Much to expensive. Bittorrent is still burning bandwidth like crazy. All that is needed is a desktop client with a (mobile) web gui. Your own crawler that inde

Much to expensive. Bittorrent is still burning bandwidth like crazy. All that is needed is a desktop client with a (mobile) web gui. Your own crawler that indexes stuff however aggressively you like. A list of good nodes/friends. And a discoverable html tag embedded in web pages so that you can subscribe to the site search engine API. (using a slightly different protocol) Giving the favicon.ico a fixed path was a good idea. Sometimes looking for /feed.rss or /rss.xml works but it offers no guarantees. Lets do /site-search.xml or /search.json or /search-project-name-here/index.bla?q=%s The only challenge is ranking results. If you can figure that out the rest is easy. (after a few days a yacy instance already becomes amazingly useful with very little data - try it)