If patent workers are overworked, why don't we replace them with a certificate of origination. You file it, it takes two weeks, it's dated, and it describes the work. Then, if you need to sue someone, you then have to pay to prove that they are infringing. Reverse the burden of proof and delay the cost of approving patents.
Second, my biggest problem with software patents is that you can't build the thing from the description. If you look at a patent for a device, there are schematics and descriptions of how it works. Software patents just say "the user can click a button and it automatically buys a book". Where are the detailed drawings of how the web server connects to the payment system which connects to the distribution center? How do you scale that up to millions of users? That's where the magic is.
Patents are supposed to give the inventor a reasonable period of time to profit from his work. After that time period is over, it becomes public knowledge, so everyone can do it for free. Software patents do not give away the knowledge that it actually takes to build the system. Instead, they typically just describe the tiny bit that is apparent to the user.
The problem that I see with the certificate of origination idea is that then the patent trolls will file for literally millions of certificates of origination. Then they'll just send you threatening letters asking for a settlement and you have to do the costly vetting process to decide whether or not their certificate of origination has merit. Most likely you'll just settle.
A variant on this contributes to the problem. The mechanism is that person 'a' 'invents' X while working on something doesn't think its all that special, person 'b' also 'invents' feature X but they work at a place that has incentivized filing patents so they file a patent on it.
Now some time later person 'c' also invents X and gets sued by person 'b's company. (or the holding company that bought out person 'b's company's IP portfolio).
Engineers have been told to literally "file a patent on everything you do, regardless of how novel or patentable you think it is, let the lawyers figure that out." further, to protect themselves they pay 'bonuses' for filing. Its a strategy and it pays dividends in the current climate.
So I understand Marco's and others frustration here, I would be interested in clearly bogus patents that have been filed post 2005 (my thesis is that starting in 1995 the PTO was hit by a zillion new patents and it wasn't until 2005 that the examiners had become trained enough to recognize the obvious ones and discard them) I've done some expert witness work and found that the prosecution histories of patents show a significant improvement in the examiner's responses as you move from 1995 to 2005.
If there are many defendants, they could work together to find prior art. Crowdsourcing might also be an option - for example, ArticleOnePartners.com pays cash rewards to individuals who find prior art.
Mandatory arbitration (prior to trial) might also be an interesting alternative, especially with the burden of proof on the plaintiff.
The lawyers for the patent trolls are smarter than that, they aren't going to start with dozens of defendants, they are going to start with one or two, bully them into settling, and then use those settlements as precedent when they go after 4 other people... and hey, their patent must be valid, because look at these two guys who are licensing it already.... then they go after the next group (expanding the net each time like they did in those old 'she'll tell two people' shampoo ads).
Precedent counts for a LOT when it comes to resolving any legal issues, even if the precedent is settling, and the trolls know that and use it.
Settlements might make it easier to bully others later, but they do not provide legal precedent. Only an actual court ruling does that. And as senior/superior a court as possible.
Annoying that Marco seems to have taken many of his talking points from This American Life's "When Patents Attack" episode -- without attribution, and mangling the facts in the process!
Toast patent -- featured in the episode.
Saying IV shell companies only sue people -- misrepresentation of something in the episode; companies affiliated with IV sue people, but some IV shell companies may only be used to acquire and license patents (thus their shell companies do not "only sue people" as Marco says).
Marco does not address the popular notion that software patents should be wholesale dropped as a way to fix the problem.
Also more substantial and less nieve (ie. not from somebody who listened to the TIL episode, read wikipedia for 15 minutes, and then decided to write a blog post)
I can't understand why you think Patel's article is better. I will grant that it is well written, but the argument is practically non-existent.
First, it's nowhere near a similar argument. Patel argues that software patents are fundamentally sound, they're just implemented poorly.
This article instead argues a position closer to my own which is that software patents are fundamentally broken because they can't be implemented correctly. The assumption in Patel's argument is that it's possible for the USPTO to police software patents effectively. Thus far this has not been shown to be the case. The argument in this blog post is that because of the nature of governmental regulation, this is impossible to do effectively. After reaching this conclusion, the conclusion must be that software patents are fundamentally broken.
Patel can make an argument that this is possible, but he doesn't.
We have two software patent articles on the front page of HN. One from a developer saying they are unfixable; another from a lawyer saying they aren't broken. I think this succinctly describes the situation we are in.
The answer apparently is 'because the US legal system doesn't work'.
>In practice, therefore, an issued patent is a valid patent //
Well no. An invalid patent is still invalid even if you use it to badger someone in to paying for a license. If that person knows that it's invalid, eg because they know of certain publicly available prior art, then it is their choice whether to pay for a license or follow a legal remedy.
If a legal remedy is uninviting because of problems with the legal system this does not mean that "software patents are not fixable" it means that the legal system in general is poor.
> If a legal remedy is uninviting because of problems with the legal system this does not mean that "software patents are not fixable" it means that the legal system in general is poor.
Rather, he assumes that the legal system is poor, and reasons from that viewpoint. It's correct today, it'll be correct tomorrow, and it's unlikely to be different anytime soon.
It is easy to blame the patent office until you realize that most cases are a single examiner in the office up against multiple law firms who continually ask what is wrong with the patent, told that according to their interpretation of the rules it is ok, refine and re-apply until granted.
You have hundreds of millions of dollars at stake, and the resources of the best patent lawyers in the world and the largest companies in the world up against (usually) a single examiner.
What you're saying is true, but even after digesting it it is still easy for me to blame the patent office (if not individual patent examiners) for the problem.
The current brokenness of the patent system is due to a snowball effect. If the patent office hadn't let the system slide out of control to the current state of brokenness, their examiners would have more authority to push back on anyone trying to push sketchy "inventions" through the system.
Seems like it would simply delay the issue, being, once you've gotten an overly broad and generic patent through the system (however long it took) you can wield that weapon pretty broadly.
Reduce the length of time the patent is valid or remove the benefits of software patents, and you'll reduce the pressure on the USPTO pretty sharply.
That is overstating it. They are a reasonable idea, but it does not seem at all clear that anyone has ever proven them to be actually a good idea -- i.e. that they have a net benefit, even if they were properly realised and enforced.
Those rationale can be proven to be false in virtually any industry.
Patents are supposed to be an inducement, but that rationale alone falls flat on its face once you realize that invention takes place in the absence of patent protections. Worse, patent protections make it more difficult to innovate as claims are broadened.
Disclosure is a laudable objective of patents, but largely unimportant as multiple parties come to the same inventions and talented persons reverse-engineer products (some of whom document their findings publicly). Enforced disclosure is not needed, irrespective of the protections awarded.
In industries where R&D is costly, patent protections are not needed when there is a first-mover advantage. Moreover, patent protections are not needed for commercialization, provided that the inventor continues to innovate after the first sale. Providing a guaranteed advantage to the inventor causes an overall harm to the market of the product, as the inventor has no incentive to improve the product or lower prices, because the inventor has no competition. Competition is healthy, and patents neuter it, leaving the market fragile and expensive.
All of these things can be demonstrated by analyzing the history of patent litigation. I can find no reason to support these rationale.
What about the pharmaceutical industry? Third parties are likely able to reproduce and sell generic versions of drugs within months of the original version being put on the market.
Without patent protection, that's not enough time to come
close to covering the R&D costs, assuming the market for the branded drug shrinks in the face of the much cheaper generics.
It's a nice idea, but I suspect that if you were to tweak copyright law to apply to chemical compounds used as drugs (e.g. field of use restrictions) you'd end up with patents.
patents protect the mode of action rather than the compound, and the term can be extended in court with simple little changes
it isn't as cut-and-dry in the pharma business, which is why most patent lawsuits and active trial involve pharma. the patent system is supposed to be simple, but the companies and the generic manufacturers are suing each other all the time.
copyright would just apply per-compound at a fixed term. the other way to do it is for the FDA (who are already approving drugs anyway) grant exclusive periods to new drugs before generics are allowed
What I was getting at is that most(?) compounds probably exist somewhere in nature already, so copyright wouldn't apply without modification. I think that means you'd have to copyright the compound in a novel field of use e.g. "for use in treating cancer".
Further, you'd want your protection to cover many modifications, so that a competitor can't just make a change to a non-functional aspect and piggy-back off your research. Otherwise you'd have the problem that new R&D wouldn't pay off, because competitors could just devise drugs that would use whatever mechanism you discovered, even if they didn't have the same physical embodiment.
What you end up with is a "copyright" that is - in practice - a patent.
Of course, the drug companies would probably love for their protection to last 75 years!
On the "found in nature" aspect, I believe some of the patented gene sequences are indeed derived directly from existing organisms, but are considered novel 'inventions' due to the labour and skill involved in identifying, isolating and applying them. Copyright doesn't really fit for that sort of use-case.
In my humble opinion, software patents are a bad idea and should be eliminated. If this can't be done, then they should be really limited in length, say three to five years, which is a long time in the software world.
As things stand now, the only people I see benefiting from software patents are lawyers and the Patent Office itself, which I sincerely believe is more interested in collecting the filing fees than anything else.
The USPTO has not always been so eager to issue software patents. Up until the early 90s they were reluctant as they saw code as a language which could be copyrighted, but not invented. From what I understand, most software patents came flooding in after the Court of Appeals ruled in favor in the case in re Alappat, 1994. OP could have spent some more time doing research on software patent history before writing this, I feel.
One possible solution is a class-action lawsuit against the patent office.
The patent office is causing harm left and right by granting multitudes of ridiculous patents. If it won't stop by itself, everyone that has been harmed by such patents should join in suing it.
All students who have suffered through the shit-tastic Blackboard course system (protected by software patents on a client-server-based app for anything school-related, despite prior art from decades before!) could join in …
The problems he’s discussing arise from the fact that litigation is expensive, and those with lots of legal resources can use their superior position to assert their will against those who can’t afford (or may not be willing) to defend themselves. That’s a valid issue, but a broader one that exists outside the context of patents. It’s not a valid argument against the patent system itself. A powerful chemical company that poisons a town’s water table can do precisely the same thing. Does that mean we should outlaw chemical companies? No, of course not. His argument is only valid in the context of arguing for wider reforms to the legal process in general that make this sort of abuse less feasible. And that’s a different discussion entirely.
This is precisely the point the other article today by Nilay Patel—which was so lambasted here—was trying to make: the arguments being made about patents don’t necessarily point to tearing down the system, but simply fixing the things that are wrong.
"Just because my theory doesn't work in reality doesn't mean we should give up on the whole theory! Society should suffer what they must until this human experiment is refined"
Let's try to make this more objective. What's the goal of patents, how is success measured, and why are patents the only way to achieve this outcome. Furthermore, how did we establish that intellectual protectionism is more advantageous than open ideas? You seem to be at odds with the open source community.
" A powerful chemical company that poisons a town’s water table can do precisely the same thing."
Isn't it illegal to poison towns' water tables? That seems like a pretty severe flaw in your analogy, since presumably you don't think suing people for patent infringement and then offering to settle is or ought to be illegal.
There's also the fact that chemical companies provide clear social value: people buy what they make. With software patents, though, I'm unaware of any case where someone has reconstructed an invention from the description in the patent.
When you apply for a patent you have to show plans for producing the thing and and the price per license to give it out (no upper limit so if you want to only produce, set it to $10^24). For every year that you don't produce the thing and nobody licenses it, the officially recorded license price gets halved. For every year that someone does license it or you produce the thing, you get to adjust the price to whatever you want. Thus companies like Lodsys can only hold onto an unused patent for a fixed number of years before it becomes irrelevant.
Obviously the timing is not ideal: a better half-life for an unused patent might need to be adjusted. However, I think there is something to this idea.
Edit: A better way to determine the half-life would be to tie it to the initial price or the price at the time of reduction. Basically, no matter how high you start out, if you are a patent troll, your patent expires in 5 years.
I like the idea of showing financial plans. A patent should encourage innovation and nothing more. I've always thought that the inventors (or patent holders) should be entitled to all proceeds that their invention generates until they've been adequately compensated for the opportunity cost and capital they've risked to bring their idea to market. Capital costs are easy to measure, opportunity costs not so much. But that is captured in what the inventors plan to earn off their invention either through selling it or licensing it, so the patent office can figure out how long a patent would need to be valid in order for the inventors to be properly rewarded. Since a dollar value would be attached to it in that case, any purchase of the patent would probably change the time it takes for the inventors to be made whole.
In any case, 20 years of protection no matter who owns the patent is a terribly stupid idea.
> I've always thought that the inventors (or patent holders) should be entitled to all proceeds that their invention generates until they've been adequately compensated for the opportunity cost
that is worse than what we have now (damages were, until very recently, calculated on 25% of profit from the part of a product that violated a patent)
also, there are already stringent 'use it or lose it' provisions in most jurisdictions - they are easily avoided, and you only need a single state to become 'patent friendly' to screw it all up anyway
No upper limit? Why wouldn't someone set it to $2^1,000,000,000, never produce the thing, it can't be reasonably licensed at that price, and poses the patent for millenia?
If the fixed number of years is longer than the current 20-year limit, this solution is counterproductive.
There's something in this idea, as I've often toiled over the sharp edges, but it's not all there.
Take one of the article's linked examples, the "Linked List" patent [0], what are the requirements in "producing the thing?" Churn out repositories all day that print random indices of random length, linked list? This idea seems hard to enforce on software patents, which tend to be the most disingenuous, along with most UI, UX, and form/material patents. Do I have my OS create trillions of glossy chat bubbles every day, on a large distributed network, to protect my trivial patent?
The big problem that keeps coming up with software patents is the fact that most of them are bogus, because of being trivial or unoriginal. Barring cases of malicious intent by patent examiners, which are a completely separate issue from the viability of the patent system itself, the reason for that is that the patent examiners don't understand what constitutes non-obvious work in software. In fact, I would guess that now there are very few fields in which it's feasible for someone who's not already familiar with the field to make that judgment in any reasonable amount of time. Since the criteria for patent validity are based on how the work would be viewed by average people in the field, maybe we should put patent-granting into the hands of people in each field, who are actually qualified to evaluate such work. Just have the central patent office keep records and oversee devolved organizations for each discipline, who actually control which patents are granted and what the terms of patents in that field are. That way, not only can bogus patents be avoided, but fields with radically different environments won't be hampered by patent terms designed for other areas (e.g. software patents could have a much shorter duration than others).
Why don't we turn the patent review process into a public endeavor?
Hacker News / Digg style... vote up the novel patents and down the obvious ones, while commenting with conflicting patents or prior art... then the examiner can use this as input.
If the system is perfected eventually that examiner could be eliminated. This also would create jobs in the private sector as companies would be inclined to hire contributors to this process to protect their interests. You obviously need transparency to see who's providing this feedback to prevent abuse, but that's a minor aspect.
Finally, IMHO software patents specifically deserve more stringent guidelines for patentability. There's no intrinsic technological worth in a major percentage of software patents. "It's software, if you want a word processor I can make you a word processor"... if you want a button to tap and order something instantly, I can make you a button. Now, if you want a button that can be clicked by millions of people over a minute, there may be some novel technology behind that. These differences need to be isolated.
I haven't seen anyone present this sort of angle on it, so I'm wondering what y'all think of this as an idea for reducing abuse of the patent system:
If you're an entity that has bought a patent (but is not its original inventor), you get two years to implement the invention before that patent dissolves. And unless you are an active competitor to companies you aim to sue over said patent, you are not entitled to sue them. Those two things eliminate the nefarious motives both for selling, and for buying patents.
Those restrictions would not fully apply to the original inventor. If the patent resides with its originator, then the patent does not have any statute of limitation before dissolution (perhaps they reserve the right to bequeath?). But they can't litigate with that patent unless they're an active participant in the market to which the patent applies.
I'm sure there would be lots of refinement clauses, but those are the broad strokes, as far as I'm concerned. Does this seem just too pie-in-the-sky? If so, why?
I don't have a definitive answer to how to fix patents or software patents but I'm pretty sure that patents must somehow be coupled to economical counterweights. What I do know is that it's way too cheap to obtain patents and walk waving them around.
There should be a cost of ownership involved that is somehow relative to the gains from the patent and its value. Patent troll companies exist because it's too cheap and easy.
I can't imagine the details but ideally this cost would direct companies to only consider patenting ideas that would exhibit an actual novelty and also a reasonable profitability expectations. This would validate both the new and useful aspects.
Filing patents the conveyor belt way would be too expensive and even if given the money, filing patents without merit wouldn't yield profits. I don't know how to link merit and money, and I don't know how to not exclude small players from the market but cost has notoriously a pretty good chance of keeping things in some control.
Patents are both too cheap and too expensive. Too cheap for capitalized trolls, and too expensive for individual inventors.
Clearly the problem is that the novelty clause is not being enforced in any meaningful capacity.
The deeper problem of course, is that a hundreds-year-old institution is just incapable of dealing with the concept of software. The pace of development is orders of magnitude faster, and the cost is orders of magnitude lower than the type of innovation that patent law was originally designed for.
Imagine if you came up with a particularly clever idea in software and you got two years exclusivity to capitalize on it, assuming that the patent application included the actual code. If the novelty were properly enforced, I can imagine how such a system could offer an incentive to innovate and provide "open source" code to the world at large.
Of course I don't have any faith that the bureaucrats have the capacity to make such a thing happen, so honestly we're better off without software patents entirely.
One is Google sized, and the other one is patent troll business model.
The Google-sized problem has large legal teams at each others throat. i.e. if a patent doesn't have sufficient novelty, Google has sufficient resources to get it invalidated. The definition of novelty should be tightened. I would love to link it to an objective measure of brainwaves of those familiar with the arts, looking at the level of surprise and delight at a solution.
The patent troll business model is a shakedown business model, and I'm also looking at general copyright infringement over file sharing as well. This requires legislative change to require a "reverse class action lawsuit" if a firm intends to send out identical complaints to more than say, 100 defendants. This enables the defendants to band together to fight these in the courts.
I have looked through google patents beta just to see how ridiculous many software patents are, and though my sample size was small, most seemed pretty obvious ideas to me. Truly amazing that lawyers can write so much text about so trivial an idea. The core issue is that the patent office does not have "people skilled in the art" to bar these patents from being issued.
Software patents are a problem. But perhaps software engineers are biased about this particular regulation because this is the field they work in. Seems to me similar arguments can be made about all other patents, and all other IP laws, and even all other regulations period. They do more harm than good.
My grandfather was a mechanical engineer. He never said anything bad about patents, and even got one for himself. One is not automatically biased against patents in one's own field. Software engineers oppose software patents not because of bias, but because they're harmful.
I don't know the value of patents in other fields. There may be areas where patents accomplish the purpose they were designed for.
IMHO the problem with software patents is the prior art that's fundamentally attached and the degree of difficulty in discovering it. Of course, there's some other aspects:
1) In my mind, there is the disincentive to work at the US PTO [or any PTO for that matter] because the salaries suck, you're demanded to have unrealistic quota's and that means only so much time can be spent on each filing by examiners who are mostly beginning - because ones they are experienced leave to join firms who pay them a lot more because they have examined from "the other side of the table".
It's a talent glut that starts the cycle, approves poor patents, clogs the court with the attitude of "even if we are wrong [PTO] the Courts will resolve it - so we really can't make mistakes" and increases litigation costs for all involved.
2) I don't think removing patents is the solution - there have been 100's of instances where the "little guy has stopped the big guy" in all sorts of scenarios in innovation - tech or otherwise. The problem is that you have a system which is bursting because the pay cycle for PTO is low, pressure driven and this means you've inexperienced patent examiners who ultimately leave for high paying firms. Why wouldn't they ? There isn't any economic incentive to stay. Pay them more. Retain the best ones. Get better examination results, use an Open Source Prior Art system which allows community collaboration to increase prior art base and makes it harder to "prove" novelty.
3) Stopping the "patent trolls" [shell companies which no utility other than to litigate] is a completely other matter IMHO. The patent system is designed to protect innovation - I'm not sure it was designed to facilitate companies purchasing huge packets of IP protection to simply stiffle innovation. Llegislative restrictions should be placed around IP attaching to a product or service which looses it's enforceability if it abstracts itself too far from this product or service - meaning companies who acquire technology can continue to gain from their acquisition [product/service] as it is inferred they are in a similar field and therefore would satisfy any such test - but which ultimately means that trolls who acquire IP don't have anything to gain. They are too far abstracted which means their enforcability is dead.
i.e. A University developing and researching technology doesn't have an "abstract connection" to any patents they hold - they developed the technology and therefore have a direct association to the product/service - regardless of who they licence it to for further commercialization - arguably, without their initial cost of research and development the patent wouldn't exist and the world wouldn't benefit from their efforts. IF a troll acquired the patent and had no intention to develop the disclosed invention -they have no direct association and therefore it's useless. A "bad faith" test if you will.
"Trolls" - at least in my view - have no direct connection to the patentable subject matter - a so-called "abstract connection" - and therefore aren't actually "benefiting" from their effort in developing the patentable subject matter. They are essentially those entities who use patents purely as an enforceable mechanism in order to extract income from innovators and that's a sad by-product of patents unfortunately. In Trademark law, trademarks registered in "bad-faith" can be extinguished and it would not be that difficult to extend such a concept to patents such that owners who acquire patent rights in bad-faith [that is, purely to stifle innovation without any direct connection to the patentable material] - should loose them.
The revolving door between Government regulatory jobs and positions in the very businesses regulated is pretty well-established at this point. I'd like to see some evidence to back up that assessment with regards to the US PTO, but it certainly doesn't strike me as unlikely.
Wow. I know I was writing my own hand-rolled, wheel reinvented, linked list implementations, in C, well before 2006. Insanity. How can it not be obvious that the USPTO is broken with respect to software patents? Let's shut this madness down once and for all.
If patent workers are overworked, why don't we replace them with a certificate of origination. You file it, it takes two weeks, it's dated, and it describes the work. Then, if you need to sue someone, you then have to pay to prove that they are infringing. Reverse the burden of proof and delay the cost of approving patents.
Second, my biggest problem with software patents is that you can't build the thing from the description. If you look at a patent for a device, there are schematics and descriptions of how it works. Software patents just say "the user can click a button and it automatically buys a book". Where are the detailed drawings of how the web server connects to the payment system which connects to the distribution center? How do you scale that up to millions of users? That's where the magic is.
Patents are supposed to give the inventor a reasonable period of time to profit from his work. After that time period is over, it becomes public knowledge, so everyone can do it for free. Software patents do not give away the knowledge that it actually takes to build the system. Instead, they typically just describe the tiny bit that is apparent to the user.