Hello
As anyone heard or looked into something similar as
LZ differential, proposed by Jeremy de Bonet ?
http://www.debonet.com/Research/lzd/
The principles are well explained, but a single benchmark provided,
and unfortunately there is nothing else,
so that's really much too little.
The guy seems well above average when reading its other publications. so i don't doubt the results. However, if this was really interesting, someone else should have already noticed and re-use same principles in another algorithm ?