see gnu.misc.discuss for my response to the flurry of questions about the relation of 'compress' to the lempel-ziv-welch algorithm patent promoted by sperry/unisys. the title is "Sperry's patent #4,558,302 does *not* affect 'compress'" ames!jaw