Comments from Frankston, Reed, and Friends
Saturday, November 02, 2002
DanB at 8:23 PM [url]:
SATN now has an RSS feed
I finally turned on the RSS feature in Blogger. You can find a link to RSS 0.91 for this weblog on the left. It goes to the file satn_rss.xml. Thank you to all who complained and requested it. Those of you who use tools that read RSS files to determine what's current in the weblog now can track SATN.org.
Thursday, October 31, 2002
BobF at 6:39 PM [url]:
Owners add value. But we limit ownership based on a static agrarian model of property. In order to maximize society's benefit we need to be able to accept that others' need the opportunity be owners and holding bits hostage leave us with little opportunity to create value.
My wife asked me how to respond to a dialog box that popped up when she logged into one of the computers at home. I use the XP "user login" facility so I can try to keep the activities separate and relatively simple. But I can't tell applications not to harass her with questions aimed at geeks or those who care about a particular application. Part of the issue is that the personal computer has morphed into the old style time sharing systems with all the complexities. This is very bad but I don't have a better answer at the moment because it's also bad to have to have a separate physical machine for each user and for each application at each location for each purpose.
One consequence is that I am no longer in control and each of this applications and utilities is able to directly interact with (AKA harass) everyone who tries to use the computers and thus frustrate my attempt to create a reasonable environment. Of course those who provide the applications, hardware and, more to the point, the operating systems presume they are the ones defining the environment. The intent is not necessarily to prevent me from creating environments as much as, I fear, the lack of appreciation that their one answer is not really the one answer because there isn't just one answer.
I'll address operating systems and user experience issues in a separate note. The real issue here is ownership. One of the strengths of the United States has come from giving people ownership of their own land and thus giving them an incentive to invest in it and improve it. Communities are the ways owners can act collectively to enhance their environment and it requires a balance between the individual's ownership and the collective ownership. This isn't to exclude those who might rent property since they too can be part of the community. The contrast is with those who are not permitted to have a stake and thus lack this incentive.
Bringing this back to telecom (no great surprise) we can contrast the analog (POTS -- Plain Old Telephone Service) with ISDN (Integrated Services Digital Network). ISDN seems superior to POTS in everyway except, apparently, what matters to the marketplace. Modems allowed us to use the less capable analog network. And the modems kept getting better than anyone thought possible. After all, the network was just supposed to handle a relatively low frequency voice signal. They improved to the point that they were within a factor of two of ISDN service. In some cases they were better than ISDN since compression could provide improvements of more than 2x. The basic problem is that ISDN service was defined by the providers and, in the United States, was aimed at a high value (thus high priced) business applications who could also tolerate the tricky installation procedures and the special equipment required. In Europe ISDN was much more available but the high charges for all phone usage discouraged the growth of the consumer data services. In Europe the data services were defined by the providers who focused on delivering information via the television screen or special terminals (as in France's Minitel).
The modem was owned by the users and since local calls were generally not charged by the minute (after all, the copper was just there whether you used it or not) there was a strong incentive to build data services upon the commonly available and affordable analog phone network. And that's precisely what happened.
Today DSL and cable modems are owned by the service providers and there is no incentive to push the limits. In fact, there are proposals to reclaim parts of the modems in order so that the providers can take advantage of their ownership to create so-called value-added services. These are really taking capacity away from the user and reselling them at a high markup. The desire to do this is natural since, after all, they own the modems and we no longer do. This is one more reason to fear provider-defined broadband.
It's natural for a provider or author (very often a corporation) to want to exert maximal ownership rights. In Another Dangerous Idea, I argue against inappropriate policies which confuse technology with social policy. But the other extreme -- presuming that every system "just works" is also problematic. We (those in the United States) do limit the ownership of concepts. Patents and copyrights are about the ownership of specific reductions to practice and presentation. We can consider a set of bits as a kind of property. If you lease me a set of bits and I have no ownership than I have little incentive to invest and add value. But the bits are not the same as real estate. In fact, as I pointed out in A Lack of Progress, the (US) Constitution provided for ownership of intellectual property for the purpose of abetting progress and not because it was an intrinsic right.
Owners add value and encouraging ownership of real estate has been an explicit government policy. This doesn't mean that one owner has the complete right to deny others ownership. In real property there are limits such as the need to share access to resources and property may even be taken away for the public good though this must be done with strong restraint in order to mitigate the benefits of ownership. Ownership of bits too must have its limit in order to even allow others to add value let alone encourage investment. The argument that the investment of the bit authors is exclusive fails to take into account society's benefit by increasing the pool of investors. The first owner does need incentive to invest but the diversity that others' ownership is likely to create novel values and thus add to society's options. If the initial owner simply continues to collect royalties without enhancing the bits then society gains little. Even if the owner continues to enhance the bits, over a period of time that owner is much less likely to make a special contribution compared to possibilities created by a large and diverse pool of owners.
If we look beyond static bits and think of them as a pattern that I interpret or, more to the point, that I have my computer interpret we have a more direct conflict. Asserting that the original provider of the bits has unlimited ownership cedes complete control of my computer to others and that is simply intolerable. The fact that I owned my personal computer allowed me to be creative without having to answer to others. Ownership makes a difference. Circuit's City's aborting effort to sell me DVDs but only lease the content demonstrates the danger. When the service ceased those who thought they owned their content after paying for unlimited use found they were simply tenants after all.
I fear the same might be true for software that requires online authentication. The authentication is subject to the forbearance of the provider and cannot take into account my innovations such as mixing and matching pieces to make PCs and repurpose them. Microsoft seems to be saying that I am not allowed to innovate with my PC. Is it really mine?
We need to think of ownership in terms of the incentive to add value and the benefits to society. We need to be wary of a tradition based on a static agrarian model in a society that is increasingly dependent upon discovering value through innovation rather than simply exploiting finite resources.
For a tutorial on ISDN you can go to http://www.ralphb.net/ISDN/defs.html. But, as with all Internet links, this one is doomed once the author loses interest or simply forgets to renew. So I apologize if the link fails or you find something entirely different at that site. The inability to own your static handle is, itself, an example of the problem. There is a need for stable link that I own and that is incompatible with leasing a .COM name. That's a strong reason for separating the mechanisms.
Wednesday, October 30, 2002
BobF at 10:45 PM [url]:
Another Dangerous Idea
Charles Darwin's evolution (calling it a theory is like calling gravity a theory) has been called a Dangerous Idea, in part, because it forces us to accept our role as part of a larger process. The key to the Internet lies in a simple idea -- giving those at the end points the ability to define services and to decide for themselves which ones are successful and which ones fail to meet their needs. This is generally called the End-to-End argument. Attempts to govern such a system by prejudging what are good and bad services violate this basic mechanism.
The idea that the more one governs a system the worse it behaves seems paradoxical but the Internet provides us with a very simple case study in how such a system works.
We tend to think of systems as being brittle because every part is an essential element. If any part fails then the system fails. If each part is 90% reliable, by the time we have six elements the system will fail more than half the time.
Yet the Internet has millions of systems and billions of pages of information and continues to work despite numerous failures. This is because each element takes responsibility for protecting itself and thus instead of failures propagating they are quickly quenched. If an idea is successful it can quickly be shared and adopted. To really understand evolutionary systems we twist this around and say that an idea that is shared and adopted is successful. Success is determined by what thrives even if it is very disruptive! And therein lies the dilemma of governance.
This doesn't mean that we can't have social policies and preferences but we must recognize the difference between establishing rules about how society should function and the underlying technology that has been a source of innovation and economic value. By using the Internet as a case study we can also learn how individual behavior is able to give us a coherent society rather than anarchy. The Y2K fears were exacerbated by the assumption that our social fabric is brittle rather than resilient. Understanding the resilience is vital in putting today's fears of terrorism in perspective as we understand how resilient our society is.
Today's turmoil in telecom is a direct result of the power of the end-to-end argument. Historically telecommunications has been treated as a service. By shifting the definition of new services to the end points and commoditizing the transport, the Internet has changed the fundamentals of telecommunications and many people in the FCC and the industry understand this. But they are locked into a regulatory system based on premises that no longer make sense given what we've learned from the Internet and the realities of its deployment. Reading through a recent speech by FCC Commissioner Kathleen Q. Abernathy one can see the difficulty of having to tweak an untenable system.
In speaking to members of Congress (the Senate and the House) I appreciate the difficulty they face when they are put in the position of voting on complex issues such as telecommunications regulation. Last week I spoke to a Congressman who commented that maybe three people in Congress even understand what a stem cell is yet they are asked to set guidelines for such research.
The Internet has created a crisis in telecom and Congress is asked by the telecommunications industry and people in general (AKA constituents) to "fix it". Lacking deep technical knowledge they have to use their people skills to decide which experts to believe. The issues are framed in terms of the current regulatory rules. I use the term The Regulatorium as a way to be able to step outside the system of rules and talk about them. Most discussion is premised on the implicit framing defined by familiar terms like telephony and television even though they are not longer fundamental. This leads to the irresistible opportunity to do us good by giving us a lot of "broadband". The term "broadband" is now so familiar that few see the need to define the term. In practice broadband is a service define by the regulatorium and thus we don't get the opportunity to build our own (disruptive) services upon an unfettered transport and we don't get the opportunity to have the marketplace drive the supply well beyond the seemingly ambitious goals of broadband policy.
I signed the "Fail Fast" letter that was sent Michael Powell, Chairman of the FCC as part of the effort to provide counter expertise but I recognize that the real challenge lies in the realm of politics and policy. Ultimately the FCC is answerable to a constituency that wants fish not fishing. When people want dinner it is hard to explain why it is more important to learn to fish than accept immediate gratification.
Our effort to bring telecommunications in line with reality has its counterpart in those who see the Internet as vital infrastructure and see that they have a duty to assure that such a vital infrastructure is made safe and gives us what we need. These efforts generally fall into the realm of governance. These are attempts to recast the Internet in the mold of other social systems without recognizing the fundamental differences. Such efforts are understandable but threaten to deny us the benefits of the Internet. Both sides have something to contribute but I'm concerned about the lack of full engagement. While Zoë Baird raises some interesting points in Governing the Internet (in Foreign Affairs) she quotes John Perry Barlow's observation that Internet finds its own order as if it were an anarchist fantasy rather than an important insight. She is not alone in her criticism of ICANN's process. Many technologists share her concerns, but one of the fundamental problems with ICANN is in its mission which confuses the management of names (meaning) which are social policy with stable links which are necessary but lost in the concern about the commercial (.COM) names.
Confusing the names with the underlying linkages and then imposing social policies (such as Trademark) on the underlying plumping of the Internet is not a benign misunderstanding. The result is that every URL on the Internet is defined to expire and this assures that the Web must unravel. The tragedy is that imposing governance on these names frustrates efforts to provide to provide stable linkages and frustrates efforts to allow for far more flexible and responsive approaches for providing people with meaningful ways to find resources.
Understanding the Internet requires rethinking basic assumptions about how to manage society and it is complicated because those who chose careers in setting policy are unlikely to presume that the policies have fundamental limits.
The Internet is an exciting idea and to those who feel that their common sense understanding is being undermined, it is also dangerous. Even as we address specific technical and policy issues we also need to be aware of the larger changes being wrought by the Internet.
Understanding the concepts behind the Internet as well as the concepts of computing is becoming a fundamental part of our literacy. As with evolution, over the next few generations the concepts will become the new accepted wisdom.
But first we need to try to understand the concepts ourselves. Trying to explain them to others is, in itself, a way to learn.
This is not a battle between clueless policymakers and aware technologists. Technologists can indeed be clueless and making policy doesn't require a lack of understanding. Understanding the concepts require recognizing the interplay of hard edged technology and inherent ambiguities.
Sure it will take generations but we must act as if it will only take the right explanation.
DPR at 1:49 PM [url]:
Disparaging learning, death to metaphors
I read an interesting story in USA Today, entitled Powell takes path to free up airwaves by Paul Davidson. Though the story was mainly about Powell's upcoming announcement of a new spectrum policy, what really annoyed me was that the article's sloppy "popular" style was a perfect illustration of how the press refuses to deal with technology in a clear and honest way.
I'm one of the technology guys leading the Open Spectrum advocacy, but that's not the main point here. The issue is with the technology depth of the writing.
I know technology is difficult for most readers. But it does them no good to pretend to explain things by writing in a way that disparages learning and plays into the idea that bad metaphors are all you need to understand.
From the article: "Academics have long argued that more bands should be set aside for
unlicensed services and that they could even share certain frequencies with
licensed services without interfering."
I love this use of the word "academics". Where did that come from? The Open Spectrum advocacy has some folks from "academy" (Benkler, Lessig, Lippman, Shepard). But most of us have been doing business in the "real world" (me, Dewayne, Werbach, Hughes, ...). And the wider support of unlicensed radio is doing quite well as a business, thank you. Better competitive business people there than at the top of the ILEC, cable, and broadcaster megacorps. Even Intel and Microsoft support unlicensed bands where industry works together to set standards.
Academic, as in "purely academic", I suppose. Just like Szilard (the holder of the patent on the atomic bomb) was an academic. Or like the people who invented the Internet because AT&T could not bring itself to imagine a world where they weren't in control were academics.
Maybe people who think critically and creatively actually know things that are worth hearing about?
Heaven forbid that mere academics are influencing the FCC. We the people wouldn't want to know anything that makes us aware that this is not the best of all possible worlds, or aware that the CEOs in telecom and broadcasting learned everything they needed to know about technology in Kindergarten. Of course they have honorary Ph.D.s in market power.
Metaphors are dangerous
From the article: "the USA's fast-dwindling stock of airwaves"
Metaphor has replaced the need for a thoughtful population. Metaphor's become a crutch. A well-constructed metaphor can stretch the mind, but only when both the author and the reader understand the tension between metaphor and the reference - it's that tension that creates the opportunity for insight, not the substitution.
Those who coin metaphors for "simplification" do not add to the conversation, they create a faux sense of understanding, where there is actually none at all.
Sadly, when I talk about radio, I am reduced to using metaphors (though I try to point out the tension). So, for example, I talk about "ripples on a pond" as a metaphor for radio waves. But it's always amazing to me when I ask a radio engineer what the problem with ripples as a model for radio might be. When I hint that it might have something to do with the propagation speed, they still don't know that surface water waves have a velocity that is an increasing function of wavelength (square root of wavelength) - really long waves travel much faster than ripples (tidal waves travel at near the speed of sound in air). So "wideband" communications in water is very different than "wideband" in radio, where all EM waves travel at the same speed, c.
Most radio engineers work in a particularly inapt metaphor - but they don't know it. That metaphor still includes the essence of the idea called the luminiferous aether (except they call it the "spectrum"). The metaphor includes the idea that a "bit" is a unit of energy (rather than what Shannon defined it to be - which is something that represents correlated probabilities among parts of a system). This confuses the thing (bit) with one possible instance of the thing (a coded pattern of energy or matter).
Communications regulators work in an even more inapt metaphor. That's the one where you regulate services and physical media, and pretend the bits in different services and media are somehow different. You let Congress make up laws based on those inapt metaphors, and turn the police power of the state to the job of enforcing those metaphors, even when they can't possibly be right.
Metaphor is one of the most dangerous tools in use today. But we conspire and pretend metaphors are reality, rather than investigating and exposing their limits. When I talk to reporters, the first thing they want is a good metaphor. When I say it's not that simple, they lose interest. We admire the "clever metaphor" too much.