Oh no! I sketched up a script to gzip the chunks, hashsum them, and then find out how many collisions there are before the real occurrence starting from an approximate address in the PI digits chain, so that I could have: ($address*1e12)$hash$collisioncount
The resulting string is 10% of size of the gzipped string, at the expense of CPU. But when I read you achieved 100% compression I just deleted the script and got out to get a beer. :-(((
The resulting string is 10% of size of the gzipped string, at the expense of CPU. But when I read you achieved 100% compression I just deleted the script and got out to get a beer. :-(((