|
|
|
| Authored by: Anonymous on Monday, June 11 2012 @ 01:10 PM EDT |
The "changes" made to a computer by software are intended by the CPU
designer.
Just like putting people in the car is intended by the car designer.
I really like the idea of patenting a car *with four people in it*. I bet you
nobody has done that yet. Perhaps I can prevent all car companies from
advertising their cars as holding four people, once I patent the idea of
"creating a different machine" by putting four people in a car.[ Reply to This | Parent | # ]
|
| |
| Authored by: PolR on Monday, June 11 2012 @ 01:31 PM EDT |
> Any computer is in a different state when the memory has
> been changed and, thus, it is a different computer.
Memory is changed billions of times per second in a computer. When a change
makes a new computer and when it doesn't? If your answer is "always makes a
new computer" then none of the machines are practicing all the steps of the
software method recited in the claim. If your answer is not "always" I
want to hear on which criteria we distinguish a change in memory which makes a
new computer from one which doesn't.
There are programs which are being changed in memory at that rate while they
execute. These programs don't compute anything because as soon as the memory is
changed it is a different computer. This is where your argument leads.
> Superchargers, turbo-chargers, nitrous oxide injectors
> etc., are all ways to modify the car engine that is
> running. So [a] running program can be modified in memory
> while it is run is no different.
A patent on a turbo-charger is not a patent on the car engine. No one argues
that the engine constantly becomes a new machine because it is modified as it
runs but still these multiple engines are patented as a single machine
invention. But this is the sort of situation we see in software patents.
Consider this scenario. Someone patent a new video codec. The invention is a new
way to compress video. It is not a new way to make a self modifying program.
Then a developer comes along and write the codec in a language which uses a tree
of closures for execution. This particular video codec is modified in memory
millions of times per second as the computation of videocompression progresses.
Where is the machine structure for the codec?
You say changing the memory makes a new computer. This means there are billions
of machines which come and go as the codec runs. All these computers are
different in a portion of memory that is part of what is required to deliver the
codec function. I don't just talk about data modifications here. I take about
modification of bits which are used to describe the program function. But none
of these billions of computer execute all the steps which are claimed in the
patent because the computer is constantly changed.
So what is the patented machine here? Do you say the changes in memory make a
new machine? Or do you say they don't? It can't be both. But if you want the
patent to read on this codec it has to be both. You need the change in memory to
make a new machine when the program is initially loaded in memory and you want
the machine to stay the same for the whole execution in order to practice all
the steps.
> But there is one huge provision, which also is the
> theme that both Mark and you agree on, the change has to
> be sufficiently meaningful. Thus, the really hard part is
> avoiding the idea and focus on the actual invention.
No. The hard part is to accept that the concepts of idea and invention in
software don't correlate to machine structure the way patent attorneys say it
does. This view of machines does not sustain critical analysis without raising
paradoxes like the videocodec example above.
You are presenting a theory of machine where identical changes from the point of
view of physics are sometimes meaningful and sometimes non-meaningful depending
on legal context. Sometimes writing the same bits in memory makes a new machine
and sometimes it doesn't. This means the invention is not a machine. I don't say
there is no invention. I say the invention is something which is not a machine
and it should not be treated by the law as a machine invention.
[ Reply to This | Parent | # ]
|
| |
| Authored by: Anonymous on Monday, June 11 2012 @ 03:42 PM EDT |
Turbocharges, nitro, etc, are capabilities the car already had before running.
BIG DIFFERENCE! Just because they get used as designed does not change the car
in any way.
Similarly, software does not change a computing machine because it is simply
executing what it was designed to do.[ Reply to This | Parent | # ]
|
|
|
|
|
