RFC 3724






Network Working Group                                      J. Kempf, Ed.
Request for Comments: 3724                               R. Austein, Ed.
Category: Informational                                              IAB
                                                              March 2004


          The Rise of the Middle and the Future of End-to-End:
       Reflections on the Evolution of the Internet Architecture

Status of this Memo



   This memo provides information for the Internet community.  It does
   not specify an Internet standard of any kind.  Distribution of this
   memo is unlimited.

Copyright Notice



   Copyright (C) The Internet Society (2004).  All Rights Reserved.

Abstract



   The end-to-end principle is the core architectural guideline of the
   Internet.  In this document, we briefly examine the development of
   the end-to-end principle as it has been applied to the Internet
   architecture over the years.  We discuss current trends in the
   evolution of the Internet architecture in relation to the end-to-end
   principle, and try to draw some conclusion about the evolution of the
   end-to-end principle, and thus for the Internet architecture which it
   supports, in light of these current trends.

Table of Contents



   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  2
   2.  A Brief History of the End-to-End Principle. . . . . . . . . .  2
   3.  Trends Opposed to the End-to-End Principle . . . . . . . . . .  5
   4.  Whither the End-to-End Principle?. . . . . . . . . . . . . . .  8
   5.  Internet Standards as an Arena for Conflict. . . . . . . . . . 10
   6.  Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . 11
   7.  Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 11
   8.  Security Considerations. . . . . . . . . . . . . . . . . . . . 12
   9.  Informative References . . . . . . . . . . . . . . . . . . . . 12
   10. Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . 13
   11. Full Copyright Statement . . . . . . . . . . . . . . . . . . . 14








Kempf & Austein              Informational                      [Page 1]

RFC 3724                  Future of End-to-End                March 2004


1.  Introduction



   One of the key architectural guidelines of the Internet is the end-
   to-end principle in the papers by Saltzer, Reed, and Clark [1][2].
   The end-to-end principle was originally articulated as a question of
   where best not to put functions in a communication system.  Yet, in
   the ensuing years, it has evolved to address concerns of maintaining
   openness, increasing reliability and robustness, and preserving the
   properties of user choice and ease of new service development as
   discussed by Blumenthal and Clark in [3]; concerns that were not part
   of the original articulation of the end-to-end principle.

   In this document, we examine how the interpretation of the end-to-end
   principle has evolved over the years, and where it stands currently.
   We examine trends in the development of the Internet that have led to
   pressure to define services in the network, a topic that has already
   received some amount of attention from the IAB in RFC 3238 [5].  We
   describe some considerations about how the end-to-end principle might
   evolve in light of these trends.

2.  A Brief History of the End-to-End Principle



2.1.  In the Beginning...

   The end-to-end principle was originally articulated as a question of
   where best to put functions in a communication system:

      The function in question can completely and correctly be
      implemented only with the knowledge and help of the application
      standing at the end points of the communication system.
      Therefore, providing that questioned function as a feature of the
      communication system itself is not possible.  (Sometimes an
      incomplete version of the function provided by the communication
      system may be useful as a performance enhancement.) [1].

   A specific example of such a function is delivery guarantees [1].
   The original ARPANET returned a message "Request for Next Message"
   whenever it delivered a packet.  Although this message was found to
   be useful within the network as a form of congestion control, since
   the ARPANET refused to accept another message to the same destination
   until the previous acknowledgment was returned, it was never
   particularly useful as an indication of guaranteed delivery.  The
   problem was that the host stack on the sending host typically doesn't
   want to know just that the network delivered a packet, but rather the
   stack layer on the sending host wants to know that the stack layer on
   the receiving host properly processed the packet.  In terms of modern
   IP stack structure, a reliable transport layer requires an indication
   that transport processing has successfully completed, such as given



Kempf & Austein              Informational                      [Page 2]

RFC 3724                  Future of End-to-End                March 2004


   by TCP's ACK message [4], and not simply an indication from the IP
   layer that the packet arrived.  Similarly, an application layer
   protocol may require an application-specific acknowledgement that
   contains, among other things, a status code indicating the
   disposition of the request.

   The specific examples given in [1] and other references at the time
   [2] primarily involve transmission of data packets: data integrity,
   delivery guarantees, duplicate message suppression, per packet
   encryption, and transaction management.  From the viewpoint of
   today's Internet architecture, we would view most of these as
   transport layer functions (data integrity, delivery guarantees,
   duplicate message suppression, and perhaps transaction management),
   others as network layer functions with support at other layers where
   necessary (for example, packet encryption), and not application layer
   functions.

2.2.  ...In the Middle...

   As the Internet developed, the end-to-end principle gradually widened
   to concerns about where best to put the state associated with
   applications in the Internet: in the network or at end nodes.  The
   best example is the description in RFC 1958 [6]:

      This principle has important consequences if we require
      applications to survive partial network failures.  An end-to-end
      protocol design should not rely on the maintenance of state (i.e.,
      information about the state of the end-to-end communication)
      inside the network.  Such state should be maintained only in the
      endpoints, in such a way that the state can only be destroyed when
      the endpoint itself breaks (known as fate-sharing).  An immediate
      consequence of this is that datagrams are better than classical
      virtual circuits.  The network's job is to transmit datagrams as
      efficiently and flexibly as possible.  Everything else should be
      done at the fringes.

   The original articulation of the end-to-end principle - that
   knowledge and assistance of the end point is essential and that
   omitting such knowledge and implementing a function in the network
   without such knowledge and assistance is not possible - took a while
   to percolate through the engineering community, and had evolved by
   this point to a broad architectural statement about what belongs in
   the network and what doesn't.  RFC 1958 uses the term "application"
   to mean the entire network stack on the end node, including network,
   transport, and application layers, in contrast to the earlier
   articulation of the end-to-end principle as being about the
   communication system itself.  "Fate-sharing" describes this quite
   clearly: the fate of a conversation between two applications is only



Kempf & Austein              Informational                      [Page 3]

RFC 3724                  Future of End-to-End                March 2004


   shared between the two applications; the fate does not depend on
   anything in the network, except for the network's ability to get
   packets from one application to the other.

   The end-to-end principle in this formulation is specifically about
   what kind of state is maintained where:

      To perform its services, the network maintains some state
      information: routes, QoS guarantees that it makes, session
      information where that is used in header compression, compression
      histories for data compression, and the like.  This state must be
      self-healing; adaptive procedures or protocols must exist to
      derive and maintain that state, and change it when the topology or
      activity of the network changes.  The volume of this state must be
      minimized, and the loss of the state must not result in more than
      a temporary denial of service given that connectivity exists.
      Manually configured state must be kept to an absolute minimum.[6]

   In this formulation of the end-to-end principle, state involved in
   getting packets from one end of the network to the other is
   maintained in the network.  The state is "soft state," in the sense
   that it can be quickly dropped and reconstructed (or even required to
   be periodically renewed) as the network topology changes due to
   routers and switches going on and off line.  "Hard state", state upon
   which the proper functioning of the application depends, is only
   maintained in the end nodes.  This formulation of the principle is a
   definite change from the original formulation of the principle, about
   end node participation being required for proper implementation of
   most functions.

   In summary, the general awareness both of the principle itself and of
   its implications for how unavoidable state should be handled grew
   over time to become a (if not the) foundation principle of the
   Internet architecture.

2.3.  ...And Now.

   An interesting example of how the end-to-end principle continues to
   influence the technical debate in the Internet community is IP
   mobility.  The existing Internet routing architecture severely
   constrains how closely IP mobility can match the end-to-end principle
   without making fundamental changes.  Mobile IPv6, described in the
   Mobile IPv6 specification by Johnson, Perkins, and Arkko [7],
   requires a routing proxy in the mobile node's home network (the Home
   Agent) for maintaining the mapping between the mobile node's routing
   locator, the care of address, and the mobile node's node identifier,
   the home address.  But the local subnet routing proxy (the Foreign
   Agent), which was a feature of the older Mobile IPv4 design [8] that



Kempf & Austein              Informational                      [Page 4]

RFC 3724                  Future of End-to-End                March 2004


   compromised end-to-end routing, has been eliminated.  The end node
   now handles its own care of address.  In addition, Mobile IPv6
   includes secure mechanisms for optimizing routing to allow end-to-end
   routing between the mobile end node and the correspondent node,
   removing the need to route through the global routing proxy at the
   home agent.  These features are all based on end to end
   considerations.  However, the need for the global routing proxy in
   the Home Agent in Mobile IPv6 is determined by the aliasing of the
   global node identifier with the routing identifier in the Internet
   routing architecture, a topic that was discussed in an IAB workshop
   and reported in RFC 2956 [9], and that hasn't changed in IPv6.

   Despite this constraint, the vision emerging out of the IETF working
   groups developing standards for mobile networking is of a largely
   autonomous mobile node with multiple wireless link options, among
   which the mobile node picks and chooses.  The end node is therefore
   responsible for maintaining the integrity of the communication, as
   the end-to-end principle implies.  This kind of innovative
   application of the end-to-end principle derives from the same basic
   considerations of reliability and robustness (wireless link
   integrity, changes in connectivity and service availability with
   movement, etc.) that motivated the original development of the end-
   to-end principle.  While the basic reliability of wired links,
   routing, and switching equipment has improved considerably since the
   end-to-end principle was formalized 15 years ago, the reliability or
   unreliability of wireless links is governed more strongly by the
   basic physics of the medium and the instantaneous radio propagation
   conditions.

3.  Trends Opposed to the End-to-End Principle



   While the end-to-end principle continues to provide a solid
   foundation for much IETF design work, the specific application of the
   end-to-end principle described in RFC 1958 has increasingly come into
   question from various directions.  The IAB has been concerned about
   trends opposing the end-to-end principle for some time, for example
   RFC 2956 [9] and RFC 2775 [12].  The primary focus of concern in
   these documents is the reduction in transparency due to the
   introduction of NATs and other address translation mechanisms in the
   Internet, and the consequences to the end-to-end principle of various
   scenarios involving full, partial, or no deployment of IPv6.  More
   recently, the topic of concern has shifted to the consequences of
   service deployment in the network.  The IAB opinion on Open Pluggable
   Edge Services (OPES) in RFC 3238 [5] is intended to assess the
   architectural desirability of defining services in the network and to
   raise questions about how such services might result in compromises





Kempf & Austein              Informational                      [Page 5]

RFC 3724                  Future of End-to-End                March 2004


   of privacy, security, and end-to-end data integrity.  Clark, et al.
   in [10] and Carpenter in RFC 3234 [11] also take up the topic of
   service definition in the network.

   Perhaps the best review of the forces militating against the end-to-
   end principle is by Blumenthal and Clark in [3].  The authors make
   the point that the Internet originally developed among a community of
   like-minded technical professionals who trusted each other, and was
   administered by academic and government institutions who enforced a
   policy of no commercial use.  The major stakeholders in the Internet
   are quite different today.  As a consequence, new requirements have
   evolved over the last decade.  Examples of these requirements are
   discussed in the following subsections.  Other discussions about
   pressures on the end-to-end principle in today's Internet can be
   found in the discussion by Reed [13] and Moors' paper in the 2002
   IEEE International Communications Conference [14].

3.1.  Need for Authentication



   Perhaps the single most important change from the Internet of 15
   years ago is the lack of trust between users.  Because the end users
   in the Internet of 15 years ago were few, and were largely dedicated
   to using the Internet as a tool for academic research and
   communicating research results (explicit commercial use of the
   Internet was forbidden when it was run by the US government), trust
   between end users (and thus authentication of end nodes that they
   use) and between network operators and their users was simply not an
   issue in general.  Today, the motivations of some individuals using
   the Internet are not always entirely ethical, and, even if they are,
   the assumption that end nodes will always co-operate to achieve some
   mutually beneficial action, as implied by the end-to-end principle,
   is not always accurate.  In addition, the growth in users who are
   either not technologically sophisticated enough or simply
   uninterested in maintaining their own security has required network
   operators to become more proactive in deploying measures to prevent
   naive or uninterested users from inadvertently or intentionally
   generating security problems.

   While the end-to-end principle does not require that users implicitly
   trust each other, the lack of trust in the Internet today requires
   that application and system designers make a choice about how to
   handle authentication, whereas that choice was rarely apparent 15
   years ago.  One of the most common examples of network elements
   interposing between end hosts are those dedicated to security:
   firewalls, VPN tunnel endpoints, certificate servers, etc.  These
   intermediaries are designed to protect the network from unimpeded
   attack or to allow two end nodes whose users may have no inherent
   reason to trust each other to achieve some level of authentication.



Kempf & Austein              Informational                      [Page 6]

RFC 3724                  Future of End-to-End                March 2004


   At the same time, these measures act as impediments for end-to-end
   communications.  Third party trust intermediaries are not a
   requirement for security, as end-to-end security mechanisms, such as
   S/MIME [15], can be used instead, and where third party measures such
   as PKI infrastructure or keys in DNS are utilized to exchange keying
   material, they don't necessarily impinge on end-to-end traffic after
   authentication has been achieved.  Even if third parties are
   involved, ultimately it is up to the endpoints and their users in
   particular, to determine which third parties they trust.

3.2.  New Service Models



   New service models inspired by new applications require achieving the
   proper performance level as a fundamental part of the delivered
   service.  These service models are a significant change from the
   original best effort service model.  Email, file transfer, and even
   Web access aren't perceived as failing if performance degrades,
   though the user may become frustrated at the time required to
   complete the transaction.  However, for streaming audio and video, to
   say nothing of real time bidirectional voice and video, achieving the
   proper performance level, whatever that might mean for an acceptable
   user experience of the service, is part of delivering the service,
   and a customer contracting for the service has a right to expect the
   level of performance for which they have contracted.  For example,
   content distributors sometimes release content via content
   distribution servers that are spread around the Internet at various
   locations to avoid delays in delivery if the server is topologically
   far away from the client.  Retail broadband and multimedia services
   are a new service model for many service providers.

3.3.  Rise of the Third Party



   Academic and government institutions ran the Internet of 15 years
   ago.  These institutions did not expect to make a profit from their
   investment in networking technology.  In contrast, the network
   operator with which most Internet users deal today is the commercial
   ISP.  Commercial ISPs run their networks as a business, and their
   investors rightly expect the business to turn a profit.  This change
   in business model has led to a certain amount of pressure on ISPs to
   increase business prospects by deploying new services.

   In particular, the standard retail dialup bit pipe account with email
   and shell access has become a commodity service, resulting in low
   profit margins.  While many ISPs are happy with this business model
   and are able to survive on it, others would like to deploy different
   service models that have a higher profit potential and provide the
   customer with more or different services.  An example is retail
   broadband bit pipe access via cable or DSL coupled with streaming



Kempf & Austein              Informational                      [Page 7]

RFC 3724                  Future of End-to-End                March 2004


   multimedia.  Some ISPs that offer broadband access also deploy
   content distribution networks to increase the performance of
   streaming media.  These services are typically deployed so that they
   are only accessible within the ISP's network, and as a result, they
   do not contribute to open, end-to-end service.  From an ISP's
   standpoint, however, offering such service is an incentive for
   customers to buy the ISP's service.

   ISPs are not the only third party intermediary that has appeared
   within the last 10 years.  Unlike the previous involvement of
   corporations and governments in running the Internet, corporate
   network administrators and governmental officials have become
   increasingly demanding of opportunities to interpose between two
   parties in an end-to-end conversation.  A benign motivation for this
   involvement is to mitigate the lack of trust, so the third party acts
   as a trust anchor or enforcer of good behavior between the two ends.
   A less benign motivation is for the third parties to insert policy
   for their own reasons, perhaps taxation or even censorship.  The
   requirements of third parties often have little or nothing to do with
   technical concerns, but rather derive from particular social and
   legal considerations.

4.  Whither the End-to-End Principle?



   Given the pressures on the end-to-end principle discussed in the
   previous section, a question arises about the future of the end-to-
   end principle.  Does the end-to-end principle have a future in the
   Internet architecture or not?  If it does have a future, how should
   it be applied?  Clearly, an unproductive approach to answering this
   question is to insist upon the end-to-end principle as a
   fundamentalist principle that allows no compromise.  The pressures
   described above are real and powerful, and if the current Internet
   technical community chooses to ignore these pressures, the likely
   result is that a market opportunity will be created for a new
   technical community that does not ignore these pressures but which
   may not understand the implications of their design choices.  A more
   productive approach is to return to first principles and re-examine
   what the end-to-end principle is trying to accomplish, and then
   update our definition and exposition of the end-to-end principle
   given the complexities of the Internet today.

4.1.  Consequences of the End-to-End Principle



   In this section, we consider the two primary desirable consequences
   of the end-to-end principle: protection of innovation and provision
   of reliability and robustness.





Kempf & Austein              Informational                      [Page 8]

RFC 3724                  Future of End-to-End                March 2004


4.1.1.  Protection of Innovation



   One desirable consequence of the end-to-end principle is protection
   of innovation.  Requiring modification in the network in order to
   deploy new services is still typically more difficult than modifying
   end nodes.  The counterargument - that many end nodes are now
   essentially closed boxes which are not updatable and that most users
   don't want to update them anyway - does not apply to all nodes and
   all users.  Many end nodes are still user configurable and a sizable
   percentage of users are "early adopters," who are willing to put up
   with a certain amount of technological grief in order to try out a
   new idea.  And, even for the closed boxes and uninvolved users,
   downloadable code that abides by the end-to-end principle can provide
   fast service innovation.  Requiring someone with a new idea for a
   service to convince a bunch of ISPs or corporate network
   administrators to modify their networks is much more difficult than
   simply putting up a Web page with some downloadable software
   implementing the service.

4.1.2.  Reliability and Trust



   Of increasing concern today, however, is the decrease in reliability
   and robustness that results from deliberate, active attacks on the
   network infrastructure and end nodes.  While the original developers
   of the Internet were concerned by large-scale system failures,
   attacks of the subtlety and variety that the Internet experiences
   today were not a problem during the original development of the
   Internet.  By and large, the end-to-end principle was not addressed
   to the decrease in reliability resulting from attacks deliberately
   engineered to take advantage of subtle flaws in software.  These
   attacks are part of the larger issue of the trust breakdown discussed
   in Section 3.1.  Thus, the issue of the trust breakdown can be
   considered another forcing function on the Internet architecture.

   The immediate reaction to this trust breakdown has been to try to
   back fit security into existing protocols.  While this effort is
   necessary, it is not sufficient.  The issue of trust must become as
   firm an architectural principle in protocol design for the future as
   the end-to-end principle is today.  Trust isn't simply a matter of
   adding some cryptographic protection to a protocol after it is
   designed.  Rather, prior to designing the protocol, the trust
   relationships between the network elements involved in the protocol
   must be defined, and boundaries must be drawn between those network
   elements that share a trust relationship.  The trust boundaries
   should be used to determine what type of communication occurs between
   the network elements involved in the protocol and which network
   elements signal each other.  When communication occurs across a trust
   boundary, cryptographic or other security protection of some sort may



Kempf & Austein              Informational                      [Page 9]

RFC 3724                  Future of End-to-End                March 2004


   be necessary.  Additional measures may be necessary to secure the
   protocol when communicating network elements do not share a trust
   relationship.  For example, a protocol might need to minimize state
   in the recipient prior to establishing the validity of the
   credentials from the sender in order to avoid a memory depletion DoS
   attack.

4.2.  The End-to-End Principle in Applications Design



   The concern expressed by the end-to-end principle is applicable to
   applications design too.  Two key points in designing application
   protocols are to ensure they don't have any dependencies that would
   break the end-to-end principle and to ensure that they can identify
   end points in a consistent fashion.  An example of the former is
   layer violations - creating dependencies that would make it
   impossible for the transport layer, for example, to do its work
   appropriately.  Another issue is the desire to insert more
   applications infrastructure into the network.  Architectural
   considerations around this issue are discussed in RFC 3238 [5].  This
   desire need not result in a violation of the end-to-end principle if
   the partitioning of functioning is done so that services provided in
   the network operate with the explicit knowledge and involvement of
   endpoints, when such knowledge and involvement is necessary for the
   proper functioning of the service.  The result becomes a distributed
   application, in which the end-to-end principle applies to each
   connection involved in implementing the application.

5.  Internet Standards as an Arena for Conflict



   Internet standards have increasingly become an arena for conflict
   [10].  ISPs have certain concerns, businesses and government have
   others, and vendors of networking hardware and software still others.
   Often, these concerns conflict, and sometimes they conflict with the
   concerns of the end users.  For example, ISPs are reluctant to deploy
   interdomain QoS services because, among other reasons, every known
   instance creates a significant and easily exploited DoS/DDoS
   vulnerability.  However, some end users would like to have end-to-
   end, Diffserv or Intserv-style QoS available to improve support for
   voice and video multimedia applications between end nodes in
   different domains, as discussed by Huston in RFC 2990 [16].  In this
   case, the security, robustness, and reliability concerns of the ISP
   conflict with the desire of users for a different type of service.

   These conflicts will inevitably be reflected in the Internet
   architecture going forward.  Some of these conflicts are impossible
   to resolve on a technical level, and would not even be desirable,
   because they involve social and legal choices that the IETF is not
   empowered to make (for a counter argument in the area of privacy, see



Kempf & Austein              Informational                     [Page 10]

RFC 3724                  Future of End-to-End                March 2004


   Goldberg, et al. [17]).  But for those conflicts that do involve
   technical choices, the important properties of user choice and
   empowerment, reliability and integrity of end-to-end service,
   supporting trust and "good network citizen behavior," and fostering
   innovation in services should be the basis upon which resolution is
   made.  The conflict will then play out on the field of the resulting
   architecture.

6.  Conclusions



   The end-to-end principle continues to guide technical development of
   Internet standards, and remains as important today for the Internet
   architecture as in the past.  In many cases, unbundling of the end-
   to-end principle into its consequences leads to a distributed
   approach in which the end-to-end principle applies to interactions
   between the individual pieces of the application, while the unbundled
   consequences, protection of innovation, reliability, and robustness,
   apply to the entire application.  While the end-to-end principle
   originated as a focused argument about the need for the knowledge and
   assistance of end nodes to properly implement functions in a
   communication system, particular second order properties developed by
   the Internet as a result of the end-to-end principle have come to be
   recognized as being as important, if not more so, than the principle
   itself.  End user choice and empowerment, integrity of service,
   support for trust, and "good network citizen behavior" are all
   properties that have developed as a consequence of the end-to-end
   principle.  Recognizing these properties in a particular proposal for
   modifications to the Internet has become more important than before
   as the pressures to incorporate services into the network have
   increased.  Any proposal to incorporate services in the network
   should be weighed against these properties before proceeding.

7.  Acknowledgements



   Many of the ideas presented here originally appeared in the works of
   Dave Clark, John Wroclawski, Bob Braden, Karen Sollins, Marjory
   Blumenthal, and Dave Reed on forces currently influencing the
   evolution of the Internet.  The authors would particularly like to
   single out the work of Dave Clark, who was the original articulator
   of the end-to-end principle and who continues to inspire and guide
   the evolution of the Internet architecture, and John Wroclawski, with
   whom conversations during the development of this paper helped to
   clarify issues involving tussle and the Internet.








Kempf & Austein              Informational                     [Page 11]

RFC 3724                  Future of End-to-End                March 2004


8.  Security Considerations



   This document does not propose any new protocols, and therefore does
   not involve any security considerations in that sense.  However,
   throughout this document, there are discussions of the privacy and
   integrity issues and the architectural requirements created by those
   issues.

9.  Informative References



   [1]  Saltzer, J.H., Reed, D.P., and Clark, D.D., "End-to-End
        Arguments in System Design," ACM TOCS, Vol 2, Number 4, November
        1984, pp 277-288.

   [2]  Clark, D., "The Design Philosophy of the DARPA Internet
        Protocols," Proc SIGCOMM 88, ACM CCR Vol 18, Number 4, August
        1988, pp. 106-114.

   [3]  Blumenthal, M., Clark, D.D., "Rethinking the design of the
        Internet: The end-to-end arguments vs. the brave new world", ACM
        Transactions on Internet Technology, Vol. 1, No. 1, August 2001,
        pp 70-109.

   [4]  Postel, J., "Transmission Control Protocol", STD 7, RFC 793,
        September 1981.

   [5]  Floyd, S. and L. Daigle, "IAB Architectural and Policy
        Considerations for Open Pluggable Edge Services", RFC 3238,
        January 2002.

   [6]  Carpenter, B., Ed., "Architectural Principles of the Internet",
        RFC 1958, June 1996.

   [7]  Johnson, D., Perkins, C. and J. Arkko, "Mobility Support in
        IPv6", Work in Progress.

   [8]  Perkins, C., Ed., "IP Mobility Support for IPv4", RFC 3344,
        August 2002.

   [9]  Kaat, M., "Overview of 1999 IAB Network Layer Workshop," RFC
        2956, October 2000.

   [10] Clark, D.D., Wroclawski, J., Sollins, K., and Braden, B.,
        "Tussle in Cyberspace: Defining Tomorrow's Internet",
        Proceedings of Sigcomm 2002.

   [11] Carpenter, B. and S. Brim, "Middleboxes: Taxonomy and Issues",
        RFC 3234, February, 2002.



Kempf & Austein              Informational                     [Page 12]

RFC 3724                  Future of End-to-End                March 2004


   [12] Carpenter, B., "Internet Transparency", RFC 2775, February 2000.

   [13] Reed, D., "The End of the End-to-End Argument?",
        http://www.reed.com/dprframeweb/
        dprframe.asp?section=paper&fn=endofendtoend.html, April 2000.

   [14] Moors, T., "A Critical Review of End-to-end Arguments in System
        Design," Proc. 2000 IEEE International Conference on
        Communications, pp. 1214-1219, April, 2002.

   [15] Ramsdell, B., Ed., "S/MIME Version 3 Message Specification", RFC
        2633, June 1999.

   [16] Huston, G., "Next Steps for the IP QoS Architecture", RFC 2990,
        November 2000.

   [17] Goldberg, I., Wagner, D., and Brewer, E., "Privacy-enhancing
        technologies for the Internet," Proceedings of IEEE COMPCON 97,
        pp. 103-109, 1997.

10. Author Information



   Internet Architecture Board
   EMail:  iab@iab.org

   IAB Membership at time this document was completed:

      Bernard Aboba
      Harald Alvestrand
      Rob Austein
      Leslie Daigle
      Patrik Faltstrom
      Sally Floyd
      Jun-ichiro Itojun Hagino
      Mark Handley
      Geoff Huston
      Charlie Kaufman
      James Kempf
      Eric Rescorla
      Mike St. Johns











Kempf & Austein              Informational                     [Page 13]

RFC 3724                  Future of End-to-End                March 2004


11.  Full Copyright Statement



   Copyright (C) The Internet Society (2004).  This document is subject
   to the rights, licenses and restrictions contained in BCP 78 and
   except as set forth therein, the authors retain all their rights.

   This document and the information contained herein are provided on an
   "AS IS" basis and THE CONTRIBUTOR, THE ORGANIZATION HE/SHE REPRESENTS
   OR IS SPONSORED BY (IF ANY), THE INTERNET SOCIETY AND THE INTERNET
   ENGINEERING TASK FORCE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED,
   INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE
   INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED
   WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Intellectual Property



   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at
   http://www.ietf.org/ipr.

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at ietf-
   ipr@ietf.org.

Acknowledgement



   Funding for the RFC Editor function is currently provided by the
   Internet Society.









Kempf & Austein              Informational                     [Page 14]