The Econimist - Obituary for software patents
- Thread starter proufo
- Start date
Finally, a thread I can contribute to! I've been a patent lawyer since 1982, both in corporate practice and then in private practice representing multinationals, domestic companies, startup companies and universities. I've never represented a patent troll. To me the Economist article is riddled with mistakes and the connection between the issues at hand for the US Supreme Court and the issues of patent trolls are tenuous at best. Here's my take:
The Supreme Court will be once again tackling the issue of patent eligibility of computer software. Big emphasis on patent "eligibility". Under US patent law, there are various criteria under which to judge whether an invention is deserving of the exclusionary rights that are granted by a patent. In order to get a patent, the invention must be novel (new- can't rediscover a lost idea whose time has come), unobvious (the invention must go beyond what would be expected of a skilled worker in the field) and adequately disclosed (described in sufficient detail to teach others how to practice the invention and to show that the inventor really had possession of his invention).
Patent eligibility is a different matter. The courts have long noted that certain things are not even eligible for a patent, so you don't even get to argue novelty, unobviousness and adequate disclosure. These things are laws of nature ("I hereby patent gravity or calculus"), natural phenomena ("I hereby patent a new rock I found") or abstract ideas ("I hereby patent hedging").
What does this have to do with software? In the first modern software patent case in 1972, the Supreme Court held that a method of converting signals from binary coded decimal to binary using a reentrant shift register and steps of shifting, masking and adding was not patent eligible. Meaning (again) that there was no argument whether it was new, obvious of poorly disclosed- rather that is was not the kind of thing you could even apply for a patent. Their reasoning- this patent is math, math is a law of nature or abstract idea and can't be patented.
From this start, the Supreme Court went even further in 1978 saying that a method of updating an alarm limit that was used in a process for catalytic chemical conversion of hydrocarbons was not patent eligible (regardless of whether it was new, unobvious and well disclosed), because the equation on the computer it used to compute the alarm limit was math, and math is assumed to be a law of nature or an abstract idea.
Finally in 1980, the Supreme Court found a software invention to be patent eligible. In this case, rubber was cured by putting rubber in a molding press, measuring its temperature during curing and opening the rubber mold press when an equation running on a computer was satisfied.
As you can imagine, it has been difficult for the lower courts to figure out how to reconcile these cases, because they stem from erroneous premises- all math already exists in nature and a computer running software is just a math machine. There is a federal appeals court (the Court of Appeals for the Federal Circuit - CAFC) that was created in 1980 to hear all patent appeals from the federal distinct courts. A specialized patent appeals court. Yet, in the latest case that the Supreme Court will be hearing the CAFC was totally divided and issued six (6) different opinions. So the specialized patent court just threw up its hands.
How does this relate to today's technology? There have been significant innovations in semiconductors and telecomm to bring us the internet, personal computing and smartphones. But a large part of these innovations have been in the area of software. Just think about your use of the internet, your iPhone or your iPad. Much of the functionality comes from software running on a general purpose processor. Yet the Supreme Court's early decisions that software running on a computer is just math, a law of nature or an abstract idea has now put a cloud on all this software innovation. This is particularly distressing when you consider that a lot of electronic manufacturing takes place overseas, but a lot of software development takes place in the US. Yet, depending on how the Supreme Court rules, much of this software innovation could be ruled as not being patent eligible, regardless of whether it is new, unobvious and well described.
How does this relate to patent trolls? A patent troll is a non-practicing entity, like a University or an independent research company that doesn't make products but rather licenses its inventions to others. But unlike Universities or research labs, a patent troll typically buys patents from defunct companies, strains their interpretation to the nth degree ("We invented the fax machine", "we invented email", "we invented interactive phone response systems"…). Then, rather than sue the maker of the device, they sue thousands of innocent users of the device, no one of which has the knowledge or incentive to fight rather than settle.
Some troll patents relate to software but many don't. Software or not is irrelevant as to what you think of patent trolls. Making software patent ineligible will not solve the patent troll problem. It will just place a big blow on a US industry. There are other things that are being considered to stop patent trolls. One thing would be to make sure the Patent Office is strictly enforcing the legal requirements of new, unobvious and well described. Other things can be done and are being considered. But to try solve the troll problem by saying software is not patent eligible throws out the baby with the bathwater.
All of the above in my opinion, no legal advice intended.
The Supreme Court will be once again tackling the issue of patent eligibility of computer software. Big emphasis on patent "eligibility". Under US patent law, there are various criteria under which to judge whether an invention is deserving of the exclusionary rights that are granted by a patent. In order to get a patent, the invention must be novel (new- can't rediscover a lost idea whose time has come), unobvious (the invention must go beyond what would be expected of a skilled worker in the field) and adequately disclosed (described in sufficient detail to teach others how to practice the invention and to show that the inventor really had possession of his invention).
Patent eligibility is a different matter. The courts have long noted that certain things are not even eligible for a patent, so you don't even get to argue novelty, unobviousness and adequate disclosure. These things are laws of nature ("I hereby patent gravity or calculus"), natural phenomena ("I hereby patent a new rock I found") or abstract ideas ("I hereby patent hedging").
What does this have to do with software? In the first modern software patent case in 1972, the Supreme Court held that a method of converting signals from binary coded decimal to binary using a reentrant shift register and steps of shifting, masking and adding was not patent eligible. Meaning (again) that there was no argument whether it was new, obvious of poorly disclosed- rather that is was not the kind of thing you could even apply for a patent. Their reasoning- this patent is math, math is a law of nature or abstract idea and can't be patented.
From this start, the Supreme Court went even further in 1978 saying that a method of updating an alarm limit that was used in a process for catalytic chemical conversion of hydrocarbons was not patent eligible (regardless of whether it was new, unobvious and well disclosed), because the equation on the computer it used to compute the alarm limit was math, and math is assumed to be a law of nature or an abstract idea.
Finally in 1980, the Supreme Court found a software invention to be patent eligible. In this case, rubber was cured by putting rubber in a molding press, measuring its temperature during curing and opening the rubber mold press when an equation running on a computer was satisfied.
As you can imagine, it has been difficult for the lower courts to figure out how to reconcile these cases, because they stem from erroneous premises- all math already exists in nature and a computer running software is just a math machine. There is a federal appeals court (the Court of Appeals for the Federal Circuit - CAFC) that was created in 1980 to hear all patent appeals from the federal distinct courts. A specialized patent appeals court. Yet, in the latest case that the Supreme Court will be hearing the CAFC was totally divided and issued six (6) different opinions. So the specialized patent court just threw up its hands.
How does this relate to today's technology? There have been significant innovations in semiconductors and telecomm to bring us the internet, personal computing and smartphones. But a large part of these innovations have been in the area of software. Just think about your use of the internet, your iPhone or your iPad. Much of the functionality comes from software running on a general purpose processor. Yet the Supreme Court's early decisions that software running on a computer is just math, a law of nature or an abstract idea has now put a cloud on all this software innovation. This is particularly distressing when you consider that a lot of electronic manufacturing takes place overseas, but a lot of software development takes place in the US. Yet, depending on how the Supreme Court rules, much of this software innovation could be ruled as not being patent eligible, regardless of whether it is new, unobvious and well described.
How does this relate to patent trolls? A patent troll is a non-practicing entity, like a University or an independent research company that doesn't make products but rather licenses its inventions to others. But unlike Universities or research labs, a patent troll typically buys patents from defunct companies, strains their interpretation to the nth degree ("We invented the fax machine", "we invented email", "we invented interactive phone response systems"…). Then, rather than sue the maker of the device, they sue thousands of innocent users of the device, no one of which has the knowledge or incentive to fight rather than settle.
Some troll patents relate to software but many don't. Software or not is irrelevant as to what you think of patent trolls. Making software patent ineligible will not solve the patent troll problem. It will just place a big blow on a US industry. There are other things that are being considered to stop patent trolls. One thing would be to make sure the Patent Office is strictly enforcing the legal requirements of new, unobvious and well described. Other things can be done and are being considered. But to try solve the troll problem by saying software is not patent eligible throws out the baby with the bathwater.
All of the above in my opinion, no legal advice intended.
Great post indeed Dylan. I still remember the time when people thought the idea of software patents was absurd. It was not an "apparatus" and therefore it couldn't possibly be something you could patent. Then the first software patent was granted and the floodgates opened.
I have been on both sides of this and it is an extremely tough situation. I think we need a reform but what that is, is hard to define. Patent positions are abused in so many places and a lot of innovation is stopped because the underlying patent is too expensive, or the area is considered to have patent land mines.
I have been on both sides of this and it is an extremely tough situation. I think we need a reform but what that is, is hard to define. Patent positions are abused in so many places and a lot of innovation is stopped because the underlying patent is too expensive, or the area is considered to have patent land mines.
Similar threads
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Co-Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |