[zalas]

In case you haven't realized, it's using a pretty standard implementation of LZSS and the "key" you describe is simply a set of flags (one for each bit in the byte) telling the decompressor what types of token follow, whether they be verbatim byte-copy tokens or window backwards-lookup tokens. That being said, yes you can simply create a byte-stream that's all 'verbatim-copy' tokens, but if you want to learn how to write a real compressor for it, look up LZSS.