On Thu, May 10, 2001 at 05:28:57AM -0500, Steven Hazel wrote:
> Dev Random <[EMAIL PROTECTED]> writes:
>
> > Hi all,
> >
> > (Note, I'm crossposting to devl and tech lists - please reply to tech...)
> >
> > I've been thinking about the issues surrounding file sharing and anonymous
> > communication in general. One difficult issue with Freenet is that merely
> > running the protocol on the Internet may open people to attack in some
> > jurisdictions. Additionally, file sharing is currently a controversial legal
> > topic (e.g. Napster trial).
> >
> > I would like to propose a solution to this issue - a layer between
> > Freenet and TCP/IP. This layer would be encrypted, anonymous and
> > reusable. Such an infrastructure would have general applicability and
> > would have "substantial non-infringing uses". If such a layer is widely
> > adopted, it would shield controversial algorithms from direct attack.
>
> ...you mean something like SSL? I think Freenet-over-SSL is a great
> idea. If I had my way, SSL would replace 90% of the connection-layer
> crypto, unless it turned out to really be too slow for our needs.
>
> But, failing that, a Freenet connection-layer crypto RFC and library
> for use in other applications would be pretty good.
SSL is nice, since a naive scan would say "he is running SSL, and that is not a
protocol we care about". But a sophisticated party would connect by SSL and
then start talking Freenet and then attack you (through legal means and
otherwise) when you respond correctly.
My proposal anonymizes both client and servers. Nobody would be able to
prove that an IP address is running any particular protocol (except the
R-Net protocol itself, which has general applicability).
PGP signature
