From: Susumu Harada (harada@cs.washington.edu)
Date: Sun Oct 10 2004 - 10:37:39 PDT
"Tussle in Cyberspace: Defining Tomorrow's Internet"
David D. Clark et al.
This paper presents the notion of "tussle" among Internet users, namely
the presence of diverse and sometimes conflicting user groups that drive
to shape the use of the Internet to meet their respective objectives. The
paper then presents several guiding principles on what factors should be
considered in the future evolution of the Internet to best accomodate for
this inevitable growth of the impact of the tussle.
The paper paints a fairly straightforward landscape of the various issues
where Internet tussles can and do exist, and offers a general set of
design principles for how one should address them.
However I find the paper to be lacking of any concrete proposal for how to
actually go about implementing the services and protocols to address the
various tussles. There are suggestions scattered throughout the paper,
but they are quite vague and do not provide much insight into how one
might actually implement the solution. For example, in addressing the
issue of tussle over user identity visibility, the author states "We
suggest that such a framework could usefully share and arbitrate
information across many layers of the protocol stack." This seems to be
an obvious statement that doesn't invite any counterargument but at the
same time provides very little substance for guiding network engineers.
Another example is in the section on competitive wide area access, where
the author proposes "The Internet should support a mechanism for choice of
source routing that would permit a customer to control the path of his
packets at the level of providers." Sure, that is a very useful idea, but
the issue of how to go about implementing such a mechanism appears to be a
whole new research topic on its own.
I did find the notion of modularization along tussle boundaries to be a
useful way of looking at the problem of competing demands and the
framework within which one should design the solution. However all in
all, I would have liked to see more concrete proposals for addressing each
of the tussle spaces than just oversweeping suggestions.
This archive was generated by hypermail 2.1.6 : Sun Oct 10 2004 - 10:37:40 PDT