Re: [freehaven-dev] Re: [Freenet-chat] MojoNation

[Roger Dingledine <arma@mit.edu>]

On Thu, Aug 10, 2000 at 11:43:26AM +0100, David Hopwood wrote:
> > * Free Haven requires publishers to provide reliable space of their own
> >   if they want to insert documents into the system. This economy of
> >   reputation tries to ensure that people donate to the system in
> >   proportion to how much space they use.
> 
> Another interesting model is Intermemory (www.intermemory.org), which has a
> space overhead of n times due to use of an information dispersal algorithm,
> but only allows each user to publish documents totalling an n'th of the space
> they contribute to the system. In practice n would be about 5. The problem
> with this is that it doesn't allow any way to benefit from the generosity of
> people who are willing to contribute space, but don't take advantage of
> their full publishing quota.
 
This is actually very similar to Free Haven's model. We also use
an IDA, and our n may well end up around 5 (we haven't done enough
research/simulations to decide on a good initial value).

In Free Haven, people who are willing to contribute space and don't
publish much generally become known as 'nice' and 'trustworthy' people
in the free haven trust network, since they're sometimes willing to make
uneven trades (such trades increase their reputation more strongly).

This would benefit them down the road if they wanted to publish, since
people would clamor to be the ones to make the trade and get a good word
from such a reputable server.

The reason why Free Haven has the added complexity of trading and
reputation metrics is because of the difficulty of accountability in
the face of the anonymity we want. Specifically, if a server drops a
piece of data on the ground, we want some way of realizing that and not
relying on that server as much in the future (and also not allowing it
to publish as much).

What form of accountability does intermemory provide to address this
issue?

--Roger