We Have Connectivity!
We have an infrastructure that is fully capable of giving everyone always-on connectivity but the FCC and the carriers acting as if it is still a phone and cable TV network. We must recalibrate and focus on removing the impediments of connectivity instead of just mediating feuds that no longer server any larger purpose.24-Jun-2003

Preface

My message is very simple. It may seem radical but I am doing nothing more than pointing out what is rapidly becoming obvious as I read stories entitled "Make your PC a Home Entertainment Center", use my VoIP phone from Vonage to call around the world without using the phone network and find myself simply borrowing any connectivity I find lying around.

Instead of trying to force the carriers to provide connectivity, we need to recognize that the carriers are deploying connectivity for their own needs. In fact, the telecommunications system is Internet-ready because of our ability to take advantage of existing infrastructure using a combination of new technologies and the opportunistic nature of the Internet itself.

Our policies must be aligned with this reality instead of trying to continue to micromanage competition in a pre-Internet world and preempting the marketplace by mandating "broadband" as a static objective while presuming that the business of telecommunications will persist. Our policies seem to be based on a fundamental misunderstanding of the nature of the Internet. The Internet is not defined by fibers or wires or even not-wires (wireless). The Internet simply happens when we understand how to define the services outside the network itself by using any available transport as a way to exchange fungible packets of bits. Each packet is just like any other and, in itself has no meaning. We don't even ask that the infrastructure guarantee that the packets will be delivered.

We simply take advantage of any opportunities. Because we can use the most available and least expensive connectivity the costs are low. It doesn't matter how we add capacity as long as it becomes available. While it may be expensive to lay down one fiber, additional fibers cost very little more and once the fiber is available we've been able to greatly increase the capacity of the existing fibers by upgrading the electronics at the end points. We can do the same for other mediums such as the copper already in the ground and any radio transports that are available. A packet can use any combination of transports available.

One can take advantage of some of the benefits of the Internet without understanding how it works. Indeed, the carriers now find themselves enthusiastic consumers of the Internet as a way of delivering their own products. To Congress the Internet seems to be one more goodie that people demand more of while being seemingly unaware of the conflict between encouraging Internet availability and mandates like HDTV which are now counter-productive.

I can sympathize with the difficulty of understanding how the Internet can make no promises yet do work far better and more efficiently than the current telecommunications industry while allowing for services that weren't available before. Yet it's the same dynamic that has yielded such rapid growth in the personal computing industry.

For the carriers' stakeholders ignorance comes at a very steep price. The Internet Packets travel over whatever route is available and attempting to contain them simply for the purpose of creating billable entities works no better than a toll booth in the middle of a parking lot. The carriers cannot afford to have their content or services limited to the pipes they own and they cannot afford to subsidize the pipes out of other revenues.

Despite their progress the carriers seem to still be in denial and assume that the users need them to create services despite the availability of alternatives and articles on how to do it yourself. The reality is that we have connectivity now and the FCC should focus on removing impediments (such as restrictions on municipal ownership of transports). The plight of the carriers is reminiscent of the Savings and Loan failures in the early 1980s though. The sooner we face up to this the sooner we will be able to get the benefit of the connectivity we already have.

It is Simple!

A sculptor can see a beautiful statue where others see only a big stone. We need to see the connectivity locked within the copper and glass of the telecommunications infrastructure.

The first step is understanding the concepts. In order to keep this essay short I'm assuming that the reader has a general understanding of the Internet and telecommunications,

The last time I faced off with the carriers over connectivity was when I was at Microsoft. I came to Microsoft in 1993 with over 25 years of experience in computer networking going back to the 1960's. The carriers were staking their future on a delivery platform called "Residential Gateway". In contrast I had a simple goal of making it easier to network computers within the home and for that network to interconnect with the rest of the Internet. As it turned out it wasn't even a fair fight since the Residential Gateway/Interactive TV trials were public failures and home networking is now the norm and is sold at the corner store. The carriers' problem was that their approach amounted to an all-or-nothing bet that the users would tell them exactly what they wanted and then would buy it as a bundle. By providing (Internet) connectivity and not saying how it should be used, I avoided that and simply let the marketplace discover all sorts of interesting applications--at least until they had to deal with connectivity via those carriers -- the telephone and cable TV companies.

The carriers have now staked their future on building out an IP infrastructure. The good news is that this connectivity can easily be extended through the first (or last) mile of connectivity via the existing copper, fiber and wireless transports by adding some simple electronics. Current DSL and Cable Modems are still first generation technologies. The same kind of rapid development we've seen in computing has given us the ability to run DSL at much higher speeds over the existing wires and at greater distances at lower costs. Cable services such as Comcast's "On Demand" pay per view acts like a remote TiVo hosted at the head-end connected via reserved per-subscriber channels demonstrates that we are not limited to using the cable as a broadcast medium.

Since we only need a few hundred kbps (thousand bits per second rather than a million bits) to participant in the richness of the Internet, it should be very easy to provide always on connectivity everywhere. There is an expense to such a build-out but the cost savings from simply eliminating the telephone switching gear which isn't even a concept in Voice over IP (VoIP) would more than offset it. This may not be obvious because many people think of VoIP as PSTN (Public Switched Telephone Network) emulation and presume there is a cost. But the only cost is associated with bridging to the existing PSTN. VoIP calls between Internet end-points and don't use any PSTN gear at all and don't require any service provider in the middle.

The carriers' own analysis argues that the economic value is enough to justify building a completely new "broadband" infrastructure. Upgrading the existing infrastructure costs only a small fraction of that and avoids the all-or-nothing bet inherent in having to guess what people want. Providing connectivity over the existing infrastructure is a far more prudent business choice. Demand will fund further build-outs as needed. In the US we already see 30% (see Nielsen-Ratings) of users (growing rapidly) opting for always-on connectivity with speed being a secondary issue.

What's wrong with this picture? Nothing!

A Culture of Complexity

The problem is that the carriers and the FCC have a different picture in mind. It is a picture that is far more complicated and problematic. The simple picture is viewed with suspicion because the existence of a complex regulatory system, the Regulatorium, seems to prove that there must be something complicated otherwise why would you have such a complicated system? It reminds me of a time when I started hyperventilating and found it hard to accept that the cure was to just stop. The hyperventilation itself made me feel as if I couldn't breath -- it was a closed system that perpetuated itself.

The simple picture cannot emerge as long as the focus is on isolated problems and the problems are exacerbated when they are viewed in isolation. The micromanagement of DSL and policies to enforce line sharing are necessary as long as the carriers' incentive is to limit connectivity. But such management perpetuates the existence of analog telephony as an expensive service and makes it hard to upgrade the infrastructure.

The carriers themselves were born of the same culture and engage with the FCC on a common ground as they tweak the rules rather than recognizing that they are fighting over spoils of past glories while the world is passing them by.

The Escape from Complexity

In 1995, I looked at the tangled mess of modems, ISDN, online services, corporate networking and wondered what happened to the simple networking concepts I learned about decades before and I set out to find it. It was fortunate that I was working at Microsoft which meant I had both the lever of Microsoft's position and the fulcrum of the Windows platform. David Reed, who co-authored the seminal paper which laid out the end-to-end principle, helped me see the inherent simplicity of the Internet Protocol. I was able to take advantage of my opportunity to address the problem of connectivity as a whole: I made sure that Windows systems (whether at home or on the road) could connect with each other and share a common connection with the rest of the network without special setup, I coined the term No New Wires Networking and developed phone wire networking as an economical example and made sure that the elements were independent so that the elements were synergistic and open. The timing was also fortunate because DSL and Cable Modems were just becoming available at promised speeds of 10 megabits per second.

Though having a home network that is part of the Internet as a whole seems almost normal these days, it would've been considered the least likely outcome then. As far as I know I was the only one trying to make networking a consumer product and the only one who happened to be in the right place and the right time to make it happen. We were supposed to be starting at our Interactive Televisions and the carriers were supposed to provide the "Residential Gateway" for delivery of content and as a platform that other companies could use (at a fee) to deliver their services such as metering and managing the use of electricity. This "future" can still be seen in the set top box which is bedeviling us with its quirks and limitations.

When I left Microsoft in 1998 I wondered if I had succeeded in my goal and was pleasantly surprised to start to see today's inexpensive routers and other components appear in the marketplace. Companies like LinkSys, Netgear, D-Link and others got the message and realized that the product was IP connectivity. Companies that tried to sell added value networks failed not just because of price but because they tried to define what the services were rather than letting the users themselves discover what worked for them and what didn't.

The Illusion of Victory

For Cable TV companies looking for additional revenue, the Cable Modem was a great opportunity. I was fortunate in that I still lived in Massachusetts while working in Redmond and had the opportunity to be one of the first cable modem customers when Continental Cablevision started to offer the service. Working at home was important since my commute was almost 3000 miles. A decade earlier the telephone companies were worried about competition from Cable companies and responded by "discovering" a way to send video over normal phone wires. This was ADSL and with ITV in the ever-distant future, they repurposed this technology as DSL for connecting to the Internet.

Despite this enthusiasm it quickly became obvious that the Internet was a threat as it kept growing and growing in capacity and the idea that the Internet would subsume the phone system was becoming accepted wisdom, though the question of when was left open. The carriers also talked about a future in which they would provide a hundred megabits to each home. They never said exactly when and they still keep trying to figure out what the "product" is that they are supposed to deliver because they still think in terms of television rather than the Internet which is defined by the users.

It's now 2003 and the carriers seem to be competing with the Internet as if it were another business. The FCC is playing its normal role as mediator in setting the rules for competition. The Internet is now part of their product mix as if it were another television channel or a high speed phone service. The problem is that the Internet is not another business. It is a concept: the users define the value at the edges of the network and that the network itself is just a commodity transport.

The Phantom Escapes

The Internet is based on the simple idea that the meaning of the network is defined at the edges and the transport is a very simple commodity that cannot be contained. The carrier cannot distinguish between a single computer and a shared network. Attempts to "improve" the network actually work against the users by biasing towards the carriers presumed use. Even if they are right at a given moment, the dynamics of discover will create usage patterns at odds with the favored uses.

The implication is simple and stark -- a telecommunications industry based on containing the pipes so they can be the sole provider of the service can no longer control the transport -- the communications "pipe". If the pipe is a commodity then they have very strong competitive pressure that limits how much they can charge for the service, especially when the services can be built by the users themselves at a one-time cost for software.

To recover the cost of the pipes themselves they would have to contain the traffic flows. This would be like trying to bill for the grass each horse eats by fencing it in rather than taking advantage of being able to share rangeland. Even new startups trying to sell hotspot services are going to find that creating billable entities doesn't make sense. Instead the connectivity will be a courtesy service like a water fountain or a municipal service.

We already see buildings and whole cities deploying fiber and buying Internet connectivity in bulk. The cost is low and can be part of the standard bundle of community services just like the roads and trash collection. It's important to distinguish between the connectivity itself and services. You might still have to pay HBO or a sports league to view their products just like you have to pay for clothes you buy at a web site. A coffee shop will need Internet connectivity for its own use and, like the water it provides as a courtesy to customers, they will share their connectivity though it's like to already be available from the city itself.

Seizing the Present

We have an infrastructure that is Internet-ready and we have carriers that must quickly come to terms with this reality. They may delay it but delay just increases their debts while denying us the benefits of connectivity.

It is necessary for the FCC to play a leadership role and evaluate its policies against the goal of connectivity. While I realize that the FCC is bound by a complex matrix of regulations and Congressional oversight, I expect it is the carriers themselves that will be the most vocal advocates of such a policy.

The carriers' stakeholders will become advocates once it is very clear that they have the most to lose by not coming to terms with how the concepts that define the Internet require the companies to reinvent themselves in order to pay back the massive amounts of money they have borrowed to build an asset they cannot control. The carriers' who say they are the "broadband" companies are using the term for a bundle of services of which Internet connectivity is just another channel. The customers see it as just the Internet and may or may not choose to buy the television service. Voice over IP is already a simple application and as capacity increases so will TV will be just another video stream. Without the predetermined limits associated with today's TV distribution system, HDTV will be just another format except it will likely be higher resolution to take advantage of the capabilities of today's the screens we already use on PCs.

The carriers' "Broadband" is just an update to the Residential Gateway vision but unlike the modest failures of the residential gateway, these companies are spending billions of dollars of their shareholder's money as if there was value to be added.

This is where leadership is most vital. It is very hard for those who have learned about telecommunications within the current framework to step back and rethink the basic premises even when our understanding has moved far ahead. We've managed to continue to use the simplistic model of the tuning fork to define a property model of spectrum even though we've understood spread spectrum and other technologies since World War II. But the Internet is different. The vast regulatory system that prevents people from owning transmitters is unable to contain the Internet. After all, the carriers themselves are offering Internet connectivity and the only protection they have is the simple lack of awareness that services like Voice over IP already work or that the ability to get remote video on demand means that they can use the same pathways to and from the home to bypass the carriers and get real choice -- instead of the carriers' limited palette, they can access any video sources ranging from professional studios to sharing an evening with their friends.

If the carrot of bailing out the stakeholders isn't enough, there's the stick. There is a real cost to the carriers intransigence and claims about the problems with simple things like number portability fly in the face of the ITU's own evaluation of using the Internet's DNS for that purpose. Such tactics serve the short term goal of creating the illusion that the carriers are financially sound while hiding the real threats from their shareholders. How does this differ from WorldCom and Enron?

The FCC must confront such claims head-on and require transparency without the cover of accounting legerdemain. The FCC must also confront laws that make it illegal for municipalities to build their own connectivity infrastructure and deny them rights on the paths seized by eminent domain.

Commencement

I have written about concept of connectivity and I will continue to try to understand how I can better explain these concepts. It is a process of educating because the concepts do defy common sense--or at least the common sense of the elders. There's a generation that is growing assuming connectivity. After all, college students are very well connected and don't understand why, for example, a credit card machine makes a phone call for authorization when it should just exchange messages with a server over the Internet.

We can begin to find our way out by recognizing that the Regulatorium is a creature of its own devise and start rebuilding our institutions to match what we understand in 2003 and not the bygone world of 1934.