Subject: Re: Examples needed against Soft Patents
From: Ben Tilly <btilly@gmail.com>
Date: Tue, 28 Dec 2004 17:51:45 -0800

On Tue, 28 Dec 2004 03:05:21 -0500, Taran Rampersad <cnd@knowprose.com> wrote:
> All of this aside, if there are software patents, wouldn't they be
> derivative of hardware patents? After all, software requires hardware in
> the first place. So a patent on a software application would have to be
> derived from a patent from an operating system which in turn would be
> derived from a patent on the hardware.

Um, no.  No more than patents on hardware requires patents on the
raw materials from which you make that hardware.

The patent system is ideally meant to provide protection for people
who come up with ideas which are hard to think up, but once you've
seen fall into the, "well duh" category.

Picking a random "real patent" consider intermittent windshield
wipers.  There is a patent on a type of wiper that only requires 4
parts to build where only one moved.  (Filed in 1964, long since
expired.)  None of those parts were patented - they were
standard components - but that combination was.

All of the pain and frustration that we programmers see over
software patents on the apparently obvious apply to real patents
such as this.  The auto companies thought that the device was so
simple that it shouldn't be patentable.  But the device was not
easy to think of in 1964.  In fact automobile companies had been
trying to design it and had failed (their best wiper required 29
moving parts).  And the courts agreed with the inventor Robert
Kearnes and awarded him damages.

> Or are we saying that Software Patents can exist in a vacuum? Now, if
> people want to patent software and pay royalties to people/companies who
> have patents on the hardware, I may have to rethink my stance on
> software patents. That, after all, is a logical progression of patents.

Here is one way that software patents can be seen as a logical
progression of patents.

Devices are patentable.  There are patentable devices which can
be realized either in hardware or software.  If I have the patent on
realizing it in hardware, should you be allowed to circumvent me
by implementing it in software?  You aren't using a new technique,
you're translating my technique to slightly different "materials".  But
once we accept that patent law can be applied to things which are
implemented in software, why am I forced to take the patent out on
a physical device if I'm planning to actually ship it in software?  And
now we're to software patents.

I'm not saying that I agree with this progression, but it is my
understanding that something like this actually happened.

The problem here is that courts like to draw what they call
"bright lines".  They like rules with boundaries that are easy to
recognize and interpret.  But the line between "hardware" and
"software" isn't very bright - you can often move between doing
something in hardware or doing it in software and the boundary
dividing them is rather fuzzy.  (What do you call a system with
software built into ROM?)

> What is being pawned off as Software Patents is not a logical
> progression of patents - at least in my mind.

And many would agree with you.  But not the courts where I
live. :-(

> Now, fly this up a flagpole and see how fast companies and corporations
> who want software patents start rethinking their stances as well.

Sorry, but I don't think that it will convince anyone to change
their minds.

> This is not really about patents, is it? If it were, if it were about
> protecting the inventions of people... wouldn't that mean hono(u)ring
> patents for the equipment on which other patents run?

I'm puzzled by this statement.  Can you point out examples
where patent proponents have proposed that we should violate
patents on the equipment that they need?

Cheers,
Ben