Re: Why are software patents wrong?

[threeseas]

Paul Hovnanian P.E. wrote:
> threeseas wrote:
>
> [snip] 
>
> > why do you think software should not be patentable? and is your answer
> > cause or effect?
>
>
> Software is the implementation of an algorithm, which is not patentable.
> Copyright law is more suitable for the protection of software.
>
> You can patent a physical method or device which achieves some end.
> However,
> it is still possible to achieve that end with another means or device
> and not violate that patent. Physical implementations are typically
> constrained by practical engineering limitations, so there is value in
> protecting the innovation involved in their development. IOW, figuring
> out HOW to do something has value, not just WHAT to do. Software doesn't
> suffer from the same kind of restrictions on alternative
> implementations. Once the WHAT comes to mind, the HOW has multiple
> solutions. I'll leave the arguments about bounded vs unbounded solution
> sets to the math geeks.    
>
> Software patents effectively prevent achieving the intended end by any
> means rather than protecting the means. Patents are supposed to
> encourage innovation (i.e. better ways of doing something). Preventing
> that from happening is (or should be) counter to public policy.
>   

Alright, /. just had a story about the amazon one click patent being 
defended.

my response:
--------------
Patenting the act of programming????
==============
Programming is the act of automating the use of complexity, usually made
up of simpler complexities, so to enable the use and reuse of the
complexity by the user of that complexity, easy.

There is a world of prior art here, going back even before computers 
were invented, to the initial use of abstractions to express a more 
complex thought.
-------------

Alright, this is the essence of the physics of abstraction creation and 
use. But there is more, essence is not enough, we need details. And such 
details that provide the foundation or basis upon which to compair 
patent applications in software.

currently there is no such openly recognized foundation, sorta like not 
yet understanding the earth is not flat, that you won't fall off.

what are the undeniable actions they are used in abstraction physics, 
programming or otherwise?

* we start and stop something, change what we interface

* We keep track of where we are in doing something.

* we determine where we are getting input from

* we get input

* we determine where we are sending output to.

* We do things sequencially, one step at a time. and this includes those 
thing on this list, including sending output to where we have determined.

* we look up the meaning of things, to determine what action to take.

* we identify things to determine what action to take

* we apply constraints to what we look up and what we identify, so to 
more quickly move forward in what we are doing.

These are all things we do all the time, unavoidably, therefor they 
cannot be patented. And these things can also be coded into an 
application to enable a general automation tool that can be used for 
anything from automating the development process, including code 
generation to enabling the newbie end user to put two applications 
together in an automation that helps them improve their productivity.
Or even something simple like looking up the definition of an 
abstraction, such as the man page of a function or command, etc..

But there is an computing environment requirement to best enable this:

Nature likes the use of "three" in its application, such as primary 
color, Red, Blue and Green for Light (substractive color), and Red Blue 
and Yellow for paint (additive color), etc.. With such you can create 
all other colors, remove one and you greatly limit what can be done.

This follows thru in our use of computers. There are three primary user 
interfaces. First the Command Line Interface, the Graphical User 
Interface and the side door port to controlling or accessing 
functionality. This third goes by many names, function call, API, IPC 
port, etc., having many different tools for using it.

Again, remove any one of these and you greatly limit what can be done.

Its all about putting things together, be it code or applications or 
systems inter-connecting.

The ability to put things together is a human right, a given. Including 
the ability to create high level abstractions word/symbol = definition.

By understanding the physics of abstraction creation and use, we have a 
comparison base for which to determine patentability of software in 
general, as well as in specific cases.

Microsoft is a marketing company first and formost, a third party 
integrator of the works of others, second.

Longhorn and the philosophy of their "software factories" is such an 
integration of the works of others, the essence of programming and even 
the details above. And of course they are calling themselves innovators 
while distorting and biasing the truth of abstraction physics, towards 
themselves.

So why are software patents wrong, given the unpatentable natural 
physics of abstraction creation and use?

Imagine a calculator and those trying to patent some sequence of button 
pushes. Now imagine that calculator not being limited to just the 
mathmatical abstraction set, but open to any abstraction imagined and 
definable.

Yeah copyright is more appropriate for software, but the IP rights terms 
of infinity is not.