Add further ways to "build" library to reduce number of includes necessary - this is now partially mitigated by utf.php which works pretty well, Comparisons of utf_* fns vs. the PHP str* functions for edge behaviour (e.g. sub_str with positions outside of string)Benchmark different implementations of functions against each other Stuff to add Where functions accept an integer character count as an argument, would be good to have byte index versions as well - can be useful in some instances where you're mingling with native str functions Add the "romanize" stuff