US Supreme Court hallway
WHEN the ruling in Re Bilski came out, many assumed it was the end of software patents, but there was no official word on the subject, just continued disputes and disagreements [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]. The ruling is now being ascended to the Supreme Court, whose decision may be more explicit on this subject matter [pun intended].
In the ruling, the Court decided unanimously that the Sixth Circuit Court had improperly “derailed” an Ohio state court’s consideration of a claim by a death row inmate that he is mentally retarded, and thus cannot be executed. The Circuit Court, Justice Ruth Bader Ginsburg wrote in Bobby v. Bies (08-598), was wrong in issuing a habeas ruling that inmate Michael Bies’ retardation had already been established. Bies must pursue that claim in a pending state court case, the Justices concluded.
In an unsigned (”per curiam”) opinion, the Court by a 7-2 vote overturned a $5 million jury verdict in favor of a railroad employee who had work-related brain damage and lung diseases after prolonged exposure to a toxic chemical in the workplace. The Court found that the jury should have been instructed on what the worker must prove to show a genuine fear that he would someday develop cancer. The ruling came in CSX Transportation v. Hensley (08-1034). Justices Ginsburg and John Paul Stevens dissented.
The article gives a brief summary on the case and some implications form a pro-SwPat perspective. Indeed, in the end there's somewhat surprising summary of the effects of the ruling: "Thankfully, it does not seem as if Bilski has turned out to be the impediment to patentability that many feared, at least insofar as the Untited States Patent and Trademark Office is concerned. In fact, based on what is going on at the USPTO one could make a convincing argument that it is actually getting easier to obtain patents that relate to software and computer related processes."
Comments
saulgoode
2009-06-03 00:15:07
Roy Schestowitz
2009-06-03 05:49:10
Jose_X
2009-06-03 02:02:50
Some of the relevant text from the website: *****
The Supreme Court agreed on Monday to consider writing a new, modern definition of the kinds of inventions that are eligible for patent rights. It accepted for review a test case on when, if ever, an invention that involves a method of doing business can be patented. The decision in the case of Bilski, et al., v. Doll (08-964) will come in the next Term, starting Oct. 5.
...
The Bilski patent case that the Court agreed to hear will give the Justices the option of defining the scope of patent law to take account of the impact of the digital age on developing software that controls human activity, such as making business decisions about how to scope with economic risk. The U.S. Solicitor General had urged the Court to bypass the case, but the Justices did not take that advice. The case has generated hot, and even worldwide, controversy since two inventors in 1997 filed an application for a patent.
...
..The case is expected to draw a larger number of friend-of-court briefs on both sides of the patentability question.
Yfrwlf
2009-06-03 04:03:45
Roy Schestowitz
2009-06-03 05:48:03
Jose_X
2009-06-03 21:21:46
Post Bilski, some patent supporters have pointed to past court decisions (1952?) where different use of the same device was recognized as patentable. The idea being that a computer invention could be subject to an endless stream of patents depending on the software it is configured to run. Of course, by analogy, the pencil and paper should be subject to literature patents for every piece of literature that would violate a hypothetical literature patent that also specified its composition through either of those means.
All software is immediately leverageable/cloneable by peers. Can you take a physical invention and clone it instantaneously, modify it, and reproduce it without limit as you put it on the Internet? No. You can't do that to a building. You can't do that to a pencil.
Software, literature, etc, are very different. These fall under copyrights. They should not be patentable at all. Any exceptions considered would still have to be constitutional.
I see software patents as unconstitutional. They do not promote the progress of the sciences and useful arts.
As Red Hat mentioned in the EU response pdf found here http://press.redhat.com/2009/04/30/old-world-and-new-world-software-patent-problems/ , a lot of formal research also suggests patents are not equitable and don't promote advancement.
No matter how technical the issue in play or type of subject (legal, fiction, software, math...), putting monopolies on their research hampers, and potentially hampers very very much.. on a per incident basis. The only possible justification would be if such an idea would not be discoverable by others for a very long time and the lone genius would otherwise not share it for a very long time (the second is very likely false since there is much to be gained for getting credit to a discovery or for exploiting it quietly to get a head start on products). However, we know that any invention leverages much from its surrounding context. It uses many bits that were not patented by others and to which everyone else has access as well. We know that much value is being denied by shutting off an avenue to a single entity, especially since some of these avenues are extremely huge (very broad or basic patents). Again, not only is the lone genius idea almost laughable, but what research on a per patent basis is the USPTO doing to deduce that it would be a very long time before the rest of society could repeat the trick? Very little. They don't have the time, resources, expertise, or crystal ball necessary so they don't much try.
People that think that really great software inventions should be patentable (never mind who would be the judge of "really great" or be judge over whether the invention description is too broad) should watch the 2 part youtube video of Stallman http://www.youtube.com/watch?v=aE7BQXzDWzA . Pay attention to the analogy about Beethoven. In short, if people continue taking out patents and these were to be enforced more harshly, we would really hold back progress, as most of the patented material is needed in order to build the really great stuff. These inventions would also be rediscovered repeatedly as many times as necessary within the 20 years the patent protection runs.
Oh, and there are many more inventors today than there were in the 1800s. Education, physical resources (only a PC), access and collaboration (eg, the Internet), lower costs overall, etc, have greatly increased the number of inventors and contributors in this 21st century, especially when concerning software inventions, over centuries past (especially compared to when patent law was mostly written). Monopolies over ideas get in the way of a great many inventors. Sure, patents are a great incentive to some, but the marginal gains from those that require patents to invent and do get them is nothing compared to all that is lost from everyone else that is barred from those patented areas.
I mean, it makes no sense to give the first person to draft an application a patent. Every idea has to have a first (never mind that the actual first to invent possibly didn't seek a patent). Why is the government getting in the way by knocking down the (good, modest, ordinary, or even brilliant) ideas future engineers can rediscover and leverage? It makes no sense to punish everyone in order to reward one person or a few for that matter. Brilliant discoveries happen all the time. Tomorrow's brilliant discovery requires many prior brilliant steps (and we all know many patents even fall way short of the brilliant mark).
Patents are also very unfair to those that make many inventions all the time but don't have the time, money, savvy, or inclination/desire to seek patents. Their fruits help everyone, yet they get the short end of the stick the moment someone else patents a less worthy idea but which hampers their work and further progress. Time taken to write up patents is time not spent developing the ideas further, looking for new discoveries, nor implementing solutions and otherwise filling in the very important and time consuming details.
Imagine if Knuth, who is against patents, had been patenting all along? How many over the years have used the many inventions he first laid out and even made many great inventions in their own right? Imagine if Beethoven had not been able to build a sizable music piece because of patents? How does society benefit? What about law, literature, comics, etc? What if people had been patenting the technical aspects in those fields?
As a final other possible rebuttal, some will say that software creates technical functional specifications. "It's engineering and not literature." However, what makes "engineering" magical and automatically worthy of patents? "Process and machines" are terms used to describe technical creations in all sorts of fields, including for many forms of creative and noncreative writings. What today goes for a cooking recipe (presumed to be unpatentable), tomorrow becomes the actual program source code that gets that recipe to be generated by an oven. The recipe as English source code gets translated into the bits required to drive the machinery that creates the food. So it is the case that the source code is often hiding many of the actual details to interface with the hardware. It's only the accumulated effects of lots of software+hardware translation layers that allow a high level description of something to lead to an actual product (when combined with the physical inventions). It's relative to the adjoining technology. Of course, this is a side argument. Patent law cannot be unconstitutional, first and foremost, no matter what terminology is used and what characteristics are described in the patent application. After all, much is accomplished today as in the past through competition and not through monopolies that make it illegal for anyone else to even attempt to compete. It really is difficult to argue that giving a single entity the exclusive power to develop and idea will help society or the economy more than when everyone is allowed to participate and make their overlapping contributions.
[And, to repeat, let's not forget that copyright protections (debateable itself as it is also a type of monopoly) already applies to software.]
The reason patent law has stood the test of time is because they were applied to inventions that were out of scope for most people/inventors. It's time-wise and money-wise expensive to reproduce gadgets. These also use up limited resources. Inventions traditionally patented were not something you or I were likely to want to nor be able to help create, refine, and distribute tomorrow. These physical inventions contrast with literature, math, and many areas where patents have been kept out. It's so hard to argue that software does not belong with literature and math but instead belongs with watches, paper, pens, and book creation and other industrial processes. In almost all categories, the traits are shared with the former and not with the latter.
A word on the patentability of other things: It has been argued that all patents hurt rather than hinder. I am not taking a position on that question but simply wanted to contrast software with traditional classes of inventions that have been patented. Software has fewer pros and a greater number of negatives with respect to patenting than do material inventions.
Some "proof is in the pudding" to end: FOSS and many nonFOSS examples have shown that humans are more than willing to invent (especially within a quasi-capitalistic economy) without the (anti)incentive of extremely long-lasting patent monopolies. When we deal with ideas and products of the intellect, sharing is commonplace. It's even more beneficial to share when we talk about functionality because we contribute a little but gain back a lot that is immediately useful to help us carry out our work. Money can be made in numerous ways, for example, at the last mile as we turn these gains to solve specific problems for others. Keeping patents away from software and ideas lowers a lot of friction in the creation process. [Heck, it allows creation!] As a partial analogy, the government itself recognizes that most taxes should be reserved to the last mile. That is how you promote rather than hinder social, economic, scientific, artistic, etc, despite the government's needs for funds. We can make money on software but patents are not the way this should be done if we want to promote the progress of the sciences and useful arts.
Dale B. Halling
2009-06-04 23:43:10
Jose_X
2009-06-05 05:40:06
The US Constitution requires that any patents given must "promote the progress of science and the useful arts." This may be the case for hardware inventions (perhaps not even for circuit design that use existing technology) but certainly is not the case for inventions implemented as software.
See also the thread started by a comment with subject line "Is this really a strong argument?" "Authored by: Anonymous on Thursday, June 04 2009 @ 07:16 AM EDT" here http://www.groklaw.net/article.php?story=20090603224807259 . This thread is headed by this comment that argues a bit like your argument that there are many common points between software inventions and inventions in other areas. The focus there is wrt the dependence on mathematics, but the theme is that many or all patents share many things. The conclusion of that comment is along the lines of what I stated here that progress is hampered at least in the case of software if not also in other areas.
Jose_X
2009-06-05 05:56:00
The argument is against calling a computer running the software as the "machine" called for by US patent law. [However, patent law also allows "processes," so this argument is not complete. It's still interesting.]
>> This paragraph is failure to sort out the mathematics themselves from the meaning of the mathematical formulas. In the 1960s the engineers designing aircrafts were using sliding rules to perform their calculations. The mathematical formulas applicable to aircraft were not describing the sliding rules. They were describing the aircraft.
>> Nowadays the same engineers are using computers instead of the sliding rules. The mathematical formulas are not describing the computer any more than they were describing the sliding rules. They still describe the aircraft though. An aircraft is patentable subject matter according to section 101 of US patent law because a machine is listed there and aircrafts are machines. The mathematics describing the aircraft is not listed as patentable subject matter in section 101. Whether or not a computer is involved in designing the aircraft is irrelevant. Mathematics is not patentable while the aircraft is.
>> The mathematics that describe the computer are the boolean logic used to construct the gates in the circuitry. They are not the program running on the computer. The programs are mathematics describing whatever application domain outside the computer the program is used for, aircraft design in this example. This failure to identify the real meaning of the program is common among legal types that wish to argue that since aircrafts are mathematically described and patentable then so must be computers running software.
saulgoode
2009-06-05 02:38:42
Let me ask you a question, Mr Halling. If someone patents a mechanical invention -- for the sake of argument, let's say it's a cableless suspension bridge -- would it then be considered patent infringement for an unlicensed party to write a computer simulation of that bridge, mathematically modeling its behavior under various stresses and environmental conditions?
If your answer is "yes, a computer model of a patented invention requires a license" then surely patent examiners should need to consider real world devices as potential prior art of any software which seeks to be patented. Your solution "to add features easily" could only be patented if the "time critical" hardware solution never existed. You could not patent software which modeled our hypothetical bridge because the bridge was already designed. You couldn't patent any software solution to any problem which has already been solved with electronic circuits (those that have been published, anyway).
If your answer is "no, computer simulation of a patented invention does not require a license" then how can software which does the same thing as some patented software ever be accused of infringement? If some software was written that simulated our cable-less suspension bridge and this software was patented, another programmer's implementation of that same simulation is in reality merely "simulating" the patented software -- which was decided would not be considered infringement.