[Politech] Jonathan Weinberg on how to create "Net drivers licenses" [priv]

From: Declan McCullagh (declan@private)
Date: Fri Oct 10 2003 - 09:42:56 PDT

  • Next message: Declan McCullagh: "[Politech] SunnComm won't sue Princeton student over "shift key" paper [ip][fs]"

    Date: Fri, 10 Oct 2003 12:32:47 -0400 (EDT)
    From: Jonathan Weinberg <weinberg@private>
    To: Declan McCullagh <declan@private>
    Subject: Re: [Politech] RIAA replies to Politech over "Net drivers licenses"
    	But there's nothing so odd about the proposal; its core features
    would have been accomplished by Intel's Pentium III architecture back in
    1999.  (For more about Intel's Processor Serial Number and its function as
    a globally unique ID for Internet-connected computers, see Hardware-Based
    ID, Rights Management, & Trusted Systems, 52 Stanford L Rev 1251 (2000),
    <http://www.law.wayne.edu/weinberg/newstanford.PDF>).  Current
    trusted-systems design has moved from embedded GUIDs to more sophisticated
    controls.  See Ross Anderson's FAQ at
    Jonathan Weinberg
    Professor of Law, Wayne State University
    Date: Fri, 10 Oct 2003 11:06:23 -0400 (EDT)
    From: batz <batsy@private>
    To: Declan McCullagh <declan@private>
    Cc: politech@private, srh@private
    Subject: Re: [Politech] Record labels want mandatory "Net drivers license?"
    While I am not an inside source, I can tell you why and how
    this "Net drivers license" would be implemented.
    The Internet is unique as far as communications technologies go
    in that it is built around a stateless protocol (IP). The
    whole notion of a routed store-and-forward vs. a
    circuit-switched network, is that the reliability of packet
    delivery is based upon the feature of the network that
    specifically removes the requirement for and end-to-end
    control connection.
    The telephone networks and, most notably, cellular networks are
    built with this end-to-end control connection in mind. This
    is what allows companies to deliver phone services profitably.
    They have their infrastructure subsidized through various
    incentives, and charge users monthly fees for the connection,
    the interface (phone), a base service package rate, usage, and
    then leverage that control to be the sole distributor of
    services over that media, such as text messaging, voicemail,
    long distance, etc. Even if there are competitive service
    providers, the user much pay their source provider to access
    If you look at it from the perspective of a protocol
    stack, they have control of, and therefore profit from, each
    layer of physical, data/link, transport, session,
    presentation and application layers of the medium.
    Businesses that have profited by and based upon distributing
    online content (porn, napster) didn't need this kind of control
    because what they are distributing was as abundant as water.
    Similarly, content in telecoms is as abundant as water, but
    they do not control it.
    Bandwidth is a measure of the rate at which information can be
    transmitted across a channel, not the value of what that
    information represents. That is, bandwidth is a relatively
    constant quantity, whereas value is dynamic and variable,
    thus available for optimization.
    MP3 is an excellent example of this, in that it is a
    technology that allowed for (high value) music files
    to be transmitted efficiently with existing bandwidth
    To deliver services profitably, the provider must have sufficient
    control over the availability of that service. This availability is
    facilitated by the medium in the telecom world, as it was by physical
    media (CD's) in the recording industry. Therefore, control of the medium
    is control of the content. For you network engineers out there,
    the DoD or OSI protocol stack evidences that control of a lower
    layer of the medium, encapsulates the control of higher ones. The
    Internet changed this by abstracting the network layer to function
    autonomously from it's lower layers. Control of this abstracted
    network layer is only as broad as the control of the lower layers,
    which wasn't an issue until this network layer transcended them
    to become a ubiquitous layer.
    What the RIAA sees, is that profit is a function of control,
    and until an end-to-end control connection can be built into
    Internet Protocols, they have no means of profiting. ISP's
    will move to the cellular phone model, where they will manage
    the users device as a means to deliver services to him, as
    without value-added service delivery, they are essentually a
    fixed-income business that scales linearly with the number of
    users added. There is little room for optimization or value-add
    in the bandwidth business as it stands. They will recognize this
    and build in that control connection to deliver services, which
    will in turn, restore the music industrys media control.
    This is what the Microsoft Network experiment was. M$ leased
    dial pools from major ISP's and tried to get into the Internet
    business. They  had spotty control of the desktop, in that they
    owned it, but had no access to it to leverage it into delivering
    services. It can be considered a failure because of entrenched
    ISP's were already offering more user-centric cheaper service
    and MSN was doomed. AOL had the right idea from a technology
    standpoint, and they hedged their bets on local ISP's dying
    out because of not having any real services. The jury is still
    out on that one.
    What has happened is massive entry into the market by cable
    providers, who have a business model that is right in between
    cellular companies and ISP's. They have the wire, the network,
    the content and the services. With things like bandwidth
    caps and proxies, they are slowing closing the loop around
    the desktop, while dangling the carrot of integrated services
    like pay-per-view as an incentive to get users to give up
    control of their desktops.
    These alleged "licences" will be little more than phone numbers,
    but their value is that, like phone numbers, they provide a control
    connection, which the service providers (and law enforcement etc) can
    access the user to deliver services to them, and enforce policies
    against them as part of the agreements.
    I have purposely made ambiguous use of the words "control connection"
    as in this context, the meaning of an empirical control, (or known
    constant), an out of band link, and political control, all converge.
    After all, property is just information, but media is control.
    Politech mailing list
    Archived at http://www.politechbot.com/
    Moderated by Declan McCullagh (http://www.mccullagh.org/)

    This archive was generated by hypermail 2b30 : Fri Oct 10 2003 - 10:00:21 PDT