Tuesday, February 15, 2011

[xplgmgyz] Web search engine behind Tor

Create a search engine that spiders the public internet through Tor (The Onion Router), and serves its results from a hidden service within the onion network (".onion" pseudo top level domain).  This should be straightforward, though it will be an intense test of whether Tor can maintain anonymity.  As a side benefit, it can index hidden services, too.

The goal is to create a search engine whose operators remain anonymous; they cannot be threatened by lawsuit or criminal prosecution to remove items from the search results (as Google faces constantly).

Running a search engine takes lots of hardware and is expensive.  Trying to recoup costs while remaining anonymous seems close to impossible (numbered Swiss bank accounts?), so the next best thing is to spread the cost among many unrelated people who believe such a service is a good idea.

Affero GPL somewhat to the rescue: the back end software is published (probably again at an onion site because of MGM vs Grokster) so many people can run search engines, each indexing only a small portion of the web.  A user's user agent aggregates the results from many engines.

No comments :