Google just recently announced that the US Federal Trade Commission has begun a review of their business. Possible anti-competitive activities aside, there are much bigger problems pertaining Google’s monopoly position in Web searches. It’s a well known fact that Google stores search requests for purposes of profiling its users. These profiles are mainly used for targeted advertising, but could also be used for far more sinister purposes such as extracting information on a user’s political views or sexual orientation. Google has openly stated that it is willing to share its data with the US authorities.
Another well known fact is that Google censors its search results, and not just in China. They have a history of co-operating with any government to uphold local censorship laws. Due to its monopoly position, Google is able to a great extent to determine what Web users are allowed to see and what they aren’t.
Naturally, anyone worried about their privacy and freedom will spread their searches over several search engines. However, all the major search engines are run by private companies using proprietary software. There’s little reason to trust companies like Microsoft or Yahoo any more than Google. How to search the Web without succumbing to the whims of major corporations?
One genuine alternative to proprietary search engines is YaCy. What makes YaCy different is that it doesn’t have a central server, but is fully decentralized into a network of nodes run on private computers all over the world using peer-to-peer technology. All search requests are distributed over the network (which doesn’t store them), which makes it extremely difficult to gather data for profiling purposes. Also, since the search index is also distributed, global filtering of search results is practically impossible. YaCy is Free software, so anyone cun run a node. The program runs in the background as a daemon, servicing search requests and crawling the Web to add URLs and search terms into the local database which is then shared to the network.
Naturally the problem with YaCy is it’s relative obscurity: there are only a couple of hundred active nodes at the moment. However, the size of the global search index is already over one billion URLs (the estimated size of Google’s index is a little under 50 billion URLs). I’d estimate that with no more than a few thousand nodes, YaCy could already prove a serious challenge to the proprietary search engines. Therefore, I warmly recommend anyone with some clock cycles and memory to spare to run their own node. The program is written in Java, so it should run on most operating systems.