[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: Filtering out attacks?



On Tue, May 17, 2005 at 04:59:25PM +0100, Adam Langley wrote:
> Thus any application level filtering has to be network wide. The
> alternative is a non-deterministic "sometimes Google works, sometimes
> it doesn't" and that's a very bad user experience.

This is an interesting research problem.  Perhaps the Google experience
varies because Google is delivering content based upon IP address or to
which particular Google server a client connected.  Perhaps the Google
experience varies because certain ISPs between Tor exit nodes and Google
servers are filtering or modifying requests.  Perhaps the Google
experience varies because the Tor exit node has a personal firewall that
filters or modifies requests to Google.

In any of these cases, Tor, as an anonymizing network agnostic of which
exit nodes it chooses for a connection outside of published (or even
verified!) exit policy, is powerless to prevent these subtle differences
in content delivered.

Perhaps a potential solution would allow clients to make a more
well-informed decision about which exit nodes to choose when trying to
access particular resources.  I am not quite sure how this should be
implemented.  A naive way might be to have users try explicitly making
connections via certain exit nodes and publishing their success...

Geoff

Attachment: signature.asc
Description: Digital signature