SCOTUS Set To Examine Combinatory Patents

eldavojohn writes "The Washington Post is reporting that businesses are split on the current situation with patents in the United States. From the article: 'The court is scheduled to hear arguments Tuesday on what's obvious when older inventions are combined to create something new. The law says an invention that's "obvious" isn't patentable, but the definition isn't clear despite decades of litigation. The ambiguity, critics say, has led to an explosion of patents as companies stake claims on everything in sight, from strategies for avoiding taxes to golf ball designs. The result has been extensive and costly legal wrangling as companies of all sizes fight over who's infringing what. In some cases, small companies acquire patents not to develop new products but to sue for a quick windfall.' This sounds like some common criticism of the patent system that often pops up on Slashdot. The last part of the article mentions that most legal experts are expecting some changes to come of this. Hopefully soon we'll see some changes to the patent system for a better concept of intellectual property."

Comments

The issue with obviousness is this:

[James_Duncan8181]

What is obvious to a practitioner of the art must be demonstrated to be so to the satisfaction of a court. The judge may well have no real level of technical knowlege, and so is at the mercy of the expert witnesses. Since both sides can afford to buy suitable opinions regardless of the facts, it then comes down to the most persuasive legal team.

As a result of this, patent cases are often fairly random in terms of the results obtained. It's difficult to know what can be done about this, other then to allow the bench to call an independent expert witness.

Re:The issue with obviousness is this:

[Anonymous Coward]

> I'm sure that there is a better way of achieving their objectives

The objectives are the problem.

See also the FAT patent, a way of supporting long filenames on a filesystem that (almost uniquely) lacked this support. Why should MS be allowed to monetize a fix to a bad design through patent licensing, especially when the filesystem is only in use because of their illegal monopoly on the desktop?

How were long filenames not obvious, how was storing and retrieving data not obvious when the FAT patent was filed? That garbage patent even survived a recent appeal.

Re:The issue with obviousness is this:

[Alef]

What is obvious to a practitioner of the art must be demonstrated to be so to the satisfaction of a court.

What about this: For each patent application that might be considered obvious, a small number of people skilled in the art in question are hired for one day. They are presented with the applicant's description of the problem that the invention is supposed to solve, but are not given any information about the invention itself. At the end of the day, they get to present a number of approaches that one might try in order to solve it. If they come up with something similar to what is in the patent application, then the invention is obvious.

By the way, if any of the hired practitioners knew of the solution already, well then we have found prior art, so that is OK too.

A weak spot of this scheme would be that often describing the problem in the right way is how you find the solution to a tricky problem, so entirely separating the description of a problem from the solution might not always be possible. But that, I think, is a small problem compared to all the ones in the current patent system.

Re:The issue with obviousness is this:

[msobkow]

For software, there are some aspects of "obvious" that aren't applicable to physical device patents.

Any combination of a data object and a general purpose algorithm, data transport, data storage mechanism, or framework/library features should be declared obvious. For example, email transfers over IP. Any data transport can implement the IP functionality; using a particular transport (such as wireless) is not innovation. It's using tools the way they were designed by programmers and companies that encourage reuse -- something that is taught to first year students. Only lawyers and patent leeches can even try to argue that point with a straight face.

Reimplementing an existing solution in another language is not creating new technology, unless the languages are radically different (which would need expert opinion to determine how different programming languages are. I'm thinking in terms of procedural vs. functional.)

Any implementation using a technique, algorithm, or approach published in a textbook or research journal is not innovation. There may be a lack of research on the part of the development team if they didn't know about it, but that doesn't mean they came up with something "new" and patentable.

I firmly believe that software patents should require a functional implementation be submitted within 2 years of an initial application. Hardware patents require engineering diagrams; software patents should require very specific algorithm and data descriptions that are more easily expressed as code. The code in question should not be published by the patent office, but kept in archives to assist with any future dispute resolution by the courts.

Defense of a software patent may require comparison of implementation code. The patent holder making the accusation should be required to provide access to their development archives for the court's experts (not public.) Failure to provide a reasonable history archive should be considered sufficient to prove the patent holder did not take reasonable steps to protect their copyright implementation of the patent.

Like a trademark, failure to defend should result in loss of a patent, and it's previously protected content should become public.

Most important of all, the onus should be on the patent holder to demonstrate that they attempted a "good faith" negotiation to resolve their claim before filing with the courts. That includes demonstrating that they have prepared their evidence BEFORE court, not used the court as a fishing expedition to gain access to evidence that may or may not prove their allegations. The multi-year nightmare of SCO is a disgrace.

"Good faith" also implies the discussion of reasonable licensing terms for the patent, not using them to try to acquire, take over, or bankrupt the purportedly infringing legal entity.

Re:The issue with obviousness is this:

[ortholattice]

I firmly believe that software patents should require a functional implementation be submitted within 2 years of an initial application. Hardware patents require engineering diagrams; software patents should require very specific algorithm and data descriptions that are more easily expressed as code. The code in question should not be published by the patent office, but kept in archives to assist with any future dispute resolution by the courts.

I would propose that the code in question should be published if a patent is to be issued for it. It will already have the dual protection of copyright and patent. Isn't one purpose of patents to promote the arts and sciences by disclosing inventions, instead of keeping them trade secrets, in exchange for the temporary monopoly granted?

Re:The issue with obviousness is this:

[radtea]

Hardware patents require engineering diagrams; software patents should require very specific algorithm and data descriptions that are more easily expressed as code.

"Engineering diagrams" is a big overstatement of what hardware patents require. Most hardware patents contain sketches, but nothing any engineer would be willing to sign and stamp.

The digrams in patents are explantory, and for clarity of explanation they frequently leave out major features that would be required to actually build a device. The assumption is that anyone reading the patent has enough "skill in the art" to know those features are needed. I'm talking about things like bolts to fasten parts together and so on.

This is one of the reasons why workability is not a criterion for patentability, because the hardware does not actually have to be engineered to any reasonable level of detail before a patent can be granted. For example, you might submit a patent that to be built requires materials that are impossibly strong or light. That won't show up anywhere in your patent documents, and anyone who tries to engineer such a device on the basis of your patent will rapdily discover they are wasting their time.

Software and process patents simply should not be granted, anymore than patents should be granted on the plots of novels, and for the same reason. A novel, after all, is nothing more than a machine for inducing a particular state in the reader. But giving patent protection to particular designs of such machines does not "promote the useful arts" any more than giving patent protection to software does.

Re:The issue with obviousness is this:

[g2devi]

A problem with this approach is that many problems can't be solved in a day (unless you've experienced the same problem before), but they can be solved in a week or a month. Giving a 20 year patent for something that can be independently invented in a week doesn't make sense.

Another key problem is that if there are 6 obvious ways of solving the problem and you hire 5 experts, you'll let one slip.

Another problem is that how do you know you've hired the right experts? Some problems are dead obvious to anyone who knows graph theory, but opaque to someone that knows cryptography. If the problem is presented as a cryptographic problem, you'll hire the wrong experts and the experts will agree that it's not obvious.

Another problem is that it doesn't capture the case of patenting file formats since there are an infinite number of ways of encoding them but only one way to interoperate with them (the patent pending way).

Another problem is patenting standard business methods. Something nonobvious (because of regulations) was done "in the real world" but it gets patented because it's done in a computer.

The key problem with patents is the basic assumption. If they can be reinvented independently, then the patent added no value to society and adds a cost to society, so the patent should be invalidated. If it's required for interoperability or for legal regulations or common business practices, it adds no value and adds a cost to society, so the patent should be invalidated.

Re:The issue with obviousness is this:

[Lonewolf666]

This is one of the reasons why workability is not a criterion for patentability, because the hardware does not actually have to be engineered to any reasonable level of detail before a patent can be granted. For example, you might submit a patent that to be built requires materials that are impossibly strong or light. That won't show up anywhere in your patent documents, and anyone who tries to engineer such a device on the basis of your patent will rapdily discover they are wasting their time.

Unfortunately, not demanding workability offers another avenue for abuse:
People can now patent ideas that are not feasible yet, without actually contributing to the art. But as soon as someone else does the hard work and makes it possible, that someone can be sued for patent violation when he tries to market his design.

legal systems and false positives

[cucucu]

Most reasonable legal systems are tuned to avoid "false positives". That is, it is acceptable that some number of people who actually committed crimes walk away unpunished, as long as those who are punished are done so justly.
If a country's legal system does not attain to this rule, then it is probably not deemed safe to visit there.
It is a principle well known to legislators, law enforcers, lawyers and citizens in generals.
And I think it should be applied to patents as well. Only if you prove beyond doubt that your patent is non obvious, innovative, and all the criteria apply, it should be granted.
In exactly the same way that nobody is (or at least nobody should be) sent to jail if there are doubts. If there are doubts then you are free.
The patents system is not flawed per se. The problem is that it is being abused with a high noise to signal ratio.

The founding fathers saw patents

[rolfwind]

as a way to advance science and copyrights the arts.

In exchange for opening the information, inventors were given a limited time monopoly on said invention. What the founding fathers wanted to get beyond were secretive guilds and the hording of information - instead encouraging a free flow of ideas. One historical objective is to avoid the technological stagnation that accompanied the middle ages.

But with the original aim of inventions in mind - now that it seems that there is a lot of unintentional infringing of patents - it suggests that the original intention is outdated - the patents are protecting knowledge not worth protecting since others stumble onto the ideas at roughly the same time, negating the need to open the information. Even if it is not stumbled upon, reverse engineering can and does often provide the information needed for someone to duplicate many of the results - again negating the original need for patents.

Could it be that patents should only have been training wheels for the industrial revolution until a technological society is achieved - because once a certain level is achieved, there is no easy way of going back, of ceasing the competition for better products, advances, etcetera? That they should have been discarded after a time?

Could there be another way to promote an open exchange of ideas? Which was the original intention of patents. Because it seems that patents, in their current state, are setting us behind other countries (China) in terms of the future, dragging our economies down, and not at all fulfilling their original purpose. And from what I have read - the purpose of the patent was an ultimately social function, not to protect businesses or let them rest on their laurel in perpetuity (Atari).

Re:The founding fathers saw patents

[jackb_guppy]

This clause:

Even if it is not stumbled upon, reverse engineering can and does often provide the information needed for someone to duplicate many of the results - again negating the original need for patents.

Shows the reason why a patent is needed. To allow an intentor the right to make reasonable return on his work.

Now a days with large mega-corporations, the use of patents should may be restricted. Since pharmaceuticals will patent one drug and just as the patent runs out changes a minor element and patents the next to extend the patent and profit from it.

My personal suggestion would be patents/copyrights are restricted to living breathing persons only. They can not be assigned to corporation. Corporation can use trademark to protect their things.

Re:The founding fathers saw patents

[Alef]

Could it be that patents should only have been training wheels for the industrial revolution until a technological society is achieved

Yes, perhaps. I think the key difference between the 18th and 21th century, with regard to the usefulness of patents, is that today, through the advent of information technology, we have become a network society. It is now more important how information and ideas flow than the thoughts and inventions of each single individual. In the 18th century, a brilliant individual could invent a complex machine all by himself -- today, that is almost impossible.

It is no longer possible to isolate yourself, being secretive and hoarding information while remaining in the technology front. If you disconnect from the network (so to speak), you will die. Therefore, I would say that the incentive to share your ideas no longer needs to be provided by patents. Instead, patents have become an obstruction to the flow of ideas.

Patents should be restricted in scope.

[Anonymous Coward]

They should be restricted to practical engineering for physical objects.

They shouldn't include things like stories or tax shelters or mathmatics or software. That's just BS.

But practical sciences need them because the only way some things is going to happen is by throwing money at them. Stuff like automobile engine optimizations and safty improvements. Space flight. Drug research. Advances in electronic design. New chemistry breakthroughs, etc etc.

Patents are the only way your going to get certain technologies to remain out in the open while providing funding over it. Especially things that require huge amounts of capital such as drug research. Otherwise people will be forced to close everything and innact trade secret insanity which benifits nobody. Imagine corporations having to move to third world nations and hire private armies to protect their billion dollar secrets. That is the bullshit we'd have to look forward to. Patents are much better solution.

Also it avoids the need for government to control funding, which produces generally pretty bad results per amount of money (money also as a reflection of real time taken out of people's lives to deal with governmental bullshit).

Software patents are a very good indication of what is wrong with the system. Should be done away with and as far as reform for the rest of the system I don't know enough about the details to realy comment on it.

A favorable ruling from THIS Supreme Court?!?

[kcbrown]

Hopefully soon we'll see some changes to the patent system for a better concept of intellectual property.

From the same Supreme Court that in Eldred vs. Ashcroft ruled, in essence, that a copyright term remains "limited" and thus is in keeping with the Constitution as long as it has a stated limit, even if the limit increases over time at the same rate that time passes? You must be on some other planet if you think that same Supreme Court will make any improvements (from the point of view of actual practitioners in the various fields, as opposed to the patent attorneys) to "intellectual property" (gad, how I hate that phrase) law.

I think it's at least as likely that they'll decide that "obvious" really means "obvious to even the greatest of morons in the field", and thus that pretty much anything you care to name is patentable.

Of course, that assumes that they'll issue any sort of meaningful ruling whatsoever. It's entirely possible that they'll simply say that it's up to Congress to define in greater detail what it means, and until then leave things as they are. Just like they did in Eldred.

Why this will never get fixed

[Programmer_Errant]

This has nothing to do making patents and the whole concept of intellectual property logical, begging the fact that you could define logical in the first place. This is slashdot, remember?

This is about vested interests, corporations, not individuals, trying to maintain a system that so far has been to their advantage. Most of them got to where they are by being good at the present set of rules and they're not about to radically change those rules and lose that advantage.

The only chance you have of radically changing it would be to force it to break completely. You'd need to help the patent trolls get patents good enough to seriously threaten the present entrenched players, the big corporations. In theory the patent trolls could see through this scheme and realize that parasites that kill all of their hosts don't have long lifespans either. But trolls are greedy and they won't be able to help themselves.

Re:Patents and what is obvious

[Anonymous Coward]

the fatal flaw in your logic is assuming that an "individual" would find a cure to cancer or some other similarly complex task.

no corporation is going to want to rely on the goodwill of the public to "keep them alive" bla bla socialism bla bla bla

If discoveries cannot, why should inventions

[CDMA_Demo]

Seriously. If E=mC^2 cannot be patented why should anything else be allowed to be patented? If it wasn't for patents we'd have two different types of telephones, hence revolutionizing telecommunications altogether. Maybe we'd have arrived where we wanted to earlier that way.

Things that cannot be patented.

[3seas]

Natural Law (laws that govern above the level of mans ability to enforce any law he may make against it)
Physical Phenomenon (also something beyond mans reasonable control and ability to enforce)
Abstract Ideas (that which has no physical property, though it may be expessed/communicated thru physical media).
Mathmatical algorithyms are a fourth but in essence math is an association, not a thing, between abstraction and accounting of the concrete.
Also that which is obvious to anyone skilled in the art of which the product is used.

Patents also require inventorship and originality. You can't patent something someone else did.
Same with copyright, if you are not the author, you can't copyright it, though originality is a little more flexiable with copyrights
(not its only because you can get a copyright on something not so original, but prior art can later overcome your copyrights - the copyright office does notr require you to submit prior art and leves originality up to anyone interested in challenging it - of course in court)
Exception to these is if the person of creation is employed to do so, for you. Still their name is attached.

You cannot patent what has been made public for a given time before a patent application is applied for.
Prior art issue apply in defence against originality.

Now there is also the need to be novel and useful, though useful today may mean using it's to earn you an income from an infringment lawsuit (which should be grounds to dismiss such a claim as it does not contribute to the founding fathers heart intent.)

Patents do not give you the right to produce the object of patent, as you can invent a bomb and it can meet all the requirement of patentability and you can even eran royalities off of it (perhaps selling a license to the government if they don't just take it from you), but you are not given teh right to produce it by being granted a patent. hat a patent does is give you the right to say "NO, you canno use" to others.

Its this grant of saying "NO" that is also the basis of the above things that cannot be patented. If you cannot reasonable enforce "NO" then what is the point of applying for a patent (regardless of those who think getting a patent to protect an idea, that it stay open and free.... well there are easier and less expencive ways to do such as mentioned above.)

The manifestation of the concept of granting a right to say "NO, you cannot use" has reached its apex of being more a tool to extract value, from others, then it is to earn directly off the product and sale of the product of the patent. Leave it to man to abuse others through the distortion and manipulation and interpretation of his own creation of laws, laws that other before him probably created.

About that which is not patentable.

Software!

It is by the natural laws of human capabilities to create and use abstractions. It is also by the human rights of men to communicate with each other via such abstractions and that it is through the use of abstract communication that man is enabled to create a great deal, from building physical things such as homes, roads, vehicles, cloths, medicine and treatment of illnesses, even laws that dictate how a traffic light is to work, etc.. These all are examples of physical phenomenon. Not the traffice light, cloths, homes, etc. but the physical phenomenon of abstraction application, the conversion from abstract communication to physical movement. It is a human characteristic, to some degree found in other animals but only of a lower level of abstraction ability.

Software is often argued to be pure mathmatical algorithyms and that this is reason enough for it to not be patentable. Though many might not grasp so easy the ideas that words can somehow be numbers too. But no-matter math and all other abstractions fall into the scope of what is abstraction.

Natural law, physical phenomenon, abstract ideas, mathmatic......All of these non-patentable facets and more apply to software, so wh

The Medicines issue...

[itsdapead]

One of the frequent objections to patent reform seems to come from the drug companies which (if you reconfigure your brain to enable "feeling sorry for poor impoverished multinational companies mode") is sort-of understandable, given the amount of work needed to establish what might be an easy-to-replicate chemical as a drug. Its also a field where disclosure of information is likely to advance science.

Couldn't some of the protection currently offered by patents be made part of FDA and similar approval processes instead? I.e. make the investment needed to get a drug FDA approved: get an N-year monopoly on its sale?