Link to the brief, because Subby's source sucks. It's an Amicus Brief in the CLS Bank v. Alice Corp. case.
Better article at TechCrunch.

Elandriel: Theaetetus: Elandriel: I think patents should be more specific; patent trolling has become too much of an industry.

Without disagreeing with either statement, they're not related. Patents should be more specific, both as an aid to searching and to reduce litigation costs with disputes over what the meaning of "is" is. Patent trolling is unrelated to that, as it's about income through licensing and litigation, and will continue even if patents become more specific.
Both are worthy avenues for discussion, but addressing one will not resolve the other.

I dunno, I find that patent trolling regarding things like "online aggregation" or whatever the situation was that Fark got hit with, could be resolved by a bit more stringent standards for patent application. But then I'm not in the industry so more or less armchair quarterbackin' here. I'm probably way off.

Well, again, that's sort of a third issue. Basically, you've got:
1) "The USPTO keeps issuing obvious/not novel patents" - this is a problem about examination quality
2) "The patents are too vague or tough to understand" - this is a problem about the sufficiency of the description
3) "Patent trolls grrrrrrr!" - this is a problem about venues, aggregation of defendants, speculative licensing amounts, the high cost to defend a suit, etc.
and there's also:
4) "Software is just math/an abstract idea/etc." - this one is about whether the most nonobvious, novel, super specific invention by a good and practicing company should still be unpatentable, because of its industry.
5) "Patents last too long in X industry" - computers move so fast, so shouldn't software patents have shorter term? But drugs are slow, so shouldn't they have a longer term? Do we need specific terms for every industry, and for Congress to issue new term guidelines every year?

Part of the problem is that people confuse these issues... So they'll say "patent trolls are bad, because software patents last too long" or "software is just math, so it's all obvious". Even if you tell that to someone who is anti-patent, you come off as a kook. And it's really not going to sway anyone in Congress who is pro-patent.

Theaetetus: midigod: Teiritzamna: Assume Alice invents a new widget, and it is great. Alice discloses that invention and gets a patent. Alice decides however she doesn't want to make the thing. Its too much trouble, or she has a pathological fear of factories. Whatever.

It reads to me like you're conflating two different things. Making something one time in order to gain the patent is very different from making a million of them in a factory. A patent is supposed to be the registering of a real-world embodiment of an idea, not just the idea itself. (I agree that current patent laws do not require a working model, but my opinion is that that should be changed.) Certainly no one should have to set up a manufacturing line in order to be able to patent something. But you should have to build one of them, to prove that it's possible, and to automatically create intellectual borders for that idea.

What if I'm a super-brilliant but impoverished physicist working in my garage, and I come up with a way to make cold fusion practical and efficient... but to even make a "small" 1KW reactor will require a few million in materials and deuterium? No patent for me because I'm poor?

Further, as Teiritzamna said, what if I can give you a hundred pages of detailed schematics, a 3D CAD model of the reactor, and affidavits from a dozen engineers and other physicists saying "yeah, I could build this, and yeah, it would work," but I still can't build one for lack of resources?

entropic_existence: Not really. Patent law has a fairly specific definition they sue for novel/non-obvious that this doesn't necessarily run afoul of for #1. For #4, this isn't the idea that software is itself is abstract.

Ok i am confused. Not being attacky, but could you unpack further by what you mean under: vague (too broad) other than #1: "it is not a new invention or it would be an invention that is obvious to a person in the industry,"* or #4 "it is an abstract idea and thus unpatentable subject matter regardless of its novelty or embodiment.

*35 U.S.C. 103
** found as a negative implication under 35 U.S.C. 101

midigod: Teiritzamna: Assume Alice invents a new widget, and it is great. Alice discloses that invention and gets a patent. Alice decides however she doesn't want to make the thing. Its too much trouble, or she has a pathological fear of factories. Whatever.

It reads to me like you're conflating two different things. Making something one time in order to gain the patent is very different from making a million of them in a factory. A patent is supposed to be the registering of a real-world embodiment of an idea, not just the idea itself. (I agree that current patent laws do not require a working model, but my opinion is that that should be changed.) Certainly no one should have to set up a manufacturing line in order to be able to patent something. But you should have to build one of them, to prove that it's possible, and to automatically create intellectual borders for that idea.

What if I'm a super-brilliant but impoverished physicist working in my garage, and I come up with a way to make cold fusion practical and efficient... but to even make a "small" 1KW reactor will require a few million in materials and deuterium? No patent for me because I'm poor?

rolladuck: I find it hard to comprehend that we as a society can agree on rules that govern the timing of football and basketball, but can't properly nail down how to implement patent and copyright timing.

Hey, leave copyright out of this one. That's a whole separate issue. The problem there is that all of the interests of copyright owners are aligned in extending copyright. Universal doesn't care about Mickey Mouse. Disney doesn't care about Harry Potter. They all agree on wanting extended protections for their own creations, and are aligned against pirates who want no protection... and all the money of that equation is on one side, which is why you end up with lifetime+70 year terms.

Contrast that with patents... Everyone wants their own patents to be longer, but they also want everyone else's to be shorter. Apple does care about what IBM has patented, because they want access to it. Similarly, Google cares about what Apple has. As a result, all of the interests (and money) are evenly divided, and patent terms have only ever been extended twice - one of which was a negligible extension to comply with an international treaty.

entropic_existence: Theaetetus: 2) "The patents are too vague or tough to understand" - this is a problem about the sufficiency of the description

There is also the issue, which I think this particular case/brief may be referring to, which isn't vague in the sense of unclear, like you posted here, but vague as in too broad.

Totally. This brief is actually more about (4), that software patents are just claiming an abstract idea. Subby's headline is inaccurate (though due to bad journalism in the linked article).

Such that the patent is for a pretty broad or abstract idea "on a mobile device" or "on the internet". It isn't that such patents are hard to understand or impenetrable, they are the exact opposite, but they are so broad that they cover too many different ways to achieve the same outcome.

IMHO patents should bear some sort of specificity to implementations.

They are - mobile devices, or the Internet are both very specific things. But that's really not the issue - it's that the patents are claiming abstract ideas... and that that shouldn't be cured merely by reciting a non-abstract element, such as a mobile device or the Internet.
So, in other words, it's not about the mobile device. This would be patentable:
1. A method, comprising: [concrete idea] executed on a mobile device.
While this would be unpatentable:
2. A method, comprising: [abstract idea] executed on a mobile device.

The distinction isn't in the mobile device part, it's the abstract vs. concrete idea, and Google is saying that the presence of a mobile device doesn't make an abstract idea concrete.

... problem is that no one, including Google, knows how to define an "abstract idea". It's like obscenity... you know it when you see it.

Teiritzamna: That would be covered by issues #1 & #4

Not really. Patent law has a fairly specific definition they sue for novel/non-obvious that this doesn't necessarily run afoul of for #1. For #4, this isn't the idea that software is itself is abstract.

Teiritzamna: Assume Alice invents a new widget, and it is great. Alice discloses that invention and gets a patent. Alice decides however she doesn't want to make the thing. Its too much trouble, or she has a pathological fear of factories. Whatever.

It reads to me like you're conflating two different things. Making something one time in order to gain the patent is very different from making a million of them in a factory. A patent is supposed to be the registering of a real-world embodiment of an idea, not just the idea itself. (I agree that current patent laws do not require a working model, but my opinion is that that should be changed.) Certainly no one should have to set up a manufacturing line in order to be able to patent something. But you should have to build one of them, to prove that it's possible, and to automatically create intellectual borders for that idea.

Theaetetus: 5) "Patents last too long in X industry" - computers move so fast, so shouldn't software patents have shorter term? But drugs are slow, so shouldn't they have a longer term? Do we need specific terms for every industry, and for Congress to issue new term guidelines every year?

Whenever possible, the law should be written so as to be event-driven, not calendar-driven, in order to avoid abuses, or "bugs" in the law such as what you describe. For the most part, however, calendar-driven expiration for patents isn't the problem, so much as a proper selection for what event starts and stops the clock. Drug manufacturers should not have the clock started until they can legally sell the drug in the market. Software patents (or copyrights) should also expire at a certain time when the product is abandoned, if properly defined. There will still be those who game the system, but if the system is designed properly, such gamesmanship will be quite obvious, and the necessary PR burden can be applied.

I find it hard to comprehend that we as a society can agree on rules that govern the timing of football and basketball, but can't properly nail down how to implement patent and copyright timing.

entropic_existence: There is also the issue, which I think this particular case/brief may be referring to, which isn't vague in the sense of unclear, like you posted here, but vague as in too broad. Such that the patent is for a pretty broad or abstract idea "on a mobile device" or "on the internet". It isn't that such patents are hard to understand or impenetrable, they are the exact opposite, but they are so broad that they cover too many different ways to achieve the same outcome.

That would be covered by issues #1 & #4

theresnothinglft: IMO a patent holder should be required to maintain a working prototype of their invention in order to hold the patent. Having patents without a product that they're intended to protect does not provide any benefit to anyone.

Well, to offer an opposing view. The purpose of patents are to disclose information that would otherwise be kept secret and to incentivize invention. Assume Alice invents a new widget, and it is great. Alice discloses that invention and gets a patent. Alice decides however she doesn't want to make the thing. Its too much trouble, or she has a pathological fear of factories. Whatever.

At this point you may say that Alice should be prohibited from having that patent. But remember that patents are property - so why shouldnt Alice be able to merely license the invention to others. This is likely what we would all rather, as the invention gets made, Alice gets paid, no problems all around.

However, note that there is a rather anti-american style streak to your initial statement. Imgaine that Alice decides that her widget is in fact terrible, for some non-utilitarian reason. she realizes that it is being used 100% as part of a weapons system, say, and she has moral issues with that. Should she be forced to license her property out? Is not this akin to one landowner looking at the property of another and saying "hrm, i could totally use that land better, so i should be the one who owns it"

Personally, i would figure it was not the end of the world if we instituted types of mandatory licenses for certain industries/invention classes (transforming patents from rights to exclude into rights to revenue) - but it is not so simple to say "you must make X to patent X."

Theaetetus: Link to the brief, because Subby's source sucks. It's an Amicus Brief in the CLS Bank v. Alice Corp. case.
Better article at TechCrunch.

The TechCrunch article IS much better.

Theaetetus: 2) "The patents are too vague or tough to understand" - this is a problem about the sufficiency of the description

There is also the issue, which I think this particular case/brief may be referring to, which isn't vague in the sense of unclear, like you posted here, but vague as in too broad. Such that the patent is for a pretty broad or abstract idea "on a mobile device" or "on the internet". It isn't that such patents are hard to understand or impenetrable, they are the exact opposite, but they are so broad that they cover too many different ways to achieve the same outcome.

IMHO patents should bear some sort of specificity to implementations.

IMO a patent holder should be required to maintain a working prototype of their invention in order to hold the patent. Having patents without a product that they're intended to protect does not provide any benefit to anyone.

Marcus Aurelius: We'll get patent reform right after we kill all the patent lawyers.

Or maybe after we get patent examiners who don't think adding two numbers to produce a third is worthy of a patent.

Many computer-related patent claims just describe an abstract idea at a high level of generality and say to perform it on a computer or over the Internet. Such barebones claims grant exclusive rights over the abstract idea itself, with no limit on how the idea is implemented.

... except that if you don't perform the idea on a computer or over the Internet, you don't infringe.