Software patents and FreeBSD

[Bryan Maynard]

On Saturday 09 July 2005 08:31 pm, David Kelly wrote:
> On Jul 9, 2005, at 8:03 AM, Bryan Maynard wrote:
> > If patents are applied to software, at what level of astraction do
> > we stop
> > allowing patents? As Kodak has, so elequently, show - you can
> > obtain a patent
> > for: "two machines communicating in an Object Oriented way". This
> > means that
> > it is not only possible, but plausible and adventageous to patent
> > the for
> > loop and other computational constructs. Within a year all
> > consrtucts of
> > computing could, very realisticly, be patented.

> An invention implemented in software has long been patentable. Your
> "Within a year all consrtucts of computing could, very realisticly,
> be patented" is already demonstrated as unrealistic.

There are two key factors you are failing to take into consideration:

1) The recent populatiry of Open Source software
2) The need for corporations to make money

So far neither of us have answered the inital question: "How would software 
patents effect an Open Source project like FreeBSD?"

The answer can found by examining exactly what a patent would be applied to. 
As other posters have noted, patents are useless for "software" because 
"software" is a consistantly chainging organism. Patenting software would be 
like trying to patent a person at a point in time; the second the patent is 
obtained it becomes obsolete. Also, what about the person is being patented? 
THe structure of their atoms? Their knowledge, experience, world views, or 
their voice singature?

This proves that the only things patents are good for is static objects that 
do not change over time. Software DOES change over time; improvements are 
made, features are added, and bugs are (hopefully ;-) ) fixed. Either every 
time the software changes a new patent would need to be obtained or the 
underlying constructs would need to be patented so that their use and structe 
would become irrelevent (for patenting purposes). This is where my comment 
"within a year all constructs of computing could, very realistially, be 
patented". Your rebuttle to this comment shows some level of ignorance: 
softwere is not, in itself, an invention. Software uses language constructs 
(if. . .then, foreach, for, do. . .while) to produce funstionality. It is the 
language constructs that are the inventions - and even this could be debated 
because language contrusts simply provide a layer of abstraction over the 
physical circuits. Software is NOT an invention - in any way. Web servers are 
not inventions - they are implementaions of ideas that use inventions to 
provide services.

There are many libraries that FreeBSD uses at a very low level (libfoo, 
libbar, etc.). Based on my previous statements, the real danger from software 
patents does not come from FreeBSD's use of these libraries, or even the 
libraries themselves. Intead, the danger comes when the underlying constructs 
used to build the libraries is patented. This will not hapen with the current 
generation of languages (C, C++, Perl, etc.) sonce the "prior art" factor is 
plain and evident. However, when NEW languages emerge, which happens all the 
time, the problem of software patents will become instantly terrifying. Note 
all the attempts Microsoft has made to "patent the internet". Two employees 
at Microsoft developed a protocol very similar to IPv6 (the next generation 
addressing standard for the internet). When these two employees left 
Microsoft, they contributed a lot of their protocol concepts to the w3c and 
IETF - and laid the groundwork for what is now IPv6. Once Microsoft saw that 
IPv6 was beginning to pick up steam, they immediatly tried to patent the 
addressing scheme claiming they created it first and that IPv6 was a 
"dirivative work" based on their "invention".

This is an example of the dangers FreeBSD, and indeed all software designers, 
face: the toold used must remain usable by the genral public under sensible 
and appropriate licensing terms. Patents would make it impossible for FreeBSD 
designers to use any new languages, unless they were patented in such a way 
that the new languages remained freely usable - which would defeat the entire 
purpose of obtaining a patent in the forst place.

> Just because something can be abused there is no reason to throw the
> baby out with the bath water. Its pretty well documented that money
> can buy lawyers to make up == down, set Simpson, Jackson, and Scrushy
> free.

With all dew respect, this view is both nieve and dangerous: there is VERY 
GOOD reason to be both cautious and wary of ANYTHING that can be abused. I am 
in no way suggesting we "throw the baby out with the bath water". I am, 
instead, pasionatly emploring anyone involved in the decision-making process 
about software patents to look long and hard the ramifications of software 
patents.

> The problem is not "software patents" but that fool patents are
> issued for existing art. Likely that is the case with the Kodak
> patent mentioned. It certainly is the case in "Method for Exercising
> a Cat."

You are exactly right! Fool patents are a reality and have already caused much 
damage (Kodak vs Sun). This is precicely the danger that needs to be avoided.

> Another which comes to mind claimed a patent on numbering the frames
> for use as an index in a multimedia data file, as if film hasn't been
> marked with footage during manufacture the past 50 or 80 years. As I
> said earlier the wrongness is in considering software to be any
> different than any other implementation, just as in the past a 10"
> disk platter was considered fresh territory where 12" platter patents
> did not apply.

You keep proving my point. . . Using the financial resources availible to 
them, corporations could - and would - find ways to obtain all sorts of crazy 
and resrtictive patents. Remember this: a company's first order of business 
is ALWAYS to make money - without exception. This does not make companies 
evil. It simply means that they do whatever it takes to make a buck. This is 
in DIRECT opposition to us Open Source developers. We have neither in 
resources nor the inclination to "horde" all our creations to ourselves to 
turn a profit

> > Before I close, I would like to return to my point on software
> > licensing.

> > One question: Why are movies not patented? Why are "types" of
> > movies not
> > patented? I am not talking about patenting "action" movies or
> > "love" stories
> > (although that would be possible using the logic of those who
> > currently
> > support software patents). I am talking about the patenting of
> > documentable,
> > movie formulas:

> > 1.a) Boy meets girl
> > 1.b) Boy falls in love with girl
> > 1.c) Boy losses girl
> > 1.d) Boy gets girl back
> > 1.e) Boy and girl live happily ever after

> The answer is very simple, "prior art." The screenwriter did not
> invent "boy meets girl." Or "boy meets boy." Or "boy meets dog." Or
> "boy meets alien." Copyright protection last much longer than patent
> and is easier to get. Especially when heavily borrowing from other's
> ideas. Yet when one borrows too much one must hire lawyers for
> defense. How much is too much? Well, the rich guys have to settle
> that with battling lawyers. Exactly the same as with patents.

Prior are, as you so often fall back on, actaully sums up the reason software 
patents are useless: there is nothing new under the sun. As I stated earlier 
- software does not fit the definition of an "<a 
href="http://en.wikipedia.org/wiki/Invention">invention</a>". Software does, 
indeed, employ techniques to solve problems but, again, if these techniques 
(design pattern, rapid prototying, etc.) were patented FreeBSD and other Open 
Source projects would be doomed.

> The wrongest solution is to deny a patent simply because it is
> implemented in software. The ideal solution would be to have better
> patent examiners, but "better" is anathema to government. USPO
> examiners are patent lawyers perpetually showing off to prospective
> employers at taxpayer's expense.

Again you prove my point elegantly: we cannot say that software patents 
"should" be ok, or even that they "will" be ideal "when" something happens. 
NEVER make decisions based on the "ideal" solution, unless evry aspect of 
that solution has already been fully realized - that spells disaster.

> --
> David Kelly N4HHE, dkelly at HiWAAY.net
> ========================================================================
> Whom computers would destroy, they must first drive mad.

In the ned I return to my point from my previous post: Find the right tool for 
the job, and use it. Software patents are simply not the right tool for the 
job of protecting creators and artists. Software grows, changes, and matures. 
Patents cannot protect growth and change. They are intended to protect static 
object and ideas.

Licensing is the right tool for the job of artist protection and recognition 
for software. Licensing can be verisoned and swapped as software grows, 
changes, and matures. Licenses provide all the protection software needs and 
does not bring any extra baggage.

Again, thank you for your time.

Bryan
-- 
Open Source: by the people, for the people.