NoSoftwarePatents.com Industry Campaign Launches

Halo1 writes "The NoSoftwarePatents.com campaign has officially launched today. It has industry support from 1&1, Red Hat and MySQL AB. The website is already available in 12 EU languages (more to be added soon), and contains a ton of information about the dangers of software patents, including the myths that surround them. Hopefully, more large companies will join this campaign in the future."

Comments

Main Page

[zoobab]

Would be nice if this article can move to the main slashdot page, and does not stays only in the YRO section.

Re:Main Page

[Halo1]

Someone can simply submit it again when CmdrTaco is editor on duty, so it can be posted as a dupe :)

The History of Software Patents

[Mstrgeek]

This is a great site dealing with The History of Software Patents

http://www.bitlaw.com/software-patent/history.html

It has some great information hope you enjoy it and find it to be a help in supporting the need for software patents

Re: The History of Software Patents

[EzInKy]

The Supreme Court stated that in this case, the invention was not merely a mathematical algorithm, but was a process for molding rubber, and hence was patentable. This was true even though the only "novel" feature of this invention was the timing process controlled by the computer.

Hopefully the Supreme Court today realizes the ramafications of the decision they rendered over twenty years ago. Here's an excerpt of the courts opinion:

Although their process employs a well-known mathematical equation, they do not seek to pre-empt the use of that equation, except in conjunction with all of the other steps in their claimed process. A claim drawn to subject matter otherwise statutory does not become nonstatutory simply because it uses a mathematical formula, computer program, or digital computer.

And how the Patent Office construed that ruling as meaning that any "invention" with the word computer in it gets a rubber stamp is absolutely mind boggling.

Re: The History of Software Patents

[Alsee]

http://www.bitlaw.com/software-patent/history.html
supporting the need for software patents

Buahahaha. I don't think there is anything in there supporting software patents. It merely documents how the 5 member majority on the Supreme Court screwed up in Diamond v Diehr when they ordered the US patent office to issue an invalid patent. The entire case was about the patent office rejecting it as invalid.

If you actually read Diamond v Diehr, the 5 member majority supporting such a patent discussed at great length that they did *NOT* intend the ruling to lead to the patentability of software. The other four members wrote a dissent explaining exactly how the majority screwed up. The majority essentially ruled that 2+2=5 and claimed they had absolutely no intention of allowing 2=1. However once you allow 2+2=5 it is impossible to aviod a sequence of valid steps reaching 2=1. Our current software patent system is exactly what the 5 member majority stated they did not intend to permit.

There is no "need" for software patents. Software is protected by copyright. Why the heck should software be the only thing on earth with broken double coverage and broken double restrictions of both copyright and patents?

-

Re: The History of Software Patents

[tambo]

There is no "need" for software patents. Software is protected by copyright. Why the heck should software be the only thing on earth with broken double coverage and broken double restrictions of both copyright and patents?

If either form of protection should get canned, it's copyright. A functional set of instructions, or a work of art - which better describes software? In which context is it overwhelmingly applied?

I'll give you two other reasons for software patents over software copyright:

• Patents expire 20 years after filing. That term is locked in, and it is virtually unmovable. Copyright in America lasts up to 95 years after the life of the author, and continues to grow (thanks to a corporation-owned Congress.)
  
• Let's say you spend several years of your life designing a new compression algorithm, and that it is completely awesome. Now, which would you, as its inventor, prefer: to protect any use of that compression algorithm, or only to protect your particular implementation of it - i.e., your code? Someone can take your C# code, or even assembly, and re-implement it in Perl or Python or Visual BASIC or whatever, and you, the inventor of the code, have basically lost all of the commercial value of your work.

(I fully expect the responses to include: "But software wants to be free, maaaaaan! Patents are just ways that The Man oppresses us coders!" I'll address those in a separate post.)

- David Stein

Re: The History of Software Patents

[tambo]

And how the Patent Office construed that ruling as meaning that any "invention" with the word computer in it gets a rubber stamp is absolutely mind boggling.

Huh? I don't think you're understanding the history.

The court's ruling was that patent claims for algorithms could not be categorically rejected as non-statutory subject matter - i.e., as non-patentable inventions. In addition to all of the other tests of patentability (novelty, utility, non-obviousness, etc.), the algorithm also had to have at least one of two connections to the real world:

• It had to have a "real-world impact" (i.e., it controlled a machine), and/or
  
• It had to operate on "real-world inputs" (i.e., sensors.)

And the USPTO consistently applied those restrictions. That made sense, since the USPTO fiercely opposed the concept of software patents in the 1990s, even after the State Street Bank decision that overrode their opinions. So it's unfair to state that the USPTO construed the court opinion with overbreadth; on the contrary, it sought to limit it as much as possible.

- David Stein

Re: The History of Software Patents

[Alsee]

If either form of protection should get canned, it's copyright.

So long as software patents exist I would not oppose you if you want to fight for exempting software from copyright. It's a dumb idea, it would be seriously broken, but it would be an improvement and less broken than the US's current broken double coverage.

which would you, as its inventor

Since when have mathemetians and programmers called themselves "inventors"? As a programmer I am a software author. Every programmer I know of considers themselves authors. Should Isaac Newton been granted a patent for "inventing" calculus?

I fully expect the responses to include: "But software wants to be free, maaaaaan! Patents are just ways that The Man oppresses us coders!" I'll address those in a separate post.

Go hang your strawman in a cornfeild.

I could quite easily go on about how software patents are harmful, but there's no point. Arguing software patents are bad is just as pointless as arging they are good. Such arguments are entirely moot - software patents are simply invalid.

First I'll cover how we got software patents, and that will lead into why they are flat-out invalid.

The US and the entire world all agree that algorithms are not inventions and are not patentable. Math is not an invention. Yes, the US still says that algorithms are not patentable.

The problem is that in a 5-4 desision in the 1980's the Supreme court goofed and changed the rules for evaluating patent applications. It was Diamond v Deihr. The US patent office argued that the patent application in question was not an invention and was not patentable. Four Supreme court justices gave a detailed and correct explanation of why the application was not a valid invention/patent. The four Supreme court justices gave a detailed and correct explanation of how the five member's order to issue that patent was broken, and how such a ruling would blow the doors wide open to invalid patents. The five member majority even gave an extensive discussion of how their ruling was *not* intended to open the door to software patents!

Well, the five member majority was blatantly mistaken. Their redefinition of evaluating patent applications was indeed broken. The patent office and lower courts had no choice but to follow that broken majority definition. Under that broken definition someone could get a patent on a non-invention merely by playing WORDGAMES in drafting the patent application. They had no choice but to start issuing software patents. Patents which the five member majority *explicitly* stated in their ruling that they had no intention of permitting!

It was a screwup. A screwup that went contrary to the entire history of US patent law, contrary to the consistant and proper rulings of the US patent office, contrary to the patent standards of every country on earth.

Prior to that broken ruling the US patent office consistantly and properly rejected any attempt to patent software or algorithms. They did so based on the Mental Steps Doctrine. That doctrine states that you cannot get a patent for mental steps, and that includes math and calculations. All softwarer is nothing more than mental steps. You do not need a computer to run software. Software can be "run" within a human brain, in pure thought. In principal any software can (eventually) be run mentally. Not only that, but there are many software patents that can in fact be carried out mentally within a matter of minutes.

Lets take a list of all supposedly valid software patents. I then look through that list and select the most convient "easiest" one. I then proceed to preform a live demonstration and in fact carry out that patented process purely mentally. Can you seriously claim that certain thoughts can be against the law??

And if you try to avoid that issue by narrowing the patent just to doing it on a computer then it is still invalid. There is nothing inventive about t

Re: The History of Software Patents

[tambo]

So long as software patents exist I would not oppose you if you want to fight for exempting software from copyright.

I'll take that compromise. Double coverage is nothing but a fencepost perch.

Since when have mathemetians and programmers called themselves "inventors"?

I doubt that the creators of engineering schematics would prefer the term "inventor" to "author." Nevertheless, a schematic is a strictly functional document - certainly some aesthetic concepts might be taken into account, but its primary purpose is solely functional.

I'll set aside for now the facts that (a) this is a really bad categorization argument, and (b) most programmers have little understanding of the commercial value of their code, so it would be better to ask their employer (who actually owns the code) about its characterization.

All softwarer is nothing more than mental steps. You do not need a computer to run software. ... Patents are for physical objects and physical processes. They were never intended to cover mental processes.

I think it's disingenuous to lump together software and mental processes. Do you anyone who's ever sat down with a pencil and paper to decrypt a 128-bit RSA-encoded message according to the patented algorithm? Or who sat down with pencil, paper, and a long printout of binary data to compress it for better transmission over a modem?

Let's look at this in a technical fashion. Any process implemented in software can also be implemented in hardware, and vice versa. You can take any patented or patentable software process and write a circuit that will perform the same process. Indeed, that's how the first true software patents were granted - they were shoehorned into a circuit or general-purpose computer and claimed as a novel device. Same result, but crowded around inconsistencies and baggage.

And then there's the practical angle: Eliminating software patents would force all inventors of truly useful software to release them only as dedicated hardware, which they could patent (along with the process.) EEs have evolved to the point where modern ROMs are undumpable and unmonitorable - the circuit boards are loaded with suicide chips and sensors, so you can't just emulate the circuit in hardware. How's that for impeding innovation? Others can't even practice the invention for the purpose of experimentation, because they can't determine how the circuit works. Do you really want to encourage a return to (gulp) hardware dongles? I don't.

- David Stein

Re: The History of Software Patents

[Alsee]

Sigh. Every single time I debate software patents, the software patent advocate completely ignores the fundamental example and question. Usually I explicity ask a certain question and throw a neon sign around it saying I will simply repeat the question if it is ignored. For some reason I forgot to draw attention to it this time and it was almost hidden in my post. I was also phrased more like a retorical question. It's quite frustrating having the same point and a direct question simply ignored time after time after time, but in this case I won't place blame on you because I did bury it this time.

The example is if I do in fact carry out a software patent mentally. An effort to directly and intentionally infringe the patent - using my brain. Is that an infringment? Does patent law actually claim to prohibit those thoughts? Yes or no?

There are much simpler/easier examples of software patents than RSA encryption, but since you suggested it lets go with that example. You suggested 128 bit RSA. Actually a 128 bit RSA key could be cracked on a home PC in a matter of minutes or days. Common RSA keysize is 1024 or 2048 bits. However an important point is that keysize is irrelevant to the patented RSA process. You are still violating the patent if you use single digit keys! And obviously it becomes vastly easier to preform a live demonstration running and infringing the patent purely mentally when you decide to select a perfectly valid key like 6 or 10. Sure it would be an insanely weak key, but it is still covered by the patent.

So if I do an actual demonstration mentally RSA encrypting the text "Hello World!" with the key 6, how do you respond to the fact that those thoughts are patent infringement and prohibited by law?

-

Trolls? TROLLS!

[Dark Lord Seth]

Intel's [chairman] Grove derides such patent holders for showing little interest in producing goods with their inventions in favor of demanding licensing fees from others. "We call them trolls," he said.
The Washington Post

As an occasional troll ( Think of it as a hobby ), I am insulted. >:(

Money

[Muda69]

This far-fetched idea of no software patents will never fly because of one reason, money. Large software development houses derive a large percentage of their income from royalties from intellectually-based patents. That's the way the system works and we don't need to change it.

Re:Money

[Anonymous Coward]

Nonsense.

Software houses can offer licenses for their reference libraries of substantial works at a cost far lower than reimplimenting the functionality. There's no reason to put up roadblocks on technical innovation simply because it helps some companies maximise short term returns to the detriment of a competitive marketplace.

Re:Money

[Halo1]

This far-fetched idea of no software patents

It's only far-fetched for IP-attorneys that don't think that the economy should be taken into account when deciding about patent policy.

Large software development houses derive a large percentage of their income from royalties from intellectually-based patents.

Most software patents are not owned by software developers, but by hardware developers (they got patents in the past for their hardware, so they simply continue that tradition). Anyway, several large companies also lose a lot of money because of software patents. Examples are the companies sponsoring this campaign, but another is e.g. Cisco Systems.

You're very naive if you think all large companies favor software patents. And you're even more naive if you think that per definition means that Europe will legalise software patents.

Re:Money

[nerdlyone]

It's only far-fetched for IP-attorneys that don't think that the economy should be taken into account when deciding about patent policy

The economy is exactly why software patents should exist. Not all large companies may favor software patents (particularly those who have not invested heavily in IP protection); but lots of money has gone into developing software *because* of software patents. It is good for startups too, who can protect an idea and attract capital from investors. Investors like property rights. These facts should not be ignored in any debate over IP rights.

It seems to me that patent rights in software can co-exist with open source. They co-exist now.

Re:Money

[Halo1]

The economy is exactly why software patents should exist.

Then why do you think those IP-lawyer organisations state that you should not take into account the economy? It's quite simple: because software patents do not help the economy and innovation at all.

Not all large companies may favor software patents (particularly those who have not invested heavily in IP protection);

There is a world of a difference between the catch-all "IP protection" and software patents. Particularly companies who use other means to protect their "IP" are indeed opposed to software patents. The reason is that even if you do not think that software patents are needed to protect your investments, you still need them if they are legalised (since otherwise you are an easy prey for all those software patents owners out there).

but lots of money has gone into developing software *because* of software patents.

Prove it. Almost all studies I've seen claim that the incentives to innovate in software is competition.

It is good for startups too, who can protect an idea and attract capital from investors.

Of course, that's exactly why all these small companies are protesting the legalisation of software patents in Europe.

Investors like property rights.

Finally something that's true. But software patents are akin to the "millions of visitors" in the dot com bubble age: they by no means show that a company will be able to put out a solid product or that it will be very innovative. At best, they allow a company to go for an EOLAS or so. Did you know that enforcing a software patent in court in the US costs between US $1,500,000 and $2,000,000 (if you're lucky)? Same for defeating one in court. How many small companies can afford this, do you think? Those patents are worthless if you can't afford to enforce them.

It seems to me that patent rights in software can co-exist with open source. They co-exist now

It seems to me you're totally missing the point. This has nothing to do with open source.

Re:Money

[nerdlyone]

Then why do you think those IP-lawyer organisations state that you should not take into account the economy?

Respectfully, I am not sure what organizations you are referring to. It sounds strange, since the whole concept of IP is based on economic incentive.

'but lots of money has gone into developing software *because* of software patents. '

Prove it. Almost all studies I've seen claim that the incentives to innovate in software is competition

Actually, I was talking about investment incentive, not incentives to innovate. Competition indeed promotes innovation. But property rights promote investment.

You also say "almost all studies...." I assume this means you have read some studies that support the arguments for patents in software.

Did you know that enforcing a software patent in court in the US costs between US $1,500,000 and $2,000,000 (if you're lucky)? Same for defeating one in court. How many small companies can afford this, do you think? Those patents are worthless if you can't afford to enforce them.

Litigation is not the only way patents are used. There is a large market in licensing, and lots of money changes hands based on patents without ever litigating. Your argument, respectfully, sounds like this: "Civil tort litigation can cost millions. Therefore we should do away with tort claims." While the current administration in DC is pursuing this logic with respect to medical malpractice torts, I disagree with it.

The point remains that lots of capital goes into software based on property rights. Remove those rights and you will weaken the industry by reducing the amount invested in it. I am not saying it would destroy the industry; but I disagree that patents in software "do not help the economy and innovation at all," as you argue. It seems patently obvious. ;)

Re:Money

[Halo1]

Respectfully, I am not sure what organizations you are referring to.

I posted links in my original reply to you. For clarity's sake:

• AIPLA statement: see point 10.
• IPO statement: see point 10 (and 6)

It sounds strange, since the whole concept of IP is based on economic incentive.

That's the theory, yes. In practice, not all types of IP are able to meet that goal. There is no economic law that says "more and stronger exclusion rights per definition result in a healthier economy". And there's also no juridical law that says "lawyers will always defend what's in the best interest of the economy as a whole".

Actually, I was talking about investment incentive, not incentives to innovate. Competition indeed promotes innovation. But property rights promote investment.

One person's rights are another person's limitation. You always need a proper balance to get good results on the whole. You cannot look at single entities in the market and conclude from that it must be beneficial on the whole (or harmful for that matter).

Case in point: offering one company a monopoly on selling things via the internet, makes investment in all online shopping companies a lot less attractive.

You also say "almost all studies...." I assume this means you have read some studies that support the arguments for patents in software.

Indeed, there is also one economical study in the entire world (that I know of, I really don't know any others) that claims that software patents are generally beneficial, and that patent thickets are a myth. You can find it here. And here's a rebuttal.

Litigation is not the only way patents are used. There is a large market in licensing, and lots of money changes hands based on patents without ever litigating.

Yes, but that's not productive. In fact, every time such a transfer occurs, you extract money from the software business and invest it in lawyers. Something like that is only defensible if the effects on the whole are positive, e.g. if such transfers are necessary to spread knowledge throughout the industry. That is not the case with the software sector (see the studies I pointed you to).

Your argument, respectfully, sounds like this: "Civil tort litigation can cost millions. Therefore we should do away with tort claims."

My argument was simply that small companies do not have the means to enforce software patents, and that therefore arguing that they are beneficial for small companies is at least very doubtful. That's not even specific to software patents, it's a general recognised problem of the patent system.

However, in some sectors this negative effect may be offset by other, positive effects that patents have (e.g., if you need tens of millions of dollars to even start doing your product development, the cost a few patents on top of that can be disregarded).

The point remains that lots of capital goes into software based on property rights.

You keep talking about "property rights" as if software patents are the only form of property rights in the software industry. They're not by a long shot, as you probably know. I'm also not asking for abolishment of copyright or trademarks or anything else.

Remove those rights and you will weaken the industry by reducing the amount invested in it.

I fail to see how you can weaken an industry by stopping to invest in something which hampers that same indust

Re:Money

[nerdlyone]

You cited the AIPLA and IPO statements as supporting your view that "software patents do not help the economy and innovation at all." But your citations from these organizations don't really make that point. From the AIPLA statement, in the body of the document:

AIPLA believes that Congress, and not the PTO or the courts, is the proper authority to consider economic theory and competition policy-oriented principles. For the reasons discussed below, the PTO and the courts should not inject these theories and principles into their decision-making.

They are talking about uncertainty in the law, which is a blight. Uncertainty restrains economic behavior by increasing risk. The statement states that allowing the PTO and courts--rather than congress--to inject economic considerations into their decision-making would be bad, because it would not be in actual legislation, but in rules promulgation--making it ad hoc and slightly arbitrary. But their point is definitely not that patents hurt economic activity. Their point is that uncertainty in rule application would hurt economic activity. And I agree with them.

You also cited the IPO statement:

6. IPO does not support the FTC recommendation concerning considering potential harm to competition in deciding upon the scope of patentable subject matter. 10. IPO does not support expanding economic considerations in patent law decision making.

#6 seems to contradict your position, because the IPO is arguing that considering the scope of patents should not take "potential harm to competition" into account. Your position is based on the idea that patents harm competition. So they seem to directly contradicting you there.

#10 reiterates the AIPLA argument made above, which does not apply to this discussion, as I explained.

"There is no economic law that says "more and stronger exclusion rights per definition result in a healthier economy". And there's also no juridical law that says "lawyers will always defend what's in the best interest of the economy as a whole""

I am not arguing that patent protection should be extended to absurd levels; I haven't even argued the current level of protection is best (I don't think it is--patents in software should be harder to get than currently). I did argue against your point that "software patents do not help the economy and innovation at all."

Case in point: offering one company a monopoly on selling things via the internet, makes investment in all online shopping companies a lot less attractive

Your example monopoly is a bit broad, and would indeed stifle the market; however, patents are not nearly as broad. You essentially describe a patent on the entire industry, an absurd example, and the only one in which your point holds. But the fact that monopolies can be obtained at all provides great incentive to invest in those companies which have monopolies.

Yes, but that's not productive. In fact, every time such a transfer occurs, you extract money from the software business and invest it in lawyers. Something like that is only defensible if the effects on the whole are positive, e.g. if such transfers are necessary to spread knowledge throughout the industry. That is not the case with the software sector (see the studies I pointed you to).

You are focusing on the monopoly aspect, and accurately pointing out the downsides; but you're ignoring the upsides to property rights. The software may never have been written to begin with without an investor to pay the programmer.

And your idea that "every time such a transfer occurs, you extract money from the software business and invest it in lawyers," ignores the fact that the licensee gets to sell the product too--which makes money for the industry and employs programmers. Consider Apple's refusal to license their product. You don't see many Macs around today compared to IBM and its clones. That's because IBM licensed their technology out, while MAC refused to. Licen

Re:Money

[Halo1]

You cited the AIPLA and IPO statements as supporting your view that "software patents do not help the economy and innovation at all."

I initially cited them while saying "It [having no software patents]'s only far-fetched for IP-attorneys that don't think that the economy [typepad.com] should be taken into account [typepad.com] when deciding about patent policy." Afterwards, I indeed inferred the above in your reply.

They are talking about uncertainty in the law, which is a blight. Uncertainty restrains economic behavior by increasing risk. The statement states that allowing the PTO and courts--rather than congress--to inject economic considerations into their decision-making would be bad, because it would not be in actual legislation, but in rules promulgation--making it ad hoc and slightly arbitrary. But their point is definitely not that patents hurt economic activity. Their point is that uncertainty in rule application would hurt economic activity. And I agree with them.

The point the FTC made is that this economic balancing requirement is basically a foundation of the patent system. Patents are monopolies offered by society. They're not a natural right, they are something you can get if society thinks it's in its own advantage to grant you that monopoly. Some filters are subject matter (although there's not much left of that), novelty, non-obviousness and usefulness, and all introduce uncertainty (you can't know all prior art, you don't know what the patent examiner will think is obvious to a skilled person in the art).

I do not think that considering whether or not something in a field where until now no patents have been granted should be patentable or not, introduces that much extra uncertainty. I think you should read the recommendation of the FTC in that light, not that each patent should also be subjected to an "economic impact" assessment. You always have to look at both sides, especially since it were in fact the PTO and the courts that extended the patentable subject matter to software-related innovations (and not Congress, although they also did not forbid it afterwards).

Given Machlup's famous quote "If we did not have a patent system, it would be irresponsible... to recommend instituting one. But since we have had a patent system for a long time, it would be irresponsible, on the basis of our present knowledge, to recommend abolishing it", erring on the safe side seems the prudent choice here.

6. IPO does not support the FTC recommendation concerning considering potential harm to competition in deciding upon the scope of patentable subject matter.

#6 seems to contradict your position, because the IPO is arguing that considering the scope of patents should not take "potential harm to competition" into account. Your position is based on the idea that patents harm competition. So they seem to directly contradicting you there.

I'm afraid I don't follow you here. My basic point was that IP lawyers do necessarily recommend what's best for the economy. And the reason for that is (possibly, in my opinion) that software patents are bad for the economy. Competition is good for the economy. So them arguing against taking into account competition means that they (again) give precedence to the holy principle that everything should be patentable, regardless of economical consequences.

"There is no economic law that says "more and stronger exclusion rights per definition result in a healthier economy". And there's also no juridical law that says "lawyers will always defend what's in the best interest of the economy as a whole""

I am not arguing that patent protection should be extended to absurd levels;

I'm not talking about absurd levels of patents. I'm saying that the (according to you probably completely reasonable) extension of

Re:Money

[Halo1]

Sorry, I just noticed I forgot to reply to one of your remarks:

Case in point: offering one company a monopoly on selling things via the internet, makes investment in all online shopping companies a lot less attractive

Your example monopoly is a bit broad, and would indeed stifle the market; however, patents are not nearly as broad. You essentially describe a patent on the entire industry, an absurd example, and the only one in which your point holds.

Unfortunately, that patent is quite realistic (at least in Europe, but I assume there are US equivalents). You may want to look at claim 1 of both patents mentioned next to the first patent at webshop.ffii.org. Of course, they don't claim "selling over the internet", but what they do claim corresponds to the way you more or less have to implement this. Those claims are extremely broad, despite the number of words they use.

FWIW, the first claim of the MP3 patent mentioned on that page is extremely broad as well, and doesn't just cover MP3 compression by a long shot. In fact, I've still to encounter a software patent with narrow claims. And yes, I know the first claim of a patent is always as broad as possible and that the rest refines on this, but you still have to go to court to defeat the broad claim (or take a license, which is often much cheaper and thus the preferred option).

Re:Money

[Alsee]

Aside from the fact that patenting calculations is absurd, your entire premise that they have driven investment in software is entirely empty. Prior to the 1980's the US consistantly and properly rejected any and all attempts to patent software. Such patents were only granted in the US in signifigant numbers in the 1990's. Such patents are still virtually never enforced in the US, and most of the world still does not allow or enforce software patents. The EU has been issuing a few software patents, but they are almost never upheld in the few rare cases where someone attepts to use them.

The entire explosion of the computer and software industry and the internet has been entirely free of software patent "protection". Actually it would be more appropriate to say those explosions were free of software patent constraint.

Today patent threats are generally either used to crush small innovators, or traded in entirely unproductive blanket crosslicening deals between entrenched industry players.

If software patents were to actually be enforced to any signifigant extent the US software industry - and all industries that use software, would be crippled or collapse. Software development would flee overseas, and US industries would suffer the handicap of being denied the use of vast quantities of software. For example it is estimated that Linux infringes almost 300 US software patents. The US economy would absolutely collapse if you were to suddenly deny US companies the use of Linux and all of the software running on top of Linux.

I also guarantee you that Windows infringes countless patents. There is no reason such a patent holder would have to settle with Microsoft. The patent holder could not only sue Microsoft for millions, not only could they get an injuction prohibiting the distribution af any new copies of Windows (or machines with Windows), they can go after end users. Software patents are process patents, it is the end user that is carrying out that infringing process. They can not only sue for money, they can get an injuction prohibiting any further use of that software.

But the issue of whether software patents are good or bad is moot. They are invalid. In the 1980's the Supreme Court goofed in a borderline 5-4 decision. Prior to that decision the US patent office consistantly and properly rejected any attempt to patent software on the Mental Steps Doctrine - that mental steps are not an invention and cannot be patented. Any software can (slowly) be carried out mentally. Running software mentally is a normal part of the programming and debugging process. Granting software patents is either granting a patent making certian thoughts against the law, or it is granting a patent on the blood obvious step of using an ordinary computer simply to speed up that calculation.

-

Re:Money

[tambo]

Almost all studies I've seen claim that the incentives to innovate in software is competition.

Correct. But that is today. Software development in twenty years will likely look very different.

Consider: None of these arguments is, well, novel - they have thirty years of dust on them. They were all made regarding biotech, too. Thirty years ago, the instigators of the uproar over biotech patents said that:

• Biotech patents would prevent normal researchers from using basic, ordinary benchtop lab practices.
  
• Biotech patents would basically centralize all of biotech within two or three massive companies that would no longer need to innovate.
  
• The USPTO could not even hope to staff its examiners with biotech experts, and therefore could not conduct a valid examination, leading to a torrent of facially invalid biotech patents for all eternity.
• Biotech patents would basically cause the end of biotech innovation. No more medical therapies; no more lab research; no more industrial biochemistry.

Fortunately, those arguments were declined. Obviously, the predicted cataclysm has not materialized. Today, our biotech industry is basically causing a golden age of medicine - we're creating far more disease cures, far faster, than ever before in history.

So it's pretty simple:

• Biomedical innovation requires funding.
• Funding of early-stage biomedical innovation comes from investors.
• Investors only invest in companies that have a strong business plan.
• A strong business plan that hinges on conducting expensive research to create a new market must mention a way to gain a return on investment by securing that market for a limited time.

Hence, biotech patents. Hence, biotech innovation.

Software is not markedly different. In a nutshell, I will offer the following insights:

• Software technologies and techniques that are routine and ubiquitous today will still be routine and ubiquitous tomorrow - free of patents.
• Bad software patents expire 20 years after their filing date, just as do valuable patents. Thus, even bad patents today - those that cover routine techniques - will expire and become public knowledge in a few years.
• The industry will become more selective about filing software patents. It must, since absurdities like patents on hash tables will never be useful to anyone. Even large companies cannot afford to throw away vast sums of money on patent portfolios that are not enforceable.
• By approximately 2020, we will have an incredibly well-documented record of the state of software development - both in the form of 50 years of programming journals, and in the form of all previously-filed software patent applications. The patent system will improve, as it has before.

- David Stein

Re:Money

[Halo1]

Correct. But that is today. Software development in twenty years will likely look very different.

Will it? It doesn't look that different today than it did 20 years ago. There are some new concepts (OOP, AOP, etc), we have RAD's, but in the end it's still programmers and designers thinking of new algorithms, debugging stuff, and integrating everything into a stable and usable whole.

Consider: None of these arguments is, well, novel - they have thirty years of dust on them. They were all made regarding biotech, too.

I doubt it. Did anyone ever claim the biotech industry is a "cottage industry"? (see the last paragraph) That you barely need any investments to start a new biotech company? That everything underlying biotech innovations in pure maths? That biotech patents pave the way for patents on business methods? That biotech patents could be used to prevent publications of new biotech techniques and not just their use? (program claims) That biotech is pervasive throughout all economic sectors going from grocery shops to space stations, and as such is an "enabling technology" of which hindrances have very broad reaches?

Fortunately, those arguments were declined. Obviously, the predicted cataclysm has not materialized. Today, our biotech industry is basically causing a golden age of medicine - we're creating far more disease cures, far faster, than ever before in history.

To be fair, I've never followed the biotech patent situation. So I just searched for "biotech patents" on Google. The fourth link contains several links which seem to show the controversy is still far from settled. So does the sixth and the eighth. There are of course other views as well (such as the tenth link), but claiming everything is happy happy joy joy with no downsides seems just a tad misleading.

There's also a bit about it in the recent FTC report on patents and innovation. They note that the fact that biotech includes quite a bit of consequential innovation (as opposed to traditional pharmaceuticals) causes some problems. You are presumably aware of the fact that software development is almost nothing but consequential innovation (and lots of reuse as well). The solution proposed by the panel members regarding biotech is what is currently already done in the software world: extensive cross licensing. Of course, you need a lot of patents to be able to join that game.

The industry will become more selective about filing software patents. It must, since absurdities like patents on hash tables will never be useful to anyone. Even large companies cannot afford to throw away vast sums of money on patent portfolios that are not enforceable.

Of course they are useful for those companies! They are strategic assets, used as trading cards or litigation tokens. Enforceability is generally not even a concern, as many small companies can simply not afford the litigation costs (if you have the choice between a $50,000 license or a $2,000,000 lawsuit, what do you pick?)

By approximately 2020, we will have an incredibly well-documented record of the state of software development - both in the form of 50 years of programming journals, and in the form of all previously-filed software patent applications.

And there will be tons more of programming legacy which is not documented in this way at all, but just available as source code (which is also a publication, given that source code

Go Software Patents!

[Creepy Crawler]

Patent mathematics.
+ is .0002$ per use
- is .0002$ per use
* is .001$ per use
/ is 1.00$ per use
mod() is 10$ per use
sqrt() is .009$ per use

We acknologe that these mathematics are the only ones you can use, and all derivatibe works are subject to this cost times amount of operations done. EG: Limits are infinite +, so you are incessantly in the hole.... (enter lawyer BS)

Re:Go Software Patents!

[bizpile]

Patent mathematics.
+ is .0002$ per use
- is .0002$ per use
* is .001$ per use
/ is 1.00$ per use
mod() is 10$ per use
sqrt() is .009$ per use

Now I'm boned, I keep writing infinite loops...

Re:Go Software Patents!

[Weh]

Hi, I own the patent for the use of the lowercase letter "a", rates are currently $.005 for first use and $.001 for each subsequent usage. You would owe me $0.023, however since I used some of your patented mathematics in calculating what you owe me this amount is reduced by .0002 for one use of + plus .001 for use of * which makes the total, eh .0002 + .001, which brings the total on eh .0002 + .001 + .0002, which brings the total on eh .0002 + .001 + .0002+ .0002.... Ah well screw it...

Re:Go Software Patents!

[Anonymous Coward]

Don't give up, it may not generate any economic wealth but it could make you rich. That's what it's all about, making bling; who cares if the quality of life for millions of people is potentially reduced so long as you can make a quick buck? Right?

Re:Go Software Patents!

[hackwrench]

I've patented the use of the letter e in conjunction with a computational device. that "in conjunction with a computational device" makes my patent infinitely more defensible in a court of law than yours, so there!

Cheap Ass Mod Operation...

[mopslik]

+ is .0002$ per use
- is .0002$ per use
* is .001$ per use
/ is 1.00$ per use
mod() is 10$ per use

$10 for mod? Bah!

int cheapAssModOperation(int X, int Y)
{
int A, B, Z;
Z = X / Y;
A = Z * Y;
B = X - A;
return B;
}

Total cost = 1 division, 1 multiplication, 1 subtraction = $1.0012.

Re:Cheap Ass Mod Operation...

[gsasha]

Act quick! Patent it!

Re:Cheap Ass Mod Operation...

[evilmousse]

Hmm, what about getting savings on the
division and multiplication too?

1 mult = .001
1 add = .0002

so if you're multiplying by less than 5,
you might as well use addition instead. ..of course, if you were to automate that
detection, you'd lose the savings...

what about patenting the if statement?
patenting howver-many-million color
variations on 1 pixel?
patenting "on" and "off"? ^^ ..of course all of the above are
"in conjunction with a computational device".

-evilme

Re:Cheap Ass Mod Operation...

[Anonymous Coward]

of course, if you were to automate that detection, you'd lose the savings

I assumed that the >, <, and similar operations would be patented along with the other operators. Plus, simplicity is king.

The extension is left as an exercise to the reader...

Sanity

[Mark_MF-WN]

Maybe if Europe banned the practice of software patents, our leaders here in North America will wise up and do the same.

I'll be if we required software makers to publish the source code of any project containing patented code, that software patents would die a quick and decisive death.

Re:Sanity

[Anonymous Coward]

Just because you're not in Europe doesn't mean you can't help raise awareness. Link

Re:Sanity

[Anonymous Coward]

Link that you need to manually fix like so:

data:image/png;base64,

Isn't slashcode great?

Can we get a US effort like this started?

[borgheron]

All of the anti-patent stuff I'm seeing is from the EU, which is frustating since it does nothing for those of us here in the United States.

GJC

Re:Can we get a US effort like this started?

[Halo1]

Actually it does, otherwise the USPTO wouldn't try to influence the European discussion. You can bet that if Europe does not legalise software patents, this will have quite a profound effect on the US situation. At least the whining "abolishing software patents is not realistic" whiners might shut up, for a start.

Re:Can we get a US effort like this started?

[Alsee]

Actually I do expect it to have an effect on the US. We are sitting on a software-patent time bomb. Once the bomb goes off, once we start seeing a non-trivial number of such patents enforced, we will see a huge migration of software development to the EU. You will also see all sorts of US industries handicapped by the inability to use all sorts of "infringing" software. Imagine when would happen if every company in the US had to stop using Linux - and thus using every peice of software that runs on Linux.

Between the exodus of software work, and the competitive handicap of US industries, you will likely see the US go go bonkers, perhaps to the extent of a trade-war. Hopefully the US government - and US business - will have some minimal leval of sanity and see that such a trade tantrum would be purely self destructive, that the only solution is to eliminate the entire broken idea of software patents. Once software patents start hurting business you'll see the message get through to the government pretty damn fast.

-

Re:Can we get a US effort like this started?

[steelneck]

Ohh, it will benefit you americans too, if we succeed that is. It will be a very strong signal to your politicians.

Software patents not inherently evil

[Rufus88]

It's not that software patents are, in principle, bad. It's that the idiots in the USPTO are letting trivial ones through the syste. Some software patents are completely legitimate. Take, for example, this patent on the "Marching Cubes" computer graphics algorithm. The paper describing this algorithm made it into SIGGRAPH's Seminal Graphics collection of most important papers in computer graphics. Not all software patents are trivial and obvious.

Skeptical

[ravenspear]

Hopefully, more large companies will join this campaign in the future.

Since it's precisely large companies that are most benefiting from the abuse of the patent system to obtain frivolous software patents, I doubt this will happen. It won't gain much corporate support outside of open source firms.

Re:Skeptical

[Halo1]

1&1 is not an open source company. And if you look at the signatories of this (Dutch) press-release of Belgian companies protesting against software patents, you'll find a bank and 9 companies, none of which is particularly open source. And at least the top two ones (EVS and The IRIS Group are quite big.

Re:Skeptical

[ravenspear]

And if you look at the signatories of this (Dutch) press-release [softwarepatenten.be] of Belgian companies protesting against software patents, you'll find a bank and 9 companies, none of which is particularly open source.

Well, the international community has been able to hold on to some sense of reality with patents/copyright far better than the US.

question Re:Money

[js7a]

Will they arrange a $50K-300K low interest loan if I promise not to patent my products?

Best anti-patent argument I've read

[dk.r*nger]

I've been searching for a really good argument against softwarepatents, and they all seen to center around some semi-marxist anti-large-corporation basis, such as "GIF and MP3 is patented and OBVIOUSLY that is bad". Really, that isn't obvious in any sense or way. Neither GIF nor MP3 is trivial, and PNG and Ogg Vorbis took years to emerge and mature. This may og may not be a good counter-argument, but nobody cared to explain why LZW and MP3 is trivial ..

Anyway, the argument, as presented:

Patents on software are just as wrong as expanding the patent system to literature.
With patents on story elements, no movie could be published without having to firstly check whether there is any general idea in the storyline that someone patented during the last 20 years. Here's an example: At first sight, Dirty Dancing and Titanic are two very distinct movies. However, if there were patents on story elements, then the makers of Dirty Dancing could have sued the studio of Titanic. Both movies have a scene in which a poor boy takes a rich girl from a party of her social peers to a dancing party of his group, and she enjoys it. Dirty Dancing came out only nine years before Titanic, so any patent would still have been in force. No one knows whether James Cameron had that Dirty Dancing scene in mind as he wrote the Titanic script. Maybe Cameron never saw Dirty Dancing but the patent (if it existed) could be used against him anyway.

Re:Best anti-patent argument I've read

[bill_of_wrongs]

You are partly right that the reasons are not always thoroghly explained. However, software patents are also wrong for other reasons than being trivial or the argument you quote. Let's look at the GIF example you mentioned: UNISYS who owned the patent on the LZW algorithm (which is used among other things to greate compressed GIFs) just sat around waiting for the GIF format to become a defacto standard on the net before they started saying they had a patent on it. From their point of view it made sense to act like that but legalized extortion is hardly a sign of a working market economy. I don't think it is fair to call someone who criticizes a system that encourages such behaviour a semi-marxist. The fact that IBM also managed to get a patent on the same algorithm later should also tell you something about how well the system works even if you for a moment assume that all software patents are non-trivial.

Re:Best anti-patent argument I've read

[Alsee]

I've been searching for a really good argument against softwarepatents

All you have to do is look back before we instituted this absurd system of granting software patents. It happened in the 1980's in a hotly disputed 5-4 Supreme Court decision, Diamond v Diehr. That ruling was a goof that squeaked through by a single vote.

Prior to that the US patent office, and essentially every other patent office in the world, consistantly and properly rejected all attempts to patent software. Specificly the USPO rejected them based on the Mental Steps Doctrine. That doctrine stated that mental steps and calculations were not inventions and could not be patented.

Any software can (slowly) be run mentally. Running software mentally is a normal and routine part of programming and debugging. Allowing software patents is either to claim that thinking certain thoughts is against the law, or is to issue a patent on the simple and blatently obvious step of using an ordinary computer to carry out that exact same calculation.

-

Re:Best anti-patent argument I've read

[torokun]

I can't believe you actually believe this. You think any software can be run mentally??!

You're not talking about running software mentally. You're talking about thinking about what software does mentally. Try running mathematica mentally sometime. It won't just be slow, I think it would be thousands of years before you could complete some fairly straightforward operations through mentally manipulating the binary code.

Your post is complete BS.

Re:Best anti-patent argument I've read

[Alsee]

You are looking at it backwards. It does not matter how long the longest supposedly valid software patent would take to run mentally. The question is how long would it take to carry out the mentally fastest supposedly valid software patent.

The answer is that many of the "mentally fastest" software patents can in fact be carried out in a matter of minutes or less.

I merely need to find a single patent I find suitable. When I do carry out that supposedly valid patent mentally you are faced with the fact that your patent is claiming my thougts were infringing and prohibited by law, or you have to admit that software is not an invention and is not validly patentable.

Sure some software would take millions of years to run mentally, but in principal any software can be run mentally. Either software patents are valid or they are not. Any attempt to put up an imaginary line between "mentally fast" software and "mentally slow" software would rapidly degenerate into pure silliness.

By the way, I realized much of this in relation to the DMCA and DeCSS. I have read DeCSS decryption code, and it is surprizingly simple. By taking one of those special memory training courses it would definitly be possible to run DeCSS decryption mentally. The DeCSS code is dead-simple, but you do have to be able to remember some numbers while doing it. Obviously it would take forever to decrypt an entire DVD, but it would most definitly be doable to decrypt a peice of a DVD, or perhaps some CSS-encrypted text. I was thinking it would be an awsome demonstration to commit DMCA circumvention crime purely mentally live in court. You could illegally read a DRM'd e-book purely mentally. Under the DMCA it would literally be thought-crime. I say that makes the DMCA blatantly unconstitutional. I defy any judge to uphold a law creating thought-crime.

-

Re:Best anti-patent argument I've read

[Anonymous Coward]

I don't think anybody claims that all software patents are trivial. Some are, and that is a symptom of a non-working system for granting software patents.

However, I think it is putting the cart before the horse to ask why patents on software should not be granted. The correct question is why we should have patents at all and why we should have them on software in particular. Is there any gain for society in granting software patents? Will the rate of innovation actually increase? Will the cost of goods and services be lower because we grant these patents? Will we have a better climate for starting new businesses?

When I ask myself these questions, I come up with the answer that society would be better off without granting software patents.

Re:Best anti-patent argument I've read

[elegie]

With the GIF format, there was a problem in that the patent was not obvious. The article that described the LZW technique did not mention the patent. In the USA, it is possible to publish a technique and then patent the technique no later than 12 months aftewards. Another problem with the GIF situation was how the patented technique became frozen into an image format which was meant to be open. Due to backwards compatibility, it was not possible to substitute a nonpatented technique afterwards.

Re:Best anti-patent argument I've read

[elegie]

Software is different from other industries in that a single individual can produce very good software without specialized machinery, etc. Consider shareware, freeware, and other such software. There is also free (as in "freedom") software and open-source software. Production of software does not require a large company. Software patents could be difficult if not impossible for an individual or a small company to deal with.

Re:Best anti-patent argument I've read

[Stevyn]

"Ogg Vorbis took years to emerge"

Can we stop making fun of gentoo people for once!

They're unnecessary and dangerous

[gidds]

Why do we need software patents, though?

If they're to protect only an implementation of an idea, as they say they are, then copyright already offers perfectly good protection.

And if they're protecting more than that, then they shouldn't be!

Either way, they're unnecessary. And given the sort of flagrant abuses we're seeing so often, what reason is there for keeping them?

Re:They're unnecessary and dangerous

[Rufus88]

Why do we need software patents, though?

Because the kind of hard work and knowledge and investigation and thought that goes into devising a solution to a difficult problem in matter-space is exactly the kind that goes into devising a solution to a difficult problem in information-space. One type of problem solving shouldn't be protected more than the other simply because you need bolts and motors and lubricant to implement it.

If they're to protect only an implementation of an idea, as they say they are,

They do not exist merely to protect an implemention of an idea. They protect the technique that the implementation embodies.

And if they're protecting more than that, then they shouldn't be!

Why not? Should Xerox not have been granted a patent on the xerographic photocopy process, but rather only a copyright on their machine's manufacturing blueprints?

Re:They're unnecessary and dangerous

[Halo1]

Because the kind of hard work and knowledge and investigation and thought that goes into devising a solution to a difficult problem in matter-space is exactly the kind that goes into devising a solution to a difficult problem in information-space. One type of problem solving shouldn't be protected more than the other simply because you need bolts and motors and lubricant to implement it.

It has indeed nothing to do with implementation. It has everything to do with the economic effects:

1. Are patents necessary to get enough investments in the IT-sector?
2. Are the positive effects of granting software patents larger than their associated negative effects?

The problem is that the answer is "no" in both cases. Most economic studies conclude either that, or "we don't know". There's only one in the entire world that I know of which claims the reverse.

And there are differences between the software world and other sectors which change the effects that patents have there, see e.g. this text (under the black box) from the US National Research Council.

Why not? Should Xerox not have been granted a patent on the xerographic photocopy process, but rather only a copyright on their machine's manufacturing bluep

In the physical world, there is the process, a description of the process and an implementation of this description. In software, the description is the implementation (since software is nothing but a description of something). Whether or not what you describe is patentable, should be entirely independent from how you describe it. Yet, an awful lot of unpatentable stuff suddenly becomes patentable if you say it's done by a computer.

Re:They're unnecessary and dangerous

[Rufus88]

Are patents necessary to get enough investments in the IT-sector?

The IT sector as a whole? Probably not, but that's like asking whether patents are necessary to get enough investments in manufacturing. The point is that there may be particular endeavors in IT whose value may be dependent on being able to be the sole supplier of the technology for a while after it is developed.

In software, the description is the implementation (since software is nothing but a description of something).

I strongly disagree. I can describe an algorithm in English, and through the use of visual aids, but further effort must be expended to create working code, just like a blueprint requires further effort to create a working device.

Whether or not what you describe is patentable, should be entirely independent from how you describe it

You're making my point for me here. Protecting only the implementation through copyright should be independent of protecting the method through patent.

Yet, an awful lot of unpatentable stuff suddenly becomes patentable if you say it's done by a computer.

Agreed. Something shouldn't simply become patentable by virtue of doing it on a computer. But in cases where the thing being patented is not only being done on a computer, but also being done for the very first time, whether a computer happens to be involved or not, should be protectable.

Re:They're unnecessary and dangerous

[Halo1]

The IT sector as a whole? Probably not, but that's like asking whether patents are necessary to get enough investments in manufacturing. The point is that there may be particular endeavors in IT whose value may be dependent on being able to be the sole supplier of the technology for a while after it is developed.

Patent law is not just about the value that some monopoly has, but also whether the positive effects of granting such monopolies actually stimulates the sector in which they are granted, considering the inherent downsides of granting monopolies.

In software, the description is the implementation (since software is nothing but a description of something).

I strongly disagree. I can describe an algorithm in English, and through the use of visual aids, but further effort must be expended to create working code, just like a blueprint requires further effort to create a working device.

The effort to create working code is also protected by copyright (and not in any way by patent law, for that matter). You're still describing things in code, you're in effect translating.

A computer program is not a device, it's basically equivalent to a technical instruction manual (in a machine-interpretable form). A working device is not protected by copyright, a computer program is (for the very reason that it is a written work, and not a device).

Whether or not what you describe is patentable, should be entirely independent from how you describe it

You're making my point for me here. Protecting only the implementation through copyright should be independent of protecting the method through patent.

The description is indeed the only thing that is protected by copyright, that's how copyright is designed. But that does not mean that all new things done on a computer per definition have to be patentable.

Yet, an awful lot of unpatentable stuff suddenly becomes patentable if you say it's done by a computer.

Agreed. Something shouldn't simply become patentable by virtue of doing it on a computer. But in cases where the thing being patented is not only being done on a computer, but also being done for the very first time, whether a computer happens to be involved or not, should be protectable.

I think you misunderstood me. I wasn't talking about novelty, I was talking about subject matter. No, I do not think that all new things done on a computer have to be per definition monopolizable. There is no economic rationale for this. My point is merely that if you have a patentable invention, that it must not become "not patentable" if you say you do it with a computer.

However, conversely, I do not think that if some (new) thing is not patentable (for example a faster fourier transform), that it should become patentable if you say it's done by a computer ("to process video or sound data" or any other "useful purpose"). Why?

1. Because of the previously mentioned economic studies (the most important argument, which you did not address)
2. Because monopolies on mathematical algorithms are generally considered too broad, and saying it has to be done "on a computer" does not significantly reduce the scope in practice (who's going to do an FFT by hand?), nor does specifying a "useful purpose" (it's generally quite easy to specify most if not all things the algorithm is useful for)
3. The patent then still granted for the achievement of discovering the new mathematical method (because the work of describing it in a way that a computer understands is covered by copyright). I find it hypocrite to only allow patents on it if one says it's done by a computer, if you at the same time keep claiming such achievements are not patentable. The patent is not granted because you were such a genius to us

Re:They're unnecessary and dangerous

[Alsee]

The problem is that software is not an invention. It is a calculation. You do not need a computer to run software, an software can (slowly) be carried out mentally. Lets say I select some suitable sort example of software which you claim to be patentable. I then in fact carry out that short patent. Did I violate that patent? Are you seriously suggesting that pure thought can/should be against the law?

And if not, there's nothing patentable or inventive in the simple and obvious step of using an ordinary computer simply to speed up that calculation.

Prior to the 1980's, prior to the Supreme Court case Diamond v Diehr, we had the Mental Steps doctrin which states you cannot get a patent for mental steps, hat thoughts and calculations are not inventions and not patentable. The Supreme court screwed up in that 5-4 decision and violated and destroyed that principle. Prior to that the US patent office, and essentialy every other patent office in the world, properly and consistantly rejected any attempt to patent software.

-

Re:They're unnecessary and dangerous

[Rufus88]

The problem is that software is not an invention. It is a calculation.

1) You unjustifiably trivialize the complexity, novelty, and non-obviousness of some valuable computer algorithms by referring to them as "calculations". Even if technically correct, it is somewhat misleading to give the impression that we are talking about a mere linear sequence of arithmetic operations, like computing the tax on a restaurant bill.
2)I don't accept your premise that the terms "calculation" and "invention" are mutually exclusive.

I then in fact carry out that short patent. Did I violate that patent? Are you seriously suggesting that pure thought can/should be against the law?

First, it isn't "pure thought" unless you're not using the results in a commercial manner. Second, manufacture of a patented device often consists of pressing some metal into a particular shape and fastening a few pieces with nuts and bolts. Should mundane activities like this be against the law?

And if not, there's nothing patentable or inventive in the simple and obvious step of using an ordinary computer simply to speed up that calculation.

I agree whole-heartedly with the sentiment that a method or algorithm for solving some difficult problem shouldn't be patentable merely by virtue of the fact that you're doing it on a computer. It's the non-obviousness and novelty of the approach that warrants protection.

we had the Mental Steps doctrin which states you cannot get a patent for mental steps, that thoughts and calculations are not inventions and not patentable.

And then people realized that the bottom line is that we're trying to protect the return on investment of intellectual labor performed for the sake of real problem solving, and that the form of the final embodiment of the solution is completely irrelevant.

Re:They're unnecessary and dangerous

[Rufus88]

Patent law is not just about the value that some monopoly has, but also whether the positive effects of granting such monopolies actually stimulates the sector in which they are granted, considering the inherent downsides of granting monopolies.

When was the last time an otherwise-valid patent was rejected because the economic sector to which it pertained was doing just fine at the moment and didn't need any further stimulation?

The effort to create working code is also protected by copyright (and not in any way by patent law, for that matter). You're still describing things in code, you're in effect translating.

Not relevant. I agree that the code itself should be copyrightable. That holds true even if the code doesn't do anything unique or novel or non-obvious. But there are novel non-obvious algorithmic solutions to problems that take a substantial amount of work to devise, before a single line of code is written, and copyrighting one particular source-code embodiment doesn't protect the investment in devising the algorithm.

A computer program is not a device, it's basically equivalent to a technical instruction manual

Semantics. A computer program may or may not be a device. An algorithm encoded in a computer program may or may not be a device. And whether or not you categorize something as a "device" shouldn't determine whether or not it is an "invention".

The description is indeed the only thing that is protected by copyright, that's how copyright is designed

Agreed, and I'm not suggesting that anything other than the description receive copyright protection. I'm suggesting that the underlying method being described is deserving of a different kind of protection.

No, I do not think that all new things done on a computer have to be per definition monopolizable.

Agreed. I can write a shoot-em-up computer game that doesn't do anything particularly novel or non-obvious, and therefore doesn't warrant a patent, but the code should still be protected by copyright. The patentability rests in the fact that something solves a difficult problem in a way that is novel and non-obvious to someone skilled in the art.

However, conversely, I do not think that if some (new) thing is not patentable (for example a faster fourier transform), that it should become patentable if you say it's done by a computer

I agree completely. I'm not sure why you thought I would be disputing this.

Because of the previously mentioned economic studies (the most important argument, which you did not address)

See my first paragraph above.

Because monopolies on mathematical algorithms are generally considered too broad,

Well, any patent application can be too broad. That can be fixed.

I find it hypocrite to only allow patents on it if one says it's done by a computer, if you at the same time keep claiming such achievements are not patentable. The patent is not granted because you were such a genius to use a computer to calculate that stuff.


Again, I agree.

If you really think mathematical achievements and/or business methods should be patentable, then please say so.

I never said anything about business methods. And mathematical achievements is too vague. I'm talking about complex novel non-obvious algorithms, not the discovery of basic truths like the Pythagorean Theorm.

And if you think they should be patentable, then show convincingly how this patentability will increase wealth in the sectors these patents apply to, otherwise it makes no sense to grant such patents.

The same way a mechanical process can increase wealth: by being the sole supplier of any embodiment of that process. The patent I originally cited is one that has been licensed, thereby demonstrating its economic value.

You seem to be arguing that it should be possible to get a monopol

Re:They're unnecessary and dangerous

[Halo1]

When was the last time an otherwise-valid patent was rejected because the economic sector to which it pertained was doing just fine at the moment and didn't need any further stimulation?

You do not decide this patent per patent, but over a whole class of subject matter. If something is not patentable subject matter, you cannot have "an otherwise-valid patent" on it. It's true that patentable subject matter has been extended in the US to include software-described algorithms and business methods, but in Europe that is not yet the case (according to the law anyway).

You still have not given any economical rationale as to why this extension is desirable. There is no economic law that states that monopolies on on "complex, non-obvious algorithmic solutions to problems in information-space" help innovation and the economy in these fields (even if we would live in the fantasy world where only patents on complex algorithms were granted). Most studies show in fact the reverse.

Patent law is in essence an economic law. You focus on juridical consistency (everything new, non-obvious and useful should be patentable), but then you completely disregard the ends (economic prosperity), and consider the means (monopolies) and as an end in itself. This may look nice and easy when you're a lawyer, but is not very productive for the rest of us who have to live with the consequences. I also don't see how you can defend this practice only with a reasoning like "we've always disregarded any consequences, so why start now?"

Agreed, and I'm not suggesting that anything other than the description receive copyright protection. I'm suggesting that the underlying method being described is deserving of a different kind of protection.

It may be patentable, but that should be completely independent from whether or not you use software to describe it, and that is not the case in practice. You cannot get a patent on a mathematical algorithm (new/useful/complex or not), but you can get patents on mathematical algorithms when they are described in software (e.g. run-length encoding, which is used in JPEG).

Re:They're unnecessary and dangerous

[Alsee]

You unjustifiably trivialize the complexity, novelty, and non-obviousness of some valuable computer algorithms

No I don't. I freely acknowledge that math can be quite complex, quite non-obvious, and entirely novel.

The entire world, and the US up until the 5-4 Supreme Court goof, consistantly and properly agreed that math and calculations and mental steps were not inventions and were not patentable. It was well established, and with good reason. Any attempt to say that thoughts are against the law is just plain broken. Patent are and have always been for physical objects and physical processes. Not mental processes.

>we had the Mental Steps doctrine

And then people realized that the bottom line is that we're trying to protect the return on investment of intellectual labor

You just made up a rationalization and projected it backwards in time. That is not what happened.

The actual history and origin of software patents is the Diamond v Deihr case. As the 4 dissenting Supreme Court justices explained, as the US patent office argued in court, the patent application in question actually described the "invention" of a fairly simple equation. However the way the patent was written they also mentioned some stuff about manufacturing rubber. The applicant claimed to have invented a rubber making process. They didn't, it was the same old rubber making process as always. The US patent office said they did not invent a rubber making process, the 4 member minority said they did not invent a rubber making process. Unfortunately the 5 member majority used some flawed logic. They saw it as a patent on a rubber process and wrote brand new (and broken) rules for evaluating patent applications.

The 5 member majority explicitly stated that equations and calculations and sofware were unpatentable. The 4 member dissent warned that the majority's flawed logic and broken rules for evaluations applications would result in patents on software. The 5 member majority gave an elaborate explanation that their rules would not permit software patents. Well, the majority obviously screwed up. Lower courts were bound by the erroneous logic used by the majority. The US patent office had no choice to to implement the broken rules for evaluation patents. Under those new and broken rules anyone could patent equations and calculations and software as long as they played wordgames in drafting the patent. Patents the majority explicitly said they had no intention of permitting.

NONE of the 9 Supreme Court justices intended to permit this result. The 4 member minority tried to warn the 5 member majority. The 5 member majority simply didn't realize the actual impact and implications of the ruling they issued.

There was absolutely no intention to abandon 200 years of established US patent law and international accepted standards and to reverse the rules. Extending patents to software was nothing more than an error. An error faithfully carried out and extended by lower courts and by the patent office.

-