PCT (Patents, Copyright, Trademark) policy and Open Source

[Russell McOrmond]

On Wed, 28 Jan 2004, Alexander Terekhov wrote:

> Robert Osfield wrote:
> [...]
> > vulnerabilities and risks to our livelihood.
> 
> If you don't intend to eliminate all IP laws

  Please stop trying to prove Richard Stallman correct by abusing the term
"Intellectual Property" to suggest that you are "either for IP, or against
IP".
  http://www.fsf.org/philosophy/words-to-avoid.html#IntellectualProperty

  Patents, Copyrights and Trademarks are self contradictory (if you offer
more "protection" to one group you are taking that protection *AWAY* from
another) and are a balance between competing interests.  Sometimes, just
sometimes, the public interest is considered in this area of public policy
(but unfortunately not as often as it should be).

  Patents and Copyrights are to inventors and creators like water is to
humans:  too little and you dehydrate and die, and too much and you drown 
and die.  Only with the right balance can we survive.

  Hopefully most people will see through attempts to drown us, and instead
work with us to try to create the right balance between competing
interests that best serves the entire software sector ("software
manufacturing" and FLOSS) as well as the public good.

> (as an ultimate solution to the problem of vulnerabilities and risks),
> then something like www.pubpat.org is the way to go, I think.


  This is like thinking that the solution to nuclear weapons in the hands
of "bad" people is to have more nuclear weapons in the hands of "good"
people (recognizing of course that who is "bad" and who is "good" is an
extremely subjective and political determination).

  The problem is that "patent pools" and expensive fights against the
extremely high number of poor quality patents in "information processes"
only works for organizations with a lot of money (Like IBM) or some sort
of benefactors of those monied special interests.  Threats and chills on
innovation are not reduced by this, and the only real solution is
non-proliferation treaties to try to rid us of the problem in the first
place.

  http://www.pubpat.org/ is not a solution to the "information process"  
patent problem, but I would agree it will help in subject matter areas
where patent policy is mostly helpful but poor patents still get in.

  The first part of recognizing the problem we are trying to discuss here
is to recognize that those who are against "information process" patents
are not necessarily against patents in other sucject matter.  We just
believe that the logic and justification behind patent policy fails in
some of the recently expanded subject matter areas.  Subject-matter
independent economic analysis is needed, and this analysis is currently
not being done.

  Those who support patents in other subject matter should be helping us
with these subject-matter independent analysis as the integrity of the
entire patent system is being questioned because of improper expansion of
patent policy into bad subject matter areas.


  Try this out: Offer the software community the justifications for
patents, and we will explain why this doesn't apply to "information
processes" like software.


Here are just two of the most common ones and one possible response:

  Justification: the alternative to patents is trade secrets where we
never learn about the invention.  Patents get inventors to disclose the
invention so that it is available to the public after the term of the
patent, in exchange for a temporary monopoly.

  Response: distributed software is already published and not a trade
secret. Whether the software is Open Source or not, the right to reverse
engineer to create compatible products means that the 'invention' is never
secret.  In the case of many existing software patents the distributed
software (disassembled) still provides more disclosure of the invention
than the legalistic wording of the patents do.


  Non-distributed software embedded in internal processes (such as
software that controls a robot to manufacture something) is a very
separate situation.  In this case if the process is patentable it is
patentable regardless of the existence of the software ("software neither
adds to nor subtracts from the patentability of an invention").  That does
not mean that the software alone taken outside of the context of that
manufacturing process should be patentable -- that is what the "per say"  
talks about, and the concept is quite simple even if obfusticated by those
who wish to expand patent policy to information processes.



  Justification: long expensive up-front R&D needs temporary monopoly to
recoup capital costs.

  Response: I think FLOSS and small-business "software manufacturing"  
software authors (including shareware and freeware) have always tossed
this idea out the door.  Network effects and first mover advantage are
often far more effective than a patent can be in gaining market share, and
these effects apply equally in a "free market" sense to the entire sector
and not just the largest players.

  As we move from capital-intensive speculative supply-side software
development to incremental non-speculative demand-side software
development the lack of a need for capital for the software creation will
be realized. I consider this to be a basic part of the maturing of the
software sector, and something where old-economy thinking is slowing down
this maturity.

  There may be some software that doesn't get developed if high-risk R&D
doesn't receve the temporary monopoly.  I think the number of cases this
applies to is small compared to the innovation that is not happening today
because of the high-risk of patent infringement lawsuits due to extremely
poor quality information process patents.  I see no evidence at all that 
software patents are reducing risk for expensive R&D, and all evidence 
that patents in this subject matter area are causing the opposite.


  This is not to say that venture capital will not be needed for
non-software aspects of a business, but that it will not be needed for the
software development aspects of the business.


Note: I read in the book "OPEN INNOVATION" by Henry Chesbrough that IBM
has moved away from old-economy thinking for their hardware innovation by
opening hardware patents to licensing.  In fact Henry dedicated an entire
chapter "From Closed to Open Innovation" to IBM's moves here.  The
equivalent move for software and other information processes would be to
lobby for governments to abandon information process patents entirely
given the extremely different market dynamic that exists between hardware
and software.



..etc...etc..

  We could go on and on through the justifications for patents in a given
subject matter and all the arguments why they don't apply to software or
other information processes.  This same analysis suggests that patent
policy is still needed in manufacturing (such as IBM's hardware
manufacturing innovations), pharmaceuticals and other areas -- but what is
good for one subject matter is not necessarily good for another.


  There is an additional slide presentation referenced from
http://www.flora.ca/patent2003/ titled "Interest of software patents for a
small software publisher," by Gérald-Sédrati-Dinet which is quite
worthwhile to browse through.

> regards,
> alexander.
---
 Russell McOrmond, Internet Consultant: <http://www.flora.ca/> 
 Governance software that controls ICT, automates government policy, or
 electronically counts votes, shouldn't be bought any more than 
 politicians should be bought.  -- http://www.flora.ca/russell/

--
license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3