (Cerf is a vice president for Google.)
• One of the fears about the original idea for the internet was that the little modules of information would get queued up in an endless "traffic jam." As it turned out, this was not a problem; the depth and breadth of the internet accommodated the kind of freestream flow that permitted mass proliferation of data at the speed of electrons. Kahn drew a comparison to the natural distribution of oxygen molecules in the large ballroom. Worrying about the e-modules of information queueing up, while not unreasonable, proved to be like worrying about all the oxygen molecules in the ballroom suddenly congregating in the upper righthand corner of the room, near the roof, resulting in the suffocation of the entire audience. The internet proved to be just as amenable to the free flow of electronic information as the atmosphere is to free flow of oxygen and other molecules.
• Commercialization has added impediments to the free flow of information on the internet. Inevitably, economic interests have resulted in attempts to "control" the new environment in order to profit from it.
• Information management is an important emerging field related to the internet (archiving, meta-tagging for search engines, etc.). Kahn has been specializing in this over the past few years. A key element was the ability to recognize data structures independent of where they originated, where they are stored, or the information is file-named (digital objective identifiers, or DOIs).
• Longevity of the equipment that can read media is a growing issue. Witness the move from 5 ¼-inch floppies to 3 ½-inch floppies to zip disks to CDs to DVDs to memory sticks ... what's next? What about moving data archived in older media into the newer media?
• The long-term answer is to move towards more abstract, universal forms of electronic archiving that can transcend evolving soft media. In essence, the internet could house everything in DOIs. These, in turn, could be archived on all kinds of new media and always be compatible with the latest storage devices—universal long-term single-sourcing. Each data structure would have a unique identifier, like its "DNA," that would always allow it to be retrieved, no matter how the internet evolves over time.
• Individual internet providers might seek to "control" information by tagging it in some manner, but the inherent identifier of each data structure would always be to the universal open architecture, permitting data to be moved from one provider to another.
• What does the future hold? The internet inventors foresee new forms of collaboration devices that will enable virtual experiences at levels we cannot currently imagine.