When Patents Attack
July 28, 2011 6:59 PM   Subscribe

When Patents Attack. The team at PRI's This American Life and NPR's Planet Money bring you an hour long look into the growing "Mafia War" around software patents. Diving into the corporate filings, patent acquisitions, and office locations of Nathan Myrhvold's Intellectual Ventures and it's shell companies, Laura Sydell and Alex Blumburg uncover a disturbing protection scheme which threatens to undermine the competitiveness of the US tech industry[pdf].

Intellectual Ventures responds. Forbes suggests some solutions, including abolishing software patents.
posted by Popular Ethics (123 comments total) 61 users marked this as a favorite

 
I heard this piece on the weekend, and I'm a bit shocked it hasn't made more waves. This is a great piece of investigative journalism, and I really hope it opens a broader discussion about the failures of the patent system, and not just another "infringers gonna infringe" straw man dismissal. The consequences of inaction could be dire on the industry.
posted by Popular Ethics at 7:03 PM on July 28, 2011 [7 favorites]


(sorry for mis-apostrophing the its)
posted by Popular Ethics at 7:04 PM on July 28, 2011 [2 favorites]


We need to get rid of software patents. They are completely ridiculous.

One of the big problems with the idea of software patents is that software is all analogous to itself. A program that 'does' one thing can easily be made to 'do' another thing simply by changing the meaning of the inputs and outputs.

It seems like software patents have become more and more damaging recently, I wonder if it's going to cause a lot of innovation in software to move over to Europe, as developers decide they don't want to bother dealing with the patent mess.
posted by delmoi at 7:08 PM on July 28, 2011 [9 favorites]


I didn't see IV answer (or NPR directly put to them) the key criticism: that these patents are crap. They do not represent actual innovation. If these patents were genuinely innovative, people would want to license them for production. If they were innovative, there would not be a large number of identical "inventions". This appears to be an artifact of bad patents issues by PTO and the enormous cost of litigation. If lawyers were cheap, would this strategy work?
posted by a robot made out of meat at 7:18 PM on July 28, 2011 [4 favorites]


Regarding IV's shell company network: The Federal Trade Commission recently recommended [pdf] (among a whole host of other things) "the enactment of legislation requiring the public recordation of assignments of patents and published patent applications. To ensure that such listings provide maximum benefit to public notice, they should identify both the formal assignee and the real party in interest." Of course, given how long it took the current patent reform act to get traction I wouldn't hold my breath.

It seems like software patents have become more and more damaging recently, I wonder if it's going to cause a lot of innovation in software to move over to Europe, as developers decide they don't want to bother dealing with the patent mess.

That would be the opposite of what happened with gene patents and patents on genetically modified organisms. The US embraced them, Europe and Japan did not, and now the US biotech industry is leaps and bounds ahead of theirs.
posted by jedicus at 7:20 PM on July 28, 2011


I am reading a very interesting history (Smil, Inventing the Twentieth Century) of key inventions and commercialization of older inventions which transformed the world 1870-1910 and I was shocked how much patents, fights, priority, licensing, cutting deals, competing portfolios, were all a huge part of the fight back when we were figuring out which filament would last for an extra day in a vacuum.
posted by shothotbot at 7:20 PM on July 28, 2011


(I heard this over the weekend too. Major complements to TAL for making a patent fight gripping and important)
posted by shothotbot at 7:21 PM on July 28, 2011 [1 favorite]


If lawyers were cheap, would this strategy work?

Lawyers are unlikely to get cheaper, but overturning bad patents could be cheaper. Lowering the standard of proof for invalidity, tightening the written description and enablement requirements, and making fee-shifting more common are all ways to make the process cheaper and reduce the viability of IV's business model.
posted by jedicus at 7:24 PM on July 28, 2011 [2 favorites]


We need to get rid of software patents. They are completely ridiculous.

As the program mentions, the USPTO was originally of a similar mind, and was effectively forced to accept software patents by the courts.

Now however, the USPTO tends to grant pretty much everything and leaves it to the courts to sort it out. The courts, meanwhile, tend to assume that if the patent was granted, it is valid.
posted by ChurchHatesTucker at 7:25 PM on July 28, 2011 [7 favorites]


That would be the opposite of what happened with gene patents and patents on genetically modified organisms. The US embraced them, Europe and Japan did not, and now the US biotech industry is leaps and bounds ahead of theirs.

That's not really the whole story. The EU is extremely skeptical of the biotech industry, to the point where are outright bans on field tests of GM crops and there was a commission to investigate health risks associated with Roundup Ready plants, etc... it's not that they were perceived as unfriendly by the industry, they are in fact quite unfriendly to the industry.
posted by mek at 7:29 PM on July 28, 2011 [6 favorites]


Who (other than patent trolls) is in favor of software patents at this point? Does anybody knowledgable really think they encourage innovation more than they discourage it?

I mean, I can't imagine that companies like Apple and Microsoft enjoy paying $4 billion for ridiculous patents just to prevent anyone else from buying them first and suing them. That kind of money could have gone a long way towards lobbying for patent reform.

On the other hand, who's going to argue that patents they've spent billions of dollars on should be worthless?

"Sure, in theory, it might be nice if we worked somewhere other than a giant field of shit, but… we've worked so hard to build our pile of shit the highest."
posted by designbot at 7:48 PM on July 28, 2011 [16 favorites]


The joke's on Myrhvold. I patented the algorithm for "masturbating in public using a $600 set of cookbooks" years ago. TREBLE DAMAGES!
posted by benzenedream at 8:00 PM on July 28, 2011 [12 favorites]


i found this very interesting. Alex Bloomberg was doing the kind of question asking i wish more journalists would do.

when they were out talking to software engineers in the park at lunch, i liked how the one guy was like, "yeah, i've worked on a lot of software patents and they are bullshit."
posted by sio42 at 8:09 PM on July 28, 2011 [1 favorite]


That would be the opposite of what happened with gene patents and patents on genetically modified organisms. The US embraced them, Europe and Japan did not, and now the US biotech industry is leaps and bounds ahead of theirs.
As a software developer, the idea of making the software industry like the bio-tech industry sounds absolutely horrible. With biotech, it intrinsically takes millions of dollars to research and develop products. A $700 patent fee and a few thousands in legal fees isn't a big deal, obviously. With software, the startup costs are zero.

The threat of patents is that it's going to take tens of thousands of legal fees and patent licensing fees to develop a product, rather then nothing (other then developer time, which is free if you're a developer)

I would much rather have a 'smaller' industry open to all then a 'more successful' one controlled by Apple and Microsoft, where developers are restricted from innovating on their own, and are forced to hand over money to parasitic lawyers.
posted by delmoi at 8:10 PM on July 28, 2011 [10 favorites]


Glad this is getting mainstream press again.

An interesting historical parallel to this battle might be independent filmmakers vs. the Edison Trust. The independents not only won, creating Hollywood and the major movie studios known today, they also were responsible for an innovation known as movie stars.
posted by RobotVoodooPower at 8:11 PM on July 28, 2011


It's too bad Nathan Myrhvold is running such a terrible "business", because I'd really like a copy of "Modernist Cuisine".
posted by atbash at 8:13 PM on July 28, 2011


Not surprised Intellectual Ventures run by Nathan Myrhvold, who has been helping Microsoft rape computer and software consumers for years. What a waste of a gifted brain.

In Silicon Valley, IP law firms are popping up like mushrooms, and they're flush with cash. It's not unusual to see entire commercial building complexes taken over by one or two former senior partners of a law firm that just won a big IP case. The partners are flush with cash and are able to pick up clients on the heels of their "victory". Firms like Wilson, Sonsini (WSGR) have sleepover rooms for paralegals, so they can work late into the evening and start early the next morning, to generate all important billing hours.

What gets lost is the cost of all this patent attack and defense is that the costs of litigation gets passed directly to consumers, while (as usual) the attorney's score big time.
posted by Vibrissae at 8:13 PM on July 28, 2011 [4 favorites]


Bullies act tough until you come after them.

Silicon Valley's got a lot of newly-minted uber-rich folks hanging around these days, some of whom actually still have their consciences. Wouldn't it be nice for some of them to band together, lawyer up and turn a patent troll into a smoking hole in the ground? Then a quick round of "what a shame it would be for this to happen to you"-type conversations with some of the others, and maybe we could finally get back to making stuff.

Oh wait, crap, all the rich folks are liberatarians, who believe in the sanctity of imaginary property and that the market will magically fix the problem (and that, by definition, there can't ever be a problem).
posted by ubernostrum at 8:14 PM on July 28, 2011 [4 favorites]


There is obviously a lot of abuse that stifles industry, but developers should have some means of protecting their work. Is the system of copyright insufficient? Patent absurdity is a good documentary on this issue.
posted by MisplaceDisgrace at 8:16 PM on July 28, 2011


I got an algorithm patent, kinda by accident. I don't do any gizmo-oriented research. It's all just math and modeling. But... I was working with some shady people who I realized might patent what I was doing, and so I went ahead and did it (through my university), just to save me from getting homicidal should things go awry later.

But I was shocked that I got the patent. I mean, I felt like I was patenting calculus or differential equations or something. And I'm sure the patent is little more than a cute line on my cv: impressive-sounding to lay people, but pretty much toilet paper otherwise. It seems trivial to get around most algorithm patents. "Oh, they used unscented Kalman filters. We'll use HMMs instead." That sort of thing.
posted by mondo dentro at 8:17 PM on July 28, 2011 [4 favorites]


Wouldn't it be nice for some of them to band together, lawyer up and turn a patent troll into a smoking hole in the ground?

I vaguely recall from that This American Life show this past weekend that the patent troll firm had over 1300 shell companies (!!) and , I truly hope that I heard this wrong, was backed by a group of investors who had funded it with a billion dollars and change and that over the next decade they hoped to quadruple that investment. A Billion dollars?? If I heard that correctly wtf!?? Which probably explains why no one yet has turned them into that smoking hole.
posted by Poet_Lariat at 8:28 PM on July 28, 2011 [1 favorite]


I really liked this program... obviously, this has been an issue bedeviling us in the software industry for some time now, but it's been hard to get anyone to care. In a way, I can't blame people -- it's a tough thing to understand. I've really gotta commend TAL for making it into a compelling, accessible narrative... hopefully, this will attract a little more attention from lawmakers.
posted by ph00dz at 8:28 PM on July 28, 2011


Poet_Lariet - according to the show they received more than five billion dollars in venture capital.
posted by The Lamplighter at 8:34 PM on July 28, 2011 [1 favorite]


Oh... and for the record, I'm totally not at all against software patents. I mean, MPEG for instance, is something novel and new. It's a thing that does something... it has a corollary to the physical world in that it functions like a device. Cool. Works for me... that's why we have a patent system.

The real problem is a flood of ridiculous and competing trivial patents. I don't know what we should do to fix it, though...
posted by ph00dz at 8:36 PM on July 28, 2011


The USPTO doesn't seem to have the knowledge and/or manpower to make any reasonable decisions regarding patent applications in this arena, and hasn't for years. This is the office that granted the Amazon patent for one-click technology. I'm not saying they're wholly to blame, but simply rubber stamping software patent applications and letting the courts sort it out is something of a dereliction of duty, no?
posted by axiom at 8:36 PM on July 28, 2011 [3 favorites]


There is obviously a lot of abuse that stifles industry, but developers should have some means of protecting their work.

Copyright is good enough. If someone can re-implement my ideas, why should I care? Chances are I wouldn't be able to do anything without using ideas that someone, somewhere also came up with.

Which is the other problem with software patents, originality. If you're the first person who comes across a problem, and solves it in the obvious way, why should you get to prevent anyone else from solving it?
Forbes suggests some solutions, including abolishing software patents.
Actually writing for Forbes was Tim Berners Lee, widely known as the inventor of the WWW. He invented HTML and the HTTP protocol to transport it. But what if Douglas Englebart had enforced a patent on hypertext? TBL could never have created the web.

What if TBL had patented, and enforced patents on HTML and HTTP. Likely, those patents would have gone nowhere and no one would use those technologies. We'd be using something else instead and the development would have been held back by years.

Or, what if Pei-Yuan Wei had patented (and enforced) the idea of embedding objects in web pages? Or if Marc Andressen had patented (and enforced) the idea of using inline images? Again, the web would be dead.

Software, unlike biotech makes it easy to create and implement new ideas quickly. Days or weeks rather then years. Zero cost rather then millions of dollars.
posted by delmoi at 8:37 PM on July 28, 2011 [10 favorites]


I mean, MPEG for instance, is something novel and new.

How so?
posted by delmoi at 8:38 PM on July 28, 2011


A billion would be chump change in this fight. As mentioned up thread, a consortium of Apple, Microsoft, Nokia, Sony, RIM and few others just ganged up to outbid Google and buy a chunk of patents off the bankrupt Nortel for $4 Billion or so.

The gloves are off and phone manufacturers are suing each other left and right, if you don't have a trove of patents to counter-sue with you are toast. It is impossible to play in this industry if you don't have hundreds of millions of dollars spare.

HTC was criticized when they spent $600 million (from memory) on patents owned by 3D card company S3 (or did they just buy S3 outright, I can't remember), but they are now using those patents to counter-sue Apple or are suing everyone under the sun.
posted by markr at 8:41 PM on July 28, 2011 [1 favorite]


...counter-sue Apple who are...
posted by markr at 8:43 PM on July 28, 2011


I love how the IV blog post says that parts of the story were absurd without giving any specific examples. Their position as stated is effectively "Am not! You are, times a hundred!"
posted by No-sword at 8:44 PM on July 28, 2011 [1 favorite]


Oh wait, crap, all the rich folks are liberatarians, who believe in the sanctity of imaginary property and that the market will magically fix the problem (and that, by definition, there can't ever be a problem).

"Libertarians" are not uniform supporters of intellectual property as a concept or any particular IP regime, as a quick scan of literature by both notable big-L Libertarian writers and notable small-l libertarian writers (mostly neckbeards and other tech types) would demonstrate, but keep herping that derp.
posted by Inspector.Gadget at 8:54 PM on July 28, 2011 [3 favorites]


Microsoft could make more money from patent licensing fees from Android phone developers, than from sales of its own Windows Phone 7 software.
posted by cwhitfcd at 8:57 PM on July 28, 2011


Mondo Dentro: I got an algorithm patent, kinda by accident... I was working with some shady people who I realized might patent what I was doing, and so I went ahead and did it (through my university)

If you don't mind me asking, how much did this cost you?
From what I've read, a patent would cost me about a year's rent. Does your university just have some guys on salary that do the expensive legal work as part of their job there?

I am continually being told that I would be very successful if I were to patent even a fraction of the innovative stuff I do, but I look at the patent system and it just doesn't seem to be a system that little people like me get to use. :-(
posted by anonymisc at 8:59 PM on July 28, 2011 [3 favorites]


Yep, many universities have staff -- entire departments -- dedicated to helping students and faculty with patents.

One of the perks of paying for a degree. Or not paying for one, since I would guess most students at the level of actually patenting technology are receiving significant aid - it's a big win for the university to have excellent students, especially when the university can potentially gain income long after the student has left.
posted by lesli212 at 9:11 PM on July 28, 2011


Oh wait, crap, all the rich folks are liberatarians, who believe in the sanctity of imaginary property and that the market will magically fix the problem...

Yay! It's time for the MeFi libertard chant! Everyone now!

In reality, tech libertarian organizations like the EFF are against software patents. Meanwhile, Democrats like Obama feel "Right now we could give our entrepreneurs the chance to let their job-creating ideas move to market faster by strengthening our patent process".

The problem, as Lawrence Lessig has identified and is focusing on now, is that our political process has become corrupted and moneyed interested have too much influence on it.
posted by formless at 9:29 PM on July 28, 2011 [4 favorites]


Meanwhile, Democrats like Obama feel "Right now we could give our entrepreneurs the chance to let their job-creating ideas move to market faster by strengthening our patent process".

What a joke.
posted by delmoi at 9:35 PM on July 28, 2011 [1 favorite]


anonymisc: If you don't mind me asking, how much did this cost you?

Me? Nothing. The university IP office did it. I don't know honestly what the total cost to them is. 25K maybe? They elected to not go after international, which I believe is much more costly. They don't do everything of course. They assess each invention disclosure and decide if they want to go forward. In my case, they gave me a thumbs up, and I got a gold star, so to speak.

No way I would have done it myself. Like I said, it was sort of defensive on my part. And, really, I view patents like a Kevlar vest: it won't keep you from getting killed, but it gives you better odds. If a loan individual has a patent and has to go up against a deep-pocketed adversary, (in the vest analogy, that's someone with a big gun) the patent would have to be awfully tight to hang in under the beating it would get. So, having the university behind it made a lot of sense that way, too.
posted by mondo dentro at 9:58 PM on July 28, 2011


Here's Nathan Myrhvold blowing his own trumpet on TED. I remember watching it and thinking he seemed a bit of a phoney, a bit of a dilettante. But I didn't realise that Nathan Myrhvold was in effect king of the patent trolls.
posted by johnny novak at 10:04 PM on July 28, 2011


Oops, here's the link to Nathan Myrhvold on Ted.
posted by johnny novak at 10:05 PM on July 28, 2011


Capital tends to concentrate. That tends to screw up markets. Then economies slump, but the last ones standing are the biggest thieves who've still got some liquidity left. It's a positive feedback situation that always goes wrong unless there is some visible hand of regulation to prevent it. If the regulators are corrupted, there's no interrupting the downward spiral.

Face it, the fat boys don't care if it all goes to hell. If they are one of the last standing, they'll be able to buy up what's left for pennies on the dollar.

The software industry has long been headed in the direction of a private monopoly utility where they no longer have customers, only rate payers. Think of it as taxation without governance. Time to upgrade, suckers.
posted by warbaby at 10:20 PM on July 28, 2011 [6 favorites]


This was a great piece by TAL. And I agree that the response from Intellectual Ventures was shockingly weak. It's like they didn't even try.
posted by chinston at 10:26 PM on July 28, 2011


It really seems like civilization as we know it is collapsing under the weight of its own complexity. Well-meaning legislators, courts, regulators, and the electorate are being overwhelmed on all fronts by cynical concerted efforts to game the system, and it's largely because the issues are so mind-bendingly complicated that non-experts cannot understand them.

Intellectual Ventures, ALEC, Goldman-Sachs, GOP, Newscorp, etc. Game theoretic defection on a massive, massive scale.

This American Life has done a fantastic job with their investigative reporting here and of the mortgage crisis / sub-prime fiasco. But will it really make a difference? These issues are just too abstract for voters to get really riled up about.
posted by qxntpqbbbqxl at 10:39 PM on July 28, 2011 [10 favorites]


The picture that ran through my head of the opening scene of the show where they're describing IV's "innovation lab" was that of a Bond Villain's legitimate front; where they pretend they're running a legitimate business or lab, while in the back the evil minions are doing the stuff that makes the real money.

Honestly though, the media needs to start viewing and treating Myrvold like they do Murdoch, and stop bestowing him with status and kudos.
posted by stratastar at 10:51 PM on July 28, 2011


A lot of friends who don't enjoy This American Life were unanimously delighted to come and listen to this episode, and I've not heard a bad word against it. It's really fantastic. If I may, I suggest a different This American Life episode that unfolds and unfolds, that of The Petty Tyrant.

(If this is any indication, I enjoy This American Life a lot.)
posted by jscott at 11:22 PM on July 28, 2011 [1 favorite]


It really seems like civilization as we know it is collapsing under the weight of its own complexity.

Well, paradigm shifts do tend to look a bit like that, from the inside. What's definitely the case is that the framework of the democratic nation is being outpaced by innovation (technical, mediatic, financial, etc) and becoming increasingly incapable of regulating it. Those attempts at regulation are circumvented by increasingly powerful transnational actors, or captured by them and rendered ineffective. The endgame is the failure of the state to govern at all, a point which the USA is reaching much faster than anyone could have reasonably suspected. The Arab Spring is not dissimilar. Many others will probably follow suit.

We live in interesting times.
posted by mek at 11:47 PM on July 28, 2011 [8 favorites]


So, in business terms: if I do business in the USA I can pay a licencing fee to Intellectual Ventures of about $50,000 per annum, and then will be free from worry about software patents?
posted by alasdair at 12:32 AM on July 29, 2011


Just to make sure that we all know how awesome both This American Life and the Planet Money team are, be sure to have previously heard:

#355: The Giant Pool of Money (May, 2008)

#382: The Watchmen (June, 2009)

#390: The Return to the Giant Pool of Money (Sept. 2009)

#405: Inside Job (April, 2010)
posted by gen at 12:57 AM on July 29, 2011 [10 favorites]


Actually writing for Forbes was Tim Berners Lee, widely known as the inventor of the WWW. He invented HTML and the HTTP protocol to transport it. But what if Douglas Englebart had enforced a patent on hypertext? TBL could never have created the web.

No it wasn't.
posted by gyc at 1:08 AM on July 29, 2011


If you don't mind me asking, how much did this cost you?

Me? Nothing. The university IP office did it.


So... your name is on the patent, but any future patent licensing fees "earned" from people using/licensing the innovation would go to the university, not you, and this is a good deal because at least it means you can do your work without some asshole patenting it out from under you and making life difficult?

Or is it more like university gets the lion's share of any income, but you would still get a cut? (I wonder if my alma mater does that for ex students?)

Currently, my poor man's version of a defensive patent has been "put the details up on the web so no-one can patent it because the idea is Out There and everyone is looking into it and there is prior art". That's a small good for society, but in a game where the other guys get to use the patent-hammer while home-grown innovators don't, the lack of an even playing field undercuts my ability to put resources into developing the much more powerful stuff that I should be doing instead of just small-fry bullshit :-/

But putting the info Out There still results in watching People With Resources run with your work and prosper a million-fold over what you'll ever see. It's disheartening.

No patents at all would be great. A fair patent system would be great. But this current system is bullshit and self-destructive. More labour is being wasted on arguing over who did what first, than is being invested on actually creating the wealth that is needed to run the whole show. It's like living in Gormenghast, and that was the only book I ever read that was so depressing I couldn't get even a quarter of the way into it. /rant
posted by anonymisc at 3:17 AM on July 29, 2011 [2 favorites]


Now however, the USPTO tends to grant pretty much everything and leaves it to the courts to sort it out

Yeah, this is pretty much the real problem. It's not that software should be inherently unpatentable; it's that the patent office needs to stop allowing patents on obvious non-inventions like "one-click purchase" or "cloud computing".
posted by ook at 4:13 AM on July 29, 2011 [2 favorites]


I'm working on a software startup, and due to the nature of the product, if it ever sees any success the chance of a patent troll coming after me is, oh, something approaching 100%.

Its really a terrible feeling, knowing that if I do well I will certainly be the target of a kind of extortion where the perpetrator is acting within the law, and overpowers me by orders or magnitude. I will simply have to give in and hope for mercy, that is the only option.

This anticipation and powerlessness is a truly horrible feeling and listening to this just made me feel even more hopeless.
posted by tempythethird at 4:13 AM on July 29, 2011 [8 favorites]


Nathan Myrhvold's Intellectual Ventures

I first mis-read that as "Intellectual Vultures." Maybe not so far off.
posted by ShutterBun at 5:05 AM on July 29, 2011 [2 favorites]


Last week just before the TAL story I'd seen this update from a developer about a patent threat he'd received. He develops text adventure games. The TAL story sounded more like a wacky text game than real life to me. Our weird world just keeps getting more like the movies/games/trash conspiracy novels.
posted by sammyo at 5:33 AM on July 29, 2011 [2 favorites]


It really seems like civilization as we know it is collapsing under the weight of its own complexity.

Joseph Tainter's theory may apply here.

Whenever people talk about peak oil and resource depletion, invariably someone chimes in with pseudo-cornucopian arguments that always boil down to "people are so innovative, we'll just innovate, etc., etc.," As I was listening to this podcast yesterday, it struck me that this is the fatal flaw in the cornucopian argument. Science, like economics, is not an absolute Platonic reality. Science is a human process embedded in and shaped by human society. You cannot innovate your way out of environmental collapse if you've got a system set up - such as the emerging intellectual property framework, of which patents are just one part of - to strangle precisely the innovation that is necessary to save the broader society from eventual overshoot and collapse. Greater and greater expenses will make it harder and harder for small innovators to survive, propping up behemoths who themselves spend less and less on innovation and more and more on protecting their intellectual property.
posted by jhandey at 6:39 AM on July 29, 2011 [12 favorites]


What I took away from the program was that there are currently dozens of patents vaguely covering the same concept. If you have managed to develop a great idea, there are legions of trolls waiting to pounce. If you don't have very deep pockets, you are S.O.L.

Apple, Microsoft, et. al. have divisions of lawyers to defend their interests. To them, the system is just fine the way it is.
posted by digsrus at 7:29 AM on July 29, 2011 [2 favorites]


So... your name is on the patent, but any future patent licensing fees "earned" from people using/licensing the innovation would go to the university, not you [...] Or is it more like university gets the lion's share of any income, but you would still get a cut?


It's closer to the latter, not the former. I would get (emphasis on the hypothetical) a big fraction of the royalities (I think it's about 60%). Universities are far more generous in this way than most corporations--that I'm aware of, at least. I know (know of) guys with dozens of patents for big companies (like GE) that just got a few bonus checks! But these are older guys in older military-industrial complex industries. I have no idea what standard practice for IP sharing is in Silicon Valley, or in the software industry in general.

Keep in mind, I'm an employee of the university, not a student. I don't believe that a student has the same option, or obligation, that an employee does.
posted by mondo dentro at 7:34 AM on July 29, 2011


When I caught the show, one of the things that most surprised me was part of the lawsuit story over the photosharing company. When they were sued over the patent, the company suing them didn't even have to say what specifically they were suing over -- that this would only come out if they went to court, which at that point would cost so much in legal fees. That seems like a loophole that should be closed, and could be done pretty easily.
posted by bizzyb at 7:38 AM on July 29, 2011 [1 favorite]


Pure software patents, I agree, are kind of ridiculous. But these patent issues can be stifling on the other side of the argument as well. I mean, the lack of any really meaningful patent protection for technology based innovations is stifling for smaller companies who genuinely want to innovate and otherwise could.

I recently proposed an idea to the company I work for to develop a new line of services that includes a mobile device application component, and after meeting with the patent attorneys recently, I walked away with the sinking feeling that there's increasingly no safe way for a smaller company to even consider getting into an innovative new line of business if it involves technology in some way. We could forego trying to get a patent and just try to bring the idea to market, but then we run the risk of getting sued anyway by one of these patent troll outfits. On the other hand, for a smaller company like the one I work for (that depends on billable hours for its cash flow), it's hard to make a business case for sinking tons of money into trying to get a patent when even in the best case scenario the odds of getting the patent are 50/50 (after four or five years and at a minimum tens of thousands in attorney and patent fees).

And then, even if you get the patent, what have you got? A piece of paper somewhere that anyone can still dispute with the patent examiners office or in court forever anyway? You might as well just issue yourself a patent, post it on the web, and then lawyer up for all the good an actual patent will do these days.

I'm genuinely flummoxed. Everyone who I've approached with the idea (including the patent attorneys and the executives at my company) agree that it's a great idea that could really help our company expand and increase its prestige, and that we've got the resources to make the technological side of the idea happen. And yet, here I am, stuck in code monkey land, and even though I'm still optimistic we could make this thing happen, the prospect of dealing with all the potential legal issues is almost too much to contemplate.
posted by saulgoodman at 7:39 AM on July 29, 2011 [3 favorites]


The problem is not "software patents" (which none of these articles even try to define: Is a patent on a fuel injection method that is implemented by embedded software running in an engine's ECU a "software patent"? What about one for an image compression algorithm used in a video camera?), nor really the USPTO (while it used to be quite lax in the late '90s and early noughties, it is rather obstructive these days, as saulgoodman suggests).

Nope, the big problem is that patent litigation in the US is extremely expensive (we are talking about several million dollars in legal costs in average), time-consuming and absorbing, and lacks adequate safeguards against frivolous lawsuits. This both makes it difficult for genuine innovators to enforce their valid patents, and easy for trolls to run legal protection rackets by sending mass mailings to companies who may, or may not, be infringing their mostly bogus patents, but will almost certainly rather pay 50,000 dollars in license fees to a troll than 2 million in legal fees to their own lawyers.

In my opinion, the problem, where it exists, would be better tackled downstream, by litigation reform, than upstream, in granting practice. Only a very small percentage of all granted patents is actually involved in actual litigation.
posted by Skeptic at 8:37 AM on July 29, 2011 [1 favorite]


Apple, Microsoft, et. al. have divisions of lawyers to defend their interests. To them, the system is just fine the way it is.

That is not true. Microsoft, for example, tried very hard in two different cases (most recently the Microsoft v. i4i Supreme Court case) to make it easier to invalidate patents by lowering the standard of proof, at least when offering evidence not considered by the Patent Office. Microsoft is a defendant in far, far more patent suits than it is a plaintiff. I suspect the same is true of Apple, even if it is more litigious overall.

the USPTO was originally of a similar mind, and was effectively forced to accept software patents by the courts. Now however, the USPTO tends to grant pretty much everything and leaves it to the courts to sort it out.

These facts are related. The USPTO fought tooth and nail against software patents and as a result it was woefully underequipped to deal with examining software patent applications. And it still is. Part of the problem is that being a patent examiner means working in Alexandria, Virginia. Not exactly a hotbed of software development. The Patent Office needs to open satellite offices in places like Silicon Valley and Boston.

Another problem is that the Patent Office makes it difficult for people with CS backgrounds to become patent agents and patent attorneys, compared to people with hard science backgrounds. But once you're a patent attorney you can draft any kind of patent application. So you get a lot of crappy applications written by people without CS backgrounds that are then examined by examiners who may have very little clue what the application is about.

Yet another problem with the Office is that the turnover rate is horrific. Currently, about 80% of examiners have 3 years or less experience and only 7% have more than 10. Given that a patent application typically takes well over 3 years to work its way through the office and you have a situation where it's extremely likely that a patent application won't be handled by the same examiner all the way through. That is not a recipe for solid examination.

The Patent Office budget should be dramatically increased, and multiple satellite offices should be established in tech hubs. It's the only way to improve examination on the front-end.
posted by jedicus at 8:50 AM on July 29, 2011 [7 favorites]


Actually writing for Forbes was Tim Berners Lee, widely known as the inventor of the WWW. He invented HTML and the HTTP protocol to transport it. But what if Douglas Englebart had enforced a patent on hypertext? TBL could never have created the web.
No it wasn't.
*blink*. Um, whoops. Well, the author has a similar name to TBL and anyway what I said about TBL and patents is still accurate, if somewhat superfluous.
The Patent Office budget should be dramatically increased, and multiple satellite offices should be established in tech hubs. It's the only way to improve examination on the front-end.
It would be simpler and cheaper to get rid of software patents.
posted by delmoi at 11:18 AM on July 29, 2011


delmoi--but then what small company would ever take the risk of trying to bring something genuinely new to market, knowing full well how what will probably happen when they do is that, immediately, bigger companies will simply push them out of the market, having unfair competitive advantages due to their size?

Patent law should be designed to protect the small players who could genuinely be bringing innovative goods and services to market, but the instinctively nihilistic attitude toward the patent system so many seem to have nowadays reminds me a lot of the Tea Partiers' hostility to the IRS.
posted by saulgoodman at 11:25 AM on July 29, 2011


Einstein was smarter than you, and he believed in the patent system.
posted by saulgoodman at 11:25 AM on July 29, 2011


Einstein was smarter than you, and he believed in the patent system.

Einstein was not a software developer.

I liked TAL more when they spent more time on individual stories of eccentric Americans. Their big "The State We're In" stories are interesting and well researched but I don't find them as compelling listening. Their "teamed up with Planet Money" stories to me always end up feeling oddly told split between two reporters simultaneously and, extra bonus, they get stretched out through All Things Considered during the week reaping weak dividends to listeners of both shows. Listening to the broadcasts twice over three days does not improve the listenability.
posted by Ogre Lawless at 11:59 AM on July 29, 2011


what small company would ever take the risk of trying to bring something genuinely new to market, knowing full well how what will probably happen when they do is that, immediately, bigger companies will simply push them out of the market, having unfair competitive advantages due to their size?
A patent is no use in these cases, as it takes millions of dollars to prosecute them. If you got $2-5 million in capital to develop your device, you have no money left over to defend yourself in the courts.

I've seen small companies entirely eaten up by the cost to defend their true, not frivolous inventions, with their only recourse being to sell to a larger company.

Today, patents are a tool of the big and wealth, not the small innovator. Einstein would surely agree if he were here today; the situation is not remotely similar.
posted by Llama-Lime at 12:02 PM on July 29, 2011 [2 favorites]


delmoi--but then what small company would ever take the risk of trying to bring something genuinely new to market, knowing full well how what will probably happen when they do is that, immediately, bigger companies will simply push them out of the market, having unfair competitive advantages due to their size?
Has that ever been a problem in the past? Name one popular software product that only succeed because they sued competitors out of existence using software patents. We got to where we are today without small companies having their innovations protected due to software patents

This is what's so bizarre about the software patents thing: that never actually happens in the software world. The only people who bring software patent suits are patent trolls who sell no products on their own. After all, if they did sell products they would be open to reverse patent suits, which is exaclty what happens when big software companies sue each other, they get counter-sued because the sue-er is invariably violating some other patent.

And that's another point. You cannot sell software that uses only a single "idea". You invariably have to use multiple ideas that have existed for a while in your entire product. So the idea of a small company doing 'truly innovative' things and being protected by patents is absurd. More likely you'll see the reverse. Truly innovative companies shut down (or never started) due to parasitic patent trolls like Intellectual Ventures.

Again, give me one example of a software product that exists today because someone enforced software patents and prevented competition. Just one!
posted by delmoi at 12:16 PM on July 29, 2011 [1 favorite]


Oh and the Einstein thing. Where does it say he believed in the patent system? He worked as a patent examiner, because he needed to eat. That doesn't mean he 'believed' in them.

And anyway, patents had to be a lot more concrete in that era.
posted by delmoi at 12:17 PM on July 29, 2011 [1 favorite]


(thought the Einstein thing was a pretty obvious appeal-to-authority joke, but then these topics always get so serious.)
posted by saulgoodman at 12:33 PM on July 29, 2011


Has that ever been a problem in the past? Name one popular software product that only succeed because they sued competitors out of existence using software patents.

No, I'm not saying they will necessarily sue them out of business w/out patent protection, but that the small company won't be willing to invest the capital needed to get into the new line of business at all, because even if they manage to bring a new product or service to market first, they'll stand no chance of competing on a resource-basis with larger, predatory competitors who can easily copy the innovation and then outmaneuver the smaller player on the basis of their superior financial resources and other competitive advantages of being a larger player (like the de facto collusive market practices that wonks euphemistically call "vertical integration").

And I'm actually talking about both new software technologies and new software technology supported service models here.
posted by saulgoodman at 12:39 PM on July 29, 2011


The only people who bring software patent suits are patent trolls who sell no products on their own.

Oh now that's just completely false. Apple and Microsoft, to cite two examples, have sued over the allegedly infringing use of technologies that they implement in their own products. TiVo did the same thing in its long-running case against Echostar. i4i, a company that sued Microsoft, makes various software products related to XML. z4, another company that sued Microsoft, makes DRM software. Oracle has sued Google over software patents. I could go on and on. Of those cases, the Microsoft cases did not end in a cross-license, so it is also not the case that patent infringement suits between practicing entities invariably ends in a cross-licensing deal.

Again, give me one example of a software product that exists today because someone enforced software patents and prevented competition

First, it's impossible to say that a given technology would never have been developed were it not for software patents. Software patents are neither necessary nor sufficient for the development of new technology. However, one can argue that they encourage investment in new technology.

Second, most patent infringement issues never go to litigation. Either people realize something is patented and steer clear (e.g. there is very little commercial competition to the MPEG video standards), or they get a cease & desist letter and stop infringing / take a license, or the case settles before it becomes well-known in the media (and settlement terms are often secret). So this is a very difficult question to answer, empirically.

Anyway, one possible answer: Amazon's 1-Click technology. Amazon has successfully enforced its patent against competitors but also licenses the technology as a product (e.g. Apple licenses it).

If you're the first person who comes across a problem, and solves it in the obvious way, why should you get to prevent anyone else from solving it?

You shouldn't. That's the purpose of the nonobviousness requirement.
posted by jedicus at 1:36 PM on July 29, 2011 [1 favorite]


Amazon's 1-Click technology

Isn't this the perfect example of patenting something that isn't even a technology? It's just a really basic and not even complex idea, which can be solved in any number of ways, the technical side being almost irrelevant to the appearance to the end-user. eg. Apple's one-click solution would be quite different to Amazon's, and one-click on PC and one-click on handhelds are two totally different kettle of fish, yet lumped under the same patent. Basically Amazon is earning licence fees on something that has existed in vending machines for decades.
posted by mek at 2:00 PM on July 29, 2011 [3 favorites]


No, I'm not saying they will necessarily sue them out of business w/out patent protection, but that the small company won't be willing to invest the capital needed to get into the new line of business at all, because even if they manage to bring a new product or service to market first, they'll stand no chance of competing on a resource-basis with larger, predatory competitors who can easily copy the innovation and then outmaneuver the smaller player on the basis of their superior financial resources and other competitive advantages of being a larger player
I think it's well established that small companies are much more concerned about being crushed by patent trolls. But again, do you have a specific example of anything like that happening, ever? Because it just doesn't sound like how software companies actually work.

The silicon valley/venture capital model is based on 9/10 companies failing. If one company fails you just start another one using the contacts you made in the first round. Failure isn't that big of a deal and you're expected to have lots of 'good ideas'. It's not the idea that counts, it's the execution and at the bottom of it there are no "original ideas" Every idea is going to be similar to something someone else thought up the 'winner' is just the group that does it best.

Anyway, again, lets hear of a specific, concrete example of a small company being willing to invest in an idea simply because they knew they could get a patent and prevent competitors from entering the same market space.
Apple and Microsoft, to cite two examples, have sued over the allegedly infringing use of technologies that they implement in their own products. TiVo did the same thing in its long-running case against Echostar. i4i, a company that sued Microsoft, makes various software products related to XML. z4, another company that sued Microsoft, makes DRM software. Oracle has sued Google over software patents.
All of those examples (except for i4i and z4, I guess) are examples of huge companies suing eachother. Not small independent companies being able to create a successful product on their own due to patent protection. And I'd never heard of i4i or z4 other then in the context of them suing Microsoft for patent infringement.

And of course these are not at all 'the norm' for software development. You guys keep pushing this bizarre idea of independent software development where 'innovators' come up with ideas, patent them, and then bring their product to market with no competition. That's just not how it actually works at all. Instead you bring to market a solid product that works well for the given task, and people use it.
You shouldn't. That's the purpose of the nonobviousness requirement.
Which the patent office doesn't seem to bother with in the case of software. How can you even tell if something is 'obvious' or not if you haven't come across the same problem? Everything seems obvious in retrospect.
posted by delmoi at 2:23 PM on July 29, 2011


Patent Lawsuit May Boot Spotify From USA
posted by Mister Fabulous at 2:30 PM on July 29, 2011 [1 favorite]


Haven't listened to the show yet, but the related written article has one really remarkable point.. No one from the tech industry would talk to the journalists on the record. No one except Chris Sacca
"I tried to put you in touch with other people in this community to talk to you about this and they almost uniformly said they couldn't talk to you," Sacca told us. "They were afraid to." IV has the power to "literally obliterate startups," Sacca says.
posted by Nelson at 2:45 PM on July 29, 2011


In other frivolous patent news, an appeals court decided 2-1 that gene patents are valid. Gene patents like the one Myriad held are just as much an affront as obvious software patents. Gene patents previously on Metafilter.
posted by Llama-Lime at 3:22 PM on July 29, 2011


The silicon valley/venture capital model is based on 9/10 companies failing.

One of the useful functions of patents is that they provide a way for investors to recoup their investment in a company even if the company fails before it has much in the way of tangible assets. This is one of the ways that many so-called patent trolls acquire their patents (IV, notably, does not do this much): they buy them at a bankruptcy sale and then try to turn the IP into cash via licensing and (more rarely) litigation.

All of those examples (except for i4i and z4, I guess) are examples of huge companies suing eachother.

So by "all" you mean 4 out of 6. Another example is Mangosoft, which sued Oracle. You claimed these examples didn't exist at all. The facts are otherwise.

And I'd never heard of i4i or z4 other then in the context of them suing Microsoft for patent infringement.

I'm not sure how that matters. How many small software companies can you name, as a percentage of all small software companies?

You guys keep pushing this bizarre idea of independent software development where 'innovators' come up with ideas, patent them, and then bring their product to market with no competition.

I don't know who "you guys" are, but I push no such idea. I'd be the first to tell you that patents are not a license to print money, nor do they eliminate competition. A sound business model and good execution are still required.

How can you even tell if something is 'obvious' or not if you haven't come across the same problem?

You break the invention down into elements, find prior art containing each of these elements, and then find some reason to combine them, even if it's just common sense or common creativity. That's more or less the modern test for obviousness.
posted by jedicus at 3:43 PM on July 29, 2011


What I don't get about a lot of these patents is that they don't seem to be patenting an invention, but rather what an invention does.

You shouldn't be able to patent the concept of updating software over a network. You should be able to patent the specific way you do this, but not the idea.

It would be like you invent the light bulb, but instead of patenting 'electricity passed through a tungsten filament suspended in a vacuum' you patent 'device for producing light' and then go after all the candle makers.
posted by Dreadnought at 3:48 PM on July 29, 2011


You shouldn't be able to patent the concept of updating software over a network. You should be able to patent the specific way you do this, but not the idea.


But where you draw the line between 'the way you do it' and 'what it does'. With software, it's pretty nebulous.
One of the useful functions of patents is that they provide a way for investors to recoup their investment in a company even if the company fails before it has much in the way of tangible assets.
Right, so failures have the opportunity to act like parasites and prevent anyone from using ideas similar to theirs. Exactly the problem that will be solved by eliminating software patents.
posted by delmoi at 3:57 PM on July 29, 2011


What I don't get about a lot of these patents is that they don't seem to be patenting an invention, but rather what an invention does.

You shouldn't be able to patent the concept of updating software over a network. You should be able to patent the specific way you do this, but not the idea.


Have you ever actually read a patent? This is indeed how this is supposed to work. However, journalists reporting on patents are hardly going to recite the full, quite specific content of the patent claims, and generally just give an overly broad description of the subject-matter of the patent.

For sure, patents are granted with badly drafted, "free beer" claims, but they are the exception rather than the rule, and they'll hardly withstand litigation. Most patents mentioned in the press are a lot more specific than the very abbreviated accounts that the press gives of them.
posted by Skeptic at 5:01 PM on July 29, 2011 [1 favorite]


Have you ever actually read a patent? This is indeed how this is supposed to work.

How do you tell the difference between "what it does" and "how it does it" exactly? Like for example if JPEG uses discrete cosine transforms to compress images in 8x8 blocks, can I write an image compression library that uses *9x9* blocks instead? Or what about if I use the 3-norm rather then cosine similarity? Where do you draw the line?
posted by delmoi at 5:07 PM on July 29, 2011


I thought this piece was really biased. Among other things:

(A) The use of the derogatory term "patent troll" to refer to the subject of the piece; and

(B) Arguing that IV's lawyer's failure to disclose information about his client to a radio show was a sign that IV is ominously hiding something.

As an IP lawyer, I thought TAL took 40 minutes to get to the big picture idea: patents for things like software and business "methods" are coming much too close (and often crossing the line) into ideas. Patents are supposed to be for methods and devices.
posted by pollex at 5:40 PM on July 29, 2011


Actually the spotify link is a good example of patent trollery preventing companies from operating in the U.S. This company has no problem operating in Europe, but now that they're trying to launch in the U.S they're getting sued by a parasitic company over what sounds like a patent basic music streaming services.

The big part of the problem is that it costs orders of magnitude more to fight a patent-troll lawsuit then it does to develop the software itself. On the other hand the benefit from a patent for a 'legitimate' developer is pretty low. Only if you fail and become a parasitic patent troll is the patent worth any money. If you produce a good product and make money you should be fine.
posted by delmoi at 5:44 PM on July 29, 2011


If you're the first person who comes across a problem, and solves it in the obvious way, why should you get to prevent anyone else from solving it?

You shouldn't. That's the purpose of the nonobviousness requirement.


But the fact that no-one has solved it before is accepted as the proof that it is non-obvious.

The non-obvious requirement is precisely what ensures that stupidly obvious and straightforward solutions are "novel" and patentable.

USPTO is broken on so many levels in so many ways that suggestions that patents can be fixed, that something useful can be salvaged, seem crazy optimistic. This is why so many of the genuine innovators would be happy to see the whole edifice burned to the ground.
posted by anonymisc at 6:16 PM on July 29, 2011


The use of the derogatory term "patent troll" to refer to the subject of the piece

There isn't a non-derogatory term for inherently disgusting behaviour. You could instead say, quite factually and in most cases uncontroversially "company that does not produce goods or services, but maintains an income by requiring license fees or litigation from productive companies that are not in a position to come out better off via the alternatively-threatened litigation, despite having the stronger case", and it's still derogatory, simply because when described accurately, this business model is reprehensible.
posted by anonymisc at 6:30 PM on July 29, 2011 [3 favorites]


But the fact that no-one has solved it before is accepted as the proof that it is non-obvious.

By this I mean that you can demonstrate that there is no prior art. There is no prior art because no-one has ever done it before. No-one has done it before (even though thousands have mused about doing it before, because it's so obvious) because until X and Y happened, it wasn't remotely technologically feasible. In the software world, this happens every day.

posted by anonymisc at 6:38 PM on July 29, 2011


the use of the derogatory term "patent troll" to refer to the subject of the piece

If everyone they interview uses the phrase, are they then not supposed to use it?

They were also using it as a framing device to lead through from Peter Detkin coining the phrase at Intel to his later co-founding of Intellectual Ventures. (I thought that was a lovely reveal, myself.)
posted by We had a deal, Kyle at 6:46 PM on July 29, 2011 [1 favorite]


I was thinking people should come up with an 'open source pool' of patents. Free to use in any open-source app, and free to use in closed source apps provided that the companies are not paying off any other patent trolls. Essentially, make it impossible for companies to sell software and pay patent trolls at the same time, thus destroying the industry.
posted by delmoi at 7:47 PM on July 29, 2011 [1 favorite]


The non-obvious requirement is precisely what ensures that stupidly obvious and straightforward solutions are "novel" and patentable.

Wow. This statement falls in the "not even wrong" category. There are two different patentability requirements: novelty and non-obviousness (in Europe we call the latter "inventive step"). The idea is that, for something to be patentable, novelty is not enough. It must also not be "stupidly obvious and straightforward", as you put it. Judging this, however, is far less than straightforward. With the benefit of hindsight, many solutions may seem obvious which weren't so obvious before someone came up with them. Because of this, the courts and patent offices have developed a number of legal tests to try to determine non-obviousness as objectively as possible.

Until 2007, the USPTO generally applied the so-called TSM test ("teaching, suggestion or motivation") according to which, combining two elements from different disclosures would be obvious if there would be, for the person having ordinary skill in the art, a teaching, suggestion or motivation to combine those two different disclosures. However, in KSR vs. Teleflex, the Supreme Court gutted the TSM test, basically saying that non-obviousness couldn't be judged using such rigid tests and that "common sense" ought to be applied. Unfortunately, common sense is rather uncommon, and the immediate result is that the percentage of rejections at the USPTO has soared, patent examiners making often somewhat bizarre combinations to allege that something is obvious (I saw one case in which the examiner combined an e-commerce website, a multi-CD box, and a perfume flask).
posted by Skeptic at 1:32 AM on July 30, 2011


I was thinking people should come up with an 'open source pool' of patents. Free to use in any open-source app, and free to use in closed source apps provided that the companies are not paying off any other patent trolls.

There's already something quite similar called the Open Invention Network. The "not paying from trolls" condition which you suggest is however hopelessly impractical. Because:

a) How do you define what is a patent troll? An entity which doesn't itself practice the invention? That would cover most universities and research centers. An non-practicing entity without its own R&D? That would cover most legitimate patent pools, but not IV.

b) What if the "troll" has a good patent and wins in court?
posted by Skeptic at 1:57 AM on July 30, 2011


Actually the spotify link is a good example of patent trollery preventing companies from operating in the U.S. This company has no problem operating in Europe, but now that they're trying to launch in the U.S they're getting sued by a parasitic company over what sounds like a patent basic music streaming services.

Which shows how the problem lies less in the patents that are granted that in the cost of litigation in the US. Contrary to what is often alleged, software patents are granted in Europe and upheld in court (certainly in this field: Spotify itself has filed European patent applications).
posted by Skeptic at 2:06 AM on July 30, 2011


How do you tell the difference between "what it does" and "how it does it" exactly? Like for example if JPEG uses discrete cosine transforms to compress images in 8x8 blocks, can I write an image compression library that uses *9x9* blocks instead? Or what about if I use the 3-norm rather then cosine similarity? Where do you draw the line?

That depends on how the patent claims are drafted. A good patent drafter will normally not specify the size of the blocks in the broadest claims (although it will be specified in the description of the "best embodiment"). However, it will be more difficult to avoid specifying that a cosine transform is used, for two reasons:

Firstly, the broadest the claim, the more likely it is to be anticipated by prior art.

Secondly, there is, besides the novelty and inventive step requirements, also an "enablement" ("sufficient disclosure" in European patent-speak) requirement in patent law. Basically, in the patent specification, the invention has to be sufficiently disclosed to enable the person skilled in the art to reproduce the invention over the whole width of the claim. While disclosing the use of 8x8 blocks would appear to enable the skilled person to reproduce the invention with any block size without a particular effort, just disclosing the use of a cosine transform does not appear to enable the skilled person to apply any alternative method instead.
posted by Skeptic at 2:23 AM on July 30, 2011


The software industry was doing just fine before patents applied.

Now, a multi-billion dollar industry has appeared playing a game of legal extortion and you can get sued for daring to use Apple's own system to do in-app purchases.

Broken. Stop now.
posted by Devonian at 3:40 AM on July 30, 2011


The software industry was doing just fine before patents applied.

And this was when, exactly?
posted by Skeptic at 3:58 AM on July 30, 2011


-There isn't a non-derogatory term for inherently disgusting behaviour.

-If everyone they interview uses the phrase, are they then not supposed to use it?

If TAL wants to be unbiased, it shouldn't use a derogatory term to refer to a subject of its reporting. Lots of TAL pieces are (obviously) biased, and that's fine. But this episode aspires to investigative reporting, and it was completely one-sided. I agree that many (most?) software patents do not promote the progress of arts and science, but I was amazed at the blatant one-sidedness of this piece.
posted by pollex at 5:21 AM on July 30, 2011


With the benefit of hindsight, many solutions may seem obvious which weren't so obvious before someone came up with them. Because of this, the courts and patent offices have developed a number of legal tests to try to determine non-obviousness as objectively as possible.

As objectively as possible meaning "anything flies, let the courts sort it out".

You can point to the tests that are done, and that some of the crap gets rejected, but this doesn't address the elephant in the room, which is that mindblowingly obvious stuff is patented daily. Stuff that was so obvious that people would be talking about it ten years previously "when this happens, we'll finally be able to do that", and then when it happens ten years later, the "that" gets patented as if it was something other than the logical next step.

If the USPTO is trying so hard to apply non-obviousness filters, and failing so spectacularly and clearly, then maybe the system is not merely broken, but fundamentally unworkable at the deepest levels.
posted by anonymisc at 5:22 AM on July 30, 2011


a) How do you define what is a patent troll? An entity which doesn't itself practice the invention? That would cover most universities and research centers. An non-practicing entity without its own R&D? That would cover most legitimate patent pools, but not IV.
It would be more then just patent trolls, any company that makes it's money by threatening to sue people for patent infringement, and settling would be targeted.
b) What if the "troll" has a good patent and wins in court?
The pool would, itself be a patent troll. It wouldn't produce any product, so it couldn't be shut down. If the 'troll' wins in court, then the product that infringes on the trolls patent, and our patent, then the product can't be sold. The end. The point is to destroy the software patent industry, and people will have to choose or move to Europe.
And this was when, exactly?
First of all, that patent clearly shows an electronic circuit, that patent is very clearly for a hardware device, and references hardware features like registers. You do understand the difference between hardware and software, right?

Second of all, software patents actually are new. And enforcement is certainly new.
If TAL wants to be unbiased, it shouldn't use a derogatory term to refer to a subject of its reporting.
Why would they want to be unbiased? Non-bias and dishonesty can't be decoupled. If one side is right, and the other side is wrong, then the only way you can be 'non-biased' is to lie. "Patent Troll" is a common term
posted by delmoi at 5:59 AM on July 30, 2011


Pollex,

I don't know about you but my listening of the piece pegged it as pretty standard libertarian-side TAL fare. And I'm ok with that. A fair number of the IV quotes that were supplied which were based on questions about how IV handles defending their patents, or who they sell their patents to, were heavy on the weasel words. And I would think that even if the term "Patent Troll" wasn't used to describe what IV and it's shell companies do.

There was a strong whiff of "We don't have to talk to YOU, little person" in every one of the plaintiff side responses to the questions set out by the TAL investigators. And since Ira, TAL and Planet Money have never made a claime (either way I might add) of being unbiased fact reporters or Muckrakers/gonzo journalists I think their use of "biased language" is just calling a spade a spade.

Did you catch the part where they revealed that shell companies with no employees beyond a murder of lawyers are distinctly linked to IV? And that doesn't make you think 'gee, that sure seems shady to me'.

Personally I'm glad they didn't use soft language in this piece. Because it unecessairly confuses the mind.


As the Philosopher Carlin once said "shell shock, simple direct language".
posted by Severian at 6:44 AM on July 30, 2011


Intellectual Ventures has had plenty of one sided softball articles written about their amazing innovative business and the dinner parties where they spool off 30 patentable inventions over wine and nitrofried cheese foam. Why, right here on Metafilter. The only issue of balance is it took so long for a journalist to go dig into their extortionist business and put the lie to the claim they are somehow funding innovation. It's a shame it only aired on NPR where only smug lefties will hear it.
posted by Nelson at 9:42 AM on July 30, 2011


(A) The use of the derogatory term "patent troll" to refer to the subject of the piece; and

Yeah Pollex... this is the sort of "smug logical punch in the teeth" statement that wins lawyers all that love. I hope that if one day an entire industry appears that makes its money by acting as a parasite on your profession/love/craft, you maintain your lawyerly equanimity.
posted by tempythethird at 9:47 AM on July 30, 2011


You can point to the tests that are done, and that some of the crap gets rejected, but this doesn't address the elephant in the room, which is that mindblowingly obvious stuff is patented daily.

Can you point me to that elephant, please? Have you ever read one single such patent?

It would be more then just patent trolls, any company that makes it's money by threatening to sue people for patent infringement, and settling would be targeted.

So, that would include most R&D companies which prefer to offer licenses, rather than ruin both sides in litigation. Also all standard-related patent pools. Goodbye Mpeg, jpeg, DVD, Blu-Ray, HDTV, GSM, UMTS, WiFi, Bluetooth...

If the 'troll' wins in court, then the product that infringes on the trolls patent, and our patent, then the product can't be sold.

Sorry, I can't quite parse this sentence, and I believe it's simply because you just still don't understand what patents are. Patents are negative rights. A patent is a right to forbid others from exploiting what's covered by the patent claims. Whether these others hold a license for a different patent which also covers at least part of their product is completely irrelevant, if they are infringing on your patent. Also, the consequences of being found to infringe a patent aren't just that you can't sell the product anymore. You also then have to pay damages to the patentee for the goods and services you already sold in infringement of the patent. Triple damages in the US if you are a willful infringer.

The point is to destroy the software patent industry

Your plan would destroy the software industry, period.
posted by Skeptic at 10:27 AM on July 30, 2011


So, what would it take for the Supreme Court to abolish software patents?

Can people present it as a proposition for the next California elections?
posted by The ____ of Justice at 12:47 PM on July 30, 2011


an appeals court decided 2-1 that gene patents are valid. Gene patents like the one Myriad held are just as much an affront as obvious software patents.

More on the Myriad case: Appeals Court Says Genes Are Patentable, Because They're 'Separate' From Your DNA
posted by homunculus at 1:24 PM on July 30, 2011


Your plan would destroy the software industry, period.

Exactly! But only in the U.S, and only until software patents are abolished, which is the goal.

The fact that even you think the software industry can be 'destroyed' through the use of software patents proves that even you think software patents can be used in a harmful way. The argument being put forward is that the way software patents are currently being used is already damaging the industry.
posted by delmoi at 2:49 PM on July 30, 2011 [1 favorite]


And actually I'm being a little flippant, for most companies it wouldn't be a problem. They could sell all the software they want and continue to ignore software patents unless they get sued by a patent troll.
Sorry, I can't quite parse this sentence, and I believe it's simply because you just still don't understand what patents are. Patents are negative rights. A patent is a right to forbid others from exploiting what's covered by the patent claims. Whether these others hold a license for a different patent which also covers at least part of their product is completely irrelevant, if they are infringing on your patent.
What are you talking about? I understand how patents work. I don't think you understood what I was explaining, even though it's quite simple.

It works like this. I hold patents p1, p2, p3 and so on. And companies A B and C are infringing, or possibly infringing. At this point we do nothing at all. A, B, and C are free to use the ideas so long as they never pay off any other patent trolls.

So patent a patent troll company, call them Intellectual Vultures comes along. They threaten to sue A and demand a settlement over patent qx. If A settles, then we come in and say that A is also infringing on our patents and has to stop doing so -- which means they have to stop selling the software. And if they do that, they won't have any money to pay Intellectual Vultures. IV gets drained of money, and goes out of business. The end.

(In some cases we would have to take settlements in order continue to fund operations, I suppose, also if I were actually running this, Open Source software would be exempt)

Now according to you this will destroy the software industry. But that's not true since it's still not the case that most software companies pay patent license fees. But it will destroy the part of the industry that does. And that's the whole point! Use the patent system against itself the same way GPL-style free software licensing uses copyright against itself. Rather then destroy the software industry, Open Source licensing was responsible for creating a huge ecosystem of free software.

But what if patent licensing becomes endemic to the software industry? Then this plan would, in fact destroy the industry. In which case the U.S would have to either give up on having a software industry, or give up on having software patents -- or do something like require reasonable licensing fees.

But here's the thing. If you really think this plan would work, then you obviously do believe that software patents have the capacity to destroy or damage the software industry. The argument being put forward is that software patents are already damaging the industry. If patents weren't harmful, this wouldn't be possible.
posted by delmoi at 3:09 PM on July 30, 2011


from the front page of reddit
posted by delmoi at 4:10 PM on July 30, 2011


delmoi A few problems with your scheme. First, you'd need to hold patents. Relevant patents. Contrary to what you assume, this is easier said than done. Much easier.
Second, patent litigation is expensive, even more so in the software field where discovery may involve poring over many, many lines of code. It is expensive for both parties, not just the defendant. I wonder were you'd get the funding for your scheme: even if you won every case (very unlikely), you probably wouldn't break even.
Finally, "only in the US"? Software patents are granted and upheld elsewhere, including in Europe. I should know, I have drafted some myself.
posted by Skeptic at 4:31 PM on July 30, 2011


Skeptic: links to EU pure software patents being upheld? Given then fact you linked to a hardware patent earlier and called it software, I'm a little skeptical.
posted by delmoi at 4:45 PM on July 30, 2011


delmoi Here you are.

As for my previous link, first rule of patent interpretation: RTFC. Read the fucking claims. Although the claims of that patent are nominally directed to an "apparatus", they are drafted in quite functional terms. You want a more unequivocally "software" patent? What about this one from 1975?
posted by Skeptic at 5:14 PM on July 30, 2011


The german patent requires that 'technical' things outside of the computer be effected for the patent to be valid.

The 1975 patent does seem to be an actual software patent, but that's kind of beside the point. The number of pure software patents has exploded recently and so has software patent enforcement. It doesn't matter if the patents were sitting in a filing cabinet somewhere -- and the problem
Second, patent litigation is expensive, even more so in the software field where discovery may involve poring over many, many lines of code. It is expensive for both parties, not just the defendant.
Exactly. Patent litigation is far more expensive then actually writing the software in the first place, thus, software patents are economically distortion and do damage to the software industry by increasing the costs of 'innovating' and actually bringing a product to market.
posted by delmoi at 6:08 PM on July 30, 2011


Sorry and the problem is that that software patent claims against small companies is pretty new, and only recently has it been doing a lot of damage.
posted by delmoi at 6:09 PM on July 30, 2011


Software, unlike biotech makes it easy to create and implement new ideas quickly. Days or weeks rather then years. Zero cost rather then millions of dollars.

delmoi: Let me repeat myself, since I'm obviously not being clear enough. I'm more or less in the situation that I was describing to you. I'm trying to persuade the company I work for to develop and bring to market a new line of business services supported by specialized software. I'm not sure if it's what you would call a "pure" software patent.

It's a suite of services, business processes, and a software app. Together, I think these things constitute an innovation (in the sense that they offer something valuable that isn't currently on the market and that isn't obvious). Obviously, I can't disclose too much about it, since it's still ongoing, and out of spite, you would probably just take the idea from me, turn it into an open source project without any of the related services, and then sue me later if I actually managed to get my idea underway.

But smaller companies (particularly, billable hours based ones) don't necessarily have enough financial wiggle room to sink a bunch of money into a major new venture if there's little chance of getting anything concrete back out of the effort, even if it would lead to genuinely important innovations, and eventually, more revenue. It doesn't make you a parasite to take a risk on a new idea and fail, does it? I could see that, in a lot of cases, smaller companies trying to create a new line of business in this way would want at least some modicum of legal assurance that their new business model couldn't just be copied by some established software/services behemoth just to drive them out of the market. Bigger companies can afford to lose money for years developing a new product and/or related services; not necessarily so smaller companies.

Also, your claims above about software are as unrealistic as some of my clients' expectations. I know you're a developer, but I also get the sense you might work more in the business development/marketing/management side of things, because a lot of developers I know are pretty tired of the exaggerated claims so often made about the speed, ease of implementation and reliability of software systems. Sure, the tech has made huge strides, but for really customized software systems, you very, very seldom ever see development cycles measured in terms of "days or weeks." I find it hard to believe you do either, assuming you work in custom software development, and not just web design.
posted by saulgoodman at 6:51 PM on July 30, 2011


Ah--and looking at your profile, I see you are a programmer. And probably a good one. Nevertheless, I seldom see major new system implementations, involving complex business rules and lots of special requirements, that don't require at least a few months of development time, for a proper implementation. Isn't your experience the same? "Days or weeks" seems a little unrealistic for anything but the simplest, most vanilla custom software development efforts (not that legislators ever seem to grasp that).
posted by saulgoodman at 7:53 PM on July 30, 2011


What kind of APIs do you develop in OAuth? Or is that what you--oh, never mind. That's totally off topic.
posted by saulgoodman at 8:24 PM on July 30, 2011


I'm trying to persuade the company I work for to develop and bring to market a new line of business services supported by specialized software. I'm not sure if it's what you would call a "pure" software patent.

It's a suite of services, business processes, and a software app.
Well, that's actually kind of worse if it's a 'business process' patent. If it was a hardware device, with a software component that would be an edge case. A pure business process without even software to back it up is the worst kind of patent. A mix of software and business process stuff is pretty bad. Again, that's just my opinion.
It doesn't make you a parasite to take a risk on a new idea and fail, does it?
It does if you then try to prevent anyone else from trying to do the same thing, even if they think it up on their own, IMO. Or prevent them from trying without giving you a cut.

Think about it his way. Have you actually searched for any prior art? How would you feel if, a year or so later it turns out that someone patented something that's vaguely analogous to whatever it is you thought up, and demands you license their patents. You say that your idea is 'innovative' but everything is at least a little bit like something else that's already been done. This is the biggest problem with software patents, is that in software everything is analogous to everything else in at least some ways (or even isomorphic).
I seldom see major new system implementations, involving complex business rules and lots of special requirements, that don't require at least a few months of development time, for a proper implementation.
Well, when it comes to software a new idea in software you would have a core idea, which should be pretty simple, and then a lot of extra - but straight forward - stuff around that to make it into a product. It should be pretty simple to code a prototype of the 'patentable idea' The rest of the stuff should be pretty straight forward: Database access code, web wrappers, UI code and so on.

Suppose I wanted to develop a new video codec using a new idea. I would have to write a bunch of code to import raw video, probably working with an existing library. I would have to write plug ins for various media libraries and browsers if I wanted people to be able to watch them. Then, in order get people to actually create files of this type I would need to write plug-ins for authoring software, etc.

So it only takes a little work to implement an 'idea', but it takes a lot of work to create a 'product'. But it's only the 'idea' that gets the patent, not the rest of the stuff.
What kind of APIs do you develop in OAuth? Or is that what you--oh, never mind. That's totally off topic.
I was working on a facebook app, which uses OAuth to let users authorize applications.
posted by delmoi at 9:06 PM on July 30, 2011


My position is not that I would like to be able to patent, by default. I'd prefer to just try to do the thing. But the legal landscape is so sketchy with patent trolls in the equation, too, that it makes it seems like a riskier business than it needs to be. That's unfortunate for everyone, too.

It does if you then try to prevent anyone else from trying to do the same thing, even if they think it up on their own, IMO. Or prevent them from trying without giving you a cut.

You're so ideologically blinkered, it seems, that you're still missing my point. In the scenario I'm describing, the patent would ideally be unnecessary. It's expensive, it's pointless (because you basically always have to defend it yourself, in court, so you get no real protection), and it's time-consuming. In no way would I in an ideal world want to immediately patent an ideas and squat on it for rent. On the other hand, I might like some reasonable window of time to develop the idea and bring it to market with a protected status (2-3 years), in order to have a reasonable chance to .

Think about it his way. Have you actually searched for any prior art? How would you feel if, a year or so later it turns out that someone patented something that's vaguely analogous to whatever it is you thought up, and demands you license their patents.

Yes. That's exactly what I had done with the attorney. We had an IP search done, and no, the idea is not seemingly in conflict with any prior art.
posted by saulgoodman at 10:01 PM on July 30, 2011


"chance to ."

"chance to [make a go of it.]" Maybe? I don't know. I need sleep. Stressful week. Debt ceiling. Gak.

Anyway, I probably wouldn't want to collect fees (excluding any licensing arrangements we might enter into with some marketing partner, since we're primarily a custom software development shop for state government, not an ad firm) other than for the services and custom development we provide, if that wasn't clear.
posted by saulgoodman at 10:07 PM on July 30, 2011


Hmm... I guess I am missing your point. I think software patents should be abolished, which would take care of the problem with patent trolls and so on.
On the other hand, I might like some reasonable window of time to develop the idea and bring it to market with a protected status (2-3 years), in order to have a reasonable chance to .
So do you want patent protection to prevent someone else from creating a similar service while you start up? Even if no one does something exactly like what you're doing, you're still going to need a lot of extra code just to make it work, and you could still get hit by a patent troll. How could your product work without violating the claims of this patent for example, the owners of which recently tried to sue a Cake PHP developer. The patent seems to cover pretty much anything on the web that uses forms (in this case the suit was dropped later)

But yeah, like I said, I'm not sure what your point is. I think everyone, other then the patent trolls, would be better off without software patents.
posted by delmoi at 11:00 PM on July 30, 2011


Like I said, ideally, we would just be able to do this thing and not have to worry about running afoul of any existing patents. We would also have a reasonable chance to develop the business model and at least have a decent shot at reaping some benefits from taking the risk of bringing it to market. But without any assurance of even minimal legal protections, why should anyone do anything innovative, if it will probably just incur them a lot of costs down the road and create a new opportunity for, say, Microsoft or some other bloated giant to force its way into your new market? I don't know. I'm definitely not gung-ho about software patents, FWIW. But it seems to me patents, in general, can sometimes be a good thing.
posted by saulgoodman at 10:56 AM on July 31, 2011


But without any assurance of even minimal legal protections, why should anyone do anything innovative, if it will probably just incur them a lot of costs down the road and create a new opportunity for, say, Microsoft or some other bloated giant to force its way into your new market?

Well, since software patents are pretty universally despised among developers, it doesn't seem like most of them are too worried about it, and most of the patents that companies do get tend to be defensive patents, which they only get to fight off patent trolls and the like.

And keep in mind, Software patents were pretty rare when the computer industry really started to get going. It's true that the company that made VisiCalc got beaten out by competitors, but would the world as a whole be better off if no one would be able to make any competing spreadsheet programs for 14 years or whatever?

And, clearly there are lots of software companies in Europe, despite the fact that you can't really patent software (maybe with some exceptions, but the EU patent directive specifically states you can't patent "Computer Programs")

The argument that "without patents, what motivation will people have to write software" seems pretty strange because clearly people are motivated to write software, and have been motivated in the past to write software, without patent protection. On the other hand, concern over patent trolls does prevent people from creating new things, even if those things are unrelated to the patent.
posted by delmoi at 1:29 AM on August 1, 2011


Tech patent lawsuits discourage innovation, an opinion piece in Sunday's San Francisco newspaper. "By the end of the 1990s, the cost of patent litigation for public companies in most industries exceeded the profits earned from their patents by around 4 to 1, Bessen and co-author Michael Meurer wrote in their book 'Patent Failure.'"

I finally listened to the full hour radio program. It's terrific. It's also markably balanced: a lot of time is spent explaining the pro-patent argument and interviewing Intellectual Ventures people, particularly Peter Detkin, one of IV's founders. Yes, the term "patent troll" is used, but it's explained as a derogatory term from the beginning and in various places the appropriateness of the term is disputed. And yes, the piece has an overall opinion, but that just makes it TAL-style journalism. Journalists are allowed to use opinion in stories.
posted by Nelson at 3:13 PM on August 1, 2011 [3 favorites]


And, clearly there are lots of software companies in Europe, despite the fact that you can't really patent software (maybe with some exceptions, but the EU patent directive specifically states you can't patent "Computer Programs")

Delmoi, there is NOT a "EU patent directive". The proposal for a EU directive on patents for computer-implemented inventions was voted down by both software patent proponents and opponents as it became an ugly, contradictory monster in the amendment process at the European Parliament. This was wrongly but self-servingly hailed as a victory by Florian Müller, then leader of FFII Europe, and nowadays accused of being a Microsoft shill. In effect, the legislative abdicated to the courts its power to draw the line between what can and cannot be patented.

The European Patent Convention (which has nothing to do with the EU) and most European national patent laws indicate that patents can't be granted for computer programs "as such". What this "as such" means has been subject of a long debate both before the Boards of Appeal of the European Patent Office and before national courts. Basically, the consensus that appears to have been reached these days, after cases like "Aerotel vs. Macrossan" in the UK or the abovementioned German cases, is that software is patentable if it brings about a technical effect, going beyond the normal interactions between a computer program and the computer running it. That technical effect does not necessarily have to be "outside of the computer", as you wrongly claim in a comment above. Faster processing is already considered a "technical effect".

This is, BTW, quite consistent with the letter of the European Patent Convention. The "as such" waiver of Art. 52(3) EPC does not only apply to computer programs, but also, i.a. to "discoveries, scientific theories and mathematical methods", and it was introduced to allow the patenting of novel and inventive technical devices and methods using such "discoveries, scientific theories and mathematical methods".

What the EPO and European courts refuse is that a business method or other pure "mental acts" meet this "technicity requirement" just by being performed on a computer. But if the software has some novel and inventive feature that results in quicker processing or lower memory or bandwidth requirements than the prior art, for instance, it may well be patentable. This current state of affairs is not much different from that of the US, where the "machine-or-transformation" test applies.
posted by Skeptic at 2:59 AM on August 2, 2011


The Very Basis Of Our Patent System... Is A Myth
posted by homunculus at 2:35 PM on August 2, 2011 [1 favorite]


From homunculus' link:

The theory of patent law is based on the idea that a lone genius can solve problems that stump the experts, and that the lone genius will do so only if properly incented.

Erm, nope. Not at all. Really, really, really not.

The basic principle behind patent law is reward for disclosure. Before patents, somebody who invented anything had a strong incentive to protect it as a trade secret. It's still a valid alternative in some cases (Coca Cola's "secret formula" comes to mind, although that is a bit of a myth: think rather of proprietary code). With a trade secret, not only does the inventor keep an effective monopoly until somebody comes up with the same idea, but also, until the idea is disclosed, nobody can build up upon it.

With a patent, there's still a monopoly, but it's limited in time, and the invention is disclosed, so that others may improve it, or use this disclosure to seek alternative solutions around the patent. The whole purpose of the patent system is the transmission of knowledge, and it has served that purpose quite well throughout its existence.

Patent law is certainly not based on the idea of the "lone genius". Firstly, the inventor does not necessarily have to be alone: there may be many co-inventors in a patent application. But also, the inventor certainly does not have to be a genius. Although the expression "spark of genius" was once used in US patent law, it has long been disallowed. You get a patent for disclosing, in an enabling manner, something which is new and would otherwise not have been obvious at the time for the person having ordinary skill in the art. This "something" does not have to be genial: it can be quite humdrum and yet substantially benefit the public by its disclosure.

In short, Mark Lemley builds up a strawman far removed from the theory or reality of patent law in order to demolish it. He may do a thorough job of demolishing it, but it's still a strawman.

A more relevant question is whether all inventions deserve the same duration of protection for this disclosure. Already now, quite a few countries grant, besides the regular 20-year patent, also shorter-term "petty patents" or "utility models", usually granted without examination. Also, in some countries, there are varying forms of patent term extensions, especially for medicines which could only be brought to market after a long approval process. It has thus been suggested to have shorter patent terms for fields such as IT. While I find that the idea has merit in principle, I'm opposed to it because it is quite difficult to decide where a technical field ends and another one begins, and also because the decision of which fields would get which length of protection would be an infernal lobbying battleground. A one-size-fits-all scheme may not be optimal, but at least offers less scope for lobbying.
posted by Skeptic at 4:41 AM on August 3, 2011 [1 favorite]


« Older Industrial Scars: The Art of Environmental Polluti...  |  This is an art called “Life is... Newer »


This thread has been archived and is closed to new comments