Re: Strawmen and Urban Legends

[Stefaan A Eeckels]

On 28 Dec 2006 06:59:04 -0800
dt@soundmathtech.com wrote:

> 
> Stefaan A Eeckels wrote:
> > Of course, a device that peforms a particular function, like a
> > vocorder, can be patented even if it uses a DSP chip and software.
> > You patent the complete device, not its components.
> 
> That's a clever suggestion :) Patent a complete device e.g. cell
> phone, HDTV...Boeing 747...

> No matter that those "devices" are combinations of hundreds, if not
> thousands, of patented technologies developed over decades...

You don't patent a technology, you patent an invention, which has to be
described in such a way that a person "skilled in the art" can make and
use the invention (which BTW happens to be one of the problems of
so-called software patents, because they are never described using the
terms or methods used by software professionals to describe or document
software). 

The fact that an invention uses components that are not protected by
patents (though at one time they might have been patented) does not
preclude it from being patented. The fact that a device is patented
does not imply that all its components are themselves patented. 

> > Unless the patent offices manage to understand how writing software
> > works, we're better off with copyright protection.
> >
> My dear little friend... I can assure you that the USPTO has enough
> people who understand software, they just need to spend more time on
> each application.

In any case, the results aren't stellar. 

> The EPO is already doing a great job as far as "software patents" are
> concerned.

And you are an authority on these matters. 

> > Most programmers don't write software like poems. It's written much
> > like one would "write" mathematical formulae.
> 
> Oh yeah, just like mathematical formulas... When was the last time
> that you wrote a "mathematical formula" ? High school calculus ?

I am not a mathematician (I am a crystallographer by training and a
programmer by trade), so I hardly ever engage in writing down theorems
or proofs. That, however, has nothing to do with the fact that software
is not written as a poem, but rather as a set of logically coherent
statements in a restricted language, much more like mathematics.

> > And quite clearly, we
> > would not have seen much progress in mathematics if patents would
> > have been granted on Newton-Raphson or FFTs.
> 
> Oh yeah, a patent on a fast practical method of computing well-known
> Fourier transform would have a catastrophic impact on the progress of
> the entire field of mathematics.
> Your logic is flawless...

I think you're missing the point - progress in certain fields of
endeavour occurs through sharing of ideas and discoveries, rather than
by granting monopolies on them. Finding the right balance between the
ability to benefit (economically) from software inventions and the
stifling effect of a monopoly is not easy.

Specifically in the case of software patents, the chasm between the way
software is described, specified and written, and the language into
which a software invention is cast for the sake of patentability makes
it very difficult to know whether a particular algorithm is patented,
and to develop workarounds. In addition, the value of the patent after
it has expired is seriously reduced by the absence of formal
specifications, source code or a reference implementation.

> > One of the problems is that a lot of people discoursing on software
> > patents don't know what software is.
> >
> This is so evident from your comments.  Why don't you go and take some
> elementary CS and EE classes ?

Aparently, they didn't teach you much about the (lack of) value of
argumentum ad verecundiam and argumentum ad personam. 

-- 
Stefaan A Eeckels
-- 
The mushroom philosophy of product sales and support:
Keep your customers in the dark, feed them a lot of manure and
hope they will grow and flourish.