The following text is copyright 1998 by Network World, permission is hearby given for reproduction, as long as attribution is given and this notice is included.

The need to support failure

By Scott Bradner

Network World, 10/05/98

This column is being written on a Saturday afternoon in San
Francisco while I'm waiting for the University Corporation for
Advanced Internet Development (UCAID) member meeting to start.
UCAID, better known by the name of one of its projects, Internet 2,
is meeting here to bring its members up to date on the many projects
that are now under way.

UCAID's basic mission is to help facilitate the development of the
next generation of Internet applications ( In
addition, using Project Abilene - a nationwide, high-speed network
that will soon interconnect many U.S. higher education institutions -
UCAID will create a proof-of-concept network that will support
emerging quality-of-service and multicast technologies.

UCAID and projects such as Project Abilene are funded by UCAID
member institutions and many major high-tech firms. In particular,
Qwest, Cisco and Nortel are providing a lot of support for Project
Abilene. There is no direct U.S. governmental funding going into any
of these activities. Instead, government activities in this general area
include projects such as the Very High-Speed Backbone Network
Service and the High Performance Connections Program, both funded
by the National Science Foundation (

UCAID represents exactly the kind of private support for continuing
technology development the U.S. government hopes to achieve when
it provides support for basic research.

Governments invested a few hundred million dollars into the Internet.
That investment, plus a few billion from universities and
corporations, has produced an infrastructure that is worth hundreds of
billions of dollars per year and is doubling every few months.

Success stories such as the Internet are a good thing, but we should
remember to remember all the history.

Corporations around the world provide significant support to
university researchers to help them develop technologies that could be
of use to corporations in the future. UCAID's Project Abilene is only
one of the more visible of the current projects - things tend to be
somewhat visible when the vice president of the U.S. is on hand to
announce the start of the project. But this sort of corporate support
tends to be late in the game. By that I mean that corporations tend to
support technologies that are getting close to mature. They cannot
generally afford to provide support early on for new technologies that
may fail.

Supporting failure, which I admit sounds funny, is a good role for
government. Venture capitalists assume a significant failure rate
among the start-ups they fund, but venture capitalists generally only
fund ideas well past the research stage. Someone needs to be ready to
fund ideas that may fail. The lesson of history is that we very much
need the government to keep funding early research, so the
government cannot back off just because the private sector is funding
projects such as Abilene.

Disclaimer: Failure and Harvard does not parse as a concept, and the
above are my own ramblings.