Commentary: what does telecom have to do with the Internet anyway?

Putting so much effort into issues like Network Neutrality and broadband deployment diverts us from recognizing that the Internet has nothing to do with telecom except insofar as we let telecom
control our connectivity. The entire telecom industry is just another application like the web, but we give it too much power over us. In trying to write about these issues I realized that the metaphor of first or even last mile misses the point. That’s the way you talk about accessing something far away.Words have a way of reflecting and reinforcing our mental models. We think of telecommunications in terms of content being delivered (as with TV) so we often hear about the “last mile” or even the “first mile”. We should think about connectivity within our neighborhoods — the first square mile to contrast with the first mile traveled.

As Andy Lippman observed — networking is something we do, not a service we buy and the Internet is not a destination but an idea. We use the physical infrastructure — copper, fiber and radios (or, collectively, wires) to exchange packets just like we use local roads to visit friends. We don’t think of local streets as merely onramps to the nearest highway.

Because we own the wires in our homes, we can add another printer or computer without having to ask permission or pay a fee. This may be obvious but in the early days of home networking, service providers expected to be paid for each device in our homes! Instead of buying “networking” we use a copper wire and each device takes its turn sending messages. The rate has gone from 10 megabits per second ten years ago to a billion bits per second today. The speed for wireless (Wi-Fi) has also improved though not quite as fast because they are subject to the limitations imposed by the FCC.

We can interconnect the networks in our neighborhood in the same way we use routers and switches to extend the networks within our homes. By using bundles of copper or fiber we can take advantage of the large capacity to exchanges messages at gigabit speeds. By interconnecting the wires just like we interconnect the roads the network continues to operate even if there are failures. Without the need to bill for each packet, we can open up the access points to provide full wireless coverage (while still shielding the networks within our homes).

The high capacity means we don’t need special broadband connections and the reliability means that we can rely on it for municipal communications including emergency services. This should result in a net savings. Saving money is important but the real value comes from having local control. Why do we need to go to the Internet “out there” just to find out that the schools are closed due to snow? Without having to pay a toll for each message, we can connect homes and schools. We are free to bring the school to students with the ability to watch lectures from home. Instead of old style community access television, we can have real video services with live feeds as well as archived video. Without the need to limit access so that users can be billed we’re free to take advantage of full wireless coverage — your medical bracelet can call for help without having to establish a billing relationship for each device in each town or office.

At the heart of the power of the Internet is simplicity. With traditional consumer electronics, you have to make sure each wire is plugged into just the right connector and video wires are different from speaker wires and phone wires. There are many kinds of video wires and you have to make sure you install them just right. It’s far simpler to just plug everything into a common network. This hasn’t always been easy but because the solutions are defined in software we can continue to improve the experience over time using the existing transport. We can each find our own solutions rather than wait for a service provider to deliver solutions.

More insidious is the assumption that we are all a market for the provider who will meet our common needs but not any of our individual needs. There is a vast difference between serving us all on the average and giving everyone the ability to find what best meets their needs. On the average no one is average — the Internet gives each of us opportunity to shine and contribute. Telecom companies are beholden to their shareholders. We mustn’t confuse serving their customers with meeting each of their needs; they do so to maximize overall sales rather than trying to serve each individual.

For the telecom industry, network complexity represents a barrier to entry and makes us dependent upon their expertise. If the marketplace demands simplicity they can deliver it. This is why we have dials on phones. In early phone systems, you had to ask an operator to place a phone call and it became more complex as you tried to reach people further away and almost impossible to reach people in different countries. Today you just dial a number because that allowed the phone companies to offload that complex task onto the customers.

Operating digital packet networks is even simpler: you don’t even need to dial a number because it allows you to describe what you want instead of remember long strings of numbers. Today’s computers devote much of the capacity to improve usability. The same is true of networks. The high capacity the network goes towards making them function smoothly without needing highly trained individuals whose mistakes have caused so many problems in the past. The biggest source of complexity may negotiate the complex relationships among the service providers as they seek to preserve the scarcity that creates value.

We will see the same simplicity as we interconnect neighborhood networks within a city and the between cities and nations. Getting a packet from City A to City B isn’t much different than getting a packet from A to B because we can take advantage of techniques like backhauling which use dedicated paths to carry packets across long distances without complexity of going via many routers.

For example, when I worked at Microsoft I had an apartment just off of I-90 in Bellevue, Washington and a home in Newton, Massachusetts at the other end of I-90. All I would need to know is that I get on I-90 at one end and get off at the other. This may be an oversimplification for driving but this is the way we can and do networking.

Admittedly I’m being a bit idealistic as I describe these scenarios because today’s software and protocols have worked out all the kinks. This is similar to the situation I faced advocating home networking in 1995 when only professionals could install and operate a network. Because the networking itself is done in software, we were able to remove the impediments and give users the ability to install and operate their own networks. And then we could address other problems like finding network printers. The same dynamic that has given us faster network is giving us simpler networks.

The dynamic works as long as our incentives are aligned. As owners we want to improve the network even if we aren’t sure of what is important. Wireless coverage may be more important than speed and efforts to extended wireless coverage result in a better network by that measure.

The problem with today’s telecommunications industry is that it is a service industry in which the providers’ incentive is to increase their profit by selling us more services. “Internet” (without “The”) is a recent addition to the product mix. The more Internet access they provide the less valuable the services are because we can create our own solutions. The other problem with the Internet is that bits are simply bits and they can take any path.

I compare trying to make money by selling bits with trying to operate a canal across an ocean. In other words, the carriers must limit us to using narrow paths across the sea of bits. This is why we are laying fiber along all our highways but only a tiny fraction of fiber is actually lit up. And even then only a small portion of the potential capacity is available.

We’re trapped in the Regulatorium, that is, the FCC’s regulatory system that was established during the great depression in 1934 when the marketplace was not to be trusted. Changing legislation requires a political consensus but you can’t get that consensus until you have an agreed upon alternative. This is difficult when we lack examples and when the very premises that define the Regulatorium are threatened by the idea that networking is something we do ourselves. It’s like asking the railroad regulators to tolerate unregulated car driving.

The inconsistencies of the Regulatorium provide opportunity for change. We got control of the wires in our homes as a byproduct of the breakup of ATT in the early 1980′s. The Internet’s end-to-end principle itself was a solution to the pragmatic problem of creating solutions when you don’t have control of the network itself. In 1995 the service providers expected to extend their service model to make the STB the center of the digital home and their wires would also be used by other utilities for billing and managing their services. The phone and cable companies had the same idea. The phone companies developed DSL for delivering video and other services while the cables would offer what we now know as the triple play.

Home networks are very simple — they just carry IP packets and don’t know anything about the applications. But this little leak on the side of broadband was enough to set back the triple play plans for ten years and scale them back so that only traditional telecom services are defined and even then the content providers such as broadcasters are increasingly shifting to bypassing the service providers gates by going direct using downloading and other techniques.

We can wait for this dynamic to play out as the service providers continue to lose control of their transport and service revenues decline. The billions we spend on redundant broadband deployments are trivial compared with the cost of the lost opportunities in both money and quality of life. The good news is that there is an increasing understanding of the distinction between infrastructure and services.

There is going to be continued tension between the Regulatorium and other laws governing the use of common facilities. One tactic is to ban local government funding of connectivity infrastructure because it represents unfair competition for the service providers. Such laws recall attempts to limit the speed of cars to the pace of a horse. More complicated are the laws governing the rights of way which cede control to utilities and service providers. While the idea of a city being banned from using its own rights of way is indeed offensive, things get more complicated when facilities pass through multiple communities.

I may be a cynic but I believe that, as whole, legislatures are trying to be helpful but they are acting on the premise that telecom is indeed a service industry and more broadband means more Internet. Ideally we’ll soon be able to point to examples of effective infrastructure. But we also need to start taking control of the framing of the discussion.

This is why I’m so concerned about using the word broadband and the ideas of first and last mile. A phrase like First Square Mile can highlight the difference and let us follow up by creating infrastructure for our neighborhood. We need to emphasize that this has little to do with telecom. This is why I coined the term Regulatorium — if we can name it we can talk about it and then we can talk outside the Regulatorium.

The value of a connected infrastructure should be obvious, but the value comes from the opportunity it creates. Until that opportunity is realized people can’t imagine the benefit any more than they could have foreseen the benefits of the Web. Even when I was advocating home networking while at Microsoft, I found it difficult to get people to understand what it was, let alone appreciate its importance.

Examples do speak louder than words and the real value in explaining the concept of neighborhood connectivity is to reach those who are in the position to implement local connectivity. These examples are sticky in that it will be hard to go back to dependency once you’ve gotten the benefits of connectivity.

Today’s underserved rural communities may provide the test beds we need. They may not have “broadband” but they do have phone service and those copper wires have a high carrying capacity if you use the right electronics. Today those wires are not available because the FCC Universal Service Fund (USF) collects billions by adding a fee for legacy phone service and then uses the money to assure that the wires are used for phone service. I should say “wasted” since that can leave each wire running at one millionth of its potential capacity.

If the community had real ownership coupled with honest and transparent funding it could use those wires to jumpstart neighborhood connectivity. While traditional DSL service is fairly slow we can use back-to-back DSL units to extend the reach and new technologies to run each wire at 100 Mbps or more.

State regulators and commissioners have an opportunity to play a leadership role recognizing that their mission has changed. They can and must serve their community rather than presuming that anything good for the service providers is good for the community. It isn’t true because the telecom model serves the mythical average and not any of us. With analog signaling we may have had to subsume our individual needs to the restrictions of the technology, but digital technology frees us from having to have a special wire for each purpose. Bits are just bits.

As long as we think of networking in terms of being at the last mile of a service delivery pipe, we will have to settle for what happens to arrive. If we look at the first square mile around us – our neighborhood – we will get the opportunity to be participants who can meeting their own needs while also contributing to the common good.

- – - – About the author – - – -

Bob Frankston has been working with computers since 1963. He graduated from MIT (undergrad) in 1970 and continued in graduate school. He worked on the Multics projects and used the predecessor of the Internet beginning in 1969. Commercially he has supported online services since 1966. In 1979 he went from the mainframe world to the PC industry and co-founded Software Arts with Dan Bricklin where he implemented his concept of VisiCalc. He was with Lotus Development from 1986 to 1990 where he created Lotus Express (and started Lotus.com though it was before the Web). At Microsoft from 1993 to 1998 he championed “IP Everywhere” with phone wire networking being one result.

Bob is now on his own pursuing a number of projects among them, trying to explain the larger concepts of IP everywhere.