-> Update: I checked something out and if we can get an algorithm as -> good as zip, then we can get about 46% compression. I have a big zip -> of a lot of .85s and .85g files, that has an average compression of -> that. That would make a savings of about 12k in the above example. I know of an easy to implement compression algorithm that's better than Deflate ( which is what ZIP uses ) and is not patended.