Subject: Re: Examples needed against Soft Patents
From: "Stephen J. Turnbull" <>
Date: Mon, 20 Dec 2004 14:09:31 +0900

>>>>> "Bernard" == Bernard Lang <> writes:

    Bernard> I am just coming out of a difficult meeting, where I was
    Bernard> alone to defend the position that free software is
    Bernard> incompatible with software patents.  The people I was
    Bernard> opposed to are major pro-patent lobbyists at national
    Bernard> level ... i.e. the kind of people who make the arguments
    Bernard> presented to government officials.

Talk to Jacques Cremer and Jean Tirole at IDEI ( for contact
info).  These guys are in the front rank of software/internet
economics.  You might want to go to the conference they're having in
January (Conference on Internet and Software Economics in Toulouse,
Jan 20-21 or so, see the IDEI home page); there will be a lot of hot
economists in the field there.  In general, these days I think
economists (who know anything about the field) are sympathetic to free
software, all but the hardest right of the hard right.  That doesn't
mean they are opposed to IP law, just that they wish free software
could be really true.  :-)

The problem is coming up with convincing arguments.  To convince
experts in law and economics or industrial organization, leave it to
the pros.  This is hard stuff, and not at all part of the orthodox
curriculum yet.  Find some sympathetic ones (I think Cremer and Tirole
are, although I gather Cremer is not 100% convinced), and task them to
talk to the legislators' technical staff, some of whom will at least
know enough about economics to be impressed by those names.

For general economists, you can talk to them yourself.  Say
"transactions costs" and point out how difficult it is to overturn a
bad patent.  Then point out how bad the bad patents are, and the
institutional reasons why it's likely to stay that way (primarily,
patent offices can't afford to hire even apprentice-level engineers,
so they have no clue about "novel" or "obvious"; in the US, the
court-order bias toward awarding patents, which has a knock-on effect
in Europe because of conventions about honoring other nations awards).
Have a list of bad patents ("bad" doesn't mean "harmful", it means
"obvious" or "long-known to practictioners"), the longer the better
(Greg Aharonian probably has one), and summarize a few of the worst.
Point out that it is very expensive to search a patent compared to
inventing around a software problem.  That means that if developers
respect patents by searching for them, most patented "inventions" in
software are a disastrous social loss.

Furthermore, if it weren't for the systemic problems mentioned above,
it would be possible to filter out most of the bad patents---in
theory, you don't have to take the bad patents with the good.  But
reform is very difficult, because the system needs to be burned to the
ground and redesigned from scratch.  Tell 'em I said so ... on second
thought, phrase that "it's better to avoid further entrenching the
problems by _reducing_ the power of (software) patents as much as
possible until a viable system is proposed."  Not that I really expect
one to be proposed, but that sounds "reasonable" to folks trying to
balance the pro- and anti-software patent arguments, and allows them
to focus on the evil of the current system.  It's a pretty safe thing
to concede: there are very few "good" software patents that pro-patent
types can point to---unlike the situation in drugs where "rich men's
diseases" abound in treatments, while we still can't pump out enough
anti-malarials to stamp out a disease that should be as rare as
smallpox now.  So on balance it seems plausible to get rid of software
patents (or, better, not allow them in the first place).

All of this won't absolutely convince, but if you treat doubters as
human beings with a right to their own opinions, they will leave more
than half-convinced.  And later they may convince themselves.  Those
arguments are _strong_, they worked on me, after all.  :-)

Also, point out that because of the requirement that patents be
published, and in order to maximize claims, they are written as
obscurely as possible, typically in turgid legalese that is useless to
an engineer.  Find one of the ones where the "specification" is a dump
of machine code!  Patents are not educational tools, in fact, quite
the opposite.  By contrast, free software doesn't merely encourage
transparency, it is transparency in action.  Legislators and their
non-economist staffers will appreciate this argument, too.

Be careful about contradicting the corporate spokespeople, because
they are correct, from a certain point of view.  Software patents are
not incompatible with creation and distribution of free software, only
with its use (not a joke!), and pragmatically, only for _organized_
free software (business or GNU, eg).  Any company that tried to sue
enough ordinary people to eliminate "samizdat software" would generate
incredible amounts of antipathy, just as the RIAA has by threatening
music fans and destroying Napster.  So disagree with them, if you
like, but avoid being too strident---to the undecided, you'll just
mark yourself as an extremist, and maybe even lacking in clue.  Focus
instead on the damage of _extensions_ of IP law that will be demanded
to cover "samizdat software", and on the benefits to giving a little
extra protection to free software businesses.

Eg, "samizdat tapes and CDs" then led to the DMCA etc, and other
abuses such as those the Eldridge lawsuit was about.  Point out that
was horrid law: a couple of dozen Nobel Laureates, including Milton
Friedman, signed a brief supporting the good guys in Eldridge.

Maybe you can find some lawyers to say that it's horrible law, too.
Talk to the Larrys (Lessig and Rosen), maybe they can give some
pointers to European colleagues to put it in regional context.

Point out that free software businesses tend to be small/medium size,
which (at least in the US) is where job growth is, especially for the
entry-level employees who are badly hurting in many European economies
(in Spain 15-25 year old males have an unemployment rate 3~4X as high
as the national average).  Point out that free software businessmen
tend to have extremely strong senses of social responsibility (which
should be a very attractive argument in Europe).  These are businesses
that should be encouraged, in preference to protecting large faceless
corporations and enabling patent sharking by lawyers.

IMO, another no-no is referring to the alleged economic justifications
for abolishing "intellectual property" etc on the GNU "Philosophy"
page.  That list was collected with reference to a litmus test, and
contains very controversial stuff like the paper by Boldrin and Levine
(who are now considered to be unbecomingly partisan by most of the
professional economists I know) and "junk" written by people with no
professional credentials whatsoever.  Many of those papers are
valuable (especially Boldrin-Levine and the followups by Danny Quah of
the London School of Economics, and including some of the apparent
"junk"), but you need to have a good professional reputation and sense
to use them effectively---give the list to the pros you enlist, and
let _them_ present it.

Institute of Policy and Planning Sciences
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.