|
Authored by: Anonymous on Tuesday, June 12 2012 @ 08:45 AM EDT |
Software is math, and math is supposed to be unpatentable.
Software is
speech, and speech is covered by copyright already.
Patent lawyers are
shameless hypocrites, because their livelihood
depends on making everything
patentable. So they refuse to see the
universe as it really is, and instead
come up with their own contorted
definitions of things like "machine" and
"algorithm" and "capability", so
they can read broad swaths of software
discoveries as patentable.
People like you are driving innovation away from
the U.S. Supporting the
patentability of software is supporting parasites and
trolls over actual
wealth-creators, and supporting giant ossified
megacorporations over the
actual innovators.
If I were dictator for life,
the first thing I would do would be to scrap the
patent system and order that
patent lawyers must find some *useful*
career. Although the temptation to
punish them for the decades of their
extremely harmful parasitic activity would
be nearly overwhelming.
Honestly. I'm ashamed that patent lawyers come from
the same species
as software developers. I don't want to have to share a
planet with such a
despicable form of life. I resent them for wasting my
oxygen. [ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Tuesday, June 12 2012 @ 08:58 AM EDT |
Why should it matter that the machine won't ever exercise a particular
capability unless and until a human instructs it to?
Thats true of most
machines in most situations!
My microwave doesn't start cooking food until I
push some buttons on it,
to tell it what to do.
My car doesn't start
accelerating until I press the gas pedal.
My television doesn't turn on, or
change channels, until I instruct it to, with
the remote.
My car radio was
always capable of receiving the "Rock FM" station, even
if I had never turned
the tuning knob to instruct it to do so.
Similarly, my telephone-computer
was always capable of playing Angry
Birds--it was designed to run any and all
software programs for that
hardware, including all software programs that had
not yet been
discovered and written down. It's a general-purpose computer plus
phone, camera and touchscreen.
Loading software into the device is not really
any different from pushing
buttons on a microwave--just more complicated. [ Reply to This | Parent | # ]
|
|
Authored by: jesse on Tuesday, June 12 2012 @ 09:04 AM EDT |
"we don't get paid enough".
Not relevant.
You pay a mathematician for the time spent writing programs, or discovering more
mathematics.
You don't pay them for the mathematics.
You pay them for the time.
Written forms of the mathematics may be copyrighted.
Software is written forms of the mathematics and may be copyrighted.
Patents? Nope, there is nothing there to patent. No device present.
CPU? A device for the purpose of carrying out mathematical operations,
patentable.
Memory? A device for the purpose of holding written data and software,
patentable.
Input output peripherals? Devices for the purpose of translating symbols
into/out of written form in memory, patentable.
The software is the mathematics... and is non-patentable.[ Reply to This | Parent | # ]
|
|
Authored by: jesse on Tuesday, June 12 2012 @ 09:18 AM EDT |
<blockquote>
But I am just not buying into the view that because it CAN
be done then when it actually IS done nothing of import has
happened. The universe is made up of atoms. Atoms CAN be
chained into any structure we want. But only SOME structures
have real value, and SOMEONE must combine them.
</blockquote>
Umm - you sure about that?
I think people consider the plot of land they live on has value. And air...
people really consider that as having value.
And farmers in particular consider land to have real value.
But in none of those cases did SOMEONE (meaning human) combine them. They
evolved out of the primitive element hydrogen. And 13 billion years or so.
[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Tuesday, June 12 2012 @ 09:24 AM EDT |
The general-purpose computer is one of the MOST capable, and amazing,
machines
ever conceived of and built!
You say, that when you buy one it just sits
there like a lump. You believe
it has no capabilities until software is added
to it. But you are mistaken.
(Your belief inevitably leads to the conclusion
that loading software--
flipping electrical states in the computer's
memory--creates a new
"machine" with capabilities it didn't have before, but
this is nonsense. :)
Let me tell you a story. When I was six years old, my
mother (a teacher)
brought home a Commodore 64 for me to play with over the
summer. My
father showed me how to write extremely simple programs for it in
BASIC.
And as I experimented, it gradually dawned on me how amazing it was.
Here was a machine that could do almost anything I could think of, as
long as
I could figure out how to ask it to do it! Well, anything in the
abstract
world, anyway. Obviously it couldn't do physical things like
snapping my legos
together. But it *could* make sounds and music, show
colors and pictures, in
almost infinite variety, limited only by my human
imagination and my skill at
programming it. It wasn't nearly as exciting to
the adults around me, because
they (like you) did not really comprehend
the amazing capabilities that were
already built into this machine.
So I knew then, at six years old, that I
was going to spend the rest of my
life programming these machines. The truth
you remain unconvinced of,
has been so obvious to me since I was a child, that
I find it difficult to
even comprehend a view like yours.
Now, decades
later, I work as a programmer in the video game industry--
still struggling
against the same limitations of human imagination and
ACTUAL machine
capabilities, that I got my first tiny glimpse of at the
tender age of six.
And I'm telling you, software patents do not help us at
all and are actually
quite harmful. They're one of the few things that I
have no control over that
still worry me quite a lot. More than once, I've
turned aside from a research
direction after being informed that some of
my ideas were already encumbered by
software patents.
To practitioners, these things are like land mines strewn
across our work
environment. They add no value, and they constantly threaten
us with
great harm. Please help us get rid of them!! [ Reply to This | Parent | # ]
|
|
Authored by: PolR on Tuesday, June 12 2012 @ 09:24 AM EDT |
I hear the pushback from all these comments, and I
understand it,
but I'm just not buying it. I do not believe
that because a computer CAN do
something that it necessarily
WILL without human intervention telling it to. So
when I say
"capability" what I mean is that some human has created a
combination
of tasks for the computer that no one else had
asked the computer to do
before.
This is an argument for saying the program is a new
process. This is not an argument for saying it is a new machine. What you have
not explained is why conflate the terms? They are not synonymous.
Patent law
states that processes is patentable. So why not claim the innovation as a
process and legally analyze it on this basis? Why do we have to use a meaning of
the term machine which bears no connection to the reality of what a machine is?
It is this sort of things that brings disrespect to patent law and those who
practice it. The connection with reality is lost when lawyers do this kind of
things.
The universe is made up of atoms. Atoms CAN be
chained into
any structure we want. But only SOME structures
have real value, and SOMEONE
must combine them.
Then show us the structure which provides the
software capabilities.
When I have explained why programming a computer
makes no such structure you have told us in no uncertain term that structure
doesn't matter, capabilities do. So which is it?
The "make a new machine"
thing is a fiction which implies some structure made of atom is being made. But
this is not the case. No such structure is made. We know this because when we go
analyze the principles of operation of the computer and track the connection
between capabilities and structure we find the connection is not there. The
innovation is something which is not a structure made of atoms and we can
justify this assessment by referring to the principles of computer science.
When I buy a general purpose computer, it is going to sit
there
like a lump unless I pay someone (or do it myself) to
write programs for it. And
most of the time the paycheck is
enough to get the programs we need. But
sometimes we want a
bigger reward, in large part to compensate for all the
money
spent writing programs that are a complete bust.
Why would
this reward need be driven by software patents? There are plenty of other ways.
We may use copyrights and a lot of successful companies went that route. We may
sell service for FOSS like Red Hat does. Monetary reward does not equal patents.
Besides the patents harm those who chose the other ways because the liabilities
are unilaterally imposed on those who have no use for the patent system.
But most importantly there is subject matter which is not patentable no
matter how meritorious. Whether or not the inventor wants or should have
monetary reward is not relevant to the patentability of the subject
matter.
Is this really what this makes a new machine thing is all about? If
patent law admits that software is not a new structure made of atoms, then
software is not patentable. People who want the patent money don't like this. So
they say software makes a new machine and pretend it is a structure made of
atoms. And when the experts in the field say this is not the case, the attorneys
come tell them they don't understand the law. [ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Tuesday, June 12 2012 @ 09:57 AM EDT |
So a guitar is not capable of playing music until somebody
strums it?
If I pick up a guitar and strum a completely novel tune,
have I invented a new musical instrument?
After thinking about it all night, I finally think I see
your view. You think software is different from music,
because hardware + software = a complicated and useful tool
= "a machine".
(I'm pretty sure that's not the meaning of "machine" that
the Supreme Court had in mind, but ignore that for now.)
You think that if somebody designs a system (hardware +
software) for some novel purpose ("novel" in the ideal
patent-academic sense: totally unanticipated in the prior
art, and not obvious to a PHOSITA), that system constitutes
a special-purpose machine and should be patentable. It's a
fairly reasonable position, actually.
It does bring to light a further objection or two to
software patents: lack of specification and too much
generality. You're not supposed to be able to patent the
idea of "a machine that does X". You should only be able to
patent a particular, non-obvious way of doing X. In my
experience, software patents are not sufficiently described,
because they do not require actual reduction to practice, so
they are almost impossible to design around. I don't have
time at the moment to review your Swype example patent so
I'm not sure whether there's a specific implementation
given, but I suspect it describes a very obvious algorithm
in very general terms.
The same basic algorithm could be used to do lots of other
things (like language recognition), do we really want "field
of use" patents? When there's no chance anybody ever
bothered listing all the possible fields of use in the prior
art, and the supreme court has basically abolished the
PHOSITA obviousness test?
[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Tuesday, June 12 2012 @ 10:03 AM EDT |
"So when I say "capability" what I mean is that some human has
created a combination of tasks for the computer that no one else had asked the
computer to do before. "
So are you arguing that just because a particular asteroid hasn't ever bumped
into anything, it doesn't have the capability to bump into anything?
Or that you can get a patent on a problem specification?
Just because I haven't asked you to sit down means that you lack the capability
to sit down, so you can't do it?
You really aren't making sense here. And you are starting to sound like you are
trying to redefine words to mean what you need them to mean for your argument,
rather than what they mean to the experts in the field.[ Reply to This | Parent | # ]
|
|
Authored by: PJ on Tuesday, June 12 2012 @ 11:10 AM EDT |
Certainly mathematics is important. But is
it patentable? You keep sliding off that
rock.
Many important things are not patentable, and
they should not be. Lawyers' arguments are
very important. But they are not patentable,
and they should not be.
You assume that no one will do important things
unless they can get a patent, but you live in
a world that has an entire operating system that
can be used in everything from watches to Androids
to supercomputers, and it was written by people
who did it for fun and because they wanted a free
operating system, Linux. So your theory falls
off the cliff of reality. People do important
things every day without patents.
Groklaw is a small way changed the world's course
of events in the SCO saga, and nobody here did it
for money. There's no patent on it.
Why in the world would you think that corporations
are the only ones who can innovate? They are generally
the very last to do so, actually, since monopolies
invariably cause stagnation. And yet you think
that everything is patentable, which is to say that
there should be patent monopolies on everything.
It was Bill Gates who stated clearly that had patents
on software been in existence when he started Microsoft,
he couldn't have built his company.
Yet in the face of all this evidence, you want
*more* software patents?
Are you sure you haven't been thinking too much
of corporations and not enough about innovation and
the advancement of science and practical advancements?
If you think patents are what make innovation happen, you
surely are not seeing the big picture.[ Reply to This | Parent | # ]
|
|
Authored by: PolR on Tuesday, June 12 2012 @ 11:12 AM EDT |
I want to point out that there are things in this universe which have practical
utility and which are not physical. They are not made of atoms, electrons,
energy or anything physical.
Consider for example a dictionary. This is very useful. We may say that the
paper and ink is made of atoms but this is not the dictionary. This useful thing
is the collection of word with their definitions. Neither the words nor the
definitions are made of atoms. You may look up for the word "molecule"
and find this is made of atoms, but the physical molecule is not the definition
of the word.
Consider a simple arithmetical formula. Someone may do a pencil and paper
calculation. There is a process for manipulating the pencil and leaving marks of
lead on paper. This is not the calculation. Numbers and algebraic formulas are
not the atoms of lead. The algorithm for the calculation is not the process of
moving the hand and the pencil. If the calculation describes the flight of a
rocket the atoms of the rocket are not on the paper.
In a computer the memory plays the part of the paper. The CPU plays the part of
the human. This is a simple explanation of why the software is not made of atoms
or electrons or anything like this. Software is the manipulation of the symbols,
the bits, and this is different from electrical charges in capacitors,
electrical current or voltages etc. If the computer is used to maintain the
inventory of a grocery, the atoms of the food are not in the computer.
The error in your argument is that you don't care to show the physical structure
associated with capability, You presume that if there is a capability that
nobody thought of before then there is automatically a structure that nobody
thought of before. This is not true. You can have capabilities that are newly
discovered without a corresponding new structure. The capabilities was there, it
is just that nobody realized it was there. This is what happens when we write
software. A new structure us never made no matter how new and nonobvious the
capability may be.
This sort of things can never be patented as a machine because the subject
matter is not a machine. It may be patented as a process, although in the case
of software the process is always a mathematical computation unless we embed the
computer into something like an industrial process for curing rubber and claim
the curing of the rubber.
[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Tuesday, June 12 2012 @ 11:54 AM EDT |
Thanks for the article but even more, the many, many
replies. I
very much
agree with what you say. I just think that most of
the arguments for software
being math are wrong - as any
mathematician knows, everything is maths. (I have
yet to see
the math theory behind the Hello World program that handles
all the
different environments.)
Just take a drug patent (validity is irrelevant),
all it
involves is saying that this chemical compound does
something
(useful)
in, say, the body. Sure, there is a lot of work
finding what that compound does
and getting in a suitable
form including modifying it to avoid undesirable
effects
(not always and, hence, black labeled medicines). Producing
software
is not unlike that of drug. Like drugs, software
still needs an environment
(i.e., the computer's
body) that will cause it to become active, do it's
job, and then be removed. So, if you can get a patent for a
drug then you
should be able to get a patent for software. [ Reply to This | Parent | # ]
|
|
|
|
|