Bob Kahn may not be the most celebrated name on the Internet, but that says more about his soft-spoken nature than his considerable accomplishments as one of the central figures in the creation of the Net.
While Vint Cerf is widely recognised for his work on TCP/IP, Kahn was an equal partner in that project. In 1966, Kahn went to work for Bolt, Beranek and Newman, where he was one of a half-dozen employees to design and build the ARPAnet, the forerunner to the Internet.
Kahn then led teams that developed some of the first useful applications for the ARPAnet. From there, Kahn joined the Defense Advanced Research Projects Agency (DARPA), providing grants that eventually led to the creation of VLSI technology.
Kahn left DARPA to found the Corporation for National Research (CNRI) in 1986 and has remained involved in the groups that make the Internet work. These include the Internet Engineering Task Force (IETF), the Internet Activities Board, the Internet Society and IOPS.ORG, an association of large Internet service providers dedicated to improving Internet operations. Today, Kahn spends much of his time on the design and development of a digital object infrastructure to help solve problems of intellectual property. Like packet switching in the 1970s, it will help diverse computer systems work together.
You've been watching the standards process since Request for Comment (RFC) 1, but that game is now being played for real money. Do the standards processes need to change as a result of the Net's commercialisation?
"The standards process has evolved significantly. The RFC started as a way of sharing information in the context of the original ARPAnet development and were carried over to the Internet when we had multiple network environments and lots of commercial activity. Initially, there was the Internet Configuration Control Board, which was replaced by the Internet Activities Board. It allowed people to participate in working groups instead of coming to one centralised meeting. The number of working groups grew to be so large that one of them took over the task of managing the whole process. That became what is now known as the Internet Engineering Task Force (IETF). Over time we saw the creation of the Internet Society to provide a home for the standards process.
"We now have a mechanism that evolved from the research-oriented capabilities to encompass quite a bit of the commercial interests. In fact, it may be dominated by them at this point. We need to make sure the kind of dynamic that has existed for research communities is preserved as the Internet becomes more commercial."
Can you relate that specifically to what's going on with the domain name issue?
"The DNS issue is a different matter. The problem was exacerbated by the National Science Foundation's decision to allow InterNIC to start charging annual fees for registration in lieu of government funding. The effect was to create an unacceptable monopoly in the allocation of domain names.
"One way to resolve that is to allow additional registration functions, but there was no mechanism to decide what was official, or who should decide. The government was not taking any direct action on this, but many people believed it was the government's responsibility to do so.
"The private sector decided to do something and the Internet Society took the lead, setting up the International Ad Hoc Committee (IAHC). To other people, particularly those already involved in the process, this approach seemed arbitrary, like 'what gave them the imprimatur?' That led to another set of concerns. This has everything to do with monopolies versus non-monopolies, and the charging of infrastructure services.
"Trademark issues also played a role - who registers what domain names when trademarks come into play? That's still to be resolved because trademark has historically been handled on a country-by-country basis.
"Since the Internet spans the globe and since not all domain names are country-specific, conflicts can occur. But I don't think it's the same problem as the other one we were talking about, the standards process in general."
What enhancements are being added to TCP/IP and what kind of applications will it enable?
"It's a little early to predict exactly how that will play out. Proponents of IPv6 were running out of IP addresses, therefore, we need something with a bigger address space. But how much is enough?
"The model that says every light switch, toaster, socket and electric light bulb needs its own separate IP address pretty well argues to the fact that we need a rapid and wide expansion. Whether someone really needs to send an IP message right to a light bulb or to a control system in a house that's smart enough to know how to control the light bulb is another matter."
What are the most viable alternatives?
"IPv6 is the most viable. As for the TCP side of things, there are certain fundamental attributes of the Internet that will always require something like TCP/IP protocols. Maybe we have to deal with things like real-time quality of service performance and performance guarantees. Just where that will fit into the protocol layering hierarchy is hard to see. My guess is it has got to infiltrate many of the layers if not all of them - just like security and multiplexing. Those are the areas that will get the most attention."
What's your take on ATM vs gigabit ethernet?
"The economics of gigabit ethernet may make it far more persuasive than ATM. My guess is that gigabit ethernet will do very well in the local environment, but the actual mechanism for doing the switching may derive a lot of its horsepower from what was learned in some of the ATM switch development. Wide area networking with ATM is still going hot and heavy."
What's the mission for CNRI?
"CNRI was established in the mid 80s to foster research and development for a National Information Infrastructure (NII). We're still a long way from achieving a fully functional, coherent infrastructure where everything can interplay and applications can talk to other applications - the same kind of foundation that, for example, the automobile industry has with the road system."
What's the technological focus for CNRI and digital libraries?
"It was among the three things we had on our agenda when we got started. A second area that we hoped to focus on was the national knowledge bank where we could codify knowledge about various areas into a computer processable form.
"We were going to start with science and technology but that area just wasn't mature enough. ARPA was interested in funding that area at a technology base level and has been doing it ever since. It's still an area we're interested in and may get back into when the technology is more advanced.
"The third area was e-commerce, but we just didn't have enough resources at the time. We're starting to get into it now. We're hosting an electronic payment forum with the Financial Services Technology Consortium and CommerceNet under the auspices of the Cross Industry Working Team (XIWT)."
What's your overview of how e-commerce will develop over the next few years?
"In a nutshell, what's missing is a way of organising all the pieces of this infrastructure so people understand what's out there and how to make use of it.
"It's not clear yet how one makes money on the Net. You can use it as a transport mechanism, essentially for ordering things, but I don't consider that a new use. It's sort of a replacement for a telephone order or a telegram."
Is digital object technology the key to corporate IS in the future?
"When you talk about an object in the sense of object-oriented programming, you're talking about an entity you can only touch by the methods that apply to it. It's up to the system to figure out what goes on inside.
"I'm talking about being able to deal with the data structure part as a separable entity that becomes the common currency of the new world of distributed systems. Once you do that, it's like a packet, and the world is a network. The XIWT is just coming out with a white paper on this subject (www.xiwt.org)."
Any thoughts on Microsoft's efforts to take back the Web or on Netscape's collaboration products?
"When you say Microsoft is trying to take back the Web, it never had it to start with. If you're saying that things are more competitive now, the credit would have to go to Bill Gates and his staff for recognising the importance of it and acting on it.
"Jim Clark (Netscape CEO) is a shrewd businessman and he's up against some serious competition from Microsoft because those guys are real heavy hitters. Microsoft has got some of the greatest minds in the community on board.
"It's the young, aggressive people who understand business that do well in it.
"Technology is not a static element, and you never know where the next bright idea is going to come from. The development of a digital object infrastructure is a really important contribution."
Jeff Ubois ([email protected]) is a contributor to Internet World.
Found by calculating the strength of the material deep inside the crust of neutron stars
Can highlight in real-time the relevant regions of an image being described
Double legal trouble for Musk as he also faces civil lawsuit over renewed British pot-holer 'paedo' claims
Battery development could help boost performance of smartphones