[tahoe-dev] P2P file-sharing protocol ideas

[news.gmane.org]

Hi,

> so I am unaware if Freenet's security has improved
> since Dhamija's analysis of it [1].

A newer (Freenet 0.7) attack is described here:
	http://gnunet.org/papers/pitchblack.pdf


> My only problem with GNUnet would be on account of its lack of usability

This asks for a (more or less) talented GUI developer, not for yet
another P2P protocol.


> (my opinion) and apparent complexity (about 100,000 lines of C, 5 times
> more than libtorrent).

That's right (at least for code written in C/C++). However, this number
includes two graphical frontends (gnunet-gtk and gnunet-qt), a
configuration manager with its frontends (gnunet-setup; GTK, Qt, curses
and text based) and a few modules that exist for testing purposes
(tbench, testing, tracekit) or are completely unrelated to distributed
storage (chat, vpn).
Since GNUnet is really a framework for secure P2P applications, the
functionality it provides is divided into several modules which can be
loaded or unloaded at will.


> Provided ACTIVE_MIGRATION is set to YES, it can
> behave in a manner similar to Freenet, caching encrypted DBlocks on your
> machine to help propagate them through the network.

Right, and local content is pushed out into the network over time. The
difference between GNUnet and Freenet in this regard is that GNUnet does
not rely on this mechanism. In GNUnet FS, the primary storage of
anonymously published content is the local PC. Active migration is only
there to increase availability and provide deniability.


Best,

Nils Durner
GNUnet developer