Hi! Right now, I have my PrintTok2 program doing very well, in fact, exceptional, but I want to do better. Its main compressor is based on the technique used by somebody on this forum--his username is Toldo. He compresses whole words into 12-bit tokens and has ways to assume spaces between words. I added options to not compress rarely-used tokens, compress lower-case letters using a version of Static Huffman Codes, use a bit to determine whether a space follows a period or comma or not and shorten tokens to as many bits as are needed. I'm looking for other ideas. Does anybody here have any?