Hacker News Reader

In fact, I have yet to meet someone who opposes the current patent system who understood what patents are.

Are you talking about meeting someone on HN? How about HN user grellas, a Silicon Valley business lawyer? I doubt his arguments against the status quo are made in ignorance:

http://www.hnsearch.com/search#request/all&q=grellas+pat...

Edit to add:

Without patents, technological progress would be severely restricted-- not because there "would be no incentive to invent" (a straw-man that anti-IP people seem fond of knocking down) but because there would be vastly less communication and understanding of how other people's inventions work.

It is difficult to accept this argument when the majority of software patents we see in lawsuits are written as obtusely as possible. Software patent language is the antithesis of "communicat[ing]... how other people's inventions work."

Your response is exactly what I'm talking about. You believe, but you don't know. You can't make an argument, you point to someone else, and advocate faith.

Imagine if a religious person had used that kind of "argument" to you, and referenced a priest regarding the law?

It is ok to not know. And it is ok to take things on faith. The problem is, when people think that their faith means they are right and people who make arguments they can't counter are "Wrong" because it goes against their faith.

Claiming that nobody who disagrees with you knows what they are talking about is also a symptom of religious thinking. You sound exactly like your description. Have you made any arguments yourself that don't beg the question by saying that patents are good because they were intended to be?

You said you've never met someone who knows what patents are and still opposes them, implying that nobody who knows is opposed to patents. I provided a proof by counterexample of someone it would be difficult for you to argue doesn't know what he is talking about.

And you haven't. Instead you claim that I don't know what patents are. I do know the difference between a design patent and a utility patent (and a trademark, and copyright). I've seen good and bad examples of each. Yet I still oppose software patents, support reduced terms for all of the above except trademarks, and want to see a much higher burden of proof for a patent to be granted, and higher still for a lawsuit to proceed. I also argue that independent invention should be a defense against patent infringement that doesn't require an army of lawyers to raise. Yet somehow you will try to claim I don't know what I am talking about.

The real test to distinguish rational from religious thought is the willingness to change one's mind. I'll tell you what would change mine: a plethora of incontrovetible, scientifically sound studies that prove that the harm done to and risk imposed upon small innovators and market competition is worth the benefit to society; or a solid implementation of the changes I'd like to see, that subsequently results in a collapse of innovation (collapse of any business alone does not count as a collapse of innovation). What would change your mind?

Congratulations. You've finally convinced me that an ignore list would be a good feature on HN.

Listening to you rail against single-mindedness and groupthink when you're probably the single most self-righteous and closeminded and ideologically intolerant frequent poster on HN is surreal. It's really not easy to find people that are so reliably certain they have the answers to everything. You're certainly the first self-professed anarchist I've met that's a staunch proponent of patent and contract law.

The difference between me and you is, I make arguments to defend my positions, yet all you did is characterize me in a derogatory way.

You are free to ignore me anytime, you don't need a new feature for that. Or you can reply to me with arguments, but you don't.

In making this choice, you impeach every claim you make against me.

You will only brush off any rational argument I care to make.

But let's play: you claim to be an anarchist yet are strongly in favor of patent and contract law. Explain how you can reconcile these two beliefs. Who do you imagine will enforce these laws and how?

I'm not an expert on patents in the least, but the system feels broken when a company can be awarded $1B from a patent infringement lawsuit.

From what I've read (I'll add sources as I find them), the patent process takes more than two years on average to get approved, patent examiners are given patents that they don't have domain expertise in, and they don't get enough time to review patents.

It seems, on the outside, that lots of money and time is spent on patent lawsuits and purchasing patent portfolios and none of it really leads to the advancement of anything besides lawyers and lawsuits.

Do you think the patent system is fine how it is, or is there reform you would like to see?

Here's a counter example: ARM.

In ARM's case, patents serve a valuable purpose: they allow a separation of concerns. The company that designs the microprocessor does not have to be the company that builds things using the microprocessor. Allowing ARM to focus on what they do best--designing microprocessors, and allowing say Samsung to focus on what they do best, manufacturing things, helps everyone. The patent in this case is simply a legal device that allows people to transact in terms of a design.

Also, looking at the outlier damages award from a patent infringement isn't really meaningful, in the same way that looking at the damages award for medical malpractice isn't really meaningful. Occasionally, you'll have a jury that awards someone $10m for a stubbed toe, but if you look at comprehensive studies of medical malpractice suits you'll see that: 1) most of the time the jury is right; 2) the actual overhead of the system is 1-3% of a $1.5 trillion industry, which is tied to a profession that is otherwise almost entirely self-regulated. People tend to focus on the damages awards and think "oh that's costing the economy money" but obviously damages awards just shift money from one place to the other. The "cost" is in the litigation fees and how the law changes peoples' incentives, and that's much harder to evaluate than just looking at a big $100m judgment.

This doesn't make sense to me. Are you saying that, without patents, ARM couldn't license it's designs to Samsung? Why not?

Or are you saying ARM wouldn't be able to stay in business because Samsung would just steal their designs? I don't think that is true, for the exact "expert" reasons you mention: ARM is great at designing processors. If Samsung wants reasonable updates to stay competitive with other manufacturers, they would need to pay for those experts.

Both.

Consider how ARM might initiate a licensing transaction with Samsung. Don't you think Samsung will want its engineers to go over the design extensively to see how it'll work with their manufacturing process, etc, before committing to buy? Without IP protection, how does ARM give Samsung the blueprints without giving up any leverage they have in the transaction? What's to stop Samsung from just taking the blueprints and not paying? What keeps Samsung from turning around and selling the blueprints to other people at a lower price?

Moreover, sure ARM might be better at designing microprocessors, but once ARM does the heavy lifting, it's much easier for Samsung to hire its own people to do updates. They might not be as good as ARM's people, but its hard to argue with free.

I worked for two tech companies that sold only IP. We were in the business of developing the software that was incorporated into heavy-duty networking equipment and military radios. Our customers were companies like Raytheon, etc. Our competitive advantage was our expertise and our size. We could do experimental stuff much quicker than the big companies with more internal bureaucracy and more risk aversion. But there is no way we could match their manufacturing, distribution, and sales expertise. That wasn't up our alley. It's very difficult to conceive of working relationships like this without being able to transact in IP. The type of work we did was kind of like crypto in the sense that it took years and millions of dollars to figure out what worked, but it was pretty straightforward to verify the working solution. Without being able to patent the product of that research, we'd have no leverage against the companies that had the big manufacturing muscle. As soon as we showed them our designs they could just run with it. If we shipped them binaries, they could easily just reverse engineer them.

I think peoples' opposition to IP really stems from the fact that most of the patents people are exposed to are pretty crappy. It causes them to think of the average software patent as something they could stumble onto accidentally. But that's more of an opposition to the bar used for software patents rather than an opposition to the concept. Realize that highly complex inventions like OFDM are the subject of patents too.

Contract law is perfectly adequate in that situation - Samsung signs a contract with ARM which specifies the exact terms under which they can use ARM's designs.

It's not perfectly adequate. E.g. ARM can't let Samsung see the design until Samsung signs a contract. It makes it difficult to shop the design around. Also, a contract that provided the protections similar to a patent would be an ad-hoc informally specified version of patent law, and we'd replace patent litigation with more contract litigation.

It also provides no recourse against third parties. What if someone reverse engineers the design from one of Samsung's products? This is not theoretical--there's a major Chinese networking company that used to copy the designs of a major US networking company down to the silk screening. It also prevents ARM from publishing details in trade journals and forces them to keep documentation and designs under lock and key. It makes industrial espionage tremendously more valuable. Sub licensing, transferability, etc, become a huge hassle.

And to what benefit? Consider one suit wildly maligned on these boards: Oracle v Google. Without copyright, that lawsuit still would have happened more or less along the same lines, except it would be about whether Google violated its contract with Sun about how Google was allowed to use Sun's code. And t would have been longer and more expensive because the parties couldn't speak the common language of contract law.

It's not perfectly adequate. E.g. ARM can't let Samsung see the design until Samsung signs a contract. It makes it difficult to shop the design around. Also, a contract that provided the protections similar to a patent would be an ad-hoc informally specified version of patent law, and we'd replace patent litigation with more contract litigation.

It also provides no recourse against third parties. What if someone reverse engineers the design from one of Samsung's products? This is not theoretical--there's a major Chinese networking company that used to copy the designs of a major US networking company down to the silk screening. It also prevents ARM from publishing details in trade journals and forces them to keep documentation and designs under lock and key. It makes industrial espionage tremendously more valuable. Sub licensing, transferability, etc, become a huge hassle.

And to what benefit? Consider one suit wildly maligned on these boards: Oracle v Google. Without copyright, that lawsuit still would have happened more or less along the same lines, except it would be about whether Google violated its contract with Sun about how Google was allowed to use Sun's code. And t would have been longer and more expensive because the parties couldn't speak the common language of copyright law.

With the important difference that a truly independent reimplementation is not covered.

It also provides no recourse against third parties. What if someone reverse engineers the design from one of Samsung's products? This is not theoretical--there's a major Chinese networking company that used to copy the designs of a major US networking company down to the silk screening. It also prevents ARM from publishing details in trade journals and forces them to keep documentation and designs under lock and key. It makes industrial espionage tremendously more valuable. Sub licensing, transferability, etc, become a huge hassle.

Is this US networking company you speak of still profitable? Looks like ARM are fairly heavy-handed with defending their designs at the moment, so you'll forgive me being skeptical of patents encouraging the sharing of IP: http://www.eetimes.com/electronics-news/4042770/Student-s-AR...

How about a sitting judge on the Seventh Circuit Court of Appeals, and one of the most respected and cited voices in intellectual property law? Judge Posner [1] http://www.fool.com/investing/general/2012/07/15/can-this-le...

Your 'publication of ideas' argument applies only to things that can be hidden.

But the patent system applies to a super-set: things that are plain, too.

Therefore, by your own argument, the patent system is operating beyond its justification, and so could indeed be described as 'broken'.

In any case, the patent system is a trade-off -- of increasing production against restricting access (see: the standard economic model of it). If it is badly calibrated, the net effect is negative. And the current state of economic knowledge is that no-one knows whether it is set correctly (see: 'The economic structure of intellectual property law'; Landes, Posner; 2003. Conclusion, p422, s3.). When there is a whole big system, of various obvious costs, and you do not even know the net effect is positive, the sensible thing is to stop doing it (probably by phasing out), until something proves otherwise.

The patent system doesn't apply to things that are "plain", though this word is a bit broad, so it isn't clear if you mean it the way I take you to mean it. Further, I didn't say that this was the justification for it, I said that this was one of the advantages of it, and thus, even if you disagree with the advantage, that doesn't ipso facto, make the system broken.

>When there is a whole big system, of various obvious costs, and you do not even know the net effect is positive, the sensible thing is to stop doing it (probably by phasing out), until something proves otherwise.

Your argument also applies to government, in fact, from an economic perspective, government has a clear negative return (to the tune of $70 Trillion in the USA roughly).

Get rid of the patent system by getting rid of government, and you'll have my agreement.

If you think it is ok to have government but not a patent system, then there's contradiction in your positions.

USC 35 is freely available on the internet. You would be making a mistake to think that all the people who disagree with you do so because they know nothing about it. Especially when there are more than just design & utility patents out there -- http://www.uspto.gov/web/offices/ac/ido/oeip/taf/patdesc.htm

To be fair, most of the "trolls" that are abusing the system readily admit to not really understanding it either. They pick their targets without understanding the business of those targets solely based on the fact that they "look like companies we usually sue". They often name multiple patents with dozens of irrelevant claims hoping there's something in there that will stick.

Clearly something in there is broken. I think the real contention on HN is whether or not there's a big enough baby in all that bathwater to be worth saving.

> A lot of very vocal HNers like to go on ideological rants

Quite.

> because there would be vastly less communication and understanding of how other people's inventions work.

You must be familiar with the term 'reverse engineering'. There isn't much that can't be unwound. If secrecy of how things work was an effective protection there wouldn't be a patent system to begin with. Once you release something 'how it works' will be evident to anyone determined enough to find out.

People on HN have this view that the whole world is software. Reverse engineering may be practical and indeed trivial in software, but aren't for the circuits buried in an F-35. And publication is tremendously valuable for inventions that have cross-disciplinary value. Not many people working on artificial hearts would have the time or think to reverse engineer mechanical equipment used in a factory, even if there might be applicable inventions.

Reverse engineering has been applied to all the fields that you cite. The F-35 is a funny example in the sense that if there is one item that would be reverse engineered in its entirety down to the last transistor then it probably would be that plane if one accidentally made its way to China somehow. Nothing motivates more than a perceived gap in weapons systems ability. Reverse engineering of electronics is a solved problem, even though it may be a lot of work.

Is there some proof that someone got a useful idea on how to make an artificial heart out of a patent published about mechanical equipment used in a factory?

If you want to keep something secret, keep it to yourself. Monkey see, monkey do!

It has been applied, but it's much less practical in those fields than it is in software. I'm just saying people on HN tend to see everything through the lens of software, and other engineering fields aren't the same.