Beyond Telecom: Bob Frankston on the future we make for ourselves

It would be worse than a cliche to say Bob Frankson breaks any mold. More than anybody I know, Bob is uncontainable. For Bob there was no mold, and never will be.

Bob isn't even comfortable with the flattering descriptions history has laid on him. Widely acknowledged as the co-inventor (with his friend Dan Bricklin) of the electronic spreadsheet, Bob qualifies his contributions there, and proudly points instead to his overlooked work on home networking at Microsoft, where his energies were poorly supressed during his employ there in the 1990s.

Bob is also boundlessly expressive. My email archive contains over four thousand messages from Bob, and that's not counting thousands more to various lists on which we both participate. I doubt thre is anyone talking and writing about telecom that rivals Bob in volume, originality or provocation.

For years I've wanted -- along with other friends of Bob -- to put a refinery atop the gusher of Bob's endlessly interesting output. But there's too damn much, and it's not crude. Even in casual talk, Bob drops one liners that are gone like snow on the water if nobody writes them down. Such as, "There's a fine line between reciprocity and corruption ... if there is a line at all."

But it's hard to refine Bob's prose. If you attempt to improve it, he often resists and corrects. In that respect he resembles fellow MIT alum Richard M. Stallman. But Bob's reasons are different. RMS has a canon that lithified decades ago. Though his convictions are strong, Bob's ideas are fluid and often very new. The best you can do is try to keep up with it, and to catch a few cups from the gush.

That's what we're doing here. The interview that follows is transcribed and edited down from interviews with Bob in early February of this year. It wasn't an easy job (for either of us), but I think it's worth the effort.

Because what Bob says about telecom, the Internet -- and why we need to move beyond both -- may have the same founding importance as what RMS and Linus have also brought to the world.

- Doc Searls

DS: Lately you've started talking about The Opportunity Dynamic. Can you unpack that for us?

BF: Value comes from taking advantage of opportunity rather than limiting yourself to solving problems as stated. An opportunity gives you a new set of solutions and there are many problems waiting to be solved – all you need do is find a problem which you have, in effect, already solved using the solutions you already have. This is a recurring theme -- why solve hard problems when there are easy problems going begging and likely to be more valuable? This may seem counter-intuitive but remember that there are so many problems that the one you’re laboring to solve is probably in the middle of the pack.

It's like if all you have is a screwdriver, you’ll find there are a lot of screws -- which makes the screwdriver very useful. The best approach comes from taking advantage opportunities that match the tools you already have. Of course, sometimes you get stuck and need to solve a problem and then things get hard. We forget that those are the exceptions and not the rule.

DS: Let's go deeper into the dynamic as you're now describing it, by taking the example of frame construction. Specifically, Augustine Taylor in 1833, building St. Mary's church in Chicago. Cut nails had been invented in the late 1700s, and were cheap by then. Lumber was essentially limitless in the U.S. Lumber milling had also been invented. Wood could be milled into boards from logs, but boards were then thought more useful for finishing than for structure. For structure the defaults were still mortice and tenon, notching and securing with pegs, or working with masonry and rock. Taylor replaced those with 2x4s, 2x6s, and nails, building a frame in a way similar to what we still do today. He did it cheap and with DIY carpenters. It caught on. Suddenly one could frame anything with studs, joists, and cross members held together with nails. Old time builders called it balloon construction, because they thought it would blow away. But it didn't. Instead it revolutionized construction by letting anybody build anything cheap. You want to build Tudor, or Prairie, or an office or a cabin, you'd frame it up. As a result construction is perhaps the largest industry in the world. And nobody "owns" it.

I go into that detail because I think you're looking for a similar development, undermining telecom as we know it.

BF: Right. So let's look at how you do that.

First, you can't see the opportunity unless you de-couple interpretation. Instead of saying "I need a Tudor house", you say "I need a board". Now you've de-coupled. That way the board people can focus on doing the best job making boards. And the builder can concentrate on making the best house, whatever the style, or the other materials. And many kinds of manufacturers can concentrate on what they do best. And the market opens up.

The old way with businesses like telecom was (and still is) top down, and highly coupled. It's like deciding what the protocol is before you do anything else. All other work stops. HTML succeeded because it was good enough, and easy to adopt. If we insist and getting things right the first time, or before we start, we don’t understand the problem -- unless it's just more of the same.

Second, there is a subtle but important point here about standards, and standardization. I caught a show on the History Channel recently about the Spanish Armada. One big problem for the Spanish, the show said, was that they didn’t have a standard cannon. So the simple problem of getting the right cannon balls to the right ship was overwhelming. It might not be fair to expect the Spanish to have gotten it right the first time but they didn’t allow for a learning cycle. The British were able to build on their experience and were able to commit to a cannon suitable for naval warfare. That commitment was like the early construction industry committing to 2x4s, studs, joists and framing.

But let's go back to the sea. Americans might remember 1492 as the year a foolish sailor miscalculated the size of the Earth and headed off to China. He was lucky to find land in the middle of the sea and later others benefited from his dumb luck. It was no coincidence that was the year that Spain had finally triumphed over heresy. As David Landes noted in The Wealth and Poverty of Nations, over the next few hundred years all the gold from the Americas went through Spain but none stuck. Heresy represents the new ideas and opportunities that they had banished from the kingdom. Countries that have managed to survive turmoil have been able to get the benefits of opportunity.

The best standards arise out of experience and are sufficient rather than limiting. Who cares whether a 2x4 is exactly two by four? What matters is that you can nail boards together. Too often we tend to see standards as way to enforce arbitrary rules. Ideally they create opportunity by decoupling systems – the board supplier shouldn’t limit what I do with the board and the user can take a fresh look and not just use it for building frames. The board becomes more valuable as we discover new users. This is one example of demand creating supply.

To take a current (no pun intended) example: the USB power supply. Because power could be supplied through USB, it became widely used as a power supply. All other attempts to standardize power supplies hanging off PCs have failed, but USB was already available because it had other established uses. What mattered is that you didn’t need to wait to solve the problem. USB is not an ideal external power solution but does the job and that’s what matters. It lets us move on to other problems.

DS: So, what are the roles of the IEEE, the IETF, the ITU and the rest? What is it that allows the innovation to occur without being constrained by official bodies?

BF: Let's look at IEEE vs. IETF vs. ICANN and the ITU.

IETF says, "Here's a specification. Comply or don't”. If you use it then you help grow the community. If you don’t then fine – maybe your ideas will win out or coexist. Just try to not to create unnecessary problems for existing standards. The entire set of Internet standards arose just this way – people trying to do something different or differently.

DS: Not surprisingly, SMTP and other simple and flexible email protocols are what got adopted.

BF: Yes. Now contrast that with X.400, from the ITU. X.400 was the great intergalactic mail standard. One hundred percent of the Telcos were behind it.

DS: So were the media, and the VCs. It had buzz out the wazoo.

BF: Yes. But it was complex and rigid. Most important is that it was tied to the rigid authority hierarchy of the carriers -- the post, telephone and the rest. It had a lot of good ideas but you couldn’t take them out of context – you were supposed to run X.400 over X.25, not over IP. The biggest problem was any change took a decade.

Compare that with SMTP (Simple Mail Transfer Protocol). It could be implemented in a weekend and re-implemented next weekend. The implementations are accommodating so that you don’t need grand “bake-offs”. With a ratio of a weekend vs a decade per iteration, X.400 was a disaster. Fortunately Al Gore helped the government back off its commitment.

Alas, today SMTP seems to have become an institution itself. The problem is not SMTP but the way we think of email but that’s another topic in its own right: a reminder that we are still stuck with the Internet demo, and fail to see the Internet that can be.

DS: Let's get back to that. Meanwhile, back to standards bodies. What about ICANN?

BF: ICANN says "Do it this way, or you’ll be banished from our Internet." You have no choice, no flexibility. It doesn't support an opportunity dynamic. The business it opens up is minimal. ICANN is trying valiantly to keep the demo alive. Too bad the unintended consequence is to thwart efforts to learn and move on.

DS: IEEE?

BF: The standards are very byzantine because they arise from the world of analog electronics in which signals got easily lost so you had to be very careful. Digital electronics decouples the accidental problems of signaling from our solutions

Look at faux Digital TV standards, which is what I call "fDTV". A dozen years ago the MIT Media Lab pioneered digital HDTV (D=Definition). So-called DTV (D=Digital) takes the idea of HDTV but couples it to all these other rigid standards. It’s still analog thinking: one-way broadcast -- as if the Internet didn’t exist. Yet we still get dazzled with acronyms like QAM instead of asking why it should matter what the transport is. It really doesn't.

DS: So we're talking about telecom here. What works and what doesn't. Like standardization vs. Standards.

BF: Yes. Standardization should arise out of what you do. The process isn’t magic – we need leadership but mustn’t confuse leadership with absolute authority. And that's what we have with what I call the Regulatorium, which is comprised of the FCC and the industry it's in bed with. We have to de-couple those as well. In fact, my goal is to get rid of the FCC entirely.

DS: Do we have models for that?

BF: We have the example of decoupling hardware from software. The smart hardware companies didn’t try to tell people what to do and learned from their users. If they built a Windows machine but people ran Linux -- or anything -- on it, then, well, go with it. The good news is that this gave us the benefits of standardized hardware. Of course there is a price when one approach becomes too dominant – it limits innovation. Fortunately there are many outlets for creativity, especially in software.

DS: I had a good time talking at CES to VIA people -- both people the company and its customers -- about all the ways their mini, micro, pico and femto ITX boards are used. They're agnostic about it, and their stuff is everywhere.

BF: On the other hand with Telcos there's no equivalent. Because they do it all themselves. And, alas, they like expensive hardware because it makes them seem special. They insist on communicating on your behalf. Imagine if we had only railroads and couldn’t drive. The railroad companies could (and did) set arbitrary prices. You could go anywhere you wanted as long as it was to places that were profitable for them. After all, they are there to serve their shareholders, and it’s not your network anyway.

We tend not to see how limiting the telecom view is, because we accept their framing. Sometimes we can move behond it a bit. Cell phones are one example. Remember when cell phones were supposed to be expensive business instruments? Some people thought I was being frivolous by using it to coordinate with my wife and kids. But the botrtom line with themis that they'd rather sell what people already know – it’s very hard to explain the value of what they don't yet know.

DS: Explain to who?

BF: Exactly. Why must everyone have to justify new ideas to a telephone company?

DS: You take a dim view of what a telco is and what it can do.

BF: Yes. The only real cost of a telephone company is billing. The real cost of copper, glass and radios is far less than something as mundane as sidewalks. Imagine if sidewalks were a service. Well, you don't have to imagine it. That's exactly the model you pay for every month with your phone bill.

DS: So we need to de-couple ourselves from the frame that copper, fiber and radios -- what you call CFR -- naturally should belong only to carriers.

BF: Yes. And de-couple your thinking from finding solutions. Because if you're not trying to find solutions, you can change the problem you're solving. If it's your copper, fiber and radios -- or anybody's -- you have real freedom to solve any problem you like. You can put the opportunity dynamic to work.

DS: We've seen this proven by the Internet, which is was not created by telecom, even though we took advantage of their copper and circuits.

BF: Yes, but we've also stopped a lot of forward progress by not realizing that the Internet is just a prototype. A demo. In my case, a class project.

DS: If it's a demo, what's it demonstrating?

BF: Flexibility. Which is anathema to the old businesses that want to control it all. Where w are now is a lot like the container shipping business, just as it was starting to happen. Remember that the old shipping companies opposed it. Then look at how much business it supports today.

DS: Assignment to readers: look up container shipping and see how it happened.

BF: One book is “The Box: How the Shipping Container Made the World”. A key there is to look at how for so long it didn't happen. It was an obvious idea. Why didn't it happen sooner? Well, the shipping companies wanted to charge differentially for content. They wanted one price and method for corn and another for a different sum of stuff. Sound familiar?

The analogy today is that the Telcos want to build canals across the ocean of bits. The incumbent shipping companies couldn’t control traffic across the real oceans. Yet bit carriers have managed to do with exactly that with something as plentiful as bits. It's nutty, but it's the frame we still have.

But we have a lot of helpful examples of frame-breaking in history. The printing press liberated literacy and knowledge, for example. People were finally able to read the Bible themselves without relying on interpretations. Nearly all the ways we use print today are unintented consequences of Gutenberg's original work.

Look at wired communication. Hmm... Why wasn’t there a home telegraph?

DS: Why?

BF: We accepted the need for an intermediary – the guy who operated the telegraph key. With the advent of the talking telegraph there was no place for such an intermediary, though we still had telephone operators and a network to be managed. Remember that Al Bell was supposed to be solving the problems of the talking telegraph. There’s a good book -- The Victorian Internet -- that draws parallels with today’s Internet. It says we've been through a lot of this already. For example, the telegraph companies didn’t allow you to use your own code words because they charged by the word and viewed your own words as cheating.

Our view of telephony is heavily influenced by the railroad framing. In fact the FCC (Federal Communications Commission) was partly an outgrowth of the ICC (Interstate Commerce Commission) which regulated railroads. You can easily imagine a circuit through the phone network as a track. Thus we apply common carriage regulations. Given the opportunity, railroad owners become infamous robber barons. How different is that from today when phone companies charge you for arbitrary units of value, rather than for bit freight. SMS, for example. It's just data. Yet an SMS bit icosts a millions of times a video bit.

The carriers talk about value, and what it costs them to build and to pay off infrastructure construction. And they think the only way they can monetize that infrastructure is to call one kind of data "video" and another "telephone" and another "broadband" and another "messages" and another "minutes" and charge different sums for all of them.

DS: JP Rangaswami of British Telecom (disclosure: I consult BT on open source strategy) says the core competence of telcos is billing.

BF: That's right. And all the things they're billing for are leveraged from the old analog world -- of railroads, of RF (radio), of cable TV. These are all names of shapes they see in the sea of bits.

DS: Like charging for the constellations we see in the sky.

BF: You have to look at the history here. Analog signaling is lossy – you need to make sure you choose the right wire for each kind of signal so that you preserve the right characteristics. Radio waves had their own problems – in the 1920’s the best idea we had for sorting out signals was based on the tuning fork. The idea didn’t work for telegraphy because mechanical tuning forks were too slow. Using vacuum tubes we were able to make the idea work.

DS: As a result we still have TV "channels" that used to stand for frequencies but are now just names for sources of data streams.

BF: None of this analog baggage is necessary with digital signaling. All the wires form part of a greater whole. Perhaps more to the point. the distinction between wired and wireless bits no longer makes any sense. Why do we need megawatts to throw a signal over a distance, while a few milliwatts can reach the world via the Internet? In fact that whole idea of signaling using single frequencies is about as limiting and anachronistic as you can get – with 802.11 and Bluetooth we do very well with voluntary cooperation and design that takes responsibility for itself. We don't need a worldwide system for policing our means of communicating!

DS: As an old broadcast engineering type, it's amazing to me that we still have TV and radio broadcast transmitters on the tops of buildings and mountains, and -- in the case of AM, in arrays of towers down in the swamps and flats -- radiating on frequencies allocated between the 1920s and 1960s, putting a huge load on the electric grid... when I can listen to the same stuff, or watch it, on a computer or a cell phone or some other portable device that listens over wi-fi anywhere. And that I can originate the same material as well... if the carrier lets me. Which they often don't because the bit pipes they supply are asymmentrical, to accomodate the old broadcast model.

BF: There is some irony here. We can thank the telephone companies for pioneering the digital technologies that make it so easy to be a telephony company, so now everyone does their own telephony. Yet we are still stuck in the railroad model, and a Regulatorium that desperately works to keep protecting the railroads, I mean the Telcos, from the worst enemies – their customers. It’s as if cars were outlawed because they threatened the profitability of railroads.

Fortunately, as I note in VONRailroads, we already had horses and cars, so we couldn’t go backwards, no matter how much the railroads might have liked that. But we still think networking -- railroading for bits -- is special. We fail to notice the obvious: our home networking.

Maybe that's because what we have at home isn't even a network. It's just devices that use wires and radios to do their own connecting. It’s not a service from somebody outside.

This is why home networks can go at gigabit speeds and we don’t pay to use them. We only pay for an inexpensive router if we want to expand our "network". Why should it become an expensive, slow and crippled service once it gets into the world of 1930’s telephony and broadband?

DS: Because as soon as we leave our house we enter the Regulatorium?

BF: Right. And why should we have the Regulatorium? We don't need it.

If you go back to the1980s, UUCP (Unix-Unix Copy) is a good example of networking without a network – just cooperating computers calling each other. As with the Internet it was a learning experience. Today we can do a far better job of networking if we aren't confined to broadband pipes. But the telcos are hooked on that confinement -- and providing it as a set of "services". But it's a losing proposition. By holding on to that model they'll fail.

DS: Craig Burton told me a long time ago that the hardest and most important lesson for a company to learn is when to leave money on the table. You're saying Telcos need to do that too.

BF: They can move when they get spooked. That’s why AT&T divested itself of the local operating companies. They knew that had a problem but it took Judge Greene to get them to do what they had to do. Unfortunately they got it wrong and just recreated the problems in each mini-bell.

DS: It was like the sorcerer's apprentice. Instead of one chopped up broom, we had many brooms, all doing the same thing.

BF: They've made some progress, but not far enough. They're still in denial. Today they know that abundance is their enemy and it’s just a matter of when it gets out of control. What we need is something to wake them up so we can all move beyond telephony. I’m willing to bail out their investors because it is so important to move on.

***** Second "Tape" ******

But getting our heads straight is hard. The whole history is tangled because you've got so many things that are the result of accidental properties. And then you have solutions that treat those properties as deliberate. The "forward" progress is actually a death spiral.

So it's better to think about all the opportunity telco "solutions" close off.

It's as if the construction industry never happened. amd instead and we have the government working with a few monopoly companies that are the only ones allowed to build anything. And with the rest of us begging them to open up and let us do more. This is exactly the problem with "broadband". That's an arcane term from a monopoly industry that's using it to stand in the way of more opportunity than we can measure. (Bob: My words, but replace them with yiours.)

So we need to de-couple. If we do that, we see it's all about fungibility: replaceable pieces. (explain fungibility, and how it relates)

DS: But it's very hard because we share a conceptual framework with the telcos and cablecos.

BF: Yes. It's a hard problem. But please realize that the telcos and cablecos won't solve it. They are the problem, or a big part of it.

The way toward a solution is recognizing that there is a hierarcy of conceptual levels here. "Broadband" lives in one of those levels. If you also live there, you think you need it. But below that level the problem disappears. What you need is just to connect — any way that works. Like you do in your house. where "bandwidth" is hardly an issue at all.

We don't need "broadband" from the phone company. That, and "high speed internet", are just names for something they can package and bill.

In fact your network can be whatever you want it to be. So you start building things with the equivalent of 2x4s and screws and nails, and not by living inside the telco's dream house where all you can ever do is rent.

DS: This brings us to addressing our absent ability to understand the system we don't yet have.

BF: Exmples can help. Here's one. My college roommate wanted to make potato dumplilngs. He followed the recipe, and they disappeared. What happened?

DS: I have no idea.

BF: Because he used instant mashed potatoes. (Bob, I still don't get this analogy. If it matters, bring it home for us.)

What happened with telephony is that they developed (something) that could be used to make a telephone call. (Something about Datapoint.) But unlike Intel, Datapoint didn't (not clear) see the value of the processor. Datapoint was an early pioneer in LSI and other chip stuff, but they didn't get the real potiential. So they died off while Intel saw the processor as an any-purpose component part., and we got to see Moore's Law demonstrated.

IBM is lucky to be alive because they didn't get this either. They didn't know that they'd be making their big money off of services in 2008, and treating the operating system as just a bunch of 2x4s. But, to their credit, they finally saw that they could make their money higher up the stack. And they stayed alive until that happened.

DS: Seems to me that Google gets the abundance side of the Net, today, no?

BF: Not true. Because they believe in the tethered Net. Everything has to be tied back to Google Central. You do things their way, using their services. They do this very well, and with a high degree of openness. And a lot of their services are free in the sense that they charge nothing for them. Lots of people use those services, and as a result many of us tend to think the Net is one big Google silo. Not good.

The positive part of Google is that they have enough revenues from advertising that they can afford to be generous, and to take advantage of many benefits of the opportunity dynamic -- with the exception of complete user freedom. Because that tether is still there. The negative side is that you still depend on them -- in some cases, utterly. In that respect they seem to want to be a phone company. That's why they're in the running for this 700MHz auction.

Meanwhile I believe that for less money they can give everybody an open access point with an open face on it. And provide 100% coverage everywhere instantly -- without all the new infrastructure required at 700MHz. They'd have instant critical mass, and open huge opportunities for everybody.

DS: So what's wrong with going after 700MHz anyway? They can afford it, and they're good competition for the telcos.

BF: The problem is, 700MHz is more of the same.

DS: You mean it's yet another top-down, telco-modeled gotta-be-everywhere, only-works-in-5th-gear boondoggle.

BF: Yes. Instead, Google can be FON without the baggage.

DS: Maning FON is silo'd while what you're proposing is not.

BF: Yes. give open access points to everybody, for them to use any way they want. They wouldn't even need to do it everywhere. They could announce "We are going to spend $100 million to give 100% coverage for everybody in Silicon Valley". That would break the model of the telcos and create a whole new dynamic that would demonstrate how much opportunity there is when you get the telcos out of the picture. Instead they're playing a telco game. They're in the carrier model.

(Bob, we need to unpack this a bit more. Are we talking about just giving a wi-fi base station to everybody, or something like Fon that routes through the user's telco or cableco to something ... else? Not sure what here.)

When you're open to more opportunities, you have more chance of succeeding. A lot of the value in open source is just experimenting with cheap and simple tools. Google could provide some of those, for free, just like it provides many services for free on the Net right now -- services that still have real costs for them.

DS: Okay, you were involved in some very early history here. Let's talk a bit about that.

BF: Yes. Early history of Unix. Let's start with Multics. Problem was that Multics was captive to Honeywell's complex business processes. They kept it far more expensive than it should have been. Unix was just another operating system then, and not much of one. But it was cheap. And because it was cheap, it had a lot of innovation. Price does matter. And price is That's also why I left the big computing world. Because PCs were cheap. And I could do what I wanted with them.

And it happened with PCs, far more than Unix, back in those days. PCs didn't care what applications you ran on them. Like the chips inside them, they were intentionally not specialized. They were perfect for the opportunity dynamic.

PCs were consistent with Moore's Law. Hypergrowth is about de-coupling the marketplace, and not physics. As a source for this, (something) Bob Metcalfe and Peter DeVries (missed details here)...

DS: What's de-coupled from what?

BF: You're de-coupled from the next guy. You're not waiting for Joe. Or the big box maker. Or the telephone company. Your dependency level is low. You have freedom to do what you want. When dependency goes to zero, magic happens.

DS: I like that line.

BF: Another that's useful is from the Rolling Stones: "You can't always get what you want, but... you g et what you need." Hey, Mick Jagger did go to the London School of Economics.

DS: This is helpful if you want broadband, I guess.

BF: Yes, because you might want it, but you don't need it. By which I mean, phone company "broadband" is not the only way to connect. Any more than mainframes or minicomputers were the only way to compute.

DS: So how do you get what you need, if not from "broadband"?

BF: There are two ways to play roulette. The stupid way, and nature's way. (Lost that thread. Please detail. sounds interesting.)

DS: Let's look at the situation in respect to the carriers today then , from their side.

BF: Now that transport is fungible, silos become liabilities.

Simple example. If we separate wired and wireless bits, we have very scarce resources. Yet if you simply hop on the nearest wireless access point, a miliwatt radio will take you around the world. Why build a 50,000-watt broadcast facility with a big radio tower, just to have a signal you can only get around, say, New York City? Why have billion-dollar satellite radio systems when you can deliver the same thing over the Internet?

So the only thing that's keeping telecom alive is a desperate attempt to prevent the obvious.

To get past their death march, you have to look first at the abundance we've found through the Internet, but then to look at the Net we know as a demo. Unix was a demo. UUCP was a demo. And the Net is a demo.

DS: How is it a demo?

BF: Instead of saying "Lets take this idea of the end-to-end experiment, and move packets from piont A to piont B", we said "Let's assume that someone is making best efforts on our behalf to connect, and we'll depend on a central authority. And because this is a prototype, for engineering purposes we'll put a shim in here. That shim is called a backbone. But we can't lose track of the fact that it's still a prototype. A demo. The problem was that the demo worked too well. It's raw and imperfect and nascent, but it shows us what we can do. But it's not the real thing, which would be much more powerful. Imagine what could happen.

DS: Okay, what is the real thing?

BF: It's not having a network, but making networking something you do, just like you drive.

When you drive, you don't worry about whether the railroad would find your driving to be profitable. You just drive. It's about getting there. And if we can't right now, instead of waiting for somebody to build a track to the top of Mount Everest, we'll find a way. Without a central authority telling us how it can be done.

DS: Let's go back to the home network model that you like.

BF: That's where networking is already something you do, rather than something you have. You connect this to that, and it works.

DS: Meaning it's not AT&T or Verizon's or Time Warner's job to hook your printer to your laptop, or to hook up your RAID array.

BF: Yes.

DS: Okay, this is where my inner Richard Bennet comes in. (Richard has a long history as both a developer of networks and a contrarian toward much of what Bob — and, at other times, I — have said. On the whole he takes what we might call the pro-carrier position.) Earlier you said, "the only constraint not being disrupting the old uses". Isn't what you propose disruptive exactly that way? You've been accused of trying to destroy all of telecom. Isn't that where you're headed here?

BF: We will disrupt telecom. You pay $100 for a 10 megabit hub, and you buy a gigabit switch for $50, and you've destroyed the value of the hub, and might as well throw it out.

DS: I don't get it.

[bob, I say lose this example above, for space]

BF: Here's the key. You're allowed to decide what value you destroy and what you don't. You don't necessarily have to be destructive, but you don't have to honor all past decisions. Which may not be about the wires. I can run gigabit over Cat-3. Now, if you want to take my Cat-3, running at 10 megabits, and put in some Cat-6, you want your program running at a gigabit to not be too disrupted by the program running at 10 megabits. But at the same time if you have something running at one megabit on a gigabit network, it's disruptive. So you might want to get rid of that program running at juet one megabit. But it's up to you.

Now if I buy a computer that runs at one gigabit, I don't want to break all my computers running at 10/100 megabits. But I can say I want my computer to run at a gigabit on Cat-6. But I'm going to give up my fat coax at this point. There's always a trade-off but you have to be open to that.

Here's a better example. Modems. They were going to destroy the phone network, because they could carry voice traffic, and they displaced voice traffic. They "clogged up the network". So, what are you gonna do? Two solutions. One is to say "don't do that". And you put billions in to detecting modem calls and stuff like that. And if it weren't for Common Carriage and other laws, this "solution" would have succeeded. The other is to realize it's very easy and cheap to do the data stuff that you're already doing inside the phone system.

So it turns out that the phone companies discovered abundance inside their own system. But then they immediately worked on ways of making it scarce, and selling data services as if they were old fashioned phone services. In fact the purpose of digital inside the telcos was to increase efficiency and allow more to be billed out. Look at ISDN. Rather than do a Moore's Law calculation, they did a static pricing calculation, and charged everybody a high price for dead fish. The price of a dead fish should go down over time.

But it turned out we could use these pure data pipes to run our Internet demo. Which was lost on them. As a result they sucked value out of the market and into their network. That was another dead fish that swam very well until it reached the bottom.

DS: I look at it not as moving value, but as preventing value. How much business do they prevent by maintaining these models? I might actually be doing a lot more business with Verizon and Cox today if both didn't charge far more for "business" use of exactly the same pipes and bandwidths as I get with their home servce. Rather than enable business, they shake it down for more money. Meanwhile, I've got fiber to my home, most of which is carrying TV I don't watch.

BF: Fiber is nice, but be careful about assuming it solves all problems. People who want fiber everywhere ae missing the point of plain connectivity by the best means available. By not fixing everything with fiber, including in our own homes, we're able to use copper and get gigabit capacity at almost no cost. We haven't begun to exploit the capacity of copper. The technology for that has improved over the last twenty years.

If the local copper already installed were owned by the community, you could put in fifty dollar line cards and get plenty done. We could get creative. By treating these things as fungible, we can be creative. Copper, fiber and radios are just ways of transporting bits.

DS: So here's the pushback. For most people, the entire frame of reference is the devil we know. The Internet is bundled with phone and television, as another service. And this is seen as a Good Thing. Today I read a piece in a major magazine about "what is the best triple-play to get". And though I'm not a fan of Verizon as a company, here near Boston I'm getting 20 megabits down and 5 up with a fiber connection, and they're not blocking port 25 like my other carrier does back home in California. The HDTV picture is the best I've seen. And telephone is still its familiar old self, even though it's now running over fiber instead of analog copper. You're looking to solve a problem most people don't think they have.

BF: Of course. Lincoln wondered out loud if emancipation would be a longer chain.

You're not going to free the slaves here. You're trying to make a better, more open marketplace. Better laws. Better treatment. But bear in mind what Henry Ford said. If he'd asked customers what they wanted, they'd say faster horses. "High speed internet" is faster horses.

DS: Right. But on the other hand most of our readers are geeks who are exactly the kind of people who aren't happy being slaves, and who might not want just faster horses. These are the folks who will want to do things, start businesses, and take advantage of your opportunity dynamic.

BF: The way to look at it is, if you look at the current value, you miss the point. Don't confuse the packaging with the demo. Your readers know that if they buy a compnuter and it says Windows on the packaging, they aren't required to run Windows on it. Telephony is a Windows box. Look at connectivity choices in the same way. You can be resourceful on whatever presents itself to you as choices for networking.

A flip side, if you see an access point, don't assume you can only do what somebody's packaged purpose says about it. Look at it as something you can use any way you like. At restaurants, they have menus. Those are packaging for the convenience of the kitchen. But in most restaurants, you can make substitutions. Same with connectivity.

DS: Now let's say you're a telco company, and you don't think of yourself as evil.

BF: And I don't think of them as evil. Just forced to act that way.

DS: Okay, but let's say further that you work for one of those companies and you understand everything Bob Frankston is saying. Are there *any* benefits of incumbency beyond "triple play"?

BF: Your threat is empowered customers. they are going to make use of your capacity, and work around your toll booths, in any case.

(Something here abourt Verizon and Time Warner... too garbled. Fill in.)

To cable companies I say, go into the content business and sell off your distribution while you can still get money for it.

Telcos have a problem in that they are emulating the cable model, and that's a losing propositoin in the long run.

So basically, I think they should just sell off their assets before they go bust.

DS: I disagree. I think there are benefits to incumbency other than selling triple play. I think if the telcos and cablecos are willing to open up their capacity, and give customers all the abundance they need to open up this opportunity dynamic -- what JP Rangaswami and I have been calling "because effects" (you make money because of something, rather than with it) -- they'll be in a great position to provide new services to the companies and customers who are using that abundance. They can provide offsite backup at local places for minimal latency. Or they can lease space for that to S3 or Akamai or somebody else already in that business. They can befriend the businesses that live off of working around the carriers' own asymmetrical bottlenecks. They can provide services to help individuals move and store large amounts of media. [Bob, I may move this to my own back-page editorial.]

BF: Well, railroads made a huge amount of money selling real estate and access alongside tracks. But still, I think the situation is a lot like it would be for McDonalds if people found they really didn't like cheap burgers any more. Suddenly they're in the real estate business, and can only make money by selling it off. Think, if you're AT&T with a 50-story building that has no windows, what are you going to do? It's easier just to tear it down, You don't have free real estate. You have real estate with decommissioning costs.

Their best bet is truth and reconciliation with the feds. They can say, "You got us into this, through your own stupidity. Let's make a deal. You buy our assets (price given, garbled) ... and we'll accept the end of the FCC and we go off into other businesses.