Subject: Re: Why software patents are bad
From: Bernard Lang <Bernard.Lang@inria.fr>
Date: Sat, 18 Sep 1999 21:46:06 +0200

On Sat, Sep 18, 1999 at 12:03:10PM -0700, L. Peter Deutsch wrote:
> > Copyright protection was originaly granted as an _exchange_ for the work
> > eventually going into the public domain.
> 
> Exactly the same is true of patents.

Not always ... but basically the argument for patents is public interest,
and there essentially 3 issues:

 1- encouraging people to find innovating techniques
 2- having inventors make their discoveries public
 3- encouraging investment to bring innovation to market

  Number 1 and 3 require investment
  Number 2 requires forfeiting the exclusivity one would have
    if the technique were kept secret

  The granting of a patent based monopoly is intended to encourage 1
and ofset the cost of 1, 2 and 3. Granting a monopoly is itself a cost
for society that forfeits the optimal use of some riches for a given
period of time.
 
 However, from the pure point of view of the economic interest of
society, none of these motivations seriously apply to software:

 1- programmers and theoreticians have been innovating for years
without software patents, and mathematicians for centuries. There is
no indication that software patents has/will induce more innovation,
and every indication that it will stand in the way of independent
innovators (individuals or small firms).
   The capital investment for innovation in software is usually
extremely low.  I will grant that there is a price (not that high) to
educating the innovator, but the patents seldom goes to those who pay
for education.

 2- Most software cannot be kept secret, since it is always possible to
de-assmble code to see what it does. It is difficult, but no more so
in my opinion than unscrambling the patenting jargon.
   Or to have a converse argument, it software patenting is allowed,
giving the sources should be compulsory, so that the information
really goes out. It should also be compulsory because that is the only
effective way to check that no patent is infringed.

 The disassembling argument does not apply to software that is not
sold, such as sofware implementing services on a Portal. However, again,
how is one to check that no patent is violated.

  In other word, the only patents that can really be enforced, because
violation is visible, are those concerning exchange formats,
protocols, interfaces, ... i.e. the very areas where patents are the
most harmful to free competition, or most likely to be trivial.

 3- bringing software to market does not require heavy
capital. Programming is relatively cheap, and distribution on the
internet free. I discount advertising which brings no benefit to users
(see free software model).

(I have to shorten this for local family reasons ...)

Bernard

-- 
Bernard.Lang@inria.fr             ,_  /\o    \o/    Tel  +33 (1) 3963 5644
http://pauillac.inria.fr/~lang/  ^^^^^^^^^^^^^^^^^  Fax  +33 (1) 3963 5469
            INRIA / B.P. 105 / 78153 Le Chesnay CEDEX / France
         Je n'exprime que mon opinion - I express only my opinion
                 CAGED BEHIND WINDOWS or FREE WITH LINUX