jim@baroque.Stanford.EDU (James Helman) (08/02/90)
Back in February of 1989, Bob Lucky of AT&T gave a talk here (part of Allison's EE380 series) about data compression. Lucky concluded with a woolly tale of a bloody PC shareware war between Tom Henderson of New Jersey and Phil Katz of Wisconsin. I'd forgotten some of the details, so I reviewed the video tape. Lucky spins a good yarn. It's someplace between damn funny and tragic. The story reminds me of a cartoon with two little mice running around sniping at each other, and then the big mean cat comes in and kicks both their butts. With today's events, it seems the big cat is making his entrance, although how mean he is remains to be seen. From my notes of Lucky's talk: After Terry Welch wrote an easily understandable article for IEEE Computer in 1984, lots of people started writing LZW compression programs. One of them was Tom Henderson, who developed a shareware compress program called ARC based on LZW. ARC really took off, gaining widesread use, praise and shareware revenues. Then along comes one Phil Katz. He develops a version that runs about 10 X faster and calls it PKARC. Everyone starts using PKARC because it's so much faster. Henderson's income drops. He sues Katz (apparently based upon his trademark of the name ARC). Katz has no stomach for lawsuits and agrees to cease and desist. Henderson orders that all copies of PKARC be given up for destruction. Users, not wanting to go back to something 10X slower, rebel with comments such as "Like my GUN, they will have to pry PKARC from my cold, dead fingers!." [Perhaps he meant GNU? ;-> - jlh]. Katz has second thoughts and to get around the ARC trademark renames his program PKPACK. Henderson sues Katz for violating of the consent agreement. Henderson's growing unpopularity moves him into partial retreat. He makes ARC freely available for "all sentient beings in the universe" to use (except Phil Katz!) and states that ARC is available for royalty free incorporation into commercial products up to $100,00 profit, above that Henderson requested a 6.5% cut. He also agreed to maintain the ARC standard for the world and required that he be given full source code to any product which incorporated ARC compression!! Lucky's closing comments were: We [AT&T] apparently do have a patent on this [LZ], but nobody thinks it's worth anything.... I shouldn't say that, maybe some people do, I don't know. I don't know. But you want to know the bottom line, we bought a license, WE BOUGHT A LICENSE, from Tom Henderson. [Audience Laughter] Q. One wonders if Henderson can afford the legal fees either. Lucky: Well you see he's playing. He may be bluffing. He's playing poker. He threatens you with a suit. He may or may not have legality on his side. But it's a very powerful threat. So powerful that AT&T bought a license!. The battle was still raging when Lucky gave his talk last year. Does anyone know what's happened since to Henderson|Katz|ARC|PKARC|PKPACK? Jim Helman Department of Applied Physics Durand 012 Stanford University FAX: (415) 725-3377 (jim@KAOS.stanford.edu) Voice: (415) 723-9127
imp@dancer.Solbourne.COM (Warner Losh) (08/02/90)
In article <JIM.90Aug1125737@baroque.Stanford.EDU> jim@baroque.Stanford.EDU (James Helman) writes: >The battle was still raging when Lucky gave his talk last year. Does >anyone know what's happened since to Henderson|Katz|ARC|PKARC|PKPACK? Phil Katz stopped working on PKPACK. He then produced PKZIP which had all kinds of tm's in it. It is faster than ARC, does lots better at compression, and seems to be better built. Most places have dropped arc like a hot rock, and have adapted ZIP as the standard. -- Warner Losh imp@Solbourne.COM
karn@envy.bellcore.com (Phil R. Karn) (08/03/90)
PKZIP also makes use of a new compression scheme Katz calls "implosion", and it seems to do much better than his LZW implementation in PKARC. It even seems to outdo tar+compress under UNIX, which is remarkable given the PC's limited memory and PKZIP's compression of individual files. Does anybody know how "implosion" works, and whether it might be an alternative to LZW? It would be pretty satisfying to tell Unisys to take a hike because we've found a more efficient algorithm that's public domain... Phil
cowan@marob.masa.com (John Cowan) (08/07/90)
In article <JIM.90Aug1125737@baroque.Stanford.EDU>, jim@baroque.Stanford.EDU (James Helman) writes: >After Terry Welch wrote an easily understandable article for IEEE >Computer in 1984, lots of people started writing LZW compression >programs. One of them was Tom Henderson, who developed a shareware Minor nit: Henderson's name is Thom not Tom. Disclaimers: I sympathize with Henderson, and this bias will be evident in what follows. However, I have no connection with Henderson or his company, Systems Enhancement Associates, except as satisfied customer. >compress program called ARC based on LZW. ARC really took off, >gaining widesread use, praise and shareware revenues. ARC already existed before LZW; it used Huffman encoding. The LZW-smart version actually computed the size of the Huffman code on the fly as it was writing out the LZW data, and then if the Huffman version was smaller, it was written into the archive instead. Note for Unix weenies: ARC, like Zoo and similar programs, archives as well as compressing. Unlike tar.Z format, the archive directory information is NOT compressed. This allows different files in the archive to be compressed in different ways, with a type code in the file header to say which. This becomes important later. >Then along comes one Phil Katz. He develops a version that runs about >10 X faster and calls it PKARC. Everyone starts using PKARC because >it's so much faster. Henderson's income drops. He sues Katz >(apparently based upon his trademark of the name ARC). > >Katz has no stomach for lawsuits and agrees to cease and desist. >Henderson orders that all copies of PKARC be given up for destruction. >Users, not wanting to go back to something 10X slower, rebel with >comments such as "Like my GUN, they will have to pry PKARC from my >cold, dead fingers!." [Perhaps he meant GNU? ;-> - jlh]. This summary omits one of the main reasons Henderson decided to sue, as opposed to the legal grounds on which his suit was based. Katz released a version of PKARC which emitted >incompatible< archives that could not be processed by ARC. Specifically, PKARC added a new compression type which under some circumstances created slightly smaller archive members. This is independent from PKARC's speed advantage, which (supposedly -- details are sealed) was based on the use of judicious 8088 assembly programming in PKARC. Since the PKARC archives could not be processed by Henderson's ARC but were not otherwise distinguishable, ARC got the unfair reputation of being unable to unarchive its own output. Katz has stated, BTW, that he added the new compression method in order to beat the 13-bit LZW compression employed by Zoo. >Katz has second thoughts and to get around the ARC trademark renames >his program PKPACK. Henderson sues Katz for violating of the consent >agreement. Henderson lost this second lawsuit. >Henderson's growing unpopularity moves him into partial retreat. He >makes ARC freely available for "all sentient beings in the universe" >to use (except Phil Katz!) and states that ARC is available for >royalty free incorporation into commercial products up to $100,00 >profit, above that Henderson requested a 6.5% cut. He also agreed to >maintain the ARC standard for the world and required that he be given >full source code to any product which incorporated ARC compression!! In addition, Henderson released a new version of ARC which includes the fast PKARC product, now renamed QARC. The original ARC product is also provided, and has been updated to recognize and understand PKARC-type archives. However, neither ARC nor QARC will generate them. Katz wrote an entirely new archiver, with new formats, called PKZIP. This uses LZW compression as a fallback, but also makes use of arithmetic encoding and (in the latest version) LHARC-type LZSS+Huffman compression. Reasonable peace and quiet now prevails. Archivers in general use include ARC, Zoo, PKZIP, LHARC, and even 16-bit compress for MS-DOS. -- cowan@marob.masa.com (aka ...!hombre!marob!cowan) e'osai ko sarji la lojban
williams@umaxc.weeg.uiowa.edu (Kent Williams) (08/07/90)
In article <26BDD1B6.242B@marob.masa.com> cowan@marob.masa.com (John Cowan) writes: >In article <JIM.90Aug1125737@baroque.Stanford.EDU>, > jim@baroque.Stanford.EDU (James Helman) writes: >>After Terry Welch wrote an easily understandable article for IEEE >>Computer in 1984, lots of people started writing LZW compression >>programs. One of them was Tom Henderson, who developed a shareware > >Minor nit: Henderson's name is Thom not Tom. >Disclaimers: I sympathize with Henderson, and this bias will be evident >in what follows. However, I have no connection with Henderson or his >company, Systems Enhancement Associates, except as satisfied customer. > MAJOR NIT: I wrote the lzw compression program that Thom Henderson originally used in ARC, and he adapted it with my permission. Not only that, I sent him a copy of the Unix compress source, which he also adapted for use in ARC. There are actually four variants of LZW compress supported by ARC. 1) The code I originally sent him. 2) The same code with a quicker hash function 3) 12-bit Unix compress and 4) 13-bit unix compress. I am credited in the documentation for ARC, though my contribution isn't described. I haven't talked to Thom for a few years. I basically sympathized with PKWARE on the suit issue, but Thom always went to great pains to make sure everyone knew that he held ARC as his property. PKWARE started to make a lot of money off of something Henderson never put in the public domain. SEA and Henderson have been widely villified in the PC community -- I think unfairly. He provided a valuable utility at the right time, RELEASED COMPLETE SOURCES, and just asked for voluntary payment. I don't see how anyone -- even Philip Kearn (sp?) -- could complain too much about the way things were settled. He was goaded into writing PKZIP, which is a much superior program. -- Kent Williams 'Look, I can understand "teenage mutant ninja williams@umaxc.weeg.uiowa.edu turtles", but I can't understand "mutually williams@herky.cs.uiowa.edu recursive inline functions".' - Paul Chisholm
huggins@zip.eecs.umich.edu (James K. Huggins) (08/08/90)
So, what's the legal status of PKARC at this point? I understand that it's no longer in production. I believe I've got a copy at home somewhere -- should I destroy it, or is it ok to keep around? Enquiring minds want to do the right thing while mixing metaphors. Jim Huggins, Univ. of Michigan