\chapter{Future Works} 

There are several areas in which this Trust Module falls short.  The
Free Haven group hopes to resolve most of them over the next few
months, as we work towards our first public release.  Others require
advances in the theory of cryptographic anonymity:

\begin{description}
\item[Trust Algebra] There is no good system for formally and
rigorously talking about trust and approval.  We need one in order to
be certain that the trust system does anything like what we are hoping
for.
\item[Entrapment Defense] The current belief in referrals is somewhat
na\"ive.  We need better support for logging various referrals and
integrating their results.  The best approach is beginning to look
like some form of linear algebra to deal with the loops of trust and
metatrust.
\item[Deployment] Many of the questions we have about Free Haven and
the Trust Module cannot be answered with thought experiments.  We need
to get the system deployed and working in a test environment to see
where it is attacked and where it breaks.  This test system needs to
be clearly marked as potentially unsafe.
\end{description}