I just spoke with a friend about a time when his company was approached by a patent troll.
The patent in question was granted in 2002 [1]. He had himself worked on the technology under "patent" in 1998, and was able to find someone who had published material as far back as 1995, and who was willing to send supporting documentation and to, if necessary, testify.
He talked with his attorney, who agreed the prior art he'd found was spot on, and that in legislation they'd be almost certain to win. But that it would cost ~$3m.
The patent trolls also were going after names such as Amazon, Facebook, and other huge companies. The troll was asking for $200k for a license from each company. As far as my friend knows, they ALL settled (or at least many of them had).
There were 15-20 names on that list. If they'd simply put the $200k-$300k of "licensing fees" into a common pot and used that money to fight the troll, then it would have done a small amount of good in defeating one troll's patent.
But really the whole system does need serious reform. His suggestion was to cause the patent trolls to have to PAY for legal fees if they lose a patent challenge, though that would just encourage them to sue from individual corporations that would declare bankruptcy on losing.
But I think software patents should simply be disallowed, or the bar raised (somehow?) to exclude anything vague and hand-wavy. One thought I've had was to require an implementation of the process being patented. Then it would be far harder to claim that your patent covers something only vaguely related, and the patent would actually be useful for its original purpose, which was to put the knowledge into the public domain when the patent expires.
[1] I wouldn't swear to the dates; they're from memory. The approximate order/scale is correct, though.
If it's such an open-and-shut case, why hire a high-powered firm that'll cost $3M instead of some kid fresh out of law school who wants to make a name for himself and is willing to take ramen wages?
Can anyone here give (and defend) an example of a good patent? By that I mean one that is non-obvious, novel, and clear enough to be able to be used to build the claimed invention.
SpaceshipOne (which won the Ansari X-Prize) is patented. http://www.spaceref.com/news/viewsr.html?pid=23871. It's owned by what is basically an NPE, which serves as the investment vehicle for commercialization of the technology.
Fundamentally, you can't understand all sides of the patent debate unless you learn about how traditional engineering works. We're talking about fields where it takes $10 million to come up with a new idea, $100 million to do an MVP, and $1 billion or more to really put something complete on the market. E.g., here's a Rolls Royce gas turbine patent: http://www.faqs.org/patents/app/20130078582. I guarantee you they burned tens of millions of dollars figuring out what those numbers should be.
This is not a space conducive to startups. Thus the need to avoid a few million here in there in defending frivolous litigation is less important than the need to be able to propertize the results of R&D, so it can be traded, bought sold, shared in structured ways. Without patents, a company like Scaled Composites or ARM wouldn't exist, because they don't sell products, they sell the results of R&D. Getting rid of patents (and IP protections generally), force every company to produce tangible products, because there is no way to monetize things that aren't tangible products.
Although I no longer work for the BBC and have already received my bonuses for the filing and granting (after I left the BBC) of this patent I have a personal curiosity as the core inventive step was (I believe) mine although it was greatly extended and improved by others with greater domain knowledge in MPEG encoding. As far as I know it was a novel idea and wasn't obvious in advance to experienced video encoding engineers.
The original idea was to treat encoding losses as errors/defects that can be fixed or patched by additional data distributed separately. This was extended to arbitrary enhancements and changes and potential implementations, consequences and techniques.
I believe that there was a real non-obvious inventive step in considering encoder losses as patchable bugs. Given that idea alone a (suboptimal) implementation is relatively easy to create but was not obvious if the task being discussed is using the early 2000's Internet to deliver better video quality than broadcast (SD) TV could.
I am not aware of any current usage of the ideas in this patent or what licenses the BBC may or may not have granted.
Edit: Time for me to sleep now, will check for any replies tomorrow.
I don't read a lot of modern patents, for the obvious reasons: there are legal risks, and the patent texts are useless even if the invention isn't. But there used to be a time when a software patent wasn't necessarily an obfuscated hodgepodge of legalese with no apparent technical content.
A system description doesn't get much better than that. If somebody couldn't implement a GC based on that patent, they're not a competent systems programmer. (It was also definitely novel. Non-obviousness is harder to argue 25 years after the fact).
The phone system this was implemented for had the unique property of tightly integrating the switch with what would normally be an "outboard" IVR system. It enabled a whole class of new features that could only be implemented on a switch of that nature, where the state of a call, or any interaction other than a call, like programming a feature button, could be interposed with arbitrary audio prompts.
If you read the patent and are "skilled in the art" you could implement it.
I'm pleased with this patent because it isn't obscure, it's easy to understand why it is novel, and it isn't trying to do anything underhanded, like making something obvious for a mobile device supposedly makes it novel and non-obvious.
The patent system is being gamed in order to exploit it. If the price of ending that exploitation is to exclude software from patent protection, so be it.
You mean software patents? Because if you meant any patent, there are loads of good examples in pharmaceuticals (and chemicals more generally), and they are pretty much textbook examples of being "hard to discover, easy to reproduce from description".
Just off the top of my head: Viagra, Teflon, aspirin (the latter of which has long since expired or been seized as war reparations, depending on where you live).
The more pertinent question would be an example of a good software patent.
IMHO there isn't any. At the very least there isn't any that benefited society as a whole.
(Again, I am only talking about software patents).
As I've written before, the point is not whether the patents covers a good idea, but rather whether the covered invention would not have seen the light of day without the promise of a legal monopoly (again, all in IMHO).
There are many. Think of safety-pin? When it comes to software it is much harder. The thing I can accept are cryptographic algorithms, since as an outcome we get a product - encoded text. All these one-clicks, etc. are pure bullshit.
On a related note, this is one of the reasons I prefer to use, and make, Apache 2.0-licensed software. Its terms are similar to those of BSD and MIT but it also includes a patent licensing clause.
What if the startup uses a scorched-earth policy? Fight 'til the money runs out, liquidating your assets and closing down the business piece by piece, then if the troll wins, they get nothing because it's all been spent on lawyers and there's nothing left for them to take.
Then the startup loses. The troll assumes that the startup is rational and won't do that.
It would be interesting to see what effect a VC could have if they started a policy of not allowing their portfolio companies to settle with a troll and funding them to fight. It might be expensive at first, but in theory once they develop a reputation the trolls will know it will be an expensive and risky proposition to go after their portfolio.
Patent protection rackets have been used for more than a business model: they have been used as tools to pressure large companies into entering partnership agreements.
Given the article's opening paragraphs (small businesses and startups falling prey to the trolls), I'm surprised there hasn't been an organized attempt to inform people of the problem. Maybe I don't know about it and the author doesn't mention it, but it seems like there should be a website that categorizes the various players and their actions, with forums so victims and potential victims can share their experiences. Once you have all the data being collected in one place, then you can get the word out (how?) so that people know to check online even before checking with their lawyer. It seems like patent trolling is something that could be partially defeated with better and more widespread information.
I expect it's standard for the troll's "pay up and sign here to avoid ruinous litigation costs" deal to include an NDA that would prohibit things along those lines.
I'm fairly convinced that it's mostly marketing and hot-air. After all Barnes and Nobles managed to get 1/2 a billion dollars out of their legal fight with Microsoft, it seems unlikely that Samsung would just roll over.
But signing a bit of paper that makes it sound like a bad idea for competitors to enter Samsung's main money-making area and get some "marketing expenses" from a rich but strategy-bereft supplier makes good business sense.
What's more important that even companies which actually produce something are also engaging in patent protection racket. Such as MS, Apple and others. That's why the term "troll" (i.e. racketeer) is not exactly the same as NPE.
I see ads all the time on TV and billboards for DUI and settlement and disability lawyers. "Get the result you deserve." I think many of them are national firms, although they look local.
It seems that patent troll victims would be a great area for a firm to break into.
It is very easy to say that a company should fight a patent lawsuit instead of paying it up but that's not always possible. In many cases pragmatism is going to be better than fighting in court for a small business, especially if the likely court costs for the battle are going to exceed your businesses accumulated lifetime revenue.
[+] [-] SomeCallMeTim|13 years ago|reply
The patent in question was granted in 2002 [1]. He had himself worked on the technology under "patent" in 1998, and was able to find someone who had published material as far back as 1995, and who was willing to send supporting documentation and to, if necessary, testify.
He talked with his attorney, who agreed the prior art he'd found was spot on, and that in legislation they'd be almost certain to win. But that it would cost ~$3m.
The patent trolls also were going after names such as Amazon, Facebook, and other huge companies. The troll was asking for $200k for a license from each company. As far as my friend knows, they ALL settled (or at least many of them had).
There were 15-20 names on that list. If they'd simply put the $200k-$300k of "licensing fees" into a common pot and used that money to fight the troll, then it would have done a small amount of good in defeating one troll's patent.
But really the whole system does need serious reform. His suggestion was to cause the patent trolls to have to PAY for legal fees if they lose a patent challenge, though that would just encourage them to sue from individual corporations that would declare bankruptcy on losing.
But I think software patents should simply be disallowed, or the bar raised (somehow?) to exclude anything vague and hand-wavy. One thought I've had was to require an implementation of the process being patented. Then it would be far harder to claim that your patent covers something only vaguely related, and the patent would actually be useful for its original purpose, which was to put the knowledge into the public domain when the patent expires.
[1] I wouldn't swear to the dates; they're from memory. The approximate order/scale is correct, though.
[+] [-] binarymax|13 years ago|reply
[1] http://www.uspto.gov/web/offices/pac/mpep/s2210.html
[+] [-] shmerl|13 years ago|reply
Can you elaborate (if you have any idea) what this cost actually consists of?
[+] [-] csense|13 years ago|reply
[+] [-] jacques_chester|13 years ago|reply
[+] [-] te_platt|13 years ago|reply
[+] [-] rayiner|13 years ago|reply
Also at Bell Labs, OFDM: http://www.google.com/patents?id=9ZFEAAAAEBAJ&printsec=a... (based on a paper published that year).
If you look at the implementation of something like LTE, there are a lot of non-obvious patented technologies in there: http://blogs.wsj.com/digits/2012/10/03/plan-to-pool-lte-pate....
Tesla has a boatload of patents and some of them are quite interesting: http://www.faqs.org/patents/assignee/tesla-motors-inc.
SpaceshipOne (which won the Ansari X-Prize) is patented. http://www.spaceref.com/news/viewsr.html?pid=23871. It's owned by what is basically an NPE, which serves as the investment vehicle for commercialization of the technology.
Fundamentally, you can't understand all sides of the patent debate unless you learn about how traditional engineering works. We're talking about fields where it takes $10 million to come up with a new idea, $100 million to do an MVP, and $1 billion or more to really put something complete on the market. E.g., here's a Rolls Royce gas turbine patent: http://www.faqs.org/patents/app/20130078582. I guarantee you they burned tens of millions of dollars figuring out what those numbers should be.
This is not a space conducive to startups. Thus the need to avoid a few million here in there in defending frivolous litigation is less important than the need to be able to propertize the results of R&D, so it can be traded, bought sold, shared in structured ways. Without patents, a company like Scaled Composites or ARM wouldn't exist, because they don't sell products, they sell the results of R&D. Getting rid of patents (and IP protections generally), force every company to produce tangible products, because there is no way to monetize things that aren't tangible products.
[+] [-] josephlord|13 years ago|reply
[http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Se...]
Although I no longer work for the BBC and have already received my bonuses for the filing and granting (after I left the BBC) of this patent I have a personal curiosity as the core inventive step was (I believe) mine although it was greatly extended and improved by others with greater domain knowledge in MPEG encoding. As far as I know it was a novel idea and wasn't obvious in advance to experienced video encoding engineers.
The original idea was to treat encoding losses as errors/defects that can be fixed or patched by additional data distributed separately. This was extended to arbitrary enhancements and changes and potential implementations, consequences and techniques.
I believe that there was a real non-obvious inventive step in considering encoder losses as patchable bugs. Given that idea alone a (suboptimal) implementation is relatively easy to create but was not obvious if the task being discussed is using the early 2000's Internet to deliver better video quality than broadcast (SD) TV could.
I am not aware of any current usage of the ideas in this patent or what licenses the BBC may or may not have granted.
Edit: Time for me to sleep now, will check for any replies tomorrow.
[+] [-] jsnell|13 years ago|reply
Consider for example the (expired, so should be safe to read) mostly-copying GC patent: http://www.google.com/patents/US4907151
A system description doesn't get much better than that. If somebody couldn't implement a GC based on that patent, they're not a competent systems programmer. (It was also definitely novel. Non-obviousness is harder to argue 25 years after the fact).
[+] [-] Zigurd|13 years ago|reply
The phone system this was implemented for had the unique property of tightly integrating the switch with what would normally be an "outboard" IVR system. It enabled a whole class of new features that could only be implemented on a switch of that nature, where the state of a call, or any interaction other than a call, like programming a feature button, could be interposed with arbitrary audio prompts.
If you read the patent and are "skilled in the art" you could implement it.
I'm pleased with this patent because it isn't obscure, it's easy to understand why it is novel, and it isn't trying to do anything underhanded, like making something obvious for a mobile device supposedly makes it novel and non-obvious.
The patent system is being gamed in order to exploit it. If the price of ending that exploitation is to exclude software from patent protection, so be it.
[+] [-] SilasX|13 years ago|reply
Just off the top of my head: Viagra, Teflon, aspirin (the latter of which has long since expired or been seized as war reparations, depending on where you live).
[+] [-] linuxhansl|13 years ago|reply
IMHO there isn't any. At the very least there isn't any that benefited society as a whole. (Again, I am only talking about software patents).
As I've written before, the point is not whether the patents covers a good idea, but rather whether the covered invention would not have seen the light of day without the promise of a legal monopoly (again, all in IMHO).
[+] [-] piokoch|13 years ago|reply
[+] [-] huhtenberg|13 years ago|reply
[+] [-] NotUncivil|13 years ago|reply
[+] [-] law|13 years ago|reply
[+] [-] csense|13 years ago|reply
[+] [-] paulgb|13 years ago|reply
It would be interesting to see what effect a VC could have if they started a policy of not allowing their portfolio companies to settle with a troll and funding them to fight. It might be expensive at first, but in theory once they develop a reputation the trolls will know it will be an expensive and risky proposition to go after their portfolio.
[+] [-] brown9-2|13 years ago|reply
[+] [-] rbanffy|13 years ago|reply
Please see the second item of the list: https://news.ycombinator.com/item?id=5479381
[+] [-] 205guy|13 years ago|reply
[+] [-] AnthonyMouse|13 years ago|reply
[+] [-] mtgx|13 years ago|reply
[+] [-] ZeroGravitas|13 years ago|reply
But signing a bit of paper that makes it sound like a bad idea for competitors to enter Samsung's main money-making area and get some "marketing expenses" from a rich but strategy-bereft supplier makes good business sense.
[+] [-] Nux|13 years ago|reply
[+] [-] shmerl|13 years ago|reply
[+] [-] samspenc|13 years ago|reply
[+] [-] abfan1127|13 years ago|reply
[+] [-] nonamegiven|13 years ago|reply
It seems that patent troll victims would be a great area for a firm to break into.
[+] [-] dantiberian|13 years ago|reply
[+] [-] mehdim|13 years ago|reply
[+] [-] mathattack|13 years ago|reply