American law is patent nonsense
ft.com
ft.com
This article is akin to the lobbying that any interested party does when they want to see change... To make a comparison... half of what Chris Dodds said from the MPAA was utter balls and constantly full of inaccuracies and litter picking his facts to suit him...
I don't particularly care if the Financial Times do the same here. Not every FT reader will be clued in to the inadequacies of the patent system, and I think this article does do a fantastic job of explaining in lay terms how much of shambles technology patents are.
Back when I studied Law at University, I actually remember my first lecture on IP and Patents; it was on "Justifications for Intellectual Property Rights". We talked about Venetian/Florence Glass Blowers and the Pharmaceutical Industry...
Comparing the different industries makes it easy to see how you can justify patents for one use case. Providing a limited monopoly in pharmaceuticals (for the most part) guarantees the incentive for the ridiculously high research and development costs that goes along with bringing a product to market.
Providing a monopoly right for technology where you position your fingers over a device to interact with it, is ridiculous and just shows how antiquated and behind the times the Law is.
Yes, the article might be muddling up some things, but it does a great job of bringing to attention how broken the patent system is for technology/software/computer implemented inventions.
ADDED: One problem pharma shares with software is "patent thickets" where there are many applicable patents, applying to different layers of technology, that the current development relies on.
Note the words "To promote the Progress of Science and useful Arts..."
I would say that any necesary condition on a patent would be that it indeed "promotes the progress of Science and the useful Arts". In other words, the purpose of even sanctioning patents is so that people would have incentives to go out and invent things. The assumption is, that absent a patent, folks will sit on their bums and not invent anything. With a patent, the inventor can profit, but so does society. If there is no benefit to society, why allow the patent?
Until relatively recently (the 1980's?), software patents were virtually nonexistant. Did that stop "inventions" in the world of software? That's laughable, since that was the time of immense creativity, new companies, products, inventions. To me that is PROOF that patents in software are hardly needed to "promote the progress of Science and the useful Arts". Think of all the software companies that came of age in the 1980's and 90's. WITHOUT patent protection (for the most part). WHY allow software patents? For what purpose? The only reason I can see is to benefit the legal profession. And large companies.
In addition, software patents have a number of flaws, mostly due to practicalities deriving from the ease with which software designers can conjure up new techniques, visual elements, user-interface designs, wire protocols, languages,database designs, and the like. Software is a "soft" science - it doesn't require expensive laboratories or extensive field trials. It is playing in a virtual world. You can sit in your room and without spending millions of dollars, come up with amazing things. But so can the person across the street or around the globe! And they do! So MUCH of what is "patented" are just "obvious" takes or re-takes of what already abounds, or would abound, should someone else in a room have to solve that same problem. Many patents are not novel, complicated, unusual. Clever, maybe. But someone else in the same circumstance might come up with the same exact thing. Why should the first person be granted 20 years of "protection"?
IF we were to allow software patents at all, say, for truly novel, complex things (yet to be determined how you would define this), we should limit the term - to say, 3 years. So yeah, go patent your "One Click" or your "Bouncy Screen" - if approved (dubious for these), you would have only 3 years to profit from your exclusive deal. Far fewer patents would be applied for. Far less litigation.
And software designers and companies could create great things and compete on what they are offering, and not who they are suing. And, believe me, tech would flourish. Software patents are a TAX on everyone.
The tech industry just needs to agree that we'd all be better off with a truce, and lobby to change the laws. Personally, I'd like to see no patents for software and business methods.
But they also spent a lot of money to get (and maintain) these patents over the years, so while they might not try to fight a patent reform, there is no chance they will lobby for it: With all the pain they are enduring, the patent system STILL gives the big guys an edge on the small guys.
> Personally, I'd like to see no patents for software and business methods.
Me too. But patents (paid for already) have already been granted, so even if you stop granting them today you still have to wait 20 years for old ones to expire; and patent holders would (rightfully) demand their money back if you decide to expire them early.
The only solution (which several people have apparently come up with independently) that I'm aware of that has a chance of working, is introducing a "cost-of-carry" for patents - if patents are property, tax them like property, at e.g. 1% of decalred value per year.
Think your patent portfolio is worth $1B? That's nice. Pay $10M/year to maintain it. The value would come into play in estimation of damages (e.g., patent infringement limited to patent value per violator, or at most 3 times as much for willful infringement); and alternatively in an "eminent domain" style seizure - someone could take the patent away from you in return for 20 times declared value.
The mechanics are non trivial, but can be solved in such a way that they create an incentive not to overstate the worth of a patent, yet still provide protection to inventors - and it would put old and new patents on equal footing.
I'm in favor of a "King James I moment", revoking all prior patents. The problem with pure maintenance fees is that it does nothing to prevent overlapping patents. Also, truly innovative patents may not prove their worth for several years (this is sort of the point), but the original company may not be able to afford the huge fee during the early years. If a big company could just buy it for 20x whatever the inventor is paying for maintenance, the inventor can easily be shut out of the market.
They sued Apple for FRAND patents that, not only did they make promises on to get in to the standard, but that the license for was already aquired by Apple via Qualcomm which made the chips that used the inventions and which had a license.
So, they are abusing both the FRAND standards and trying to double dip on exhausted patents.
You can't accuse Apple of doing anything that abusive. Apple has been defending patents for inventions copied by people without a license and which are not standard essential.
Basically it seems that to a lot of people since Google needed to rip off the iPhone for android, in order to compete, therefore patents are suddenly bad.
"We think that these patent wars are not helpful to consumers. They’re not helpful to the marketplace. They’re not helpful to innovation." (http://www.forbes.com/sites/davidthier/2012/08/21/google-exe...)
All of the companies have hoards of overlapping patents on technology that was not innovative in the first place. They "defend" their patents in court because that's the game. All the major players do this, not just in technology. When do you think "Coal-fire steam locomotive" was patented? http://www.google.com/patents/US4425763 (1984)
How about three patents for technology so similar the author couldn't be bothered to change even one word of the title or abstract? http://www.google.com/patents/US6470844 http://www.google.com/patents/US6945207 http://www.google.com/patents/USRE40286
Do you really think that Apple, who has been a notoriously slow adopter of NFC (http://online.wsj.com/article/SB1000142405270230483070457749...) really created any new technology in this patent that they were just awarded? http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sec...
When there was a recent review of patent law and software patents recently Google did not come out against software patents. In no way are they setting a good example in their current behaviour, the best you could say is that they are doing their best to bring the system into disrepute by their failure to stick to their FRAND commitments.
How did you find these? If someone pointed you to them, I think they were trying to mislead you. If you found them, I think you're trying to mislead us :-)
Sure I can. Apple has been getting patents on obvious ideas that, often, they didn't event invent like:
- "data tapping" anticipated by both Netscape Navigator 2.0b1 (Live URLs in mail and newsgroup messages) and the Sidekick TSR from the 80s ( http://www.articleonepartners.com/patent-research-community/... )
- slide-to-unlock (anticipated by the Neonode N1m)
- universal search
And then Apple takes advantage of the weaknesses of the US patent system and uses these bogus "patents" as legal weapons against their competition. This is the same sort of chaos we would see if someone with a big checkbook decided to start trolling some of the many linked-list patents. What Apple is doing is just as bad (or maybe worse since Apple is more interested in market control than money).
> Basically it seems that to a lot of people since Google needed to rip off the iPhone for android, in order to compete, therefore patents are suddenly bad.
Just because you weren't aware of the history of software patent opposition doesn't mean it sprang up out of nowhere post-Android. Lots of people have been against software patents for a long, long time. Take a look at the history of the League for Programming Freedom, "Burn all GIFs", the Eolas web patent case, BTs hyperlink patents and on and on and on. Many have seen this nonsense and concluded that patents in software just don't work long before Android was a gleam in Andy Rubin's eye.
Since when?
I have not seen a single case where Google has offensively litigated any patent prior to their purchase of Motorola Mobility. And this new FRAND suit initiated by MM is the only offensive litigation Google has done since that purchase.
Or do you have a list of cases I have not yet seen?
The devil is in the details. For example:
When filing taxes for year Y (which is usually done in Q2 or Q3 of year Y+1), you have to assign a value to each and every one of patents you own, and pay the tax accordingly.
Let's say Microsoft has only 1000 patents. If they valued each at $1M, the fee would indeed be a rounding error. However, it would also be much less frightening to competitors: $1M if sued is not such a big deal for a company in business.
But they have 10,000 patents; and if they wanted to value each at $100M (as seems to be the case for things that go to court), it's suddenly $10B/year. Even Microsoft cannot afford that.
The end result would be that they would actually examine their patent portfolio yearly, and assign value only to valuable patents -- which is the whole point of this exercise.
On the other hand, if you are a small guy, and have a $1M patent, then $10K is something you should be able to easily finance. If not, then your patent is not worth $1M - or you don't have the means to enforce it under the existing system anyway (it costs at least $100K, and some people estimate upwards of $500K, to bring a patent case to court with a chance of prevailing. If you can't finance that, your patent is already worthless).
And the best way to determine the right value is exactly that what you wrote: the owner sets the price of the portfolio, the owner would pay the fees accordingly, and the licensing fee would be a percentage of the total value. And if refused a license, any potential licensee can buy the remaining years of the patent at the initial annual price (= portfolio value / portfolio lifetime).
So if the patent has value but the inventor has set the price too low to save on fees, someone can buy it and make better use of it. And setting the price too high would have to be a reflection of the inventor's trust that the patent is worth it since the inventor is going to have to pay for it each year.
If it sounds unfair between big companies and small companies and individuals, some relative pricing can be established using percentages of the inventor's turnover / annual income. Therefore, it will hurt you for 1% or something, no matter how wealthy you are or how big company you are.
By cost-of-carry, I mean that if a patent is worth $1B, it should bother you like any other $1B asset. Are you aware of any other $1B that you pay (on average) $500/year to maintain, and yet get police forces all over the US (local, feds, custom officials) to underwrite and execute said ownership for you?
Because surely all a patent troll would need to do is implement a product that demonstrates a single patent in their pool.
Or are you suggesting that every patent must be implemented as a product for it to be enforced?
And if so, as an example of similar things, the UCI (cycling body) state that all Olympic track bikes must be available to the public, and this is true of Team GB equipment that you can order, for a price in the region of GBP 100k per bike.
So would it be enough for a patent troll to offer an implementation of a smart phone for USD 5m?
We need to get to the source of the problem, which is that most software patents are either too vague or are for obvious things, and the term of protection is far too long.
Where are the objective cost/benefit studies of the patent system?
http://levine.sscnet.ucla.edu/general/intellectual/againstfi...
Chapter 3 has interesting historical case studies, including, for example, the patent-free flurry of innovation that increased the efficiency of the steam engine by 5x in the 25 years immediately following expiration of Watt's patent, in contrast to the efficiency plateau that occurred while Watt's patent was enforced. Chapter 8 has a modern macroeconomic analysis.
Besides, if you can patent the look of icons (which is design, thus, art, and not in any way an invention), who is conflating the types of IP, anyway?
Despite the name, a design patent is very different thing from a (utility) patent, which is what people normally think of as a patent. In many ways it's better to think of a design patent as a class of trademark.
If we want to reform the "patent" system, then at the very least we need to be clear what we are arguing against. Different types of IP - patents, design patents, copyright, trademarks, trade dress, etc. - were intended to achieve very different social objectives. Each has its own sets of strengths and weaknesses, and each has been distorted by regulatory capture in different ways.
I think it's important to understand these differences, and something of their history, to be able to make effective improvements.
Frankly, I've personally not yet seen an anti-patent argument from someone who actually understood what patents were, how they worked, etc. They constantly confuse terms, misrepresents what the patents they are dogging actually say, or rely solely on unsupported assertions.
Including hundreds of comments here on HN, all of which seem to be ideologically driven repetitions of talking points, and none of which come from people who have read the patents or which accurately represent what the patents actually cover. (Eg: all the commentary on the Samsung trial, not one criticizing the patents was accurate.)
Its not like patents are secret- they're published on the web for anyone to read.
Near as I can tell-- google felt they needed to rip off the iPhone for android in order to compete, and therefore, suddenly, patents are bad. Yet microsoft's metro UI shows this isn't the case....
Sticking to patents there are clearly real problems with the thicket of patents where many things can be infringing without realising it (while patents are public they are impractical to study or to get an idea of all the patents in an area like touch UIs and even if you could it is hard to correctly understand and interpret the claims to work around them).
The cost of the legal actions is a real problem (for small companies) in itself partially caused by the weakness of the initial examination that means that many of the existing patents will be found invalid if ever tested in court.
The lack of certainty over validity also harms patent holders who cannot really assume that really do have a valid patent and this will harm them in negotiations.
If patents are to be kept (and the economic case isn't clear to me although it may be there) then if the primary examination cannot be improved then there should be a second and more in depth (and more expensive) examination that should be required before legal action is brought (or optionally to strengthen their negotiating position). After this phase it should weed out sufficient invalid patents so that there is an 80-90% chance it will be found valid in court.
If you are interested... Stephan Kinsella, according to his website, has worked as a patent attorney since 1992, and is therefore likely to know how patents work. He gives talks and writes articles and blog posts and at least one book against intellectual property.
A talk that turned up quickly: http://mises.org/daily/4848/
A collection of links to his writings against IP (and, it being his website, at the top you can find pointers to more information about him and stuff): http://www.stephankinsella.com/publications/#IP
The way I am reading this article, and it can be wrong, is as a by-and-large synthesis on the author's point of view on why US patent law is antiquated and supposevily bad for the Tech Industry of this country.
I don't see any of his points as incorrect, in fact I agree with his sentiment.
Therefore, do you disagree with he is saying? Or do you just think that he is right, but fails to make any good suggestion on how to change the current flawed system?
The issue really starts with the American culture (of litigation), not American law, per se.
And of course the encouragement of litigation, the acceptance of it as a way of life and a way of doing business, and the high amount of ensuing court activity breeds a rather large body of case law, a complex legal landscape, which gives way to a sort of feedback loop leading to more litigation.
"Can we sue our way to prosperity?"
This was a title of a Congressional study a year or two ago.
I think it captures the essence of the problem. Wealth is increasingly being created _only_ through litigation. (Reminder: IP rights are often nothing more than rights to sue alleged infringers.) Is that toxic? Is it sustainable? I don't know. I don't think Congress knows either.
We should all just agree that execution is everything, and only fight when someone copies actual code etc.
Second, the problem with the patent system isn't that patents are given too easily (trust me, its not that easy), or to obvious things -- as they say, all beautiful solutions are obvious in hindsight -- but that companies that infringe patents have convinced themselves that they are within their right to do so. There's definitely a strong case for giving immunity to companies below a certain size (revenue, employee base etc) from patent prosecution, but most large companies can easily license patents but refuse to. Yes, there are still abuses of the system, but those are bound to arise no matter what system you put in place.. someone will always try to game it. And as software patents are litigated, common law will settle down on standards and businesses will know whats kosher and whats not. Much of the current excitement is because we are currently in the midst of the first explosion in assertion of software patents where common law is still evolving.
Last, many of the claims being made regarding the supposed absurdity of apple's patent claims are so fundamentally (or willfully?) uninformed, that it begs wonder (not very unlike the willful distortion of news from other countries thats so common in major western news platforms). Apple never really asserted a patent on a rectanle with round edges.. and as for its claims on UI elements -- it was a trade dress claim primarily, not patent claim. The bounce back and tap to zoom claims were patent claims -- but rest assured, micrsoft would probably have implemented the same with no bounce back and context menus to zoom. (no ms office app after 30+ yrs in mkt still has bounce back -- so maybe its not that obvious). However, the prior art arguments are legitimate and can only be settled in court.. which samsung simply failed to do.
I am for abolishing software patents - for all the reasons everyone sites. But "stealing" someone's actual code or implementation or book or music is another story.
As for abolishing software patents -- would help if you can list the reasons people cite -- I can't offer a counterargument otherwise.
Remember, the closest alternative to a patent is a trade secret. The patent bargain is supposed to be: Tell us the secret (so it won't be lost and others can build on it) and we'll give you a temporary monopoly on it.
Looking at how the bargain plays out in practice illustrates the big problems with software patents:
1. The inventions described are easy to independently invent. The public is not getting much of anything when told the secret.
2. Software patents don't have adequate disclosure. They're deliberately written in opaque language that practitioners don't understand in order to make the invention sound more significant and to make the claims as strong as possible. They also don't include source code or any guidance on the software engineering tradeoffs required (e.g. performance, flexibility, modularity, etc.) required to practically realize the invention.
3. Even if the disclosure were adequate, the incentives of the patent system (in this case, willful infringement) mean that practitioners never look at them. This is everything from the patent attorney coming to your office and telling you to never, ever look at a patent to Groklaw posts that say that if you're a software developer, you might not want to read posts that discuss patents in detail.
Patents are for the implementation of ideas.
Trademarks are for the names of ideas.
See the difference?