Over the years domain investors and website developers change DNS settings countless times, and the whole system just works. The field of domain name investment exists because standards were developed to robustly link domain names and internet addresses. This means that names that resonate with humans have great value. At the 2019 NamesCon conference, I had the opportunity to sit down with Dr. Paul Mockapetris, who invented the Domain Name System (DNS) architecture back in 1983. He provided insights on how the DNS system came to be, and how it continues to evolve. Dr. Paul Mockapetris photographed at the time of the interview at NamesCon 2019. Photo taken by Edward Zeiden. Interview BH At the time that you were developing the plan for the DNS system in 1983, did you realize how incredibly important it would become? PM I like to work on problems that other people think are not important and I think might be. My feeling at the time was that it could possibly be something very important. Most people wanted something that would just replace the single host table, but I was very fond of building distributed systems. It was more complicated than a lot of people wanted at the time, and now they ask for more features. BH I assume that at that time the idea of distributed systems was almost unheard of by many people? PM It was kind of lying around. Some people were looking at applications for distributed systems. BH This is a good time to ask about your education. PM I did a combined bachelor’s degree in physics and electrical engineering from the Massachusetts Institute of Technology. Then I headed to the west coast, where I worked at the University of Southern California while going to graduate school at University of California at Irvine. My doctorate is in information and computer science from Irvine. I was a long-term graduate student because I kept finding interesting diversions. BH Where exactly was the DNS idea developed? PM There are a lot of stories out there about where the DNS system came from. The underlying principles came from my work at the Machine Group at MIT that later became the Media Lab. I did not have very powerful hardware at the time, so I made a system combining multiple computers similar to a modern cloud. This was my early involvement with distributed computer systems. My work on distributed systems continued at University of California at Irvine where I was based when the parameters of DNS were being defined and published. BH Was your doctorate thesis on DNS? PM Not really, but it was on the related topic of distributed information systems. BH I read that you did work in the space industry early in your career? PM Long ago I worked at Draper Labs in Boston on validating flight software. One of the simplest ways to make something redundant was to have fallback with multiple providers. That idea was kicking around in my mind as I was thinking about how to define the DNS naming system. BH I guess at that early stage there was not one view of how distributed systems should work? PM There were two ways to handle distributed systems, a simple query, as well as the zone transfer method that keeps copies synchronized. Both were meant to be very simple, and they were complementary systems. BH Has the DNS largely evolved over these decades as you thought that it might? PM I had some initial ideas about how it should evolve. One of my principles was to keep it as simple as possible but have room for growth. It has grown in a number of ways, some of which I foresaw and some of which I didn’t. I always say if you design a system and can imagine all of the possible uses, it really is not very interesting. You should be building something that can be used for things that have not been invented yet. Otherwise it will become obsolete too quickly. BH What is your current position? PM We earlier talked about unintended consequences. I would contend that today there is more DNS traffic to stop things rather than to make them happen. In the early days the role of the DNS was principally to make email delivery work. Then spam started arriving and other threats arose. It is kind of interesting that although DNS-enabled modern email, now we spend much more DNS-related resources deciding what should not be delivered. I am currently Chief Scientist at ThreatSTOP. We make lists of sites that you shouldn’t talk to. We keep away traffic at the edges of your network to protect you. Our products make security work better in an automated fashion. I am also involved with several other projects, including one involving blockchain. BH I guess no NamePros interview would be complete without the following question. Do you personally invest in domain names? PM No. It seems to me like work. I probably should have, but no I do not invest in domain names. Honours Dr. Mockapetris has been awarded numerous honours. He was the recipient of the John C. Dvorak Telecommunications Excellence Award in 1997. In 2005 he was recognized by the Special Interest Group on Data Communications (SIGCOMM) with an award carrying this citation. In 2012 he was inducted into the Internet Hall of Fame as an Innovator. He has received honorary degrees and numerous other citations and awards. Over his career Dr. Mockapetris has held many different leadership positions. In the early 1990’s he was networking program manager at the Advanced Research Projects Agency (ARPA) of the U.S. Department of Defense. He has worked at numerous innovative companies including Chief Scientist at Nominum from 1999 to 2016. Among his pioneering accomplishments, he developed the first SMTP email server. Dr. Mockapetris is of course best known as the creator of the DNS system. That the DNS system Dr. Mockapetris designed has been so robust over many decades is a tribute to his intellect and foresight. Every Internet user depends on the system, and the whole field of domain name investment exists because of the effective link between domain names and Internet addresses. Even in a short interview, the bright, innovative, engaging and sensible nature of Paul Mockapetris shines through. It was no accident that it was he who invented the DNS system. Further Reading The principles of the DNS are laid down in two documents authored by Dr. Mockapetris. These were published online in the Request for Comment (RFC) series. That series is a hybrid between academic and discussion papers. The original documents provide interesting historical insights into the early development of the domain name system. The first paper defines the problem this way. The second paper, published the same year, provides the details of how the domain name system was to be implemented. Here is how that paper describes its objectives. The late Jon Postel, who was the editor of the RFC series at the time, played an important role in the development of the DNS and many other early Internet standards. He also is a member of the Internet Hall of Fame. Wikipedia has an informative article on the current domain name system, while this article from Cloudflare provides a nice description on how the DNS system works. You can read more about Paul Mockapetris various places including this biography at the time of his appointment to the Internet Hall of Fame. Dr. Mockapetris is currently involved in several ventures, including as Chief Scientist at ThreatSTOP. The company describes itself this way. Dr. Paul Mockapetris and Bob Hawkes at the time of the interview at NamesCon 2019. Photo taken by Edward Zeiden. I would like to thank Dr. Mockapetris for kindly agreeing to this interview. It was an honour to meet him.