Software Patents Compared to Hard Patents

Arie writes "The Slate discusses the obvious differences between patenting an algorithm and a drug. The article introduces the Fence test, which basically says that if you can physically protect your property, you have a case patenting it. In addition, it claims that the burden on a programmer identifying whether he is infringing on a patent or not involves excessive research burden, essentially to the inherent lack of physical boundaries. Obviously the article starts off with mentioning the patent dispute between RIM and NTL."

Comments

Investment, risk, compensation

[Toby The Economist]

The original and valid purpose of a patent is to enable people who make investments in research to be compensated for the risk they take.

This - investment, risk, compensation - is the issue.

Whether or not a fence goes around the concept patented is utterly irrelevent.

If a fence test was implemented, all investment into non-physical research would be discouraged.

Re:Investment, risk, compensation

[saxmanb]

The original intent for granting patents was to encourage people to invent things. This has nothing to do with protection of investments. All investments contain risk. If you invest in something that isn't patentable or is patentable but not profitable, then that's just tough. Welcome to captialism and the free market.

Re:Investment, risk, compensation

[Anonymous Coward]

"The original and valid purpose of a patent is to enable people who make investments in research to be compensated for the risk they take."

Wrong. Government should not, and does not, play a role in determining compensation for the risks taken by individuals or corporations.

Government does, however, play a legitimate role in establishing policies that serve the public good. In the case of patents, the public good is served by encouraging the publication of innovations that might otherwise be kept secret. It is publiction that is rewarded, not the innovation itself. Innovations may be kept secret, rewarding only the inventor, and that's fine. Making innovations available to the public is the point of patents.

The key question for patent policy, also applicable to copyrights, is whether the policy encourages making innovations available to the public. It's not about the inventor or author, it's about the public. Is the public served by granting the inventor or author a monopoly? If yes, then it's good policy. If the public is harmed, for example by software patents, then it's bad policy.

Re:Investment, risk, compensation

[kfg]

The original and valid purpose of a patent is to enable people who make investments in research to be compensated for the risk they take.

Well, no, not exactly. People already made investments in research and were commercially compensated for it before patents.

The original and valid purpose of a patent is to force them to reveal their research to the public in order to gain government protection of it.

Whether or not a fence goes around the concept patented is utterly irrelevent.

Actually, this is exactly the test applied by the founder of the American patent system and its first examiner; Thomas Jefferson. For things which you can't put a fence around there is copyright.

Patents are not ideas, they are for things.

KFG

Re:Investment, risk, compensation

[kfg]

That's completely wrong. The subject of any patent, and the core protected aspect, is an inventive and novel concept.

For which a working physical model can be submitted.

F=ma cannot be patented because it is a "pure" idea. If I tell it to you you walk away with the novel concept in its entirety and can likewise transmit to another.

A circuit board may have have a novel concept behind it, but it is the board that is patented. Something I can actually steal and physically copy. I am not allowed to steal or make a copy of it, but I can talk about it all I want. In fact, to obtain a patent in the first place I must publish everything anyone needs to know to talk about it and there are no overt restrictions on such speech unless I sign some sort of NDA.

The plans for the board, which are just much the novel concept as the board itself is, are covered by copyright, not patent, because they are a publication, not the actual device.

Only the device is traditionally covered by patent, and algorithms have only obtained patent protection by claiming that when implimented by a computer they are part of the device.

KFG

So if the boundary for patent legitimacy...

[msauve]

is to be "things you can physically protect," how does that jive with the fact that completely independent discovery/invention can run afoul of existing utility patents?

It seems to me that the most basic problem with the patent system is that patents can be written in language which no one, even if "skilled in the art" (as is supposed to be the case, but obviously isn't) can clearly understand unambiguously. That causes at least two problems - patent examiners are left befuddled, leading to undeserved patents issued. Secondly, and even more importantly, since patents are supposed to force disclosure to benefit the public (after the patent term expires), having a bunch of claims buried in obfuscating language defeats the purpose.

In the news..

[db32]

Today slashdotters were shocked when another tech writer repeated the idea that software patents are bad in yet another way. I think most people here, and throughout the industry, already know the multitude of problems with software patents. I don't think the problem is convincing people on the working side of the industry that they are bad. The problem is convincing the people profiting from software patents that they are bad. I would be interested in seeing industry leaders that can actually influence the decisions, or some politician types with the power speak out against software patents, but seeing yet another tech person do it is just getting redundant. You have to convince people that really like their money, why they shouldn't make money the way they are now. Until things get so bad that the top starts feeling the pain I don't things are going to get much better. Melancholy Elephants is an excellent short story by Spider Robinson about the end game of this situation we have these days. (It isn't terribly long, and it really is a good read) Enjoy!

Barriers to entry

[Varitek]

The basic difference between drug patents and software patents is the barrier to entry. You can buy a $200 computer, a $50 book on programming, and be infringing on software patents later that evening. It also costs nothing to release that infringing code to the world. The only people likely to be infringing on drug patents, on the other hand, are well-capitalised pharmaceuticals companies, who can afford to research patents.

This is why software patents "feel" different to the Slashdot audience. None of us think we will infringe on a patent for an anti-depressant, but we don't like the idea of infringing on some obvious patented algorithm just by writing a few lines of code.

Re:Barriers to entry

[bitspotter]

I tend to think the cost of innovation should be a major criteria for patentability. Granting monopolies on the deployment of innovations only encourages innovation when the innovation would otherwise collapse for lack of R&D funding. Such a case would indicate that the R&D is too risky for the current state of the market unless a patent can be granted to exclude competition.

This makes some sense for things like drugs that require large investments to develop. Take away the patents, and, yes, you do solve the generics problem - but you dismantle the incentives for private companies to develop these medicines.

For software, on the other hand, this is ridiculous. the costs of developing software are pretty trivial. This results in an innovation explosion for very little money. granting monopolies does nothing to encourage innovation, because the costs of such innovation are not too risky for the market - they're cheap! In this kind of environment, patents actually backfire and //retard// innovation instead, by resitricting the ability to improve existing technologies.

As technologies of all kinds develop, costs continuously decrease. Periodically, patent adminsitrations should re-evaluate the costs of innovation in each market sector, and think about changing the kinds of innovations that are permitted patents based on the costs of innovation in each segment, rather than carte blanche.

Re:Wrong

[Ekarderif]

Actually, the point of the patent was to increase information sharing. A patented product (at least in the US) must disclose all the steps necessary to create the end result; the owner then owns the exact method for a few years (which he is capable of doing anything himself but anyone who uses his method must receive his permission). This was implemented to counter trade secrets (a la Coca-Cola) by providing incentive (exclusive rights for a short period of time) to those who open up their innovations; reverse engineering a trade secret is perfectly legitimate (Pepsi), but reverse engineering a patent means nothing since everybody already knows how it's done.

The problem isn't the idea of patents. They exist for many reasons and have proven to work immensely. No, the problem lies in software patents. I'm not against software patents in general, but 20 years is far too excessive. If we reduct it to one year or some equivalent, there wouldn't be such fiascos. The software world changes quick, and thus so should the laws protecting them.