Xref: utzoo comp.arch:17362 comp.unix.wizards:23083 Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!usc!snorkelwacker!ai-lab!life!mtr From: mtr@geech.ai.mit.edu (Michael Rowan) Newsgroups: comp.arch,comp.unix.wizards Subject: Compress, LZW and LZ... (Was Re: Patents and Architecture) Message-ID: Date: 25 Jul 90 20:30:47 GMT Sender: news@wheaties.ai.mit.edu Organization: The Free Software Foundation Lines: 24 In-reply-to: wjw@eb.ele.tue.nl's message of 25 Jul 90 07:25:06 GMT [ disclaimer: if this is the wrong place for this post blah blah blah...] It turns out that the LZ and LZW (Lempel-Ziv and Lempel-Ziv & Welch) algorithms, that are used compress, are patented. Unisys Corp holds the patent and are currently making money from license fees (it is being used in current modem technology.) This patent is on the algorithm, not on code (unlike a copyright) so any program that produces compressed or uncompresses files using LZ or LZW are covered under the patent. Unisys is actively pursuing the licensing of compress. So... FSF is looking for alternative algorithms to use in lieu of compress. We were told of a new algorithm recently published that beats LZ and LZW on compression rates. Does anyone know anything about this? Also someone mentioned Shannon-Fano trees; does anyone have any references to this? Any information that is relevant to this would be highly appreciated. Drop me a line with anything at all related to this, as I don't read news that often (If I did, I wouldn't get *any* hacking in...) mtr