Talking about the way Grooveshark structure their data could be an interesting topic.
I would be surprised to test how much time it would take to find some playlists containing a song even with a trivial algorithm. The reason I'm saying that is that it would be very fast for popular songs to find a few playlists (We don't need 1000 results, just a few is great for our case). And for the less known songs, I'm happy to wait for quality playlists! ;)
I would be surprised to test how much time it would take to find some playlists containing a song even with a trivial algorithm. The reason I'm saying that is that it would be very fast for popular songs to find a few playlists (We don't need 1000 results, just a few is great for our case). And for the less known songs, I'm happy to wait for quality playlists! ;)