Informative read, seems like the author learned some things and empirically discovered that his/her solution was ultimately a case of balancing trade-offs in ge

Informative read, seems like the author learned some things and empirically discovered that his/her solution was ultimately a case of balancing trade-offs in generality vs specificity. As stated in the "Conclusions" section toward the bottom of the page, "If you know the structure of your data, you can easily do a better and faster job of compressing than a generic compression algorithm."