Hacker News Reader

> The USPTO does not receive any funding from general taxation sources. Rather, the agency is fully funded by the fees paid by patent applicants and patent owners.

This is probably a big part of why the USPTO sucks. If they had the resources to hire dozens and dozens of industry experts, and pay them aggressively competitive wages, they could probably fix most of the problems most of us have with the system as a whole.

This is why I like the idea (from someone around here I think) that patent owners should state a value of the patent when filing, and fees should be a percent of that value. Then value of patent limits maximum damage for infringement. So the amount of damages that can be sued for depends on the amount of funded attention at USPTO on checking the patent is valid (and patent owners could have the option of increasing stated value later, in which case extra fees would be used to revalidate patent). But I guess the whole culture that the number of patents = amount of innovation in economy = success of USPTO would have to be battled.

> that patent owners should state a value of the patent when filing, and fees should be a percent of that value.

In that system, only huge corporations will be able to afford to file patents that are worth anything. Everyone else gets screwed.

Are you referring to this comment of mine?

"So I had an idea the other day for a patent reform: property tax on patents.

As long as a company wants to keep the monopoly rights over a patented invention, they are taxed a percentage of the patent's market value each year. They can choose either to pay that tax or sell the patent to someone else.

The government would offer to buy any patent for its market value, using tax revenue gathered from other patents, in doing so putting the invention into the public domain.

This would discourage companies from building large 'defensive' patent portfolios, since they'd be expensive to maintain. It works for the 'lone inventor' scenario too, since the market value of a new, untried invention would be low.

Once a patent's value is proven by developing the patent, its value will rise. At some point, the benefit derived from the patent's monopoly rights will no longer be worth the cost, and the rational thing to do is sell the invention."

http://news.ycombinator.com/item?id=4558402

possibly, doesn't seem to be exactly like I remember but is quite possibly my memory fail. Although it's not clear in your comment where the market value comes from (e.g. is it essentially the government is the market maker? that seems like a bad idea as central authority is unlikely to understand and price everything correctly). What stops people from understating the patent value?

EDIT - I guess main point is - is the patent value declared by owner and does the value affect ability to litigate?

One way you may be able to do it is to make patents always available for sale at their declared value. By declaring a patent's value, anyone (including the government itself) can offer to buy it from you for that amount of money. You must then either accept the offer, or block the sale by increasing its value beyond what the other party is willing to pay.

I'm sure it wouldn't be unlike assessors for home value property tax. A bit trickier, but I'm sure it's doable.

That also completely screws the little guy, which is one of the big reasons the USPTO is said to exist.

I would also love to have 1 sq mile of land in downtown Manhattan on which to develop a high-rise.

But I'm just a little guy. The darn property tax system is set out to screw me.

This is really the whole reason that corporations were invented: so that people could pool their resources together to achieve things that a single person could not achieve. In fact, they used to just be temporary and when the original stated project on the charter was complete then the corporation disbanded.

If a little guy company wants to own a patent under this scheme either a) they have to file the patent with a lower value so they can make the annual "property tax" but that may cap their wins in the future if it's infringed on. or b) get funding from investors to make the property tax payments. Or c) sell or license the patent to a larger entity that could benefit from it. In the case of (c) if it's truly as valuable as they say it is to the patent office then they should be able to find a suitable buyer.

Flat capped taxes on things produce market inefficiencies such as you see in the domain name arena: It costs very little to hold onto things that might be valuable one day. If instead there was a sliding scale to renew a domain based on it's value then you wouldn't really see domain squatters. They'd be forced to sell their domains to someone that could make more productive use out of the resource. Same with land and property taxes. And maybe the same for patents? I kind of like this idea.

@andrewcooke: There is an analogous situation to property taxes though -- Rezoning a property. If you can acquire land zoned for something cheaper like residential or agriculture and get the city to re-zone it to commercial it's almost always worth a lot more. Overnight the property is worth a lot more and the property taxes will jump, so you'll either need to get the money to develop it yourself or sell it off to a developer who can do so.

And as far as the cost of entry, there are countless fields where patents could only conceivably be filed by someone with a lot of capital: auto, aerospace, medical, pharma, etc. There are certainly lone inventors working in these spaces trying to file patents but it's unlikely they're working on stuff where you need access to enormous wind tunnels or a medical testing population.

EDIT TO ADD: And aren't we as software developers going on and on about how software patents are worth anything? The cost of entry isn't very high at all and therefore they'd be worthless. Patents were supposed to protect the little guy from losing his large up-front research and development costs to the big established guys.

Patents were supposed to protect the little guy from losing his large up-front research and development costs to the big established guys.

I always understood the purpose of patents was to provide incentive to bring trade secrets out into the documented open for the long-term benefit of society by having a great body of knowledge that anyone can use. So understanding of great inventions didn't disappear with the inventor. A short-term exclusivity on using the technology was the way to provide renumeration for being willing to share your discoveries with the public.

Rubber band scrolling is a cute, even useful, discovery, but it seems like it could be duplicated by just about anyone without any knowledge of what would otherwise be trade secrets. Is that worth documenting for future generations in this way given the high social costs of taking such documentation?

Hmm, my comments don't really reflect this, but I agree with you. I think the end result of a system like this would be that software patents would be next-to-nothing worthless and people just wouldn't bother, except perhaps in the cases of truly innovative works of software (I can't even think of any worth patenting right now...).

Would Apple really file their rubber-band scrolling patent at a value of $1bil, then pay some multiple of that per year when it's so easily avoided by their competitors? Probably not.

And if they do then more power to them. But it would put an upper bound of the amount of silly patents that a single large company could file and maintain (think IBM, MSFT, Apple, etc). Right now it costs a company hardly nothing to file all these and maintain them but they have a huge potential upside if your competitor steps on that patent landmine.

a little guy can't buy a plot of land in manhattan, so you're not comparing like with like: it's not the taxes that stop you from buying land in manhattan, it's the original cost of entry.

there isn't the same cost of entry for having a good idea.

We're talking in the context of a court case involving two enormous multinational corporations going at it for over a year and a half with probably another half-year to go, minimum. The "little guy" can't even afford to sit in the bleachers of this fight. If some "little guy" actually had filed these patents he'd have been bankrupt a year ago.

There's probably a patent regime that actually would protect "the little guy" but it's long past time to let the USPTO wear that fig leaf over the current system. (Which is to say, I'm not saying you're wrong, I'm more saying, your point is inapplicable. And that's a shame.)

I don't think there is any basis for your assertion that patents don't protect the little guy. I worked for two small companies that had, used, and licensed patents on their key technologies. These technologies took us millions of dollars and years to develop, and were our way of competing with companies that could out-manufacture us, but not beat us at the R&D game.

If you had to pay up front on the "value" of the idea, a little guy (be it a solo person or small company) could never file a truly novel patent. That would be even worse than the system now where you could patent it but not afford to litigate (but you would be able to sell it or if it's a great case find someone to take it on contingency).

In my mind, the little guy would patent with a small declared value - so on the downside they wouldn't be able to litigate for large amounts. But they would have a defensive patent, and that patent would also have value for larger acquiring companies because if they believed in the validity of the patent they could apply for revaluing at higher price (at which point there would be extra funding to make sure it is valid).

If it's not a valuable patent then big players won't respect it (cheap to litigate) and who will want to acquire it? You'll just be run over.

Yes, while it's low-valued it's not much of a deterrent to big players which is a disadvantage (depending on how often small players realistically can afford to challenge the big players legally under current system).

But by defensive usage I mean then at least you know someone else isn't going to patent your key stuff. And once you've got funding/get acquired, the patent could be declared as higher value (and additional fees paid/validation done), and then future litigation to anyone who infringed after that date could be for higher amount.

I have suggested this before and will do so again:

Just like ROTC, Department of Education can create an Engineering Corps that allows tech students that receive government loans to pay back their debt in form of service.

I'm generally all for anything that involves government paying for college, but I don't think you're solving the problem in this case. The problem is that good engineers are expensive. Median base pay is what, about $90-100K? Which the patent office isn't willing to pay.

What you're saying is, de facto, that they should pay examiners more -- enough to be competitive with that -- in the form of student loan repayment. But then where does the money come from, and if it's available then why not reduce turnover by just paying examiners that much whether they have student loans or not?

I'm not talking about employment. It is the 21st century for God's sake, and we have the internet. )) For example, USPTO can setup a review system in cooperation with the top tier schools. Students can serve their country while they are earning their degree.

The patent system can be made more reasonable if we address the problem of bogus patents issued by an overwhelmed USPTO. This would be a step in the right direction.

All that said, if tech heads aren't willing to serve their country, even for a one lousy year, they really should stop complaining about the broken services.

Except one of the claims that I've seen about the US patent examiners is that US patent examiners are likely to leave after only a few years, whereas the EU patent examiners are far more likely to be lifers. Bringing in a ton more people with no experience would only make that much worse.

Or we could just let them keep all the money they collect in fees, and let them set the fees at a level that would let them afford to do a good job.

No, that won't work - the USPTO is already a profit center, so why aren't they doing a better job right now?

Being paid more to approve more patents is a perverse incentive so egregious it's incredible the system was setup like it was.

No matter what the fees are, decreasing the percentage of patents granted will effectively cut the budget of the agency. Even upper management that is sympathetic to the cause of increasing patent quality is going to find raising standards very painful.

Their budget needs to be completely decoupled from their revenue.

Nah, they're just doing it wrong. Let it work like this: You file a patent application, you pay all the fees you'll ever pay to the patent office, up front. No refunds if your patent is not approved, and the number of appeals/filings/whatever is limited to prevent large corporations from throwing lawyers at denials until they're approved.

Then they have an incentive for more applications, but little perverse incentive to approve them.

Of course, if they get a reputation for denying spurious claims then they won't get as many applications from spurious applicants, but that's true regardless of their funding source -- you can't produce a sensible budget without considering the number of applications they're expected to process. If the PTO's goal is budget maximization then they'll want to maximize the number of applications, which means providing incentives (in the form of low standards for patentability) for applicants to file more applications. You can't really fix that without creating some powerful incentives for not approving bad patents, like fining the head of the PTO $100K personally for every patent the PTO approved and the courts have to invalidate.

This is rediculous if true. Clearly, patents are important public policy. They should be funded as such, and if need be, taxes from profits of patent-holding companies should be dedicated to the USPTO. Taxes from fees (1) dont scale; and (2) are orders of magnitude off the economic impact, so any distortion from bad decisions is also.

I'd love to see some sort of punitive damages if a patent you sued over gets invalidated. At a minimum, Apple should need to compensate a portion of Samsung's legal fees.

Even if you stand against the current patent system, it's how it works right now, so that suggestion is a little bit of blaming the victim. Apple were told that they have a valid patent by the office which makes those determinations. How are they to blame for accepting grant of the patent?

Apple knows as well as we do how broken the patent system is—and they take advantage of that brokenness.

In no way are they a "victim."

I don't think that makes sense. It's not Apple's fault that the US patent office granted the patent. Just because they changed their minds doesn't mean Apple is responsible. If anything, the US patent office should compensate both Samsung and Apple for essentially wasted legal fees.

Disagree, it actually may make sense. The office is overwhelmed with patents and does not have the resources or people that are experts in that particular field. The author of the patent is supposed to be the world's best expert in their field (they are claiming to have crossed the boundary of known things with their patent, aren't they?) and it should be far easier for them actually check if there is prior art or the patented thing is actually not obvious to anyone, who is explained the problem that it is supposed to solve. Many low quality patents are filled and there is no risk in filling them. There should be a risk and a penalty for misusing the system.

This is my thinking. The guys pursuing these patent cases typically are well aware that their patents are not non-obvious, and quite often aren't even novel. Lodsys is my typical whipping boy; if they weren't shocked that "click for upgrade" got through the patent office, it's only due to their familiarity with patent trolling. They should suffer damages for using a ridiculous patent in the first place, and they should pay the court costs for the defendants. I'd like there to be an "invalidation with prejudice" that would allow this, if the court decides that the patent is outrageously improper. I think that would both discourage trolling and encourage defending against obviously invalid patents. Otherwise, it's almost always better (economically) to suffer the extortion than to try and fight it.

Well, that depends. If the patent was found to not be valid because of invalid intent by the "inventor" (such as known prior art, knowingly stealing an idea, knowingly patenting something that's unpatentable), it actually is the inventors fault... When submitting a patent, the inventor must sign an oath indicating these (and other) things:

> The inventor must make an oath or declaration that he/she believes himself/herself to be the original and first inventor of the subject matter of the application, and he/she must make various other statements required by law and various statements required by the USPTO rules.

Now, it would be VERY difficult to prove the oath was broken. But if it was, it's completely and 100% valid to hold the inventor (here Apple) responsible.

I'm not saying that's what's going on here (or in most cases of an overturned patent), just that it may be Apple's fault. It's not a black and white situation here...

I am not sure about legal fees, but at the very least, I think it's only fair for US patent office to refund patent fees for invalidated patents. Won't happen though.

Unfortunately I just don't see the motivation for them to not issue patents... maybe Samsung should sue them for essentially giving Apple the ammunition to put them through the year or so.

I presume that that's not possible, but still, maybe it would give their collective heads a shake.

If nothing else, delay all court proceedings until any contested patents have been reviewed.

"Successful reexaminations should be a rare, exceptional case in a properly functioning patent system."

Unfortunately some 70% of reexaminations are confirmed to have some sort of issues. This is very shaky ground to be used to routinely afford millions/billions in damages and effectively "lock up" ideas that are obvious or unpatentable.

edit: add source http://ptolitigationcenter.com/essentials/common-questions/#...

Perhaps you are right but it is much better to have the reexamination process than not to have it. There were too many computer programming patents too easily granted in the nineties and the 00s. That is a fact it already happened so saying "it would be better if this hadnt happened" is completely pointless now.

It is a good thing that a reexamination process exists to address bad patents when they are the focus of an expensive lawsuit.