Why bother implementing a completely separate search engine? In reality this only needs to be a simple program that does a reverse lookup on indices. A friend of mine already did an implementation of this as his Master's Degree Project.
I can ask him to subscribe to the list and discuss it. The downside is that his implementation is done in JavaScript in the browser. It works, but it's been decided in the past that JavaScript in Freenet pages is bad, and this disabled. Sadly, the only way around it is to have some kind of an even more crippled scripting mechanism exist in the node itself. As always, it's not possible to have a cake after eating it. IMNSHO, the way forward is not implementing applets on the node to do various one-off jobs. It would be immensly more useful to come up with a scripting engine (e.g. JavaScript with a handle on some features, but without other, purely harmful features) which could then be used for writing any kind of an interractive freesite. Gordan
