Subject: Re: Examples needed against Soft Patents
From: Ben Tilly <btilly@gmail.com>
Date: Wed, 29 Dec 2004 19:13:09 -0800

OK, I think that we can shorten this conversation considerably...

On Wed, 29 Dec 2004 19:20:01 -0500 (EST), Russell McOrmond
<russell@flora.ca> wrote:
> 
> On Wed, 29 Dec 2004, Ben Tilly wrote:
> 
> > >   a) Patent policy is *economic* public policy that exists to provide
> > > incentives for innovation.  Different subject matter exists in an entirely
> > > different economic context, and thus economic analysis and not legal
> > > analysis is needed to determine if a specific subject matter class should
> > > be patentable.
> >
> > This argument is one that could convince me.
> 
>   This is like the umbrella argument under which much of the rest of the
> discussions happen.  I can be upset about software patents not
> facilitating a full spectrum of business models (specifically
> contradicting my FLOSS business model), but that is a detail of the
> economic analysis.

A detail, but a relevant and important detail for this list. :-)

And with the growing impact of FLOSS, a growing consideration
for the value of not having software patents.

[...]
>   Much of the current lobbying work on the pro-sw-pat side is based on
> manipulating the idea that "patents exist to provide incentives for
> innovation" to suggest that "where there are patents there are by
> definition incentives for innovation, so it is unnecessary to study this".
> Once we get policy makers to ignore this rhetoric and do the studies I
> believe they will come to the same conclusions we have.

But this rhetoric is actually _correct_!

The possibility of getting a patent is an incentive for innovation.
That is dead obvious - a patent is something valuable that you
can get if you innovate.  That value is an incentive.  Attempting to
debate this is a losing proposition because you can be made to
look stupid.

The *problem* is that others can use patents in ways that inhibit
innovation.  The question is which effect matters more.

> > Microprocessors are among the most complex machines in existence and a
> > tremendous amount of effort and skill goes into their design.  They are
> > also routinely emulated in pure software during the development process.
> > Are you really saying that microprocessors are not worthy of being
> > patented?
> 
>   Lets go back to first principles again as I believe this line of
> thinking is distracting and not useful.  We are allowing the tail to wag
> the dog by bringing up exceptions to seek to define a rule.

Distracting, or merely inconvenient for you?

>   Microprocessors implemented in hardware and emulators of that hardware
> exist in entirely different economic contexts.  Any *economic* public
> policy that treated them as equivalent must obviously be flawed economic
> policy.

Tell that to Transmeta.

For another example, I've heard that cellphone companies find
themselves making very non-trivial decisions about how much
they should be doing in hardware and how much in software.
They can implement it either way, and the economic tradeoffs
are not obvious.  Thus the two options exist for them in the same
economic context.

>   Finding extreme examples doesn't prove the general case.  We must do a
> cost/benefit analysis to find out if overall offering patents in a subject
> matter class help or stifle innovation.  If the general case turns out to
> be that it stifles innovation (as I believe economic studies will show),
> and that means that there isn't sufficient incentive to implement hardware
> emulations in software, then my answer is: so be it.  On balance we are
> better off without those hardware emulators.

I guarantee you that if you try to draw a lot of fine lines between
subject matter classes, patent lawyers will try hard to get their
products classified in the most advantageous subject matter
class, and you'll get legal chaos.  Nobody likes to be the one
who wrote a law that resulted in legal chaos.

[...]
> > useable as a microprocessor while the software emulation of a
> > bridge is not a useable bridge.
> 
>   It is in virtual reality -- talk to some of the gamers who buy
> houses/bridges/etc on e-Bay for use in a virtual game universe just how
> "usable" these are to them.

Point.  But it is not useable in the same way as the real bridge is.
Emulations in software of hardware often can be.

>   This question of tangible manipulation of "real world" nature and
> emulators and virtual universes is important.  We are heading into a legal
> system where there are "virtual properties" that are bought/sold/rented
> just like physical properties.  Where there is money, there are lawyers,
> and these virtual universes will have an impact on our real-world legal
> system.

s/headed towards/are already there/

> > Where did the phrase "new learnings about the forces of nature" come
> > from?
> 
>   It comes from the considerable discussion in the EU on patents,
> including some of the TRIPS wording.  While I am a Canadian I have been
> trying to keep updated on where things are internationally as well.

Well, it just goes to show that the EU can come up with wordings
that I detest as much as any wording in US law...

[...]
> > There is no difference in the work needed to design either.
> 
>   You are stating as a fact the very thing which is up for debate.

No, I'm stating as a fact something that is a fact.  What is up for
debate is what significance to give to this fact.

When you sit down with the software that Intel uses to design a
new chip, it is capable of producing both the physical
specification of that chip and a software emulation of what the
chip will do.  There is no difference in the work needed to get
either result, the design process is identical and you get both
software and hardware out.

To me this implies a level of equivalence between that software
and that hardware.  To you it doesn't.  But whether or not you
consider that the relationship bears an equivalence, the work
to design one is exactly the same as the work to design the
other.

>   You are saying that there is a difference between a process carried out
> by a human and a machine, but no difference between a process carried out
> by a machine and by policy (information process, software).

Yes, and that difference is that we can manufacture machines
but we can't manufacture people.

When we have the technical capability to manufacture living
people with specific memories, then the question of where to
draw the line between legal people and machines will be
very relevant.  Since we can't, I'm willing to ignore that line as
a theoretical question for future generations.

> > Both will accomplish the same job in the same "way".
> > How can you not see the equivalence between these two
> > implementations?
> 
>   How can you believe there is equivalence in one situation but not the
> other?   I believe none of these three possible "ways to get from A to B"
> are equivalent (done by human, done by machine, done by software), but you
> are saying that some are and some aren't.   What is the criteria you are
> using to say that some are and some aren't?

If it can be recognized that you're performing the same
operations when you see the same input producing the
same output, then I'd say that you're clearly doing the
same thing.  If you cannot recognize a fairly literal
translation, then you're not equivalent.

[...]
>   Authoring software (information processes, instructions, etc) has a lot
> more in common with training a human (wetware ;-) than building a machine
> (hardware).

I've done all three, and I disagree.

> > The public policy question is whether it is in the interests of society
> > to reward inventors with effective patents.  You're advocating making
> > patents in certain fields of endeavour be relatively ineffective.
> 
>   This starts by stating the conclusion, which is not an argument.

Huh?

I'm saying that the question is whether we want to hand out patents
and make them work effectively (and following that would be, "if so
then to whom").  As far as I know, that really is the question.  I've
not stated an opinion on whether we should do that or not, merely
stated that it is in question whether we should.

>   I do not consider people who device information processes to be
> "inventors", do not believe there should be patents on these
> non-inventions, and thus do not believe there is any harm in what I am
> advocating.

Ah, you're quibbling with my language.  To me it sure feels just as
much like invention to come up with an idea about how to write a
program as it does to figure out an idea for what kind of device I'd
like to have.

Also note that both the dictionary and common usage says that
people who come up with specific ideas are inventors.  For
instance we say that RMS invented the GPL, Vinton Cerf and
Robert Kahn invented TCP/IP, and Tim Berners-Lee invented
the web.

Arguing against that one is like trying to convince people that
they shouldn't call script kiddies "hackers".

[...]
> > When I compare software to a chair, the comparison seems
> > ludicrous.  When I compare software to physical machinery
> > of similar size to the representation of that software, the
> > comparison becomes far less ludicrous.

>   Why do you (or others) feel the comparison is less ludicrous?  Is that
> because the chair accomplishes only one task, but a piece of physical
> machinery can be configured to do more than one task?

Consider the following sequence of comparisons.

Leonardo Da Vinci built a number of spring-wound "robots"
that could be "programmed" to do simple tricks.  This was clearly
a device and today could be patentable (were there not prior
art - which there is).  The "programming" consisted of physical
manipulations that would cause it to do different sequences of
things when released.

Some centuries later Joseph-Marie Jacquard built looms that
would make different patterns depending on punchcards that
you put in them.  Now I'd like to draw your attention to an
important detail.  It was possible for him to make the
punchcards be built into the machine or to make them
removable.  If the punchcard was built into the machine, then
the arrangement of holes on the punchcard would be part of
the design of the machine, and one could conceivably
patent the machine that was programmed in this way to
achieve a particularly useful weave.  Note that in the actual
machine that the punchcards were physical devices which
again conceivably could be patented.  (I do not know
whether any were.)

The punchcard can be viewed in two ways.  It can be viewed
in terms of the information on it, or in terms of its being a
physical device.  The former would involve looking at it as
software, the latter as hardware.

Moving forward towards the present we find a variety of
improvements on the basic idea of the punchcard, leading
to the first actual computers being built to use punchcards.
At that point there remained this visible connection, the
punchcard can be viewed as a physical object that is
meant to be part of a machine, or as a medium for storing
information.

As technology progressed, this dichotomy persisted.  We
began building better ways to store data for computers.
When you examine those storage media on the physical
scale where bits are stored, you find that they can be
viewed, like the punchcard, either as information (not
patentable) or as a physical object (patentable).  For
instance a CD literally contains a pattern of burned out
holes that is exactly analagous to the holes in a
punchcard.  A magnetic disk operates on a different
principle, it contains small regions that have been
subjected to a strong enough magnetic field to leave
them magnetized.

In either case the information is stored in a physical
form, and under normal operation this physical form, in
a complex but well-understood mechanical way, is
converted into the operation of the machine.  It can be
viewed either as information or as a mechanical
device.  Whether you think it should be patentable may
depend on which way you look at it.

[...]
> > In fact the division between what we consider a "physical thing" and
> > what we don't comes to seem to be arbitrary.
> 
>   Why?  I know there are people that have these beliefs that ideas and
> thus software have a physical manifestation (electrochemical
> configuration), but if we believe that ideas are tangible then there is
> nothing that can be said to be intangible.

Do you think that atoms are tangible?  Can you pick one up, touch
it, and verify its existence?

But I digress into philosophy. :-)

[...]
> > In fact you literally can take a single description and, through
> > automated tools, wind up either with software or equivalent
> > hardware.
> 
>   I suspect you would have a hard time getting a patent on this 4GL
> generated hardware as well given it is unlikely to be considered useful,
> novel and unobvious.

I believe that you're wrong.  More specifically, I believe that there
are a large number of patents on hardware that was designed
this way.

> > Are you accusing the lawyers representing the patent offices of
> > deliberately throwing the case because winning it is not in
> > their best interests?  That's a pretty serious accusation...
> 
>   It's not an accusation so much as a part of human nature.
> 
>   If I were a lawyer and were hired to DEFEND software patents in a court,
> I could not do so very well given I don't believe they are valid.  A huge
> sum of money might get me to bend my morals and defend something I don't
> believe in, but I'm not going to be as good as someone who believes their
> own argument.

Then it is good that you're not a lawyer!  It is common for lawyers
to take cases that they don't personally support, and then manage
to win.  In fact any lawyer will tell you that doing so is their job.

Whether they achieve that ideal all the time I cannot say.  But I'd
be careful in accusing one of not trying.

>   Now the Canadian recording industry and their recent P2P lawsuits, that
> is another story -- not providing trivially obtainable evidence, lobbying
> earlier for changes to the act, losing the case because of these two
> things, and then very publicly making false claims about the viability of
> the Canadian copyright act.  It is hard to convince me this is all
> coincidence....

If coincidence it certainly sounds convenient.

But note here that the lawyers who were misbehaving were
misbehaving in their client's interest.  I consider that
significant.

Cheers,
Ben