Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: Patents Keep Dying in the US, Interview with Mark Webbink

  • Subject: Re: Patents Keep Dying in the US, Interview with Mark Webbink
  • From: Rex Ballard <rex.ballard@xxxxxxxxx>
  • Date: Mon, 2 Feb 2009 21:16:56 -0800 (PST)
  • Complaints-to: groups-abuse@xxxxxxxxxx
  • Injection-info: z6g2000pre.googlegroups.com; posting-host=67.80.109.118; posting-account=-EkKmgkAAAAxynpkobsxB1sKy9YeqcqI
  • Newsgroups: comp.os.linux.advocacy
  • Organization: http://groups.google.com
  • References: <6200717.O6gvPHCdJE@xxxxxxxxxxxxxxx>
  • User-agent: G2/1.0
  • Xref: ellandroad.demon.co.uk comp.os.linux.advocacy:732341

On Feb 2, 10:00 pm, Roy Schestowitz <newsgro...@xxxxxxxxxxxxxxx>
wrote:

> The Supreme Court and Software Patents

> ,----[ Quote ]
> | As I noted at the time it was decided, people care about Bilski largely
> | because of what it says about legality of software patents. Software patents
> | are intensely controversial, with many geeks arguing that the software
> | industry would be better off without them. What I found striking about the
> | conversation was that both guests (and perhaps the host, although he didn't
> | tip his hand as much) took it as self-evident that there needed to be patents
> | on software and business methods
> `----
> http://www.freedom-to-tinker.com/blog/tblee/supreme-court-and-softwar...

There are three fundamental problems with software and business
process patents.  Almost since the conception of patent law, the
founding fathers wanted to create a distinction between tangible
reproducible objects, things that could be manufactured by one person
who put a great deal of research and expense into perfecting a device
such as a watch or a cotton gin, and then finds himself in direct
competition by a well-financed producer who hasn't spent any effort or
money on research and then produces identical devices even though he
may not know exactly what it does or how it works.

At the same time, the Founding Fathers also did NOT want to deal with
patents on algorythms such as how to bake a cake, or how to make a
basket.  They wanted to promote the free exchange of ideas without
fear of reprisals related to property rights.  Even Ford's assembly
line was a patent on the factory itself, not the algorythms for mass
production.

Thomas Edison was notorious for his abuse of patents, chasing film
producers from Rochester New York to California, having hired thugs
demolish cameras and projectors if the producers didn't give Edison
the royalties he demanded (which were steep).  More important was the
way he protected his patents, working desparately to prevent the
adoption of Alternating Current, which eliminated the need for a
governer used to synchronize DC generators.

Until about 1983, most software could not be patented.  It was only
when RSA filed for a patent on their encryption - one device entirely
implemented in hardware, and the other implemented entirely in
software, that the judge decided that both "black boxes" could be
patented.

Around 1994, Republicans decided that ALL software could be patented,
since software was just a bunch of black boxes connected together.
This was especally true of UNIX.  In fact, again, it was getting hard
to tell what was hardware and what was software in things like
telephone systems.

Even further blurring the lines were programmable gate arrays in which
simple gates could be programmend into a device much the way you
program an eprom or flash memory.

The problem with patenting software was that it was very hard to tell
when an idea was "original".  The problem was that software
development had been going on for almost a century since the first
tabulating machines were created for the Census, to the computers used
during World War II to decode Hitler's enigma to the computers used by
NASA to put a man on the moon, probes on mars, and even send probes to
the edge of the solar system.

There are probably 10 billion lines of code in public domain,
declassified software written for DOD, ARPA, National Science
Foundation, and hundreds of others, as well as algorythms implemented
in hardware due to the inferior components of 20, 40, and even 60
years ago (today nuclear devices need 1/10th the radioactive material,
1/20th the explosives, and can be much lighter thanks to high high
speed computer controlled switching systems that can provide
nanosecond accuracy (Little Boy and Fat Man had timing accurate to
about 1/20th of a second controlled by vacuum tubes.  Radar systems
that used vacuum tubes are now microprocessor controlled and use
multiple antennas, but the core technology was the same as 50 years
ago.

The problem with filing a software patent today is that it's a bit
like going to the dog pound, picking up a 5th generation mutt, and
attempting to provide the exact pedigree all the way back to AKC
registered dogs.

The problem, especially for companies with deep pockets to be picked,
is that there are patent lawyers who will file patent applications on
behalf of someone who claims an invention is entirely their original
work, and then attempt to file lawsuits against companies like IBM,
Microsoft, and other major software companies, demanding huge
royalties.  Even worse, the lawyers try to pack the jury full of
people who will be sympathetic to the "Little Guy" if they can get the
case to the Jury.  Microsoft lost $200 million in a lawsuit based on a
patent of technology that was based on GPL'd prior art.

Microsoft has about 300 patents on technology that could easily trace
back to Richard Stallman himself, as well as other GPL and BSD
software, but they don't dare NOT patent it because some ambulance
chaser turned patent lawyer will file a $75 application and then try
to sue based on a trivial search of ONLY software listed in the patent
archives, based on terminology used in the patent - which is
deliberately and deceptively different from terminology used by the
GPL, BSD, and Public Domain prior art.  Microsoft has done similar
"terminology substitutions" for decades, mainly to use copyrights and
trademarks based on the new terminology - even though the baseline
code might have been based on BSD or GPL code.

IBM filed something like 10,000 patent applications listing about
300,000 items of prior art, and was awarded almost 7,000 of them
anyway.  Most of the software patents were defensive patents - in some
cases, more focused on protecting almost 70 years of their own prior
art ranging from queue algolrythms to memory management, from patent
trolls.

The problem is that all of those software patents cost a lot of
money.  It can cost a few hundred thousand dollars to do a
comprehensive patent search for ALL prior art, and to do due diligence
in really doing proper research.  The ambulance chaser can get away
with doing a $200 google patent search for an hour and claim "nope, no
prior art", but IBM could get sued for filing fraudulent patent
applications if they didn't do a much more substantial search of ALL
prior art, including their own.

Grady Booch, now a technical fellow with IBM has been doing a seminar
on "IT Archeology" - discussing the importance of getting ALL of the
products and byproducts of every IT project archived and documented.
A key element of this is protecting yourself from patent trolls.

Software patents have turned out to be a huge expense to file
defensive patents that you can't enforce, primarily to protect
yourself from the patent trolls.

[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index