China radically increases patent filings

December 23, 2011 11:36 PM

China became the world's top patent filer in 2011, issuing 58% of global intellectual property filings.

Chinese companies were offered government incentives for filing patent applications, even if they were eventually rejected. Legal experts have criticized the quality of the resulting patents. It's worth observing however that, if granted, bad patents make better defensive patents and trolling patents.

Amusingly, a single day record for filing new patent lawsuits was set in September 2011.

posted by jeffburdges (49 comments total)

Legal experts have criticized the quality of the resulting patents.

These days it's a pretty low bar.
posted by Artw at 12:15 AM on December 24, 2011 [2 favorites]

That Julian Sanchez article has some very interesting points, BTW.
posted by Artw at 12:17 AM on December 24, 2011

I steal your patents, China, and I make cheap copies of those products.
posted by twoleftfeet at 12:39 AM on December 24, 2011 [2 favorites]

There are two things to keep in mind: firstly, in China "patents" designates a broader category than elsewhere. China has "invention patents" (which correspond to what is generally understood as a patent elsewhere), "utility model patents" (short-term patents that are issued without examination, first seen in German patent law) and "design patents" (also known as design patents in the US, but as "registered designs" in Europe, they don't protect a technological innovation, but the distinctive look of a product). Secondly, the Chinese government has given all sorts of incentives to encourage the filing of patent applications in China and abroad.

As a result, the numbers are quite inflated. Applications for "invention patents" make up only a small minority of patent filings in China. In my experience, they are examined quite thoroughly: the examiners of China's State Intellectual Property Office are professional and well-trained. But the vast majority of Chinese patents are "utility model" and "design" patents, and there's quite a lot of chaff there. Keep in mind that China introduced its first patent law only some 25 years ago: the learning curve has been quite steep.

This said, that China has become such a prolific filer of patents also has more solid reasons: after Western corporations have outsourced most industrial production there, and now also their R&D, China is also producing huge numbers of engineers. And, where there are engineers, there are patents. As simple as that.

By contrast, here in the West engineering and technological prowess has long been neglected for the benefit of financial creativity. Not surprising, then, if we are overtaken on a number of indicators.
posted by Skeptic at 12:51 AM on December 24, 2011 [10 favorites]

That Julian Sanchez article has some very interesting points, BTW.

Sanchez, like Lemley before him, completely misses the point of patents: patents are not just meant as an incentive to invent, but as incentive to invent and disclose.

When somebody invents something, he basically has three options:
a) file a patent application;
b) keep secret; or
c) publish.

Each one of these options has pros and cons:

Patenting costs time and money and is often uncertain: not only are granting rates lower than most people think (around 50%), and most granted patents substantially restricted in scope compared to the initial application, but enforcing the patent involves the even higher cost and uncertainty of litigation. In addition, patenting involves the disclosure of the invention. On the plus side, the patentee gets an eventual time-limited monopoly that can be traded as an asset.

Keeping the invention secret can compromise its exploitation. Moreover, the inventor runs the risk that somebody else invents parallelly and publishes it, or, worse, patents it himself. Then there's the risk of leaks. On the plus side, if the inventor can keep the secret, this ensures him a more complete monopoly than any patent, for an indefinite period.

Publishing (and the whole Open Source movement is just the latest flavour of this long-available option) does not offer any monopoly. However, if you document your disclosure well, it prevents others from validly patenting the same idea afterwards.

Most industries use a mixture of these three options. Sanchez' point (and Lemley's before him) is that most inventions may be arrived at parallelly by several people, and "defensive patents" are indeed a good proof of that. But, in a world of 7 billion people, each one with his own brain, this is quite unavoidable, I think. Patents are not meant to protect "genius" inventions, but to encourage disclosure and competition in the development of everyday solutions.
posted by Skeptic at 1:35 AM on December 24, 2011 [9 favorites]

Hmm... I find the argument that software patents contain great secrets that might otherwise be lost to the ages a little hard to swallow.
posted by Artw at 2:07 AM on December 24, 2011 [3 favorites]

I look forward to hearing multinationals who've been cramming crap like ACTA on countries via their bought-and-paid-for henchmen in the US political system squeal like stuck pigs when China starts using WIPO, ACTA, and the like to crush them like bugs.
posted by rodgerd at 2:13 AM on December 24, 2011 [1 favorite]

Hmm... I find the argument that software patents contain great secrets that might otherwise be lost to the ages a little hard to swallow.

Good thing nobody has made it. Patents encourage and accelerate the disclosure of usually small and gradual but not obvious improvements.

Also, do not judge the bulk of patents in the software field by the handful that find their way into the news, or worse, by their summaries by clueless journalists.
posted by Skeptic at 2:17 AM on December 24, 2011 [1 favorite]

I wonder if Chinese patents will be as respected by the judges that let patent trolls clog up the system here in the U.S. I kind of doubt it, but for defensive patents, these might be useful.

It's ironic that Americans have been bitching about how the Chinese aren't respecting our IP, and this seems like a good response, any bitching about patents by the U.S can be responded too with 'nu uh, you're violating our patents!'

Same thing that's going on between the big players in the cellphone market today, except between nations.
posted by delmoi at 2:44 AM on December 24, 2011

Good thing nobody has made it. Patents encourage and accelerate the disclosure of usually small and gradual but not obvious improvements.

I am a software developer. I know many software developers. The vast majority of software engineers do not look at patents, study them to learn about advances in the state of the art, or otherwise learn any techniques or practices from patent filings. With few exceptions, the only time most developers are looking at patents is if they happen to be filing one or in order to LOL at some absurdity that made the news. Indeed, I've heard some companies tell engineers not to look at patents, lest awareness of a patent claim turn the company into a knowing infringer.

If the point is to promote and accelerate disclosure, the fact that practitioners in the field tend not to ever look at the darn things is a huge red flag that indicates that the system is not serving its intended purpose.
posted by zachlipton at 3:21 AM on December 24, 2011 [11 favorites]

Haha, the King of Copy, feverishly patenting. Oh the irony!
posted by rmmcclay at 4:08 AM on December 24, 2011

I wonder if Chinese patents will be as respected by the judges that let patent trolls clog up the system here in the U.S.

They don't have to: the large majority of those Chinese patents are never filed, never mind granted, abroad.

The vast majority of software engineers do not look at patents, study them to learn about advances in the state of the art, or otherwise learn any techniques or practices from patent filings.

So, the patent system puts a trove of millions of technical disclosures at their disposal, for free and helpfully classified and indexed for their perusal and they don't use it? Maybe the problem is not the patent system, then. The Chinese certainly read them, at least according to Elon Musk.

Anyway, patents encourage disclosure in other ways than in the patent themselves, you know. For instance, many companies have long had a conscious policy of publishing new developments that they don't intend to patent themselves, in order to pre-empt patenting by third parties. This was for instance the rationale behind the IBM Technical Disclosure Bulletins, and is arguably a big reason for Google's involvement in Open Source. Also, companies have an incentive in submitting their patented technologies to standard-setting committees, which they wouldn't have otherwise: so, when you read a technology standard, you will almost certainly be reading not just one but several patents.
posted by Skeptic at 5:09 AM on December 24, 2011 [2 favorites]

I agree with Skeptic that "patents are not just meant as an incentive to invent, but as incentive to invent and disclose", but I am unaware of any modern patent that enabled/incentivized an important disclosure.

Can you name one? I thought not.  We've reached a technological point where any idea or technique that's non-trivial enough to warrant a two decade monopoly is too detailed to reasonably be contained within a patent application.. and probably uninvestably risky because academia hasn't worked over highly theoretical details yet.  We should therefore abolish the patent system.

Open source and open standards happen because the resulting code and protocols are better and more useful, i.e. developers want it to happen.¹ Linux dominates the server market because its code is simply better.² Android is open source because that makes it more useful to carriers, ensuring it's success, and protecting Google's search monopoly. etc.

We can and do mandate any disclosure we wish for pharmaceuticals already via the approvals process too, btw.

¹ I would prefer that we made open source the law by a priori denying copyright to all utilitarian product like compiled software, basically just apply the SCOTUS ruling that made fashion uncopyrightable. Instead, we'd grant copyright only to the human readable source code, but permitting compiled software to derive a copyright from the source code only when the source code is published.

² As Yaron Minsky says around #t=28m50 in his caml trading CMU talk, "you basically cannot pay people enough to code review [dull] code" (start around #t=25m30 for background), but open source provides some measure of code review.
posted by jeffburdges at 6:09 AM on December 24, 2011 [3 favorites]

Following up on Skeptic's disclosure point. Patents aren't the form of disclosure, they allow companies to show what they did and still retain much legal control over the invention. Java would be a perfect example.
posted by forforf at 6:11 AM on December 24, 2011

I should clarify slightly, patents ARE a form of disclosure, but they are not the way the invention typically disclosed.
posted by forforf at 6:17 AM on December 24, 2011

Java would not have gained the traction it did without Sun releasing the VM, libraries, etc. under an open source license. Microsoft has the install base to push a close source language like C# or VB, but afaik nobody else does.
posted by jeffburdges at 6:30 AM on December 24, 2011

Skeptic, the article you referenced was talking about aerospace-related patents. No one reads software patents. If I want to learn something new about software, I read academic papers, blog articles, or source code.
posted by demiurge at 7:46 AM on December 24, 2011

Skeptic, the problem with software patents is they are rarely about things that are helpful.

For example, there was a bit of an issue in my company because someone's patent covered having an indicator showing that there was a sub-menu in a context menu change directions if the context menu was oriented in that direction, instead of always pointing right.

And it isn't as if the patents contain the actual code or anything useful either. The end result is countless patents for inanities such as putting buttons on the screen that show ads in a unique manner.
posted by Veritron at 7:50 AM on December 24, 2011 [2 favorites]

Hmm... I find the argument that software patents contain great secrets that might otherwise be lost to the ages a little hard to swallow.

So who invented the Antikythera Device again?
posted by happyroach at 8:16 AM on December 24, 2011

I am a software developer. I know many software developers. The vast majority of software engineers do not look at patents, study them to learn about advances in the state of the art, or otherwise learn any techniques or practices from patent filings.

I am a software engineer. I know many software engineers. And none of us can write papers, post messages, or have casual conversations about techniques and practices until the company IP rights are protected via patents.

The patent system may be broken, but it is certainly fulfilling its function of allowing companies/employees to go public with new advancements. (which is often a problem, as good ideas get legs at the time of disclosure -- I.E. patent filing -- and are widespread and "obvious" by the time of granting.)
posted by Tell Me No Lies at 8:20 AM on December 24, 2011 [2 favorites]

No one reads software patents. If I want to learn something new about software, I read academic papers, blog articles, or source code.

Unless you wanted to learn about what the engineers at, say, Apple are up to, since they don't produce academic papers, blog articles, or (for the most part), source code.

And it isn't as if the patents contain the actual code or anything useful either

This is mostly a result of the Patent Office's absurd preference for flow charts and prose description of algorithms rather than pseudocode. Academic papers use pseudocode for a reason, and the Patent Office should follow suit.

I am unaware of any modern patent that enabled/incentivized an important disclosure. Can you name one? I thought not.

Pretty much any drug patent, for starters.

We've reached a technological point where any idea or technique that's non-trivial enough to warrant a two decade monopoly is too detailed to reasonably be contained within a patent application.. and probably uninvestably risky because academia hasn't worked over highly theoretical details yet.

Again, for pharmaceuticals, industrial chemicals, and mechanical devices this is just not the case. The inventions involved are non-trivial (whatever that means; I assume some combination of non-obvious and valuable) and the patents disclose everything important about the chemical or device (e.g. the chemical structure, how to make it, and often specific evidence of utility in animal studies).

Things get a little fuzzier with business methods and software, but I would still argue there are clear examples, such as Google's (technically Stanford's) patents on the Page Rank algorithm.
posted by jedicus at 8:51 AM on December 24, 2011

Also, do not judge the bulk of patents in the software field by the handful that find their way into the news, or worse, by their summaries by clueless journalists.

Considering that they're usually in the news because they're being used to hinder actual innovation, I'd say say it's fair to judge.
posted by ChurchHatesTucker at 8:54 AM on December 24, 2011

Can you name one? I thought not.  We've reached a technological point where any idea or technique that's non-trivial enough to warrant a two decade monopoly is too detailed to reasonably be contained within a patent application..

Is a Nobel Prize "non-trivial" enough for you? That's just one example off the top of my head, but there are others.
Anyway, simple != trivial. And patents aren't limited in length, anyway...
posted by Skeptic at 9:18 AM on December 24, 2011

So, the patent system puts a trove of millions of technical disclosures at their disposal, for free and helpfully classified and indexed for their perusal and they don't use it? Maybe the problem is not the patent system, then.

Maybe it is? I don't know how closely you read zachlipton's comment, but the reason they don't read the disclosures is that there's a fair threat that their employer will be found to have willfully infringed, and face up to treble damages. Even if the case never gets to trial, like most, that makes for a strong bargaining chip for settlements.

zachlipton's not some isolated example either - I used to be a software engineer, and that was official company policy for us. Whatever you do, don't read any patents. If you think you may have come up with a patentable idea, tell someone in legal and let them look into the state of the prior art, don't you DARE look yourself. I'm an attorney now, and engineers working for our (defendant) software company clients on the past several patent infringement cases have had similar policies in place.

And honestly, reading these patents wouldn't have taught them anything anyway. Every asserted patent so far in these cases has covered an incredibly basic software feature that has been well-known in the industry at the time, and most of them were already anticipated by earlier-issued U.S. patents.

So yeah, I'd call it a problem with the patent system. If the PTO cranks out so many patents covering well-known, basic aspects of software that companies actively discourage their software engineers from reading any patents lest they happen to read something (well-known, basic) that they are doing (as are many others, because it's well-known and basic but somehow successfully prosecuted anyway) which could subject their company to even higher losses to some NPE, then it's simply not going to facilitate any useful disclosure. I mean: the information will be out there, yet the target audience will have to avoid seeing it, which isn't useful.
posted by cobra_high_tigers at 9:41 AM on December 24, 2011 [2 favorites]

The Page Rank patent is an exception to the sea of useless (for reconstruction) software patents out there. That patent is actually written like an academic paper and the thing described is an algorithm for a specific purpose.

Constrast that with software patents like this one from Apple. This patent gives no more information on actually constructing the software than a non-technical person coming up with a list of specifications:

I want a program that can make avatars, and the eyes have to animate. Also, could you make it work on the iPad, and have it render 2D and 3D avatars.

When 13 pages plus images in a patent are as useful to reimplementation as that, something is wrong.
posted by demiurge at 9:44 AM on December 24, 2011 [1 favorite]

And those patent incentivized disclosure how exactly? In other words, what do we know about Page Rank that we wouldn't know without Stanford's patent? We knew it used a Markov chain model without the patent. We don't know many details of that model even now.

Pharmaceutical patents obviously cannot incentivize disclosure beyond the existing requirements imposed by the FDA. If we abolished pharmaceutical patents, we'd simply see hospitals, universities, and drug companies push drugs through clinical trials using federal grant money. It'd obviously remain illegal to sell pharmaceuticals without disclosing their contents, including contamination risks inherently arising form the manufacturing process.

I'm unconcerned about manufacturing, chemical plant, electronics, etc. patents like GMR because you need so much money to build the plant that the patent adds very little overhead. And a couple hundred years of legal tradition are preventing such patents from being as susceptible to abuse, patent trolls, etc. I'll grant that one should take a longer look than I have before screwing with that system, but the issue isn't that patents are good.
posted by jeffburdges at 9:44 AM on December 24, 2011 [1 favorite]

And then there's design patents... "Hey! We own rectangles! And the color black!"
posted by Artw at 10:32 AM on December 24, 2011 [1 favorite]

So, the patent system puts a trove of millions of technical disclosures at their disposal, for free and helpfully classified and indexed for their perusal and they don't use it? Maybe the problem is not the patent system, then.

I have worked for large software companies and small startups.

At several of them we were specifically told not to look at patents in our field because of issues with willful infringement.

These are company lawyers telling us this, via our managers.

Yeah, the problem is specifically the patent system. Patents aren't being used to "promote the Progress of Science and useful Arts" anymore, they're used to build moats around established companies products, build up IP assets to make for a quick acquisition/buyup in startups and build an entire industry for lawyers.

The fact that so many software engineers, practitioners in the field, are against them should open more eyes to the fact that the system is broken.
posted by formless at 11:23 AM on December 24, 2011 [3 favorites]

Apple engineers are specifically instructed not to research patents. When I started my own company, I set the same rules. Besides the issue of willful infringement, most engineers do not understand the patent system anyway. Infringement and patentability are legal questions best left to the lawyers.
posted by ryanrs at 1:17 PM on December 24, 2011 [1 favorite]

I am a software engineer. I know many software engineers. And none of us can write papers, post messages, or have casual conversations about techniques and practices until the company IP rights are protected via patents. The patent system may be broken, but it is certainly fulfilling its function of allowing companies/employees to go public with new advancements.

I read about advancements all day long. Very few of those are patented. I seem to read more about patent lawsuits than patented advancements, in fact.
posted by RobotVoodooPower at 1:30 PM on December 24, 2011

I read about advancements all day long. Very few of those are patented. I seem to read more about patent lawsuits than patented advancements, in fact.

Yeah, the argument that patents help promote advancements is absurd, and here are two classic examples of why from great moments in computer history:

First, Tim Paterson's QDOS, which Microsoft licensed for it's DOS that it licensed to IBM, was a complete ripoff of Gary Kildall's CP/M. Kildall had the chance to sell to IBM, but "DRI founder Gary Kildall blew off IBM to gallivant around in his airplane, and the frustrated IBMers turned to Gates for their operating system."

Regardless of Microsoft's later business practices, the separation of the OS software from the hardware led to an era of cheap computers that revolutionized the industry.

The second classic story is Apple's "borrowing" of Xerox PARC's Alto GUI. By incorporating this into the Mac, they revolutionized the computer field, making computers easier to use and introducing the importance of design and usability.

If either Xerox or DRI had patented their ideas, we would be stuck in a technological stone age. The problem with these companies wasn't that patent protection was lacking. It was that they failed to see the opportunities of their technology and they failed to execute.

I realize I'm diluting my point now, but it's the same story with Disney's abuse of the copyright system. You have a company that borrowed heavily from the public domain to create great innovative works of modern storytelling, and then turn around and use the copyright system to prevent future borrowing.
posted by formless at 1:50 PM on December 24, 2011

Derek Lowe, whose excellent pharma blog "In the Pipeline" was featured on the Blue a while back, wrote about this issue earlier this year.

The last couple of years have seen many people predicting that China would soon be leading the world in patent applications as well, which can be the occasion for pride or hand-wringing, depending on your own orientation. But there's a third response: derision.

The blog post (and relatively moron-free comments section) is worth reading in its entirety.
posted by richyoung at 1:56 PM on December 24, 2011

The United States got its industrial start by refusing to sign international intellectual property treaties. I can't link from here, but the info is easy to find. Edison and Bell made lots of money by 'stealing' foreign inventions and media, as an example.

Now that China is more and more a nation of consumers this becomes interesting. The US does not like the fact that it gets inundated by cheap Chinese knockoffs, how will it feel when it can not sell products in China because they are protected by patents? I can't wait to see how politicians and corporations will try to spin this.

Finally, Chinese engineering is getting good. The latest example I saw is a bunch of workers and engineers from one of the carbon fibre bicycle factories, one that built high end 'Italian' bikes, started their own company. At first they where making copies, but now they have their own models and a couple of improvements over the Italian designs. All the reviews I have read are very positive. And you can get your bike in ANY color, with or without decals.

There is no one more creative than a designer or engineer with something to prove.
posted by Ayn Rand and God at 2:02 PM on December 24, 2011

And those patent incentivized disclosure how exactly?

By providing a period of exclusivity, thus incentivizing the development of the technology in the first place? By creating a concrete property right around which deals could be structured?

Pharmaceutical patents obviously cannot incentivize disclosure beyond the existing requirements imposed by the FDA.

The FDA can demand disclosure, but it can't demand that new drugs be invented. The patent system promotes disclosure in areas of technology where trade secrets are an alternative (e.g. manufacturing methods, server-side software, financial methods) but it also promotes research and development even in areas where trade secrets are not a viable alternative (e.g. pharmaceuticals).

If either Xerox or DRI had patented their ideas, we would be stuck in a technological stone age.

Why? Key technologies are patented all the time. For example, patent activity from Intel, AMD, and other semiconductor companies didn't stop Moore's Law from holding for decades.

I realize I'm diluting my point now, but it's the same story with Disney's abuse of the copyright system. You have a company that borrowed heavily from the public domain to create great innovative works of modern storytelling, and then turn around and use the copyright system to prevent future borrowing.

The patent system works very differently than copyright. Whereas copyright includes the exclusive right to create derivative works, patents do not. If you patent the chair, I can patent a chair with wheels. I'll probably still need a license in order to manufacture them, but on the other hand you'll probably want a license to my invention because it's a significant improvement on yours. This kind of things happens all the time and leads to cross-licensing agreements.
posted by jedicus at 2:05 PM on December 24, 2011

The US does not like the fact that it gets inundated by cheap Chinese knockoffs, how will it feel when it can not sell products in China because they are protected by patents?

Patents are a two-way street. US companies want worthwhile patent protection in China because it's a means for stopping those knockoffs. Lots of US companies already apply for patents in China. From the US point of view the bigger problems are discrimination against foreign businesses and judicial corruption, which tend to make those Chinese patents hard to enforce.
posted by jedicus at 2:16 PM on December 24, 2011

Just like the US used to do. And we got away with it.
posted by Ayn Rand and God at 2:20 PM on December 24, 2011

If you patent the chair, I can patent a chair with wheels. I'll probably still need a license in order to manufacture them...

You don't see how that hinders rather than helps innovation?
posted by ChurchHatesTucker at 3:12 PM on December 24, 2011

The patent system works very differently than copyright.

But both are used in very much the same way, to create a monopoly that established companies use to keep competitors out of a market.

By providing a period of exclusivity, thus incentivizing the development of the technology in the first place?

I'm sorry, but I find it hilarious that the developers in this thread, the ones who create, who develop the technology, are arguing against the patent system. But the patent attorney is arguing for it. Wouldn't it be the developers who know what drives them? And it's obviously not patents.

If either Xerox or DRI had patented their ideas, we would be stuck in a technological stone age.

Why? Key technologies are patented all the time. For example, patent activity from Intel, AMD, and other semiconductor companies didn't stop Moore's Law from holding for decades.

Maybe you don't believe me, but here's a quote from Bill Gates:

If people had understood how patents would be granted when most of today’s ideas were invented, and had taken out patents, the industry would be at a complete standstill today.

You have a vested interest in the patent business, I understand that. I think the current patent system is destructive, yet in my latest venture I'm aggressively looking for patentable technologies to shore up our IP portfolio, both for investment and acquisition purposes.

That doesn't mean it's good, it just means I understand the Prisoner's Dilemma that is the patent system. It's going to lead to a diminished public commons, and we're all going to suffer as a result.
posted by formless at 3:32 PM on December 24, 2011 [2 favorites]

You don't see how that hinders rather than helps innovation?

The necessity of a license might hinder commercialization, but it doesn't hinder the invention itself.

And arguably the existence of a well-defined property right more than makes up for the hindrance. Patents solve the Arrow information paradox, which encourages negotiations, especially complex ones like standards setting organizations . The well-defined property right also makes possible complex licensing and sub-licensing schemes that are less amenable to trade secret and know-how agreements.

I'm sorry, but I find it hilarious that the developers in this thread, the ones who create, who develop the technology, are arguing against the patent system. But the patent attorney is arguing for it. Wouldn't it be the developers who know what drives them?

Most developers are not affected by the patent system in any direct way. They generally don't read patents. They don't decide whether to file for a patent on a given invention, or whether to license a given patent, or whether to enforce a patent owned by the company they work for, or whether to pursue a given area of research or product development, or whether to offer patents to a patent pool. In my experience most are not particularly familiar with the history, theory, economics, or practical realities of patent prosecution, licensing, and enforcement. So I take their opinions on patent policy with a pretty large grain of salt.

It's true that developers aren't driven by patents. Most developers are driven by their pay check (this includes a large percentage of free and open source software developers, who are paid to work on the software by companies like IBM and Google). Volunteer FOSS contributors and (some) academic developers are driven by other concerns, but patents have had virtually no effect on FOSS or academic research.

But this is beside the point. The issue isn't what drives developers, it's what drives businesses to research and develop new products and services, what turns academic research to commercially available products, what encourages negotiation, what gets deals done. That's what the patent system is for, not to be a carrot for individual developers.

You have a vested interest in the patent business, I understand that.

Not really. I do represent a few clients, but my day job is as an academic researcher. The research group I work for isn't pro-patent as such, but rather pro-"encouraging an efficient cycle of investment, innovation, development, and commercialization." If there were a viable, efficient alternative to the patent system, we'd focus on that.

And bear in mind that I have two degrees in computer science and have worked as a software developer. I spent the late 90s and early 2000s steeped in the anti-IP culture common to software developers. I have used, supported, and even occasionally developed free and open source software for over a decade. And I spent most of my law school patent class arguing against my professor. So I like to think that I come by my views honestly.
posted by jedicus at 4:11 PM on December 24, 2011

I read about advancements all day long. Very few of those are patented.

I wonder how you would know how many have patent applications associated with them? Even the universities are big into patents these days.

I seem to read more about patent lawsuits than patented advancements, in fact.

There's only so many times you can report on all of the new technology that Amazon, Apple, Cisco, Erikson, Google, IBM, McAfee, Nokia, Samsung, Skype, Sony, et al. produce on a regular basis.

Admittedly such stories don't have the staying power of day after day of coverage of the same three patent suits, but there is a hell of a lot of patented advancement happening as we speak.
posted by Tell Me No Lies at 4:15 PM on December 24, 2011

It's true that developers aren't driven by patents. Most developers are driven by their pay check.

Just as a side note, these are often linked. I've never received less than $2000 for filing a patent or less than $5000 when it was granted.

(in fact me and fourteen other guys were responsible for Cisco's change to a "if there are more than three people on the patent then you split the money" policy in the mid 1990s. Man did they pay for *that* patent)

Also, patents look very good on resumes.
posted by Tell Me No Lies at 4:37 PM on December 24, 2011 [1 favorite]

Wouldn't it be the developers who know what drives them? And it's obviously not patents.

Much like management, the legal system doesn't need developers to question its notions of what is best for developers
posted by cobra_high_tigers at 6:16 PM on December 24, 2011

In my experience most are not particularly familiar with the history, theory, economics, or practical realities of patent prosecution, licensing, and enforcement. So I take their opinions on patent policy with a pretty large grain of salt.

Discounting an entire group of stakeholders in the patent discussion like that is probably a bad idea.

For example, while I've worked and will continue to work as a software engineer, my background is in Cognitive Science (bachelors) and Information Economics, specializing in Incentive Centered Design. Designing optimal incentive systems for information and social systems was a large part of that, including study of contract theory and information asymmetries.

And my opinion, backed up by reading case studies and real-world experience, is that the current patent system in the US is not conducive to promoting technological progress. And that engineers and developers have a large stake in the game, because innovation does not always come from the top-down, it often arises in the trenches when engineers are working on a tough problem.

So I like to think that I come by my views honestly.

You do, and my attack on your arguments was frankly ad hominem and rightly shot down. My apologies.

I can understand that patents can be useful to promote innovation, but the current system practiced in the US around software patents seems to be doing more harm than good.
posted by formless at 6:29 PM on December 24, 2011

The FDA can demand disclosure, but it can't demand that new drugs be invented.

Yes, luckily the drug companies have almost nothing to do with inventing any important drugs either. Researchers paid by NIH grants develop drugs. Drug companies (a) do the marketing and (b) develop retread drugs that fit an established markets.¹

You might claim that pharmaceutical company lobbyists keep tax dollars flowing into the NIH. If that's true, then yes that's a social service they perform, but it's well established that they don't pay for the interesting drugs themselves.

If you look outside the U.S., there is a "visible efficient alternative to the patent system" for pharmaceuticals, namely national health services fund drug development, like they do already in wealthy countries, except more so, and outsource production to the lowest bidder.

¹ There are certainly cases where retread drugs reduce side effects for some people, but mostly the company spends a fortune converting doctors into prescribing said drug for people who'd never suffer said side effects anyways. And often this advertising ends up pushing a poorly understood, or even dangerous, drug over a much more well understood, and sometimes safer, off-patemt drug.
posted by jeffburdges at 7:26 PM on December 24, 2011

Discounting [software engineers] in the patent discussion like that is probably a bad idea.

Speaking as a developer who works for other people, other than reputation and spare cash I have little or no involvement with patents. I file them and the business foks go off and do businessy things with them. Occasionally I get deposed.

While I supply the seed of the patent in the form of a new idea, the lawyers do a huge amount of work to make it their own. Often when I review the final draft I have a tough time sorting out my original text.

So patents are legal documents that I have only the most basic understand of. Technology, yes. Legalese, no.

As such, as a software engineer I don't feel like a stakeholder in a patent any more than I do in any other business deal. Sure my technology was involved, but that could be said of anything the company does. And you don't want my advice about running a company.
posted by Tell Me No Lies at 7:31 PM on December 24, 2011 [1 favorite]

And arguably the existence of a well-defined property right more than makes up for the hindrance. Patents solve the Arrow information paradox, which encourages negotiations, especially complex ones like standards setting organizations . The well-defined property right also makes possible complex licensing and sub-licensing schemes that are less amenable to trade secret and know-how agreements.

To a point, I'll agree with that. Patents facilitated the whole MPEG-LA system of patent pools, and that system is probably preferable to one where a dozen competing formats would have made online video a complete disaster. On the other hand, YouTube and friends owe a huge part of their success to MPEG-LA's policy of not charging royalties on free H.264 web video.

Yet when it comes to software-related tech, patents are usually used to screw over standardization efforts rather than facilitate them. The best example of this is the W3C and the web standards process. A core value of the web is interoperability and openness: numerous competing web browsers, including open source browsers and rendering engines, run on machines as diverse as computers, tablets, phones, televisions, and alarm clocks. As a result, a web standard encumbered by non-royalty-free patents is unworkable to the extent that it probably won't actually become a widely implemented standard, defeating the entire point. We've seen this a couple times recently with Apple, such as the Widgets access requests policy patents and the W3C touch events specification.

Consider the latter for a second. Touch events are pretty important to the future of the web. I'll try to give a reasonably simple explanation. What this standard does is tell web browsers and web developers how to deal with actions the user performs on a touch screen. Current web standards allow a developer to say "do this action when the user clicks their mouse and then do this other action when the user drags their mouse across the screen." Currently, mobile web browsers turn touch events on phones and tablets into mouse events. Because the mouse events have to maintain backward compatible, this is a real limitation on mobile web applications: mouse events can't really reflect the concept of multitouch where multiple mice can have different positions at the same time. As a result, the iPhone introduced a new set of touch events in Safari starting in iOS 2.0, which allowed web developers to specify exactly what their applications should do in response to the user's touch. Google implemented a similar system in their browser for Android, and Firefox has also been adding this support to its mobile browser.

So touch events in web browsers are a pretty useful thing that is getting implemented by a couple different vendors. It would be really swell for web developers if every browser on every device could handle this the same way––i.e. the specification was standardized. That way, I could write a game like Flight Control as a web app and have it just work on iPhones, iPads, Android devices, even Windows Mobile devices if we're lucky. Because of this, the W3C has been developing the Touch Events standard.

And where do patents fit into all this? Well, the W3C's poor patent policy states: "In order to promote the widest adoption of Web standards, W3C seeks to issue Recommendations that can be implemented on a Royalty-Free (RF) basis. Subject to the conditions of this policy, W3C will not approve a Recommendation if it is aware that Essential Claims exist which are not available on Royalty-Free terms." In other words, web standards are supposed to be freely implementable by all. That might be a utopian dream inspired by some free software geeks, but the reality of the web is that free and open standards are the only way we're going to be able to live with each other.

So the standards process has been chugging along pretty well, when Apple, a W3C member, discloses four "essential" patents and patent applications and announces that it will not offer a royalty free license under the W3C patent policy. Apple is currently suing Samsung and Motorola for infringement of, among others, US Patent #7,812,828, a multitouch patent that Apple has declared to be essential to implementations of the Touch Events specification. While browser manufactures can do whatever they want (and risk lawsuits as a result), disclosures of this sort usually delay the standards process for months or years while a patent advisory group (PAG) studies the issue. In the case of the W3C Widgets spec, the PAG spent two years of hard work from formation to issuing its report, ultimately culminating in the W3C's public Call for Prior Art, which seeks evidence to invalidate, or at least work around, the patents of one of the organization's own members! A PAG will likely be formed for Touch Events early next year.

There are hundreds of millions of non-Apple multitouch devices in the world right now, with 5+ million more being added every week. I honestly don't give a crap whether Apple was first or not; I care that there's a simple universal way for web pages to work with frickin' touch screens. I can't see how the disclosures in Apple's patent filings facilitated advancement of the technology: competing devices were already on the market before many of these patent applications were ever published. There's no upside here; we haven't gained anything as a society because Apple published these patents instead of keeping their contents secret. Rather, they are simply being used to spread FUD and block interoperability, preventing technological advancement. Now what was the point of the whole system again?
posted by zachlipton at 7:35 PM on December 24, 2011 [4 favorites]

But this is beside the point. The issue isn't what drives developers, it's what drives businesses to research and develop new products and services, what turns academic research to commercially available products, what encourages negotiation, what gets deals done. That's what the patent system is for, not to be a carrot for individual developers.

I respect that, and like Tell Me No Lies says, I get that the patent system really isn't about me as an engineer at all. But my understanding is that patents are supposed to be a grand bargain: disclose the workings of your invention to the world so we can all learn from them, and in exchange, we'll give you exclusivity for a period of time. If actual practitioners in the field never look at the disclosed details (because they are usually incomprehensible, worthless, and/or the lawyers have told everyone not to look at patents), patent-holders are getting all the upside of that bargain without providing any societal benefit by disclosing their inventions.
posted by zachlipton at 7:50 PM on December 24, 2011

I can understand that patents can be useful to promote innovation, but the current system practiced in the US around software patents seems to be doing more harm than good.

I'm not sure there's enough evidence to say "more harm than good," but I agree that the current system practiced in the US (and elsewhere) has quite a few faults, many of which are exacerbated in the software context (e.g. loose written description and enablement requirements, the terrible Patent Office attrition rate).

But my understanding is that patents are supposed to be a grand bargain: disclose the workings of your invention to the world so we can all learn from them, and in exchange, we'll give you exclusivity for a period of time.

It's important to recognize that, although disclosure is required for patenting, the patent need not be the only form in which the invention is disclosed. Indeed, allowing other forms of disclosure (e.g. a product on a shelf, technology transfer to a licensee) while protecting the property right is one of the major benefits of a patent.

If actual practitioners in the field never look at the disclosed details (because they are usually incomprehensible, worthless, and/or the lawyers have told everyone not to look at patents), patent-holders are getting all the upside of that bargain without providing any societal benefit by disclosing their inventions.

In many fields it's common for the invention to be disclosed in a more comprehensible, useful form than the patent itself. For example, drugs may be patented but there are also medical journal publications and FDA disclosures. So just because nobody looks at the patents doesn't mean that patents don't result in or encourage useful disclosures.

Software suffers in this regard because a) there isn't a strong tradition of professional publications in computer science*, particularly for innovations that come out of the industry b) the closed-source nature of a lot of software prevents disclosure through the product itself and c) a lot of innovation occurs on server-side software, which prevents disclosure whether it's open or closed source.

* Compared to, say, medicine, where publication is a requirement for being taken seriously by practitioners.
posted by jedicus at 8:12 PM on December 24, 2011 [1 favorite]

We've mostly talked about U.S. patents in this thread, as much or more my fault than anyone else's, but actually I posted this for one specific reason :

There is a common argument in favor of patents that goes "Americans/Europeans are outsourcing all our manufacturing so we need some way to hang onto the income stream." I believe this gives lie to that argument.

I doubt that Chinese academics are catching up to western academics any time soon, if only because we'll continue stealing their best people, but clearly Chinese engineering is catching up quickly. I'd suspect however that Chinese patenting will overtake the west even before their engineering does, given (a) the Chinese government can effectively control this itself, without depending upon western outsourcing, and (b) that the worst patents become the best patents once granted.

I donno if Apple's patent war against Android will metastasize into a global patent based trade war, but maybe, heck maybe that's a good thing. Btw, the U.S. recently listed the Chinese government search engine Baidu as a rogue site.
posted by jeffburdges at 9:46 PM on December 24, 2011 [1 favorite]