rfc1259.txt

RFC 的详细文档！

   At the same time as the NREN is being debated and developed,
   telephone companies continue to push at the limits imposed on them by
   the "Modification of Final Judgment" (MFJ) of divestiture, the 1982
   anti-trust agreement which split up the Bell system. (14)  Under
   pressure from the D.C. Court of Appeals, Judge Greene recently lifted
   the information services restrictions on the BOCs -- despite the
   competitive tension between the telephone companies, cable TV
   carriers, and newspapers.  Thus, in the next year or so, Congress may
   well be forced to define a new set of rules for regulated
   telecommunications. (15)  Like the AT&T divestiture decision, this
   would represent a fundamental shift in national policy with enormous
   and unpredictable consequences.

   Many consumer and industry groups are concerned that as the MFJ
   restrictions are lifted, the RBOCs will come to dominate the design
   of the emerging National Public Network, shaping it more to
   accommodate their business goals than the public interest.  The
   Communications Policy Forum, a coalition of public interest and
   industry groups, has recently begun to consider what kinds of
   safeguards will be needed to maintain a competitive information
   services market that allows RBOC participation.  The role that the
   RBOCs come to play in the nation's telecommunications infrastructure
   is, of course, an issue that must be carefully considered on its own.
   But in this context, the NREN represents a critical opportunity to
   create a model for what a public network has to offer, free from
   commercial pressures.

   With all of the uncertainty that surrounds the RBOCs entry into the
   information services market, we should use the NREN to learn how to
   develop a network environment where competitive entry is easy enough
   that the RBOCs opportunity to engage in anti-competitive behavior
   would be minimized.  There is evidence that the RBOCs are resisting
   attempts to transform the public telephone system into a truly open
   public network (16) notwithstanding the FCCs stated intention do
   implement Open Network Architecture. (17)  But since the NREN
   standards and procedures can be designed away from the dominance of
   the RBOCs, a fully open network design is within reach.  In this
   sense the NREN can be a test-bed for "safeguards" against market
   abuse just as it is a test ground for new technical standards and
   innovative network applications.



Kapor                                                          [Page 10]

RFC 1259                 Building The Open Road           September 1991


   An open platform network model carrier from the NREN to the National
   Public Network would actually make some MFJ restrictions less
   necessary.  Phone companies were originally prohibited from being
   information providers because their bottleneck control over the local
   exchange hubs gives them an unfair advantage.  But on a network in
   which the local switch is open to information providers -- because
   the platform itself is so rich and well-designed -- creativity and
   quality triumph over monopoly power.  Instead of restricting
   information providers, the National Public Network developers should
   encourage the entry of as many new parties as possible. Just as
   personal computer companies started in garages and attics, so will
   tomorrow's information entrepreneurs, if we give them a chance.
   Their prototypes today, small computer networks, electronic
   newsletters, and chat lines, are among the most vibrant and
   imaginative "publishers" in the world.

III.  Encourage Pricing for Universal Access

   Everyone agrees in the abstract with universal service -- the idea
   that any individual who wishes should be able to connect to a
   National Public Network. But that's only a platitude unless
   accompanied by an inclusive pricing plan.

   The importance of extending universal access to information and
   communication resources has been widely recognized:

      In light of the possibilities for new service offerings by the
      21st century, as well as the growing importance of
      telecommunications and information services to US economic and
      social development, limiting our concept of universal service to
      the narrow provision of basic voice telephone service no longer
      services the public interest.  Added to universal basic telephone
      service should be the broader concept of universal opportunity to
      access these new technologies and applications. (18)

   The problem of disparate access to information resources has been
   recognized in other telecommunications arenas as well.  Congressman
   Edward Markey (D-Mass.), Chairman of the Subcommittee of
   Telecommunications and Finance of the House Energy and Commerce
   Committee warns that:

      [i]nformation services are beginning to proliferate.  The
      challenge before us is how to make them available swiftly to the
      largest number of Americans at costs which don't divide the
      society into information haves and havenots and in a manner which
      does not compromise our adherence to the long-cherished principles
      of diversity, competition and common carriage. (19)




Kapor                                                          [Page 11]

RFC 1259                 Building The Open Road           September 1991


   To address this problem in the long-term, there is legislation now
   pending which would broaden the guarantee of universal phone service
   to universal access to advanced telecommunications services.  Senator
   Burns has proposed that the universal service guarantee statement in
   the Communications Act of 1934 should be amended to include access to

      a nation-wide, advanced, interactive, interoperable, broadband
      communications system available to all people, businesses,
      services, organizations, and households..." (20)

   In the near term, the NREN can serve as a laboratory for testing a
   variety of pricing and access schemes in order to determine how best
   to bring basic network services to large numbers of users.  The NREN
   platform should facilitate the offering of fee-based services for
   individuals.

   Cable TV is one good model: joining a service requires an investment
   of $100 for a TV set, which 99% of households already own, about $50
   for a cable hookup, and perhaps $15 per month in basic service.
   Anything beyond that, like premium movie channels or pay-per-events
   is available at extra cost. Similarly, a carrier providing connection
   to the mature National Public Network might charge a one-time startup
   fee and then a low fixed monthly rate for access to basic services,
   which would include a voice telephone capability.

   Because regulators are concerned about any telephone service that
   might cause the price of basic voice service to rise, they are
   unwilling to approve new services which don't immediately recover
   their own costs.  They are concerned that any deficit will be passed
   on to consumers in the form of higher charges for standard services.
   As a result, telephone companies tend to be very conservative in
   estimating the demand for new services.  Prices for new services turn
   out to be much higher than what would be required for universal
   digital service.  This is a kind of catch-22, in which lower prices
   won't be set until demand goes up, but demand will never go up if
   prices aren't low enough.

   Open architecture could help phone companies offer lower rates for
   digital services. If opportunities and incentives exist for
   information entrepreneurs, they will create the services which will
   stimulate demand, increase volume, and create more revenue-generating
   traffic for the carriers.  In a competitive market, with higher
   volumes, lower prices follow.








Kapor                                                          [Page 12]

RFC 1259                 Building The Open Road           September 1991


IV.  Make the Network Simple to Use

   The ideal means of accessing the NPN will not be a personal computer
   as we know it today, but a much simpler, streamlined information
   appliance - a hybrid of the telephone and the computer.

   "Transparency" is the Holy Grail of software designers. When a
   program is perfectly transparent, people forget about the fact that
   they are using a computer. The mechanics of the program no longer
   intrude on their thoughts. The most successful computer programs are
   nearly always transparent: a spreadsheet, for instance, is as self-
   evident as a ledger page. Once users grasp a few concepts (like rows,
   cells, and formula relationships), they can say to themselves,
   "What's in cell A-6?" without feeling that they are using an alien
   language.

   Personal computer communications, by contrast, are practically
   opaque.  Users must be aware of baud rates, parity, duplex, and file
   transfer protocols -- all of which a reasonably well-designed network
   could handle for them. It's as if, every time you wanted to drive to
   the store, you had to open up the hood and adjust the sparkplugs. On
   most Internet systems, it's even worse; newcomers find themselves
   confronting what John Perry Barlow calls a "savage user interface."
   Messages bounce, conferencing commands are confusing, headers look
   like gibberish, none of it is documented, and nobody seems to care.
   The excitement about being part of an extended community quickly
   vanishes. On a National Public Network, this invites failure.  People
   without the time to invest in learning arcane commands would simply
   not participate. The network would become needlessly exclusionary.

   Part of the NREN goal of "expand[ing] the number of researchers,
   educators, and students with ... access to high performance computing
   resources" (21) is to make all network applications easy-to-use.  As
   the experience of the personal computer industry has shown, the only
   way to bring information resources to large numbers of people is with
   simple, easy-to-learn tools.  The NREN can be a place where various
   approaches to user-friendly networks are tested and evaluated.

   Technically trained people are not troglodytes; they approve of
   human-oriented design, even as they manage to use the network today
   without it.  For years, leaders within the Internet community have
   been taking steps to improve ease of use on the network.  But the
   training of the technical community as a whole has given them little
   practice making their digital artifacts appropriate for non-technical
   consumption.  Nor are they often rewarded for doing so.  To a phone
   company engineer designing a new high-speed telephone switch, or to a
   computer scientist pushing the limits of a data compression
   algorithm, the notion of making electronic mail as simple as fax



Kapor                                                          [Page 13]

RFC 1259                 Building The Open Road           September 1991


   machine may make sense, but it also feels like someone else's job.
   Being technically minded themselves, they feel comfortable with the
   specialized software they use and seldom empathize with the neophyte.
   The result is a proliferation of arcane, clumsy tools in both
   hardware and software, defended by the cognoscenti: "I use the "vi"
   editor all the time -- why would anyone have trouble with it?"

   If we have the vision and commitment to try this, the transformation
   of the network frontier from wilderness to civilization need not
   display the brutality of 19th century imperialism.  As commercial
   opportunities to offer applications and services develop,
   entrepreneurs will discover that ease of use sells. The normal,
   sometimes slow, play of competitive markets should cause industry to
   commit the resources to serve the market by making access more
   transparent.  But at the start transparency will need deliberate
   encouragement -- if only to overcome the inertia of old habits.

V.  Develop Standards of Information Presentation

   The National Public Network will need an integrated suite of high-
   level standards for the exchange of richly formatted and structured
   information, whether as text, graphics, sound, or moving images.  Use
   the NREN as a test-bed for a variety of information presentation and
   exchange standards on the road towards an internationally-accepted
   set of standards for the National Public Network.

   Standards -- the internal language of networks -- are arranged in a
   series of layers. The lower levels detail how the networks'
   subterranean "wiring" and "plumbing" is managed.  Well-developed sets
   of lower-level standards such as TCP/IP are in wide use and continue
   to be refined and extended, but these alone are not sufficient.  The
   uppermost layers contain specifications such as how text appears on
   the screen and the components of which documents are composed.  These
   are the kinds of concerns which are directly relevant to users who
   wish to communicate.  Recently independent efforts to develop high-
   level standards for document formats have begun, but these projects
   are not yet being integrated into computer networks.

   Today, for example, the only common standard for computer text is the
   American Standard Code for Information Interchange (ASCII).  But
   ASCII is inadequate; it ignores fonts, type styles (like boldface and
   italics), footnotes, headers, and other formats which people
   regularly use. Each word processing program codes these formats