First, the actual claims are considerably narrower than just 'any and all uses of sparklines.' The broadest claim is about the use of sparklines in a dynamically updated electronic document. Most of the narrower claims have to do visual effects, the handling of null values in the spreadsheet, etc. This is pretty tame stuff.
Second, this is a newly filed application. The examiner will almost certainly come back with multiple rejections based on obviousness, and the claims will likely be narrowed in response. Like most negotiations, the parties start off with extreme positions and work towards compromise.
Third, the patent application already cites Tufte (along with a dozen other pieces of prior art) in the Information Disclosure Statement. In other words: Microsoft gave the patent examiner many important pieces of prior art. The examiner will no doubt find many more. This is all publicly available through the Patent Office's Patent Application Information Retrieval system.
Fourth, there is no need for Microsoft to acknowledge Tufte as an inventor on the patent application. Inventorship in the patent context is a legal term of art with a specific meaning. The fact that Microsoft said that Tufte invented sparklines is not the damning piece of evidence many are assuming it is (and recall from point three, above, that Microsoft acknowledged Tufte in its IDS). First, Tufte invented sparklines more than a year before the filing date, so any patentable claims must be a non-obvious improvement upon or use of sparklines, not sparklines themselves. Second, Tufte clearly did not work with the Microsoft inventors, so he cannot be a co-inventor of anything claimed in this application.
Once again non-experts hear hoofbeats and scream 'Zebra Stampede!' The comments on Tufte's site, for example, are a joke, an absolute mess of uninformed speculation. Given the wealth of publicly available information on patents and patent application, the Slashdot editors should do more to fact check these stories before publishing them.
Finally, I'll just tack on that if sparklines are so great and this is all so obvious, then surely there's an open source version that predates this application. Remember, though, that this application was filed on May 7, 2008, so the open source version would need to predate that, preferably (but not necessarily) by a year or more. That would actually be an important piece of prior art.
From the essay: "Now, most of us don't have the expertise to comment on the legal technicalities"
Mr. Haselton is, as far as I can determine, not an attorney and has no formal legal education. So bear in mind that the above statement applies to the author of this essay as well.
You know how Slashdot contributors often bemoan poor science journalism written by reporters who obviously don't understand the subject matter? The same danger exists when people like Mr. Haselton, who is a freelance programmer, try to analyze and report on legal issues.
Again, from the essay: "But in the game of analogies, we're all experts, insofar as we're qualified to comment on...whether our "expectations of privacy" in the two areas are similar."
The expectation of privacy is a legal term of art. It does not simply refer to the individual's subjective feeling about whether he or she, personally, expects that a given communication, act, etc will or should be private. So, no, we are not all necessarily qualified to comment on the similarity of the expectation of privacy in two areas because there is a second, objective component of the expectation of privacy. The objective component is highly context-dependent, and its contours have been defined over the years by numerous court cases, none of which Mr. Haselton has cited, distinguished, or applied here.
And this is the glaring issue with Mr. Haselton's essay: he has analyzed the opinion in a vacuum. He does not cite or apply any supporting precedent or statutes, nor does he distinguish the facts of the case from the precedents that the judge cited. This kind of reasoning is not legal reasoning, and it can easily lead to all kinds of errors.
Note that I have, apart from the meaning of 'expectation of privacy,' refrained from critiquing the substance of Mr. Haselton's argument. It is possible that his argument could well win the day in an appeal; on the other hand, perhaps it is hogwash. I merely want the readers here not to be mislead into thinking that this is a rigorous legal argument or that Mr. Haselton is some kind of expert on the subject matter. Indeed, his lack of citations or argument from precedent would probably get him laughed out of court.
Even a cursory reading of the linked articles would show that this has almost nothing to do with patent protection, which lasts for 20 years from the date of filing regardless of the subject matter. This is all about regulatory exclusivity from the FDA. An example of regulatory exclusivity is new drug product exclusivity, which generally lasts for 5 years for completely new drugs and 3 years for new formulations of existing drugs. Another kind is the 180 day generic exclusivity for the first generic to market, which encourages generics to be made by giving them a small window of high profitability.
The issue here is whether biologic drugs should be given a longer than usual regulatory exclusivity period given that (so the argument goes) they are a new, experimental technology that is harder to develop than traditional small molecule drugs.
You might ask "if a drug is patented, then why is a (shorter) period of exclusivity even necessary?" Here's an example: inventor discovers a new compound that might be a useful drug. A patent is filed in 2000. Then 10 years go by while the inventor struggles to find the optimal dose and delivery mechanism. Now in 2010 the inventor's startup starts looking for a partner to bring it through trials and into production. 5 years later, human trials start. 3 years later the drug is approved by the FDA for sale. Now it's 2018 and the patent only has two years left. If the manufacturer has to recoup all of its costs in just 2 years, the price will have to be extremely high, which will limit the drug's availability. So the regulatory exclusivity period gives drug makers a guaranteed 5 years in which to recoup their costs.
So that's the argument for having an exclusivity period. There are also arguments against it, of course, but the main point is that all of this is only tangentially related to patent protection and has nothing whatever to do with a special patent rule for biologics.
The problem with that is that a new sorting algorithm would be novel. The application of that algorithm: not so much. Because you'll end up with "Use of bubble sort algorithm in the display of user selectable menu items", "Use of quick sort algorithm in the display of user selectable menu items", "Use of merge sort algorithm in the display of user selectable menu items".
Novelty is not the only bar to patentability. Non-obviousness is also a requirement. So the mere substitution of a new algorithm that accomplishes the exact same result (i.e., produces the list of sorted items) as the old algorithm is likely to be found obvious, particularly if it does so in a similar way in terms of space and time efficiency. The use of an existing algorithm (e.g., all of your examples) would definitely be found obvious.
For a better example, consider the problem of web search. Google's PageRank algorithm did not solve an unsolved problem, but rather represented a new, useful, and non-obvious improvement over existing solutions to the problem (i.e. it produced significantly better results). If another company develops a new search algorithm that is a new, useful, and non-obvious improvement over PageRank, then it could be patented as well, despite solving the same problem.
"And by what measure do you claim that the software arts continue to improve significantly faster than other areas of technology?"
By the sheer vast numbers of patent applications filed for inventions in the software arts.
That doesn't actually tell one much. Which is the faster moving area of technology: one with 1000 tiny incremental improvements per year or with 1 giant, paradigm-shifting leap per year? Or look at it this way, which moves faster: an area with 1000 improvements to technology used by one person or an area with one improvement to technology used by a million people? You may be right, but that evidence is insufficient.
The bar for entry into the software arena is set so low that somebody's twelve-year-old kid could conceivably infringe a patent without even realizing it.
Is that actually a problem? Patent infringement is not a crime, and suing for patent infringement is quite expensive. What business would think it worthwhile to sue a twelve-year-old kid, especially one that isn't actually trying to make money from his or her infringing program? Property owners allow low-value infringing uses all the time. This happens both in the real world (e.g., coffee shops letting non-customers hang out all day, bookstores that don't care if people read a whole magazine or paper then leave) and in the patent world (e.g., the open source patent apocalypse that somehow never happens). It's just not worth it to go after low value targets.
What if a particular software concept is so necessary in a particular subfield that being barred (legally or financially) from using that concept would substantially hinder the progress of that subfield? Such a patent would be extremely valuable, ensuring that such progress would be hindered for a full 20 years.
That's a very hypothetical argument. For it to be persuasive, this must not only come to pass but happen with sufficient regularity to outweigh the benefits of the patent system. Furthermore, it's contradictory. If a particular concept is necessary in a particular field, then how it can also be new and non-obvious and thus patentable? If someone invents a concept that creates a whole new field of activity which depends upon the concept, then surely we should encourage such pioneering inventions with a strong patent system.
An algorithm alone may be math, but the use of an algorithm to solve a problem is not math. For example, consider a sorting algorithm. I can think about it, describe it, analyze its running time and space usage, I can prove it correct. That's all math, but none of it will actually sort numbers, much less solve a problem that relies on sorting numbers. Once an algorithm is applied to the solution of a problem, it ceases being math and becomes engineering.
Engineers use math all the time to do things like design a bridge that won't fall victim to harmonic collapse. If someone finds a new, non-obvious use for an algorithm or equation that solves a problem, that should be patentable. Software is not special.
None of this will take math or algorithms away from mathematicians or computer scientists. As the saying goes, "real computer scientists don't write in anything less portable than a number two pencil." There is no particular reason to think that the patentability of software will hold back research.
I agree that software shouldn't be patentable (either directly or through the various loopholes that applicants use to get around the fact that software, when claimed directly, is not a "process, machine, manufacture, or composition of matter").
How is a software algorithm ("a process or set of rules to be followed in calculations or other problem-solving operations") not a "process?" How is a computer which has been programmed in a particular way not a "machine?"
But in my opinion, this should be a matter of policy motivated by the fact that the rate of improvements in the software arts is far too fast to permit 20-year terms of patent protection
If a patented algorithm becomes obsolete because of the fast rate of improvement in the software arts, then the patent owner will allow it to lapse in order to avoid paying maintenance fees on a worthless algorithm.
And by what measure do you claim that the software arts continue to improve significantly faster than other areas of technology? Most fundamental operations (e.g. searching, sorting, calculations, storage and retrieval, etc) either have optimal algorithms or existing algorithms have been proven to be nearly optimal. No one is talking about patenting quick sort or the hash table. Almost everything you learn in an undergraduate CS curriculum was invented decades ago. The vast majority of software patents are solutions to very specific problems (e.g. DRM) or represent tiny improvements or modifications to existing methods.
And even if that is demonstrably the case, what about other fast-moving areas of technology like biotech and materials science, specifically nanotechnology? Should those also have special rules? How fast is too fast?
Furthermore, I would suggest that the faster an area of technology moves, the less it matters how long the patent term is. A patent in such an area will quickly become obsolete. It's actually stagnant technology where a long patent term on a rare innovation is most valuable.
Knuth's argument misses the point. The distinction is not between mathematical and non-mathematical algorithms, but rather between an algorithm in the abstract and an algorithm as applied to a real world problem. An algorithm, in and of itself, lacks the utility required for patentability. Once applied to solve a problem, however, the invention is no longer the algorithm per se but rather its useful application, which should be patentable.
Suppose one invents an algorithm for efficiently solving systems of non-linear equations. Alone that algorithm should not be patentable because merely thinking about equations in the abstract is not useful. But if one applies that algorithm to, say, efficiently simulating the motion of fluids or forecasting the weather, then it becomes useful.
To avoid charges of 'thought-crime' one can always institute a requirement that the application of the algorithm be implemented on a computer, but that's a mere formality. Few new, useful, and non-obvious software algorithms can be used effectively by the human mind alone, and it would be virtually impossible to prove infringement. Furthermore it would be a PR disaster.
Red Hat's argument is also very weak. Software patents are often of low quality primarily because the Patent Office long resisted hiring patent examiners with computer science backgrounds. In fact, it is still much more difficult for someone with a computer science degree to become a patent examiner, patent attorney, or patent agent than it is for someone with a degree in chemistry, physics, biology, etc. As a result, the Patent Office is understaffed and many of the examiners are underqualified.
Another reason software patent quality suffers is because patent examiners often first look to patents and patent applications as sources of prior art. Because software has only been patentable for a couple of decades, there is a much smaller body of readily accessible prior art. Examiners can and do look to journal articles and other sources, but time and budget constraints make thorough searching difficult.
The problem Red Hat describes is not entirely confined to software, either. Patents of all stripes have dubious or overly broad claims. The answer is not a patchwork of allowed and disallowed subject matter or different rules for different kinds of technology. Better answers are to either eliminate the presumption of validity that patents enjoy or to tighten the rules on enablement and written description, which would mean that applicants could only make narrower claims closer to their specific implementation and not to the broader class of invention.
A handful of patents are not worth much against the thousands of patents the small [inventor] may be potentially violating.
Which is why Microsoft, with its thousands of patents, so easily struck down Google when all it had was a couple of patents on search technology, effectively cutting Google off at the knees and leaving Microsoft free to dominate the search engine market.
Oh wait, no, that's not what happened at all. In fact, as we've just seen with the i4i case, a single patent owned by a small company was worth millions of dollars against Microsoft and its thousands of patents.
That kind of statement sounds good if one doesn't think about it too hard, but it's not backed up by reality.
That's true, but for a lot of kinds of products it helps a great deal or is a practical necessity. This is true of anything that is easily reverse engineered and especially true of anything that is substantially cheaper to make than to develop. Pharmaceuticals are the classic example, but it's also true of commonplace inventions like coffee cup sleeves, for example.
If you are already making and selling a product, no one can jump in and buy a patent for it and try to sue you (well, they can try).
That's true, but the purpose of a patent is not to prevent others from excluding the inventor from making and selling something. It's to allow the inventor to exclude others.
Thus, a patent is a great idea if one wants to sell, for example, coffee cup sleeves to coffee shops all over the world without worrying about competitors underselling you because they didn't have to develop the sleeve, only copy it. Globalization and the internet make it very easy to sell sleeves all over the world, but only patents give the inventor the ability to exclude parasitic competition.
No more than 7 years on a patent. No extensions. No exceptions.
I'd love to see the economics research behind that number. Must have been a lot of work determining the optimal patent term. I suspect it'll be no trouble getting published. Or, as is more likely, was that arbitrary number pulled out of the air?
As for 'no extensions. no exceptions,' what about delays brought about by the patent office? Surely you wouldn't penalize the inventor for bureaucratic incompetence?
No patenting of algorithms
Algorithms are already unpatentable. An algorithm, alone, is not useful, and so it fails the requirement of utility. What is patentable is, according to the Federal Circuit, the use of an algorithm tied to a particular machine to accomplish a useful result. I suspect the Supreme Court will probably overrule the Federal Circuit and allow the patenting of the practical application of algorithms.
I can think about PageRank all day long and accomplish nothing. But if I apply PageRank to pages on the internet and use it to optimize searches, then it becomes a patentable invention.
Patents to be awarded to individuals only, not companies
This is already the case in the US. Patents can be assigned to companies, but only individuals can apply for and receive them.
I suppose you meant that companies should be prevented from owning patents at all, but that would be pointless. Employees would simply be required to license the invention to the company exclusively. It would only add transaction costs.
A unified patent system would actually benefit individual inventors, small businesses, and startups more than established players with deep pockets. Right now if one wants to file for patent protection in every country with a patent system worth the name it costs ~$200,000 in filing fees alone, to say nothing of attorney and translation costs. The lifetime maintenance fees of that single patent will be well into the millions. Even only filing in the 'big three' of the US, Europe, and Japan typically costs well over $100,000 in government fees and attorney fees.
For a big company like Microsoft, that's just the cost of doing business. But that $200,000 is a huge amount of money to a startup, to say nothing of an individual garage inventor. Globalization and the internet mean that an inventor can sell an invention to people all around the world for far less than it would have cost 20 or 30 years ago. Protecting that invention all around the world, however, is often prohibitively expensive for all but the most well-funded, established companies.
It's true that companies like Microsoft would also benefit from lower filing costs, but small companies and individual inventors will benefit much more. It will also mean less money wasted on lawyers, as a single attorney in a single country can handle the whole process instead of having to use attorneys all over the world. And of course it will mean less duplication of effort in government as patent offices share resources. Right now there is an enormous duplication of effort as each application in each country is met with the same prior art, which is overcome with the same arguments. This is a tremendous waste of both government and applicant resources.
That's a pretty absolute assertion you're making. Do you really believe that software patents are the only way knowledge of software inventions gets transferred/created in our society?
That's not what I asserted. I asserted that if an algorithm is kept a trade secret and the source code closed then the public will never find out about it (barring, of course, illegal disclosure of the trade secret or independent discovery by another party that then discloses it).
And of course software patents are not the only way knowledge of software inventions gets transferred. Open source and academic publications are two other routes. The problem with the former is that open source is not particularly innovative. A study of 500 mature, actively developed open source projects that only 1% of them contained a new technology and addressed a new market or new user need. Open source is good at providing good, cheap software, but it is not particularly good at innovating.
Academic publication is all well and good, but ultimately someone has to pay for labs and salaries. Software patents step in as a way for academic discoveries to be commercialized. For example, some of the fundamental patents underlying Google are owned by Stanford, which is where Page and Brin invented PageRank and other related technologies. Google pays Stanford a non-trivial royalty for the exclusive license to those patents. Thus, the patents on successfully commercialized technology fund more academic research.
The huge diference is that you cannot use a TRIVIAL mechanical process and patent it
The test of patentability is not "non-trivial" but rather new, useful, and nonobvious. In any event, there are lots of 'trivial' patents that aren't business methods. There are, for example, dozens of patents on coffee cup sleeves, toilet paper, plastic cups, etc. For the most part these are small, some would say 'trivial' improvements over the prior art. But as long as they are new, useful, and non-obvious, the fact that they represent only a very small improvement is not important.
A lot of those business process patent are either trivial, already in use for decenny, copy of something existing but-with-computer or just plain mathematical application. You do not have such a problem with mechanical device patent.
If the business method patents really are obvious then they are invalid under 35 USC 103. There is no need to have a specific exclusion of business method patents just become some (or even a lot) of business method patents should not have been granted. That's throwing the baby out with the bathwater.
A better approach is to remove the presumption of validity of all patents, which would make it much easier to invalidate bad patents and reduce the incentive to seek questionably valid patents.
Submarine patents are no longer a serious issue. Patents now expire 20 years from the date of filing, so keeping a patent application stuck in the patent office forever is a good way to end up with an expired patent, which is not particularly useful. Furthermore patent applications are now published after 18 months, so the world is put on notice fairly early on in the patent application process.
There may be a few submarines left from the pre-June 8, 1995 days, but as an ongoing problem they are nonexistent.
It could be something a lot of people were already doing, but if they had not documented their process in a way the patent office understands as "prior art" (which, as we've learned, is pretty much nothing), the patent can STILL stand.
The PTO, of course, limits itself to (easily accessible) printed publications, patents, and patent applications because those are all easy to use forms of prior art. But prior use is normally very easy for a defendant in a patent infringement suit to show except when the prior user decided to keep its use secret rather than disclose it in a patent application, printed publication, or through public use. The 'gap' that 35 USC 273 fills was an intentional motivator for the disclosure of otherwise secret business methods.
The 35 USC 273 prior use defense to business method patents is primarily meant to protect those whose prior use was secret. It comes up when Party A invents and uses a method secretly, then Party B invents and uses the method secretly, then Party A files for a patent more than a year later.
Normally Party B could use its prior use to invalidate the patent under 35 USC 102(b) except its use was not public. Party B could also invalidate the patent through disclosure in a printed publication or a patent application but again it chose to keep it secret. Because Party B kept the invention secret Party A gets the patent, although 35 USC 273 gives Party B a defense to patent infringement.
This presumes that the crass entrepeneur is the driver of innovation in software.
The PageRank algorithm on which the Google empire rests was invented by Larry Page and Sergei Brin. They published the algorithm, patented it, and went on to become incredibly successful 'crass entrepreneurs.' Google owns dozens of software patents on key technologies such as its Map-Reduce implementation and AdWords. Few would say that Google is not a driver of innovation in software.
Usually it's the exact opposite sort of person that drives software or scientific innovation.
That's a fairly strong claim. Would you care to provide evidence that innovators are not motivated by financial reward, which is typically either received directly through their own crass entrepreneurship or indirectly through the commercialization of their innovations by others?
A general tool of almost any kind, used in a different way, is _not patentable_. It's using the tool for what it was designed for, it does not change the physical construction of the machine nor its number of parts nor its general capabilities.
This is not entirely accurate. Suppose there is a patent on Compound X and the patent discloses the use of Compound X as a floor wax. Now suppose someone discovers that Compound X makes an excellent treatment for baldness, which is entirely nonobvious given its previous use as a floor wax. It is true that the inventor cannot patent Compound X again just because a new use has been discovered. But what the inventor can patent is a method of use of Compound X as a baldness treatment, specifying the dose and treatment regimen. This is true in the US though not true in many other jurisdictions.
The "actually implemented algorithm" is protected in its specific implementation by copyright, and should not receive the double protection of patent law against writing similar algorithms.
The protection provided by copyright is very narrow: it extends only to the literal source and machine code, not to the algorithm generally. For example: suppose one discovers a new efficient algorithm for simulating fluids. This has many potential applications: special effects, computer games, weather simulation, etc. Now suppose one publishes the algorithm and implements it in a specific program: a computer game. Nothing in copyright law stops everyone in the world from implementing the published algorithm in their own games, weather simulators, special effects software, etc. So long as they don't directly rip off the inventor's code but instead create their own implementation of the idea, they are in the clear.
So what's a business to do? A common thing to do is keep the algorithm a secret and keep the source closed. The public will never learn about the algorithm.
Software patents step in to fill the gap left by copyright. A patent on the fluid simulation algorithm protects any implementation of it in any domain. This allows the inventor to sell or license the rights to the highest bidder in each field. In the mean while, the public gets to see the algorithm published as a patent, and they are free to create and patent improvements as well as design around it. When the patent expires, the public is free to use the original idea. If the idea becomes outdated (e.g. because an open-source competitor designs a better alternative), then the patent will be worthless, the owner will stop paying the maintenance fees, and the invention will enter the public domain ahead of schedule.
You declare ownership of a new business process. Because of this you can prevent EVERYONE ELSE WHO DOES BUSINESS from benefiting from that new process.
Suppose someone invents a new mechanical device and patents it. The patentee can now prevent everyone else from making a business out of making, using, selling, offering to sell, or importing the patented device. In that sense, all patents declare ownership over a new business process. Business method patents are just applied economics in the same way that mechanical device patents are applied physics.
Others are also free to design around the patent by inventing their own substitute mechanical device or business method. Theirs may even be a superior device or method, which is better for everyone. And they have an incentive to do so: not only can they avoid infringement or licensing fees, but they can exclude or license to others in turn.
And of course, no one is actually forced to do anything. A patent is only valid if it claims something new, useful, and non-obvious, which means it can't cover anything that others were already doing. No one ever has to change their business in response to a patent: that which infringes if later anticipates if earlier.
Every patent allows the patent owner to cause trouble for everyone else for the next 17 years.
20 years from the date of filing.
The most frightening idea in all of those amicus briefs is the idea that medical procedures might be patented...Doctors invent because they take their oath to Hypocrates seriously, not because they identify with Crassius Maximus.
Medical procedures can already be patented, but practitioners and 'related medical entities' (e.g., the hospital, clinic, etc where the practitioner works) has a defense against infringement. Anyway, lots of doctors invent out of a sense of self interest as well as a desire to help people. I work with medical researchers on a regular basis, and the commercialization of their work is what ultimately funds the labs they work in.
Just FYI, it's "Hippocratic oath," and it's not sworn to Hippocrates but rather to various gods, particularly Apollo.
There is not a positive thing that has come out of software patents.
Google was able to rise to dominate search and challenge Microsoft primarily because of its numerous software patents. Without those patents, Microsoft, Altavista, Yahoo, or any other company could have used Page Rank and other algorithms covered by Google's dozens of software patents. Only part of Google's success is its good web design and marketing savvy. A huge part is the fact that it just plain does search better than almost anybody else.
Remember: Microsoft rose to dominance at a time when software was not patentable. That's one of the reasons why Embrace, Extend, Extinguish worked so well: without software patents there was nothing to stop Microsoft from embracing the new inventions of others.
The quality of software hasn't improved
Software quality (as in lack of bugs, bloat, etc) is not something that software patents address, at least not directly. What they address is the invention and commercialization of new features.
patent trolls are numerous
They are not particularly numerous, nor are they particularly unique to software. Only about 3000 patent infringement cases are filed in the US each year, and only about 300 of those go to trial. Trolls are a very small subset of that and hardly numerous. Furthermore, the existence of trolls has more to do with the presumption of patent validity and the inability of the PTO to find every piece of prior art than it does the patent eligibility of software implemented inventions.
because of software patents there are many incompatibilities in software that lead to many, many problems.
That's a very strong claim. Do you have evidence of it?
Can someone please tell me how software patents are promoting the progress of science/useful arts?
Copyright protection of software only extends to the specific implementation (i.e., the literal source and machine code). Patents protect the idea behind the implementation. This is an important distinction.
Say an inventor comes up with a brilliant new algorithm for efficiently and accurately modeling fluids on a computer. There are many possible applications: computer graphics, weather simulation, etc. The inventor would like to maximize his or her profit from the invention. Now, with copyright the inventor could write a program for each and every possible use and form a company to sell copies as copyrighted programs. But that's an awful lot of work for one person, especially if the inventor is a good inventor but a bad business person or doesn't know much about some of the possible application domains.
With software patents, the inventor can patent the invention and then either sell the patent to an established company or license the patent to different companies for different uses: one company gets the rights to use it in computer games, another to use it in weather simulation, etc. This lets each company best use its comparative advantage, which maximizes the benefit to consumers while minimizing inefficiencies.
Now you might say the inventor could keep the invention a trade secret and give it to the companies under a non-disclosure agreement. That's true, but then the public would never find out about the invention. The beauty of patents is that they let inventors and commercializers profit from an invention while still disclosing the invention to the public, which can freely use the invention after the patent expires. This is all true in software just as it is in more traditional fields of invention. Software is not special, and I say that as someone with both bachelor's and master's degrees in computer science.
Full Disclosure: I work for the team that wrote the amicus brief of Dr. Ananda Chakrabarty (he of the Diamond v. Chakrabarty Supreme Court case that established the patentability of genetically modified organisms).
But this ignores the constitutional requirement that it promote progress:
It is not obvious that business method patents hinder the progress of the useful arts, and merely stating it does not make it so. But if you have a good argument for why business methods are contrary to the constitutional purpose of patents, then by all means compose and submit an amicus curiae brief in support of the Respondent. The due date is October 2nd. You don't have to be an attorney to submit an amicus brief on your own behalf.
In this case the U.S. is one of very few nations to support software patents.
This case is not about software patents. Here are the questions presented in the case:
1. Whether the Federal Circuit erred by holding that a "process" must be tied to a particular machine or apparatus, or transform a particular article into a different state or thing ("machine-or-transformation" test), to be eligible for patenting under 35 U.S.C. 101, despite this Court's precedent declining to limit the broad statutory grant of patent eligibility for "any" new and useful process beyond excluding patents for "laws of nature, physical phenomena, and abstract ideas."
2. Whether the Federal Circuit's "machine-or-transformation" test for patent eligibility, which effectively forecloses meaningful patent protection to many business methods, contradicts the clear Congressional intent that patents protect "method[s] of doing or conducting business." 35 U.S.C. 273.
Now it is true that the case has implications for software patents, diagnostic method patents, etc, but concerns about software patents and the like will be, at most, secondary public policy considerations in the Court's decision making process.
That said, the point about international harmonization is that other countries and the EPO have broad definitions of patent eligible subject matter. The EPO, for example, says "European patents shall be granted for any inventions which are susceptible of industrial application, which are new and which involve an inventive step" with the further proviso that the invention must be of a technical character. Now, the EPO has a specific exclusion for business methods, but it is just that: a specific exclusion. The test of patentable subject matter remains broad. The argument is that the US should do the same: broad patentable subject matter with (if necessary) specific exclusions as determined by Congress rather than the courts. As it so happens, there are many good arguments that Congress has not yet excluded (and might never exclude) business methods from patent eligibility.
Full Disclosure: I work for the team that wrote the amicus brief of Dr. Ananda Chakrabarty (he of the Diamond v. Chakrabarty Supreme Court case that established the patentability of genetically modified organisms).
The 44 amicus curiae briefs that have been filed so far are only those submitted in support of the Petitioners (i.e., the inventors Bilski and Warsaw) or in support of neither party. Amicus briefs in support of the Respondent (i.e., the Patent Office) will be submitted after the Respondent's merits brief is submitted, which will occur on or before September 25. Once the merits brief is submitted, amici have 7 days to submit briefs in support of the Respondent.
While I don't expect there to be quite the same volume of briefs supporting the PTO as the Petitioner, there will probably be at least a few.
Full Disclosure: I work for the team that wrote the brief of Dr. Ananda Chakrabarty (he of the Diamond v. Chakrabarty Supreme Court case that established the patentability of genetically modified organisms).
As is all too often the case here on Slashdot, the summary has seized upon the title of the patent, which has no legal effect whatsoever, while ignoring the actual patent claims, which are all important.
If one actually reads the claims, one sees that the main new part of the invention are the 'hint elements' contained in the XML file. The written description expands upon what hint elements mean: "hints are provided within the XML associated files providing applications that understand XML a shortcut to understanding some of the features provided by the word-processor. By using the hints, the applications do not have to know all of the specific details of the internal processing of the word-processor in order to recreate a feature."
Basically, the invention here is the inclusion of information that lets third-party programs better understand what to do with the format. You can imagine, for example, if HTML included something like this. The del ('strikethrough') tag might be written:
<del hint="draw line 1px horizontal">
That code would allow a program that did not natively understand the tag to implement a simple version of it. The idea is to allow new features to be introduced into the format while enabling older versions of the software to use them without updating their code. The necessary code comes with the file.
Now, whether that's still new and nonobvious, I don't know, but it's a significantly more accurate summary of the invention than "Microsoft Patents XML Word Processing Documents."
First, a citation to the published paper: Hana Yi, et al., Selection of a variant of Geobacter sulfurreducens with enhanced capacity for current production in microbial fuel cells, Biosensors and Bioelectronics, Volume 24, Issue 12, 15 August 2009, Pages 3498-3503.
The extrapolated current density was 7.4 ± 0.1 A/m2. The individual fuel cells produced 14mA, which was sustained for 24 months.
Slashdot Mirror
First, the actual claims are considerably narrower than just 'any and all uses of sparklines.' The broadest claim is about the use of sparklines in a dynamically updated electronic document. Most of the narrower claims have to do visual effects, the handling of null values in the spreadsheet, etc. This is pretty tame stuff.
Second, this is a newly filed application. The examiner will almost certainly come back with multiple rejections based on obviousness, and the claims will likely be narrowed in response. Like most negotiations, the parties start off with extreme positions and work towards compromise.
Third, the patent application already cites Tufte (along with a dozen other pieces of prior art) in the Information Disclosure Statement. In other words: Microsoft gave the patent examiner many important pieces of prior art. The examiner will no doubt find many more. This is all publicly available through the Patent Office's Patent Application Information Retrieval system.
Fourth, there is no need for Microsoft to acknowledge Tufte as an inventor on the patent application. Inventorship in the patent context is a legal term of art with a specific meaning. The fact that Microsoft said that Tufte invented sparklines is not the damning piece of evidence many are assuming it is (and recall from point three, above, that Microsoft acknowledged Tufte in its IDS). First, Tufte invented sparklines more than a year before the filing date, so any patentable claims must be a non-obvious improvement upon or use of sparklines, not sparklines themselves. Second, Tufte clearly did not work with the Microsoft inventors, so he cannot be a co-inventor of anything claimed in this application.
Once again non-experts hear hoofbeats and scream 'Zebra Stampede!' The comments on Tufte's site, for example, are a joke, an absolute mess of uninformed speculation. Given the wealth of publicly available information on patents and patent application, the Slashdot editors should do more to fact check these stories before publishing them.
Finally, I'll just tack on that if sparklines are so great and this is all so obvious, then surely there's an open source version that predates this application. Remember, though, that this application was filed on May 7, 2008, so the open source version would need to predate that, preferably (but not necessarily) by a year or more. That would actually be an important piece of prior art.
From the essay: "Now, most of us don't have the expertise to comment on the legal technicalities"
Mr. Haselton is, as far as I can determine, not an attorney and has no formal legal education. So bear in mind that the above statement applies to the author of this essay as well.
You know how Slashdot contributors often bemoan poor science journalism written by reporters who obviously don't understand the subject matter? The same danger exists when people like Mr. Haselton, who is a freelance programmer, try to analyze and report on legal issues.
Again, from the essay: "But in the game of analogies, we're all experts, insofar as we're qualified to comment on...whether our "expectations of privacy" in the two areas are similar."
The expectation of privacy is a legal term of art. It does not simply refer to the individual's subjective feeling about whether he or she, personally, expects that a given communication, act, etc will or should be private. So, no, we are not all necessarily qualified to comment on the similarity of the expectation of privacy in two areas because there is a second, objective component of the expectation of privacy. The objective component is highly context-dependent, and its contours have been defined over the years by numerous court cases, none of which Mr. Haselton has cited, distinguished, or applied here.
And this is the glaring issue with Mr. Haselton's essay: he has analyzed the opinion in a vacuum. He does not cite or apply any supporting precedent or statutes, nor does he distinguish the facts of the case from the precedents that the judge cited. This kind of reasoning is not legal reasoning, and it can easily lead to all kinds of errors.
Note that I have, apart from the meaning of 'expectation of privacy,' refrained from critiquing the substance of Mr. Haselton's argument. It is possible that his argument could well win the day in an appeal; on the other hand, perhaps it is hogwash. I merely want the readers here not to be mislead into thinking that this is a rigorous legal argument or that Mr. Haselton is some kind of expert on the subject matter. Indeed, his lack of citations or argument from precedent would probably get him laughed out of court.
getting denied health insurance for having bad genes
The federal Genetic Information Nondiscrimination Act is designed to prevent this. There are also supplementary state laws in, for example, Oregon and New Jersey.
Even a cursory reading of the linked articles would show that this has almost nothing to do with patent protection, which lasts for 20 years from the date of filing regardless of the subject matter. This is all about regulatory exclusivity from the FDA. An example of regulatory exclusivity is new drug product exclusivity, which generally lasts for 5 years for completely new drugs and 3 years for new formulations of existing drugs. Another kind is the 180 day generic exclusivity for the first generic to market, which encourages generics to be made by giving them a small window of high profitability.
The issue here is whether biologic drugs should be given a longer than usual regulatory exclusivity period given that (so the argument goes) they are a new, experimental technology that is harder to develop than traditional small molecule drugs.
You might ask "if a drug is patented, then why is a (shorter) period of exclusivity even necessary?" Here's an example: inventor discovers a new compound that might be a useful drug. A patent is filed in 2000. Then 10 years go by while the inventor struggles to find the optimal dose and delivery mechanism. Now in 2010 the inventor's startup starts looking for a partner to bring it through trials and into production. 5 years later, human trials start. 3 years later the drug is approved by the FDA for sale. Now it's 2018 and the patent only has two years left. If the manufacturer has to recoup all of its costs in just 2 years, the price will have to be extremely high, which will limit the drug's availability. So the regulatory exclusivity period gives drug makers a guaranteed 5 years in which to recoup their costs.
So that's the argument for having an exclusivity period. There are also arguments against it, of course, but the main point is that all of this is only tangentially related to patent protection and has nothing whatever to do with a special patent rule for biologics.
The problem with that is that a new sorting algorithm would be novel. The application of that algorithm: not so much. Because you'll end up with "Use of bubble sort algorithm in the display of user selectable menu items", "Use of quick sort algorithm in the display of user selectable menu items", "Use of merge sort algorithm in the display of user selectable menu items".
Novelty is not the only bar to patentability. Non-obviousness is also a requirement. So the mere substitution of a new algorithm that accomplishes the exact same result (i.e., produces the list of sorted items) as the old algorithm is likely to be found obvious, particularly if it does so in a similar way in terms of space and time efficiency. The use of an existing algorithm (e.g., all of your examples) would definitely be found obvious.
For a better example, consider the problem of web search. Google's PageRank algorithm did not solve an unsolved problem, but rather represented a new, useful, and non-obvious improvement over existing solutions to the problem (i.e. it produced significantly better results). If another company develops a new search algorithm that is a new, useful, and non-obvious improvement over PageRank, then it could be patented as well, despite solving the same problem.
"And by what measure do you claim that the software arts continue to improve significantly faster than other areas of technology?"
By the sheer vast numbers of patent applications filed for inventions in the software arts.
That doesn't actually tell one much. Which is the faster moving area of technology: one with 1000 tiny incremental improvements per year or with 1 giant, paradigm-shifting leap per year? Or look at it this way, which moves faster: an area with 1000 improvements to technology used by one person or an area with one improvement to technology used by a million people? You may be right, but that evidence is insufficient.
The bar for entry into the software arena is set so low that somebody's twelve-year-old kid could conceivably infringe a patent without even realizing it.
Is that actually a problem? Patent infringement is not a crime, and suing for patent infringement is quite expensive. What business would think it worthwhile to sue a twelve-year-old kid, especially one that isn't actually trying to make money from his or her infringing program? Property owners allow low-value infringing uses all the time. This happens both in the real world (e.g., coffee shops letting non-customers hang out all day, bookstores that don't care if people read a whole magazine or paper then leave) and in the patent world (e.g., the open source patent apocalypse that somehow never happens). It's just not worth it to go after low value targets.
What if a particular software concept is so necessary in a particular subfield that being barred (legally or financially) from using that concept would substantially hinder the progress of that subfield? Such a patent would be extremely valuable, ensuring that such progress would be hindered for a full 20 years.
That's a very hypothetical argument. For it to be persuasive, this must not only come to pass but happen with sufficient regularity to outweigh the benefits of the patent system. Furthermore, it's contradictory. If a particular concept is necessary in a particular field, then how it can also be new and non-obvious and thus patentable? If someone invents a concept that creates a whole new field of activity which depends upon the concept, then surely we should encourage such pioneering inventions with a strong patent system.
It shouldn't be patentable because it's math
An algorithm alone may be math, but the use of an algorithm to solve a problem is not math. For example, consider a sorting algorithm. I can think about it, describe it, analyze its running time and space usage, I can prove it correct. That's all math, but none of it will actually sort numbers, much less solve a problem that relies on sorting numbers. Once an algorithm is applied to the solution of a problem, it ceases being math and becomes engineering.
Engineers use math all the time to do things like design a bridge that won't fall victim to harmonic collapse. If someone finds a new, non-obvious use for an algorithm or equation that solves a problem, that should be patentable. Software is not special.
None of this will take math or algorithms away from mathematicians or computer scientists. As the saying goes, "real computer scientists don't write in anything less portable than a number two pencil." There is no particular reason to think that the patentability of software will hold back research.
I agree that software shouldn't be patentable (either directly or through the various loopholes that applicants use to get around the fact that software, when claimed directly, is not a "process, machine, manufacture, or composition of matter").
How is a software algorithm ("a process or set of rules to be followed in calculations or other problem-solving operations") not a "process?" How is a computer which has been programmed in a particular way not a "machine?"
But in my opinion, this should be a matter of policy motivated by the fact that the rate of improvements in the software arts is far too fast to permit 20-year terms of patent protection
If a patented algorithm becomes obsolete because of the fast rate of improvement in the software arts, then the patent owner will allow it to lapse in order to avoid paying maintenance fees on a worthless algorithm.
And by what measure do you claim that the software arts continue to improve significantly faster than other areas of technology? Most fundamental operations (e.g. searching, sorting, calculations, storage and retrieval, etc) either have optimal algorithms or existing algorithms have been proven to be nearly optimal. No one is talking about patenting quick sort or the hash table. Almost everything you learn in an undergraduate CS curriculum was invented decades ago. The vast majority of software patents are solutions to very specific problems (e.g. DRM) or represent tiny improvements or modifications to existing methods.
And even if that is demonstrably the case, what about other fast-moving areas of technology like biotech and materials science, specifically nanotechnology? Should those also have special rules? How fast is too fast?
Furthermore, I would suggest that the faster an area of technology moves, the less it matters how long the patent term is. A patent in such an area will quickly become obsolete. It's actually stagnant technology where a long patent term on a rare innovation is most valuable.
Knuth's argument misses the point. The distinction is not between mathematical and non-mathematical algorithms, but rather between an algorithm in the abstract and an algorithm as applied to a real world problem. An algorithm, in and of itself, lacks the utility required for patentability. Once applied to solve a problem, however, the invention is no longer the algorithm per se but rather its useful application, which should be patentable.
Suppose one invents an algorithm for efficiently solving systems of non-linear equations. Alone that algorithm should not be patentable because merely thinking about equations in the abstract is not useful. But if one applies that algorithm to, say, efficiently simulating the motion of fluids or forecasting the weather, then it becomes useful.
To avoid charges of 'thought-crime' one can always institute a requirement that the application of the algorithm be implemented on a computer, but that's a mere formality. Few new, useful, and non-obvious software algorithms can be used effectively by the human mind alone, and it would be virtually impossible to prove infringement. Furthermore it would be a PR disaster.
Red Hat's argument is also very weak. Software patents are often of low quality primarily because the Patent Office long resisted hiring patent examiners with computer science backgrounds. In fact, it is still much more difficult for someone with a computer science degree to become a patent examiner, patent attorney, or patent agent than it is for someone with a degree in chemistry, physics, biology, etc. As a result, the Patent Office is understaffed and many of the examiners are underqualified.
Another reason software patent quality suffers is because patent examiners often first look to patents and patent applications as sources of prior art. Because software has only been patentable for a couple of decades, there is a much smaller body of readily accessible prior art. Examiners can and do look to journal articles and other sources, but time and budget constraints make thorough searching difficult.
The problem Red Hat describes is not entirely confined to software, either. Patents of all stripes have dubious or overly broad claims. The answer is not a patchwork of allowed and disallowed subject matter or different rules for different kinds of technology. Better answers are to either eliminate the presumption of validity that patents enjoy or to tighten the rules on enablement and written description, which would mean that applicants could only make narrower claims closer to their specific implementation and not to the broader class of invention.
A handful of patents are not worth much against the thousands of patents the small [inventor] may be potentially violating.
Which is why Microsoft, with its thousands of patents, so easily struck down Google when all it had was a couple of patents on search technology, effectively cutting Google off at the knees and leaving Microsoft free to dominate the search engine market.
Oh wait, no, that's not what happened at all. In fact, as we've just seen with the i4i case, a single patent owned by a small company was worth millions of dollars against Microsoft and its thousands of patents.
That kind of statement sounds good if one doesn't think about it too hard, but it's not backed up by reality.
You don't need a patent to sell a product.
That's true, but for a lot of kinds of products it helps a great deal or is a practical necessity. This is true of anything that is easily reverse engineered and especially true of anything that is substantially cheaper to make than to develop. Pharmaceuticals are the classic example, but it's also true of commonplace inventions like coffee cup sleeves, for example.
If you are already making and selling a product, no one can jump in and buy a patent for it and try to sue you (well, they can try).
That's true, but the purpose of a patent is not to prevent others from excluding the inventor from making and selling something. It's to allow the inventor to exclude others.
Thus, a patent is a great idea if one wants to sell, for example, coffee cup sleeves to coffee shops all over the world without worrying about competitors underselling you because they didn't have to develop the sleeve, only copy it. Globalization and the internet make it very easy to sell sleeves all over the world, but only patents give the inventor the ability to exclude parasitic competition.
No more than 7 years on a patent. No extensions. No exceptions.
I'd love to see the economics research behind that number. Must have been a lot of work determining the optimal patent term. I suspect it'll be no trouble getting published. Or, as is more likely, was that arbitrary number pulled out of the air?
As for 'no extensions. no exceptions,' what about delays brought about by the patent office? Surely you wouldn't penalize the inventor for bureaucratic incompetence?
No patenting of algorithms
Algorithms are already unpatentable. An algorithm, alone, is not useful, and so it fails the requirement of utility. What is patentable is, according to the Federal Circuit, the use of an algorithm tied to a particular machine to accomplish a useful result. I suspect the Supreme Court will probably overrule the Federal Circuit and allow the patenting of the practical application of algorithms.
I can think about PageRank all day long and accomplish nothing. But if I apply PageRank to pages on the internet and use it to optimize searches, then it becomes a patentable invention.
Patents to be awarded to individuals only, not companies
This is already the case in the US. Patents can be assigned to companies, but only individuals can apply for and receive them.
I suppose you meant that companies should be prevented from owning patents at all, but that would be pointless. Employees would simply be required to license the invention to the company exclusively. It would only add transaction costs.
A unified patent system would actually benefit individual inventors, small businesses, and startups more than established players with deep pockets. Right now if one wants to file for patent protection in every country with a patent system worth the name it costs ~$200,000 in filing fees alone, to say nothing of attorney and translation costs. The lifetime maintenance fees of that single patent will be well into the millions. Even only filing in the 'big three' of the US, Europe, and Japan typically costs well over $100,000 in government fees and attorney fees.
For a big company like Microsoft, that's just the cost of doing business. But that $200,000 is a huge amount of money to a startup, to say nothing of an individual garage inventor. Globalization and the internet mean that an inventor can sell an invention to people all around the world for far less than it would have cost 20 or 30 years ago. Protecting that invention all around the world, however, is often prohibitively expensive for all but the most well-funded, established companies.
It's true that companies like Microsoft would also benefit from lower filing costs, but small companies and individual inventors will benefit much more. It will also mean less money wasted on lawyers, as a single attorney in a single country can handle the whole process instead of having to use attorneys all over the world. And of course it will mean less duplication of effort in government as patent offices share resources. Right now there is an enormous duplication of effort as each application in each country is met with the same prior art, which is overcome with the same arguments. This is a tremendous waste of both government and applicant resources.
That's a pretty absolute assertion you're making. Do you really believe that software patents are the only way knowledge of software inventions gets transferred/created in our society?
That's not what I asserted. I asserted that if an algorithm is kept a trade secret and the source code closed then the public will never find out about it (barring, of course, illegal disclosure of the trade secret or independent discovery by another party that then discloses it).
And of course software patents are not the only way knowledge of software inventions gets transferred. Open source and academic publications are two other routes. The problem with the former is that open source is not particularly innovative. A study of 500 mature, actively developed open source projects that only 1% of them contained a new technology and addressed a new market or new user need. Open source is good at providing good, cheap software, but it is not particularly good at innovating.
Academic publication is all well and good, but ultimately someone has to pay for labs and salaries. Software patents step in as a way for academic discoveries to be commercialized. For example, some of the fundamental patents underlying Google are owned by Stanford, which is where Page and Brin invented PageRank and other related technologies. Google pays Stanford a non-trivial royalty for the exclusive license to those patents. Thus, the patents on successfully commercialized technology fund more academic research.
The huge diference is that you cannot use a TRIVIAL mechanical process and patent it
The test of patentability is not "non-trivial" but rather new, useful, and nonobvious. In any event, there are lots of 'trivial' patents that aren't business methods. There are, for example, dozens of patents on coffee cup sleeves, toilet paper, plastic cups, etc. For the most part these are small, some would say 'trivial' improvements over the prior art. But as long as they are new, useful, and non-obvious, the fact that they represent only a very small improvement is not important.
A lot of those business process patent are either trivial, already in use for decenny, copy of something existing but-with-computer or just plain mathematical application. You do not have such a problem with mechanical device patent.
If the business method patents really are obvious then they are invalid under 35 USC 103. There is no need to have a specific exclusion of business method patents just become some (or even a lot) of business method patents should not have been granted. That's throwing the baby out with the bathwater.
A better approach is to remove the presumption of validity of all patents, which would make it much easier to invalidate bad patents and reduce the incentive to seek questionably valid patents.
I'd rather see submarine patents be got rid of
Submarine patents are no longer a serious issue. Patents now expire 20 years from the date of filing, so keeping a patent application stuck in the patent office forever is a good way to end up with an expired patent, which is not particularly useful. Furthermore patent applications are now published after 18 months, so the world is put on notice fairly early on in the patent application process.
There may be a few submarines left from the pre-June 8, 1995 days, but as an ongoing problem they are nonexistent.
It could be something a lot of people were already doing, but if they had not documented their process in a way the patent office understands as "prior art" (which, as we've learned, is pretty much nothing), the patent can STILL stand.
The PTO, of course, limits itself to (easily accessible) printed publications, patents, and patent applications because those are all easy to use forms of prior art. But prior use is normally very easy for a defendant in a patent infringement suit to show except when the prior user decided to keep its use secret rather than disclose it in a patent application, printed publication, or through public use. The 'gap' that 35 USC 273 fills was an intentional motivator for the disclosure of otherwise secret business methods.
The 35 USC 273 prior use defense to business method patents is primarily meant to protect those whose prior use was secret. It comes up when Party A invents and uses a method secretly, then Party B invents and uses the method secretly, then Party A files for a patent more than a year later.
Normally Party B could use its prior use to invalidate the patent under 35 USC 102(b) except its use was not public. Party B could also invalidate the patent through disclosure in a printed publication or a patent application but again it chose to keep it secret. Because Party B kept the invention secret Party A gets the patent, although 35 USC 273 gives Party B a defense to patent infringement.
This presumes that the crass entrepeneur is the driver of innovation in software.
The PageRank algorithm on which the Google empire rests was invented by Larry Page and Sergei Brin. They published the algorithm, patented it, and went on to become incredibly successful 'crass entrepreneurs.' Google owns dozens of software patents on key technologies such as its Map-Reduce implementation and AdWords. Few would say that Google is not a driver of innovation in software.
Usually it's the exact opposite sort of person that drives software or scientific innovation.
That's a fairly strong claim. Would you care to provide evidence that innovators are not motivated by financial reward, which is typically either received directly through their own crass entrepreneurship or indirectly through the commercialization of their innovations by others?
Who, in your view, is driving innovation in software? It's not open source: A study of 500 mature, actively-developed open source projects found that only 1% of them included a new technology and addressed a new market or user need.
Necessity is the mother of invention, not greed.
Then why have a patent system at all?
A general tool of almost any kind, used in a different way, is _not patentable_. It's using the tool for what it was designed for, it does not change the physical construction of the machine nor its number of parts nor its general capabilities.
This is not entirely accurate. Suppose there is a patent on Compound X and the patent discloses the use of Compound X as a floor wax. Now suppose someone discovers that Compound X makes an excellent treatment for baldness, which is entirely nonobvious given its previous use as a floor wax. It is true that the inventor cannot patent Compound X again just because a new use has been discovered. But what the inventor can patent is a method of use of Compound X as a baldness treatment, specifying the dose and treatment regimen. This is true in the US though not true in many other jurisdictions.
The "actually implemented algorithm" is protected in its specific implementation by copyright, and should not receive the double protection of patent law against writing similar algorithms.
The protection provided by copyright is very narrow: it extends only to the literal source and machine code, not to the algorithm generally. For example: suppose one discovers a new efficient algorithm for simulating fluids. This has many potential applications: special effects, computer games, weather simulation, etc. Now suppose one publishes the algorithm and implements it in a specific program: a computer game. Nothing in copyright law stops everyone in the world from implementing the published algorithm in their own games, weather simulators, special effects software, etc. So long as they don't directly rip off the inventor's code but instead create their own implementation of the idea, they are in the clear.
So what's a business to do? A common thing to do is keep the algorithm a secret and keep the source closed. The public will never learn about the algorithm.
Software patents step in to fill the gap left by copyright. A patent on the fluid simulation algorithm protects any implementation of it in any domain. This allows the inventor to sell or license the rights to the highest bidder in each field. In the mean while, the public gets to see the algorithm published as a patent, and they are free to create and patent improvements as well as design around it. When the patent expires, the public is free to use the original idea. If the idea becomes outdated (e.g. because an open-source competitor designs a better alternative), then the patent will be worthless, the owner will stop paying the maintenance fees, and the invention will enter the public domain ahead of schedule.
You declare ownership of a new business process. Because of this you can prevent EVERYONE ELSE WHO DOES BUSINESS from benefiting from that new process.
Suppose someone invents a new mechanical device and patents it. The patentee can now prevent everyone else from making a business out of making, using, selling, offering to sell, or importing the patented device. In that sense, all patents declare ownership over a new business process. Business method patents are just applied economics in the same way that mechanical device patents are applied physics.
Others are also free to design around the patent by inventing their own substitute mechanical device or business method. Theirs may even be a superior device or method, which is better for everyone. And they have an incentive to do so: not only can they avoid infringement or licensing fees, but they can exclude or license to others in turn.
And of course, no one is actually forced to do anything. A patent is only valid if it claims something new, useful, and non-obvious, which means it can't cover anything that others were already doing. No one ever has to change their business in response to a patent: that which infringes if later anticipates if earlier.
Every patent allows the patent owner to cause trouble for everyone else for the next 17 years.
20 years from the date of filing.
The most frightening idea in all of those amicus briefs is the idea that medical procedures might be patented...Doctors invent because they take their oath to Hypocrates seriously, not because they identify with Crassius Maximus.
Medical procedures can already be patented, but practitioners and 'related medical entities' (e.g., the hospital, clinic, etc where the practitioner works) has a defense against infringement. Anyway, lots of doctors invent out of a sense of self interest as well as a desire to help people. I work with medical researchers on a regular basis, and the commercialization of their work is what ultimately funds the labs they work in.
Just FYI, it's "Hippocratic oath," and it's not sworn to Hippocrates but rather to various gods, particularly Apollo.
There is not a positive thing that has come out of software patents.
Google was able to rise to dominate search and challenge Microsoft primarily because of its numerous software patents. Without those patents, Microsoft, Altavista, Yahoo, or any other company could have used Page Rank and other algorithms covered by Google's dozens of software patents. Only part of Google's success is its good web design and marketing savvy. A huge part is the fact that it just plain does search better than almost anybody else.
Remember: Microsoft rose to dominance at a time when software was not patentable. That's one of the reasons why Embrace, Extend, Extinguish worked so well: without software patents there was nothing to stop Microsoft from embracing the new inventions of others.
The quality of software hasn't improved
Software quality (as in lack of bugs, bloat, etc) is not something that software patents address, at least not directly. What they address is the invention and commercialization of new features.
patent trolls are numerous
They are not particularly numerous, nor are they particularly unique to software. Only about 3000 patent infringement cases are filed in the US each year, and only about 300 of those go to trial. Trolls are a very small subset of that and hardly numerous. Furthermore, the existence of trolls has more to do with the presumption of patent validity and the inability of the PTO to find every piece of prior art than it does the patent eligibility of software implemented inventions.
because of software patents there are many incompatibilities in software that lead to many, many problems.
That's a very strong claim. Do you have evidence of it?
Can someone please tell me how software patents are promoting the progress of science/useful arts?
Copyright protection of software only extends to the specific implementation (i.e., the literal source and machine code). Patents protect the idea behind the implementation. This is an important distinction.
Say an inventor comes up with a brilliant new algorithm for efficiently and accurately modeling fluids on a computer. There are many possible applications: computer graphics, weather simulation, etc. The inventor would like to maximize his or her profit from the invention. Now, with copyright the inventor could write a program for each and every possible use and form a company to sell copies as copyrighted programs. But that's an awful lot of work for one person, especially if the inventor is a good inventor but a bad business person or doesn't know much about some of the possible application domains.
With software patents, the inventor can patent the invention and then either sell the patent to an established company or license the patent to different companies for different uses: one company gets the rights to use it in computer games, another to use it in weather simulation, etc. This lets each company best use its comparative advantage, which maximizes the benefit to consumers while minimizing inefficiencies.
Now you might say the inventor could keep the invention a trade secret and give it to the companies under a non-disclosure agreement. That's true, but then the public would never find out about the invention. The beauty of patents is that they let inventors and commercializers profit from an invention while still disclosing the invention to the public, which can freely use the invention after the patent expires. This is all true in software just as it is in more traditional fields of invention. Software is not special, and I say that as someone with both bachelor's and master's degrees in computer science.
Full Disclosure: I work for the team that wrote the amicus brief of Dr. Ananda Chakrabarty (he of the Diamond v. Chakrabarty Supreme Court case that established the patentability of genetically modified organisms).
But this ignores the constitutional requirement that it promote progress:
It is not obvious that business method patents hinder the progress of the useful arts, and merely stating it does not make it so. But if you have a good argument for why business methods are contrary to the constitutional purpose of patents, then by all means compose and submit an amicus curiae brief in support of the Respondent. The due date is October 2nd. You don't have to be an attorney to submit an amicus brief on your own behalf.
In this case the U.S. is one of very few nations to support software patents.
This case is not about software patents. Here are the questions presented in the case:
1. Whether the Federal Circuit erred by holding that a "process" must be tied to a particular machine or apparatus, or transform a particular article into a different state or thing ("machine-or-transformation" test), to be eligible for patenting under 35 U.S.C. 101, despite this Court's precedent declining to limit the broad statutory grant of patent eligibility for "any" new and useful process beyond excluding patents for "laws of nature, physical phenomena, and abstract ideas."
2. Whether the Federal Circuit's "machine-or-transformation" test for patent eligibility, which effectively forecloses meaningful patent protection to many business methods, contradicts the clear Congressional intent that patents protect "method[s] of doing or conducting business." 35 U.S.C. 273.
Now it is true that the case has implications for software patents, diagnostic method patents, etc, but concerns about software patents and the like will be, at most, secondary public policy considerations in the Court's decision making process.
That said, the point about international harmonization is that other countries and the EPO have broad definitions of patent eligible subject matter. The EPO, for example, says "European patents shall be granted for any inventions which are susceptible of industrial application, which are new and which involve an inventive step" with the further proviso that the invention must be of a technical character. Now, the EPO has a specific exclusion for business methods, but it is just that: a specific exclusion. The test of patentable subject matter remains broad. The argument is that the US should do the same: broad patentable subject matter with (if necessary) specific exclusions as determined by Congress rather than the courts. As it so happens, there are many good arguments that Congress has not yet excluded (and might never exclude) business methods from patent eligibility.
Full Disclosure: I work for the team that wrote the amicus brief of Dr. Ananda Chakrabarty (he of the Diamond v. Chakrabarty Supreme Court case that established the patentability of genetically modified organisms).
The 44 amicus curiae briefs that have been filed so far are only those submitted in support of the Petitioners (i.e., the inventors Bilski and Warsaw) or in support of neither party. Amicus briefs in support of the Respondent (i.e., the Patent Office) will be submitted after the Respondent's merits brief is submitted, which will occur on or before September 25. Once the merits brief is submitted, amici have 7 days to submit briefs in support of the Respondent.
While I don't expect there to be quite the same volume of briefs supporting the PTO as the Petitioner, there will probably be at least a few.
Here is the the Supreme Court docket for the case.
Full Disclosure: I work for the team that wrote the brief of Dr. Ananda Chakrabarty (he of the Diamond v. Chakrabarty Supreme Court case that established the patentability of genetically modified organisms).
As is all too often the case here on Slashdot, the summary has seized upon the title of the patent, which has no legal effect whatsoever, while ignoring the actual patent claims, which are all important.
If one actually reads the claims, one sees that the main new part of the invention are the 'hint elements' contained in the XML file. The written description expands upon what hint elements mean: "hints are provided within the XML associated files providing applications that understand XML a shortcut to understanding some of the features provided by the word-processor. By using the hints, the applications do not have to know all of the specific details of the internal processing of the word-processor in order to recreate a feature."
Basically, the invention here is the inclusion of information that lets third-party programs better understand what to do with the format. You can imagine, for example, if HTML included something like this. The del ('strikethrough') tag might be written:
That code would allow a program that did not natively understand the tag to implement a simple version of it. The idea is to allow new features to be introduced into the format while enabling older versions of the software to use them without updating their code. The necessary code comes with the file.
Now, whether that's still new and nonobvious, I don't know, but it's a significantly more accurate summary of the invention than "Microsoft Patents XML Word Processing Documents."
First, a citation to the published paper: Hana Yi, et al., Selection of a variant of Geobacter sulfurreducens with enhanced capacity for current production in microbial fuel cells, Biosensors and Bioelectronics, Volume 24, Issue 12, 15 August 2009, Pages 3498-3503.
The extrapolated current density was 7.4 ± 0.1 A/m2. The individual fuel cells produced 14mA, which was sustained for 24 months.