Re: [freenet-chat] Crazy idea: How trust in darknets enables secure democratic censorship

Matthew Toseland <toad-EI5O+8PHWbJeeLb3ft/[email protected]> wrote:

> > I assume that the inserting node gets more blame than the surrounding
> > nodes and that blame does not stick around forever. Maybe for a very
> > long time, but definitely not forever. 
> > 
> > So I can surround my "inserter" with one (as above) or more layers of
> > nodes which act as "blame absorbers".
> > 
> > When I insert evil stuff my "inserter" will be blamed for it. The nodes
> > next to it will get some fraction of the blame, but definitely less.
> > the nodes on the outer layer will get nearly no blame. 
> 
> They will get no blame, unless they oppose the complaint.

They could even support the complaint. The "inserter" is expendable.

> > Before the nodes on my outer layer get blamed for talking to evil nodes
> > I shutdown the core of my little "evilnet" and replace it with new
> > nodes.
> 
> If complaints are repeatedly held up against new nodes connected to a
> specific small number of nodes, it should be possible for the community
> to notice this and take action against _them_. 

They are expendable unless they belong to the outer layer.

The inner layers can be replaced every two weeks, the outer layers every
year or so. Obviously I have to keep the outermost layer.

> The network topology has to be open for premix routing to work, and the
> node which was punished is revealed when a complaint is upheld.

How long would it take for the community to identify the outer layer of
"evilnet"?
Even a single person should be able to protect the "inserter" with 4-6
layers of "blame absorbers".