Q&A with .ca ‘godfather’ John Demco

John Demco is a pioneer, in the true sense of the word. When the Internet was terra incognito, Demco helped put it on the map, and into the collective consciousness of Canadians. From his early days with the University of British Columbia's Department of Computer Science to his current roles of co-founder and director of Webnames.ca, a Vancouver-based domain registrar and provider of Internet services, John has been at the vanguard of the Canadian Internet.This week marks the 20th anniversary of John's first steps in creating a uniquely Canadian online presence: on May 14, 1987, the Internet Assigned Numbers Authority - the agency then responsible for IP address and domain name distribution - tapped John as the caretaker for the nascent ".ca" top-level domain (TLD). Thirteen years later, the Canadian Internet Registration Authority was set up to administer Canada's TLD, and the rest, as they say, is history.In recognition of his work in setting up the first registry for Canada's domain name, UBC will name a learning centre after him later this year. Canadian NEW MEDIA sat down with John earlier this month and asked him about a variety of topics, from the dwindling supply of IP addresses to the dilemma of ‘Net neutrality. How have things changed since you ran an Internet registry out of UBC?Well, I certainly have been involved with all of it but I want to say up front: this is not, in my way of thinking, about me personally. Lots of interested people freely given of their time and effort, and it just so happens that I've been associated with some of these things. I've had people that helped all the way along the way - I want to make that really clear.The two areas that I've been really interested in have, it turns out, come together: one of them was networking, and the other was naming. The .ca domain historically came about for me because...in the early 1980s - .ca didn't formally get started until 1987, but I'd been thinking about issues around this for five years or more before that - and until 1986 or '87 or '88, depending on what you measure, it wasn't absolutely clear that the Internet as we know it now was what was going to ‘win'.In the early days, and I don't know if you remember this or were involved with this or whatever, there were a wide variety of networks that were in use... Like ATM, or frame relay?Well, even before then, and not even as sophisticated as that....The people who were running the IBM mainframes that were still around at all the universities and big non-profits and government labs and so on, they came up with BitNet, and it was a way for IBM mainframes to communicate and to submit remote jobs. On top of that they put email and file transfer. At a similar time, the computer science community came up with something called CSNet, and the high-energy physicists came up with something they called HEPNet - the High-Energy Physics Network - and the energy people in the United States came up with ESNet.If you look around the world there were a few hundred different purpose-built networks growing up. And, they were by no means homogenous as far as technical capabilities were concerned. They didn't all use TCP/IP; in fact, not very many of them used TCP/IP at the time.I was manager of one of these networks, called CDNnet, which was for the Canadian research and education community. I saw an opportunity to have a common naming system, because not only did they use different naming protocols, but their email addressing wasn't compatible, and how you transferred files and exchanged email and so on wasn't compatible. I saw the .ca domain, and in particular the domain name system - the DNS - as an opportunity to bring some kind of, I'll call it a lowest common denominator, to all of this.At a similar time, I became interested in what were then the OSI protocols - the seven-layer stack - and in fact I was involved in the development of some email software. I helped to establish the first Internet link for western Canada; at about the same time, there were ones happening in Toronto and in Montreal, and in Ottawa the defence research establishment was doing something.So, that kind of sets the historical base from this. Where it's come from, then...hard to put this in words appropriately I guess. On the one hand, because I was running the .ca registry from the beginning and seeing the growth, even though the numbers were fairly small the first few years, you could tell within four or five years - maybe even less than that - if you plotted the line you could see it was curving upward, that there was the beginning of exponential growth. The Web was sort of invented in '89, didn't really hit the public until '91, didn't really hit the public's consciousness until '93 or '94. Before a lot of people had heard about it, you could see the exponential growth happening, even in the .ca domain, before then. You mentioned that BitNet, HEPNet, PhsyNet had no cohesive protocol or standard, but there seemed to be an appetite there for getting everybody rowing in the same direction. That must have taken quite a bit of people setting aside their personal ambitions...Yeah, there were these protocol wars that happened back then, so it wasn't all sweetness and light. It was a lot of fun to go through, because it was important and people passionately cared about these things, but this was also pre-commercial as well, as far as the Internet was concerned. One of the great things about it was you could actually make some changes without huge financial investments being involved. But it was the usual combination of goodwill and politics and back-room dealing and all those kinds of things, which to be honest with I wasn't involved with. I wasn't one of the power brokers here - I was really interested in the technology and in good things happening.The Internet's standards-making process, called the IETF , was highly successful, especially when you stack it up against the CCITT , the government/telco standardization process.Part of their success has been because they've latched onto a different way of doing things, but to be honest I think part of it is you have an opportunity when you're small to get stuff done. When things start to get large or when they have been in place for years and years, inertia starts to set in and potential difficult with making changes becomes more apparent. When URLs really started to become part of the public consciousness in the mid-90s, it seemed to be all .com at first - the Web was being driven by the massive commercial might of the US. Were you ever worried that the .ca domain wouldn't take off?I think...we were actually reasonably fortunate. Almost every top-level domain, especially the country-code top-level domains, has its own little quirks. There were very few that were cookie-cutters of each other, and .ca was no exception. We had more of a geopolitical structure to the namespace, rather than flat namespace. And then, the other organizations had third-level domain names, so it was something.co.uk - it wasn't something.uk directly.In Canada, the details were a little bit different. We had fairly restricted rules as to who could get a .ca domain-name and how many. Before there was a commercial network, having a maximum of one or two domain names for your organization, or some fixed number, wasn't a big deal because you weren't advertising products on the ‘Net anyway. So, if you were a commercial organization you wouldn't need a domain name for all three of your products, or for your slogans or your trademarks or whatever.So, the .ca domain name, for whatever reason, was actually not on the leading edge of the change that occurred in a lot of other domains. But then the commercial organizations started to run the registry , and about the same time, the rules for the Internet were evolving, because in the early days you couldn't use it for commercial purposes. At about a similar kind of time, the .com/.net/.org registry moved over to a commercial organization, and they started to, as part of this, the rules for these top-level domains started to change.And then, the Web started to enter the consciousness of people and started to be used as well. For .ca, we didn't start to think about changing the rules in a big way - there had been fairly low-level discussions going on, but it didn't really start to come to a head until 1997. By that time, some of the other registries top-level country codes and domains, like Germany and Holland, had started to modify their rules.Again, we were maybe a little behind the curve there. The bad thing is, we didn't have the same uptake in .ca registration as they had in .com domain registrations, even in Canada. The upside is, or some of the upsides are, that for the .ca domain, the main goal wasn't to register as many domain names as possible. The goal was to manage a public resource for the good of the country, even though it wasn't stated so boldly or succinctly in those kinds of terms. That was part of the vision that was there, even in the early days: CIRA has made that more explicit now.So, we actually had an opportunity to come to that kind of a realization. We haven't had the same problem with domain tasting and cyber-squatting and all the other kinds of things that have happened in some other top-level domains. Last summer, I think, there were dire warnings that we were running out of IP addresses, and I think they found a workaround for that to put a bit more capacity into the system. And, they've been at work on the Internet2 now for nearly a decade. Are we going to be constantly building on the ruins of an earlier Internet and trying to keep up?Yes, unless something is broken enough that you have to fix it, you will tend to build on something rather than try to reinvent it. And, the Internet...wasn't just something that happened by accident, a lot of hard work and design went into it. Nevertheless, successful as it is, these 30-plus years later, there are shortcomings. So, the community needs to figure out somehow, at what point do we decide to continue building on top of this thing, and at what point do you figure maybe we need to be thinking about a new infrastructure as well.You're right, it was something new, and Canada has CANet 1,2,3,4 - Canet and Canarie are deeply involved with that. A lot of that has been fundamentally new stuff, but a lot of it is still building on top of. Internet2 is really about using fibre and getting really high bandwidth, but it was still using TCP/IP. This clean-slate approach is about taking everything off the table, and saying if we could do all of this over again, how would we do this? I think it's really great to do that, the question though is: do you want the community to adopt this, and if you do, how do we get from where we are to there? And, can you do something that doesn't require a crisis to happen of some kind?They said we're running out of IP addresses at the latest ICANN meeting that just finished in Lisbon - the current prediction is by 2010, 2011. We have been managing it really carefully for the past 10 years or so, but even with this very careful management we're starting to run out.Is it going to take shortages of some kind before people will finally start to make changes? There's no question that if a community has the resolve to do something and a great sense of urgency, it can do it. The problem is, there isn't that urgency yet. What would the catalyst be, in the case of the Internet or the World Wide Web, do you think?Well, I think one of them you already identified - actually starting to run out of IPv4 addresses, the old-style IP addresses. If we do really start to run out of those things, then the value of those things is going to start to go up. There may be something to do with Internet domain names in the .com space. If we really do start to run out of .com domain names in a fundamental sense, is that going to trigger something? I think that's less likely, because there are a lot of other top-level domains out there.To me, you can view it as ‘are there going to be problems that will trigger something,' but another way to view it is ‘will there be new things that come along that will enable something?' To me, and I'm just speculating here, one of the things that may enable some new will be what I call lightpath-to-the-home, but some people call it fibre-to-the-home.I would think it would be really great to have cities or municipal governments manage a fibre infrastructure, just like they manage a power infrastructure for their city, or sewer infrastructure or something like that. Delivery of bits to the home, where you can choose the provider who's offering those services and then you can choose the content you get on top of that.I think the people who currently deliver those things now, the Teluses and the Bells and the Shaws and the Rogers, are going to view that as a threat. I think it's actually an opportunity for them, because they won't have to be saddled anymore with maintaining this physical infrastructure. They'll let the municipal government do that.If we can get essentially unlimited bandwidth to the home and to the business, I think that will open up vistas that we haven't even thought seriously about right now. It's boggling to imagine, but with the technologies that are happening at the same time - you've got flat-panel technology and roll-up display technology starting to happen, those kinds of things - you could imagine whole walls of your home that could be live. Would people go for that? I think they would leap for that kind of thing.We go, ‘Gee, what we've got here now is really great,' compared to what we had before. The first Internet connection to western Canada was a 19.2 kilobit line from UBC to the University of Washington, back in '88. I helped to set up that line, and it was fabulous back then. It's all a matter of perspective. People don't even think about what 19.2 kilobits is anymore. It's up in the megabits or gigabits or whatever. But can you imagine if it was up in the terabits, or tens of terabits, what kinds of things that would enable? It's ubiquitous and everywhere now, and in some sense it's all been done. At another level, we're just getting started. What do you think of the whole ‘Net neutrality debate? I know some people who have been around as long as you have are uncomfortable with the idea of access providers being able to say ‘You'll get our information, our content, or that of people we have commercial relationships with, much faster than the other stuff'...If we are going to talk about ‘Net neutrality intelligently, I think we need to come to a common understanding of what it is. I don't think we've come there yet. I think once you start to have a discussion about ‘Net neutrality even among technically competent people or among people who are in the know about this stuff, you'll still get a paragraph definition of what it is, and then you won't get consensus on that.Having said all of that, I think there is the beginnings of a consensus around what the issues are - I don't think it's a 25-words-or-less issue - and to me one of the things that I think would be helpful would be to have as much choice as possible. I think part of that would be delivery of lightpaths, or fibre, to the home, and lots of bandwidth upstream and so on, and to figure out the right place for different companies to play there.Where that sifts out as far as the CRTC is concerned, or whatever it's going to be called down the road - I don't know what Bernier's plans are - if things are too vertically integrated, then you can potentially get issues. A vertically organized industry will tend to try to keep you in the silos that have been defined.I think it would be helpful to try to figure out where a reasonable place to chop this up is, so that you have people who are providing content, and people who are providing the infrastructure from the home to the collection point - and maybe government is a good organization to help with that, or contract it out - and then you have other organizations that provide backbone-type service, and they don't necessarily have to be all the same organization. In fact, if they are all the same organization you'll eventually start to get into issues.If there is choice though, and I mean real choice, eventually I think some of these issues with ‘Net neutrality will start to go away. One of the issues with Canada though is that, relative to the US and Europe and so on, it's not that big. So, people will keep coming up with these economies of scale, which is why we do need to have more control over these kinds of things and more regulation. It could be interesting to see how it plays out. Again, I'm not the political kind of guy, but my gut tells me that offering as much choice as possible is going to be part of the solution here.

Q&A with .ca ‘godfather’ John Demco

Broadcast | May 18, 2007

Unlock all the Canadian telecom, broadcasting and digital media news you need.

FREE TRIAL

SUBSCRIBE

Inside The Wire Report

Information

Hill Times Publishing

Get in Touch With Us