Software's game of mutually assured damage

[Bart Meerdink]

REPLY TO: D66 at nic.surfnet.nl

Een artikel waarin precies staat hoe software patenten in de praktijk
daadwerkelijke investeringen en vooruitgang remmen, en slechts als wapen
ingezet worden om de winst te verhogen door mogelijke concurrentie uit
te schakelen.

http://www.smh.com.au/articles/2004/07/30/1091080437270.html?oneclick=true

Software's game of mutually assured damage
By Ross Gittins
July 31, 2004

Monopolies are the only way to make real money these days, and patents
are fantastic because they allow you to establish legal monopolies.

That's what the company lawyer told a packed room of software engineers,
including one who submitted to me a revealing confession. For reasons
that will become obvious, he prefers to remain anonymous.

This is what the software engineer related to me:

The lawyer went on to explain that since what was important was the
monopoly, it was necessary not only to patent the way we were doing
things, but also to think laterally, and patent all the ways other
people might do them as well, not so that we could actually do these
things ourselves, but so we could prevent others from doing them.

We are a large software house, part of one of the world's largest
software companies, writing leading edge security software, and this
talk went down like a ton of bricks with the local engineers.

"Don't worry about whether you think the idea is worthy of patenting -
that's for us to decide," the lawyer went on.

Normally for a patent to be worth filing, it has to be original and
inventive. However, the United States patent office is deluged with
applications, and largely rubber stamps anything that looks plausible.
The feeling is that anything contentious can be sorted out in the courts.

This is something that causes annoyance to software engineers as more
and more ridiculous patents are granted, but the lawyers are fairly
comfortable with situation. "Look, if the examiners were any good they'd
be in industry, so you don't have too much to worry about," the lawyer
told me privately.

The whole idea of software patents is a bit strange, really. A
traditional patent is for a mechanical invention that may have taken a
long time to design, produce and bring to market, and provides
protection for the original inventor while eventually allowing their
ideas to enter the public domain.

Software, however, is mercurial. A good programming idea may only be
useful for a few months and, even after the dotcom crash, it's still the
case that after a couple of years generally there is a complete
generational change in the tools, techniques and even programming
languages used.

What's more, it is usually the implementation of an idea - the actual
computer program - which is most useful, and that is already covered
under copyright law.

Patents on software often appear completely counterproductive - by
monopolising a technique, a patent can simply ensure that the technique
is never used. Rather than making money, a patent can cause the death of
an otherwise promising technology, and this is frequently the aim of
patents held by owners of threatened technology.

It's a curiosity of the industry that the areas where there are no
patents, such as the original internet, the world wide web and the "open
source" movement, usually show the fastest innovation and progression.

Today a software patent is often the modern equivalent of an
old-fashioned robber baron setting up on a public highway and demanding
a toll from all who pass. Usually it's cheaper simply to take another
road (hence the need to patent all those other solutions!)

It's not purely predatory, though. There is also a defensive element to
the patent business, and this is the slippery slope that led to my
becoming a rabid patenter myself.

Recently my firm was taken to the cleaners in a US court case by another
firm over software that we'd been developing and selling for years, but
hadn't patented. The other company did file patents, there was some
difficulty definitively proving dates, and the firm lost a lot of money.

So we got another round of emails encouraging us to file patents, as a
way of defensively time-stamping some of our work, and offering
attractive bonuses for doing so. My area is particularly "bleeding
edge", and my manager pointed out that we'd look pretty stupid if our
everyday activities were patented by a rival.

So a colleague and I sat down for a few hours one afternoon, and tossed
off six fairly straightforward ideas, of the sort that any competent
worker in our field might come up with. The lawyer was delighted, but
insisted on splitting one of our ideas in two, so then we had seven
patents being filed in our names.

Having got into the swing of things, we sat down for a couple of hours
again the other day, and tossed off another 10. So far they have all
been accepted and filed by the firm's lawyers, which on a
time-per-patent basis puts us well in front of Thomas Edison, who I
believe previously held the record.

Since we receive a bonus of $8000 per patent, if all goes well we'll
share well over $150,000. And there seems no reason we can't keep this
game up indefinitely. We should be able to manage around 50 a year, and
this nice little earner will see the mortgage paid off in no time.

Meanwhile, the firm spends between $50,000 and $200,000 on legal fees
and filing fees for each patent, so we've created over $1 million worth
of employment for our legal friends. This is money that the company
can't spend on research and development, but there you are.

The company isn't being stupid here, though. It must know many of these
patents are of dubious value. But all the large software firms have big
patent portfolios and they mainly go unused, in a sort of "mutually
assured destruction" arrangement whereby each large firm is prevented
from using their patent portfolio by fear of devastating reprisals.

This leads to an interesting paradox. The vast majority of software
patents are unused, because people who actually make things often can't
afford the risk of using them. Generally, software patents are only used
to keep down small company competition, as defence, or by firms that
have nothing to lose.

The last includes professional "intellectual property portfolio holders"
(who don't produce anything themselves, and are therefore immune to
reprisals) and companies that are facing certain demise in their
traditional business, and so try to extract value from their existing
intellectual property.

Software patents, and their accompanying monopolies, have done
immeasurable damage to the world of computer programming, and are one of
the reasons the centre of innovation has moved either to open source
software, or to corporate working groups where everyone agrees to
automatically cross-license all their patents to each other - thus
forming a patent oligopoly rather than a patent monopoly.

Ross Gittins is the Herald's Economics Editor.

**********
Dit bericht is verzonden via de informele D66 discussielijst (D66 at nic.surfnet.nl).
Aanmelden: stuur een email naar LISTSERV at nic.surfnet.nl met in het tekstveld alleen: SUBSCRIBE D66 uwvoornaam uwachternaam
Afmelden: stuur een email naar LISTSERV at nic.surfnet.nl met in het tekstveld alleen: SIGNOFF D66
Het on-line archief is te vinden op: http://listserv.surfnet.nl/archives/d66.html
**********