home of the madduck/ blog/
Alternative, innovative search engines

These days (and for quite a while now), the populace equates "search" with Google. If you're like me, you'll also use alternatives. For instance, I've been quite happy with the clustered sorting provided by Clusty, which produces pretty good search results, but it isn't the fastest of all. Also, Eurekster's Swicki technology looks promising, and I am currently investigating whether it could make a good search platform for the Debian project.

What do you use as alternatives to Google? What I am after aren't answers like "Yahoo" or "A9", but rather more innovative search approaches. The days of index searches and even "Page Rank" are over, IMHO. What's next? Let's hear it, please!

NP: Proto-Kaw / Before Became After

Update: Fabian Greffrath pointed me to Metager2, which is a meta search engine (meaning it lets others do the search), which also inspects the pages in the search result set and thus promises better quality of the results. I could not yet find more technical details except that it's Linux/Python-based. It seems to be rather centred on German contents, and the way it displays linked pages in the search results consumes too much screen estate for me.

Update: Steve McIntyre contributed Trexy, which is also a meta engine that records your previous searches or lets you profit from other people's search "trails". I am not a big fan of meta searches, since those are really good at screwing up the sorting order by the engines they use, but I might try this one.

Update: Here's a list of 100 alternative search engines.

Update: Olivier Tétard mentioned Exalead, which feature thumbnails in the results, proximity search, and some other features (see its Wikipedia page).