[freenet-chat] small worlds and time-correlation

Joel Salomon joelcsalomon at gmail.com
Sun Apr 30 19:13:34 UTC 2006


A thought regarding what "borg" on 0.7 Frost called "promiscuous peering":

Granted that you lose some anonymity to peers you are directly
connected to, you regain some deniability by being connected to many
non-trusted peers.

The real trouble seems to be that this behavior breaks 0.7's
small-world assumption --- or so I was warned.  However, I suspect
that most people trading node references on Frost or IRC are only
doing so until they get a certain number of them, then stop.  The
effect of this is to create a different sort of small-world network;
people you are connected to are likely to be connected to each other,
to a certain extent.  There may not always be a short path from one
node to another but there will be many parallel paths.

Does this much change how the network works?

--Joel


More information about the chat mailing list