I know this is just a rehash ... but it seems boggling to me that you can have a random hash of data and have *any* significant compression. As far as I know, most established compression routines work somewhat like the following: every time it sees 001001001001001001001, for example, it replaces that with 'a' or something akin to that.

It seems that most compression routines use a cipher-esque hash and in this way crush data down.

All the same ... if this is true ... I wonder if the US will up its encryption standards?