Return to NOTES CVS log

Up to [CSRG BSD Unix] / 43BSDReno / usr.bin / compress

Annotation of 43BSDReno/usr.bin/compress/NOTES, revision 1.1.1.1

1.1       root        1: From: James A. Woods <[email protected]>
                      2: 
                      3: >From vn Fri Dec  2 18:05:27 1988
                      4: Subject: Re: Looking for C source for RSA
                      5: Newsgroups: sci.crypt
                      6: 
                      7: # Illegitimi noncarborundum
                      8: 
                      9: Patents are a tar pit.
                     10: 
                     11: A good case can be made that most are just a license to sue, and nothing
                     12: is illegal until a patent is upheld in court.
                     13: 
                     14: For example, if you receive netnews by means other than 'nntp',
                     15: these very words are being modulated by 'compress',
                     16: a variation on the patented Lempel-Ziv-Welch algorithm.
                     17: 
                     18: Original Ziv-Lempel is patent number 4,464,650, and the more powerful
                     19: LZW method is #4,558,302.  Yet despite any similarities between 'compress'
                     20: and LZW (the public-domain 'compress' code was designed and given to the
                     21: world before the ink on the Welch patent was dry), no attorneys from Sperry
                     22: (the assignee) have asked you to unplug your Usenet connection.
                     23: 
                     24: Why?  I can't speak for them, but it is possible the claims are too broad,
                     25: or, just as bad, not broad enough.  ('compress' does things not mentioned
                     26: in the Welch patent.)  Maybe they realize that they can commercialize
                     27: LZW better by selling hardware implementations rather than by licensing
                     28: software.  Again, the LZW software delineated in the patent is *not*
                     29: the same as that of 'compress'.
                     30: 
                     31: At any rate, court-tested software patents are a different animal;
                     32: corporate patents in a portfolio are usually traded like baseball cards
                     33: to shut out small fry rather than actually be defended before
                     34: non-technical juries.  Perhaps RSA will undergo this test successfully,
                     35: although the grant to "exclude others from making, using, or selling"
                     36: the invention would then only apply to the U.S. (witness the 
                     37: Genentech patent of the TPA molecule in the U.S. but struck down
                     38: in Great Britain as too broad.)
                     39: 
                     40: The concept is still exotic for those who learned in school the rule of thumb
                     41: that one may patent "apparatus" but not an "idea".
                     42: Apparently this all changed in Diamond v. Diehr (1981) when the U. S. Supreme
                     43: Court reversed itself.  
                     44: 
                     45: Scholars should consult the excellent article in the Washington and Lee
                     46: Law Review (fall 1984, vol. 41, no. 4) by Anthony and Colwell for a
                     47: comprehensive survey of an area which will remain murky for some time.
                     48: 
                     49: Until the dust clears, how you approach ideas which are patented depends
                     50: on how paranoid you are of a legal onslaught.  Arbitrary?  Yes.  But
                     51: the patent bar the the CCPA (Court of Customs and Patent Appeals)
                     52: thanks you for any uncertainty as they, at least, stand to gain
                     53: from any trouble.
                     54: 
                     55: =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
                     56: From: James A. Woods <[email protected]>
                     57: Subject: Re: Looking for C source for RSA (actually 'compress' patents)
                     58: 
                     59:        In article <[email protected]> you write:
                     60:        >The concept is still exotic for those who learned in school the rule of thumb
                     61:        >that one may patent "apparatus" but not an "idea".
                     62: 
                     63: A rule of thumb that has never been completely valid, as any chemical
                     64: engineer can tell you.  (Chemical processes were among the earliest patents,
                     65: as I recall.)
                     66: 
                     67:        ah yes -- i date myself when relaying out-of-date advice from elderly
                     68:        attorneys who don't even specialize in patents.  one other interesting
                     69:        class of patents include the output of optical lens design programs,
                     70:        which yield formulae which can then fairly directly can be molded
                     71:        into glass.  although there are restrictions on patenting equations,
                     72:        the "embedded systems" seem to fly past the legal gauntlets.
                     73: 
                     74:        anyway, i'm still learning about intellectual property law after
                     75:        several conversations from a unisys (nee sperry) lawyer re 'compress'.
                     76: 
                     77:        it's more complicated than this, but they're letting (oral
                     78:        communication only) software versions of 'compress' slide
                     79:        as far as licensing fees go.  this includes 'arc', 'stuffit',
                     80:        and other commercial wrappers for 'compress'.  yet they are
                     81:        signing up licensees for hardware chips.  hewlett-packard
                     82:        supposedly has an active vlsi project, and unisys has
                     83:        board-level lzw-based tape controllers.  (to build lzw into
                     84:        a disk controller would be strange, as you'd have to build
                     85:        in a filesystem too!)
                     86: 
                     87:        it's byzantine
                     88:        that unisys is in a tiff with hp regarding the patents,
                     89:        after discovering some sort of "compress" button on some
                     90:        hp terminal product.  why?  well, professor abraham lempel jumped
                     91:        from being department chairman of computer science at technion in
                     92:        israel to sperry (where he got the first patent), but then to work
                     93:        at hewlett-packard on sabbatical.  the second welch patent
                     94:        is only weakly derivative of the first, so they want chip
                     95:        licenses and hp relented.  however, everyone agrees something
                     96:        like the current unix implementation is the way to go with
                     97:        software, so hp (and ucb) long ago asked spencer thomas and i to sign
                     98:        off on copyright permission (although they didn't need to, it being pd).
                     99:        lempel, hp, and unisys grumbles they can't make money off the
                    100:        software since a good free implementation (not the best --
                    101:        i have more ideas!) escaped via usenet.  (lempel's own pascal
                    102:        code was apparently horribly slow.)
                    103:        i don't follow the ibm 'arc' legal bickering; my impression
                    104:        is that the pc folks are making money off the archiver/wrapper
                    105:        look/feel of the thing [if ms-dos can be said to have a look and feel]. 
                    106: 
                    107:        now where is telebit with the compress firmware?  in a limbo
                    108:        netherworld, probably, with sperry still welcoming outfits
                    109:        to sign patent licenses, a common tactic to bring other small fry
                    110:        into the fold.  the guy who crammed 12-bit compess into the modem
                    111:        there left.  also what is transpiring with 'compress' and sys 5 rel 4?
                    112:        beats me, but if sperry got a hold of them on these issues,
                    113:        at&t would likely re-implement another algorithm if they
                    114:        thought 'compress' infringes.  needful to say, i don't think
                    115:        it does after the abovementioned legal conversation.
                    116:        my own beliefs on whether algorithms should be patentable at all
                    117:        change with the weather.  if the courts finally nail down
                    118:        patent protection for algorithms, academic publication in
                    119:        textbooks will be somewhat at odds with the engineering world,
                    120:        where the textbook codes will simply be a big tease to get
                    121:        money into the patent holder coffers...
                    122: 
                    123:        oh, if you implement lzw from the patent, you won't get
                    124:        good rates because it doesn't mention adaptive table reset,
                    125:        lack thereof being *the* serious deficiency of thomas' first version.
                    126: 
                    127:        now i know that patent law generally protects against independent
                    128:        re-invention (like the 'xor' hash function pleasantly mentioned
                    129:        in the patent [but not the paper]).
                    130:        but the upshot is that if anyone ever wanted to sue us,
                    131:        we're partially covered with
                    132:        independently-developed twists, plus the fact that some of us work
                    133:        in a bureacratic morass (as contractor to a public agency in my case).
                    134: 
                    135:        quite a mess, huh?  i've wanted to tell someone this stuff
                    136:        for a long time, for posterity if nothing else.
                    137: 
                    138: james 
                    139: 

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.