"I can write a patent and get rights in the same invention described in the book."
No, you can not, unless you make the actual object.
You are wrong on this. It is not necessary to actually make the object to get it patented. I am 100% certain of this.
Mental processes are real in the sense of constructs, but they do not constitute physical objects.
A computer running software is a physical object, not a mental process.
of course a computer running software is a machine, and no one is saying that isn't patentable. When the EU-parliament admented the proposol, they said as much: that (when novel, non-obvious, etc.) machines, also when running software, could be patented. But NOT merely on the basis of the software. It's the machine itself that should be novel (etc.) thus, regardless of the software, though not excluding it.
Two identical computers running two different programs are different machines. They accept different inputs, process the data differently to produce different outputs. There's really no reason to require the computer part of the invention in the patent. If you look at patents on apple sorters, for example, you will find that they probably don't make any reference to a base or a frame or other supports that the gears and levers are mounted upon, yet all assple sorters will have such a base. Similarly, there is no reason to reference the underlying physical computer when patenting software.
All our discussion on this thread really does not go to the bottom of things, however. As others have pointed out, you do not have a natural 'right' on a patent.
I completely agree, and never intended to argue a natural "right" to a patent. The point of my first post was found in the title - software is more like a machine, than a book. The useful aspects of machines are protected by patents, so it makes more sense to use patents to protect software than it does to try to protect it with copyright, which is used for protecting the particular "expression" of an idea, but not the underlying idea itself.
A patent is a monopoly for a certain time, granted by the state. Monopolies are never a good thing, and the only reason why it is allowed in this case, is with the idea that it stimulates research and further innovation.
I disagree with your phrasing, but not your underlying point. Society has concluded that a monopoly granted for a limited time is good if it stimulates R&D, or causes the inventor to disclose that which he would otherewise not disclose.
With software-patents it becomes increasingly obvious (quite some research on this has been done, lately) that it does not serve this purpose at all, on the contrary. Thus, it logically follows there is no reason to create or give softwarepatents, and in fact, for the stated goal of stimulating progress in the field, they should be outright forbidden.
So we come back to the reasons for my original post.
1)Patents make more sense than copyright for protecting software since computers running software are more like machines than books.
2) The general opinion here seems to be that software patents are bad, so I'd like to know why people think that. Specifically, what do people think is different about software technology from all the other areas of technology covered by patents.
since you seem a logical person, I'm sure you also can see the logic in this reasoning, whether you personally feel entitled to a softwarepatent or not.
I don't believe in software patent "entitlement" any more than you. I'm not sure I even believe that software patents are good for society. I simply haven't made up my mind, and I'm looking for answers. I have read some of the studies that others here have pointed me to. If anyone reads t
the RSA algorithm was patented in the United States (but nowhere else)....
The patent expired a few years later, but by then the point was largely moot, as a number of better algorithms had been developed in Europe and, as Europe had no software patents, were available for all to use freely.
But isn't it possible that if RSA *had* been patented in Europe, the "number of better algorithms" that were later developed in Europe might have been developed a few years earlier under the competitive pressure to design around the RSA patent? If better algorithms had been developed earlier, might Europeans be farther ahead now than they are?
If you'd read some of the studies I pointed, you might find out.
I did read them, but felt you should be able to summarize why you thought software was different.
You quoted the FTC study that said "Representatives from both the computer hardware and software industries observed that firms in their industries are obtaining patents for defensive purposes at rapidly increasing rates."
There are tow types of defensive patenting. One is where you patent something before someone else. That type makes little sense since you can merely publish it and tehreby prevent anotehr from getting a patent on it. The other is to patent something you think you can cross license if you are accused of infringing another patent. That makes sense, but what is wrong with that? It means that companies and individuals make disclsures of the latest technology to the patent office in order to obtain a patent.
They explained that the increased likelihood of firms holding overlapping intellectual property rights creates a "patent thicket" that they must clear away to commercialize new technology. They discussed how patent thickets divert funds away from R&D, make it difficult to commercialize new products, and raise uncertainty and investment risks.
I understand the concern, but this pejorative label of "patent thicket" can be applied to jsut about any other field of technology where there are patents. What is it about software that you think is so different?
The main point isn't so much that disclosure of how apple sorters work helps innovation or the economy that much, but that their monopolisation has less negative effects.
Society has concluded that the net effect of patents on technology is positive, not just less negative. Why is it that you think the net effect on software is negative?
And if you want to see more reasons, read e.g. the summaries of the Fraunhofer/Max Planck study of 2001 and the Digital Dilemma book of 2000 (or the studies themselves).
I will look for them. Thanks, but would love to hear your own viewpoint. I still don't see you pointing to anything about software that you think is truly different from designing a new paperclip or mousetrap or golf club or nuclear powerplant that would make patents less useful in encouraging innovation. I just don't see software as being inherently different in any way.
All countries have _limited_ patent laws. Patents rarely last for ever, for instance, and rarely apply to every type of 'invention'. These limitations are not inherent to the nature of patents, but rather are a result of the law.
A good point. For example, pharmaceuticals are not patentable in some countries. The argument is that medicine shouldn't be restricted by patents.
It's probably worth noting that medical development, however, is mostly occurring in countries where pharmaceutical patents are available, and there's reasonably good evidence that the cost of pharmaceutical R&D would often not be acceptable if they couldn't be recovered during the exclusive atent period. So what, if anything, is it about software that you think is like pharmaceuticals?
Patents are a type of industry regulation. The sort of patents a country should allow depends on the benefits for society.
I completely agree, but while it's easy to see when a patent restricts someone, it's much harder to see when the missing patent incentive casues us to lose out on a new or better pharmaceutical or software invention.
Generally, modern capitalist societies see innovation as an important benefit; they see innovation mainly as the result of healthy markets, and see competition mainly as a measure of healthy markets.
Again, we are in agreement.
The little we know about software patents seems to suggest that they are abundantly used for stark anti-competitive, anti-innovation behaviour.
And here we get to the crux. Why do you and others here think this, when it does not seem to be true for all other industries and technologies?
That is why at this point in time, our western, capitalist societies should reject software patents.
If this was true, I would join you, but I don't think it is true. Patents have served a major part in driving technolgy forward for hundreds of years. Why do you suddenly think it won't work any more?
Furthermore, an apple-sorting machine is not trivial to make. A data-sorting program is trivial to make.
But if the data sorter is trivial to make, then it's probably already either known (not new) or a simple variation on something that is already known (obvious) so it can't be patented. But let's suppose that a company decides that they could do better if they could come up with a data sorter that was better than anything out there, but it would cost them a few years of work to find out if it was even possible, and even if it was, after developing it, their competitors would immediately copy it. They have no incentive to try to develop it because you have removed that incentive. HOw do you know how many such improvements you have lost? I surely do not knkow either, but I can see historical evidence that providing the incentive for improvement pays off to society in the end.
And yes, I am for the abolishment of all patents. They were introduced in a time when inventing something meant building it. A huge investment of time and materials were required--you know the sort of thing: inventing a better lightbulb (scroll down to incandescent lamps for the English text).
There never was and is not now any requirement for the patented item to have taken immense amounts of time to have been developed. I'm not convinced that software is any easier or harder to develop than any other invention.
In addition to the reduced incentive to develop software by eliminating patents, tehre's another effect. Patents are ther to encourage disclosure of improvements to the world. Suppose our data sorter company decides to build the new and improved data sorter that they really need. Are they going to tell the world about their breakthrough? No. They keep it as a trade secret, since if they tell, they lose their competitive advantage. Professionals in the data sorting field
"you agree software is not a device per se?"
Yes I do, but it's directly transformable into an actual physical device simply by compiling it and loading it into a suitable computer.
"How is this different from me taking your book describing the apple sorter reading it, following the instructions, and operating it on data (apples)?
My point exactly! It's not at all different! If you want to protect your new design for an apple sorter, you don't write a book, send in the $30 and get a copyright, then sue someone 50 years later for copying the apple sorter design you described. That would be unfair, since no one checked to see if the design of your apple sorter in your book was really new. You have to write a patent application, send it in, have it examined to see if it's new, then you only have 20 years of protection. Right now, software is following the first model more closely than the second. It just does not make sense to me.
"You seem to be asking why software should be a special case and I agree, but why should it be a special case not covered by copyright?"
I think software should definitely be covered by copyright. Anyone who copies the source code exactly or the object code exactly, should be liable for copyright infringement. However, if the open source community wants to replicate the functions of that software by writing their own routines and without "copying" the "protected expression" of the software author, then they should be allowed to do that, unless there is a patent, because they are copying what the software *does* and how it *functions* and not copying the style or text or "expression" selected by the software author.
"What you describe is basically a special 'state' of the computer - certain bits in memory are set to a predetermined pattern described by software"
Agreed.
"... but there is nothing inherently novel about this - all the states of the machine can, given sufficient computing power, be enumerated."
True. But the number of possible states is so large that finding a specific one (by writing sotware to initiate that first state) is "novel".
This reminds me of the following question - Walk into a digital camera store and ask the salesman "How many different pictures can this digital camera take?" The first reaction is that it can take a picture of anything, surely the number of things it can take a picture of are infinite, yet the number of possible pictures is not infinite. Each pixel can have a finite number of states (colors) and there are an easily enumerated number of pixels so the number of pictures it can take is easily defined by those two numbers, one to the power of the other. Yet, copyright protection is afforded to digital pictures, just like analog pictures, despite the finite number of possible pictures.
" What you are really describing here is a discovery."
You could make this argument for any technology. The guys who design lasers, and the guys who build microelectronic devices and the guys who build appple sorters all have the same uneasy feeling that a patent might be bad for them, since they might inadvertently infringe or be prevented from designing or building something. However, most of the successsful societies in this world have reached the opoposite conclusion and have set up a patent system.
"Furthermore, this state you have discovered is one which is intrinsically linked to a function, i.e. the state describes the function. The two cannot be separated - the machine "reacts" to the state it is in along a set path, which performs whatever database sorting you described."
True, but not much different from selecting from bolts and levers and gears to produce a machine.
"You cannot create a 'new' device from software"
Surely you can. No one has ever created a machine comprising a computer with the starting state defined by your software.
"you cannot extend the bounds of computing beyond the limits of the hardware - you have simply discovered a pre-existing state"
I think the point is that it was *not* pre-existing. It was a pre-existing
"No, the computer is the machine, and the software is the instruction that the machine follows, as the poster said."
I agree, but I can build a hardware version of a computer running software that is entirely hardware with only hardware parts (firmware or hard-coded ones/zeros) instead of the software. You're pointing to a difference that has no substance.
"Is it that difficult to see?"
Let me ask the same question - Do you really think that a computer running software is somehow not "real?" That I can't build an only hardware version?
"You seem to be of the opinion books (and the content therein) can't be patented, right? "
Yes - sort of. The contents can be patented, but writing a book and obtaining a copyright on it does not give you patent-like rights in any invention you describe in the book. You have to prove that your invention is entitled to such protection, i.e., that it's new, etc.
"Now, imagine computers get smart enough to understand normal written words...then suddenly, ordinary text could be patented too?"
You can already do this. If I write a book about a new invention, copyrighting the book gives me no rights in the invention. However, I can write a patent and get rights in the same invention described in the book. This is my point. Software is an invention. If we are going to protect it, it needs to be good enough to deserve protection, and copyright isn't the right vehicle for that protection.
"Software is *not* a machine, it does *not* produce electrical charges: the electrical charges are provided by the powersupply of the computer. "
This is still a difference without meaning. The patent would then just claim a computer running a memory have ing a charge pattern as follows... and define the charge pattern that is created by the software. There's no point in doing it taht way.
"The only thing it does, is instructing these charges how to behave. Since instructions are mental processes and not physical objects, it falls outside the scope of patents."
Baloney. My apple sorter has electrical relays and mechanical arms. "The only thing it (my apple sorter) does" is tell the parts what to do to achieve the sorting desired by my mental processes used to design my apple sorter. Therefore the apple sorter should also be outside of patents. That's ridiculous. Software is simply a set of instructions that controls how the computer operates. A computer operating according to those instructions is as real as an apple sorter and as much of a machine as an apple sorter.
"Except that software patents are given for the idea of sorting a database"
We agree they should not be, and the patent law as I understand it agres with both of us.
"Further than that, proving non-obviousness is nigh on impossible. So the patent office doesn't try."
The patent office has to show that it's obvious, but they have to do that with everything else too. Why should it be harder for software?
"Yet to any experienced software engineer, many database sorting mechanisms are obvious."
Then the patent office should reach the same conclusion. This problem occurs in all areas of technology, not just software, why do you think software should be handled differently?
"Add to that, if you hold a patent for an apple sorting machine, that gives you the ability to monopolise efficient apple sorting (until someone devises a competitive alternative).
If you hold a patent for database sorting, nobody else can sort databases. You've just crippled every industry in the country that uses databases; that's the IT industry, finance, telecomms, travel, manufacturing, oil, and pretty much everything else. And nobody will devise a competitive alternative because you've patented the idea, not the implementation. So nobody will extend the sorting to optimise it, simplify it, take it into new areas - in other words, innovation is being prevented."
If the stakes are that high, the incentive to design around will be higher. Patent law specifically states that patents are not granted on ideas, but only on specific inventionsthat implement the idea.
"So no, I don't agree that you are entitled to the same protection. It just isn't needed. Look how far the software industry has come in the last 50 years, without software patents, and look what it's provided, and please, tell me exactly what benefit there is to society from introducing them?"
The same benefit that all other forms have technology have received. Increased incentives to improve, increased incentives to try other options that wouldn't be funded or tried if not for a blocking patent.
"All you have actually done is write a detailed technical manual describing a method of sorting a database... unless you can think of a way in which a computer program is something other than a detailed set of instructions."
There are two ways to look at this, which is why the copyright and patent laws have gotten so tangled up. First, you can see software as an actual physical machine. When software is loaded into memory, it produces an actual device with electrical charges that causes the processsor and display and printer to perform in a predetermined way (at lest it should:-) That's a machine and should be patentable if it's something that is new and meets patent standards. Software code is simply the way we describe that machine.
You can also see software as a written manual. It has the author's style and expression just like any other written book or manual. Different source code can compile to the same object code. The compilation process removes much of the author's expression IMHO, but not all. I would limit copyright protection more than it is now, which would lead to more freedom to write competing programs.
"The problem lies in that you'd get a very general patent that actually COULD cover any kind of database sorting."
This is a legitimate concern, but why are you more worried that the software patent would cover all types of database sorting than you are that the example patent would cover all types of object sorting, and not just fruit or apples specifically?
Sure, you're worried about overly broad patents, but that's not a concern that's limited to software. It sounds like an unjustified scare to me sinc eit applies to everything patentable. The decisiosn was made long ago that the rewards for issuing patents are greater than the downside risks. Patents are all about incentives to build the better database sorter and incentives to design around the last patented sorter to get an even better one. Why give up a system that has worked and driven technology for the betterment of all?
You need to cross a river. You construct and patent a "bridge". VOILA! Now no-one can build another bridge of *any* design ""
That's not how patents work. There are lots of patents issued on specific designs for bridges. Those patents ensure that the guy who spent the time and money to figure out a good bridge design can pay for his R&D. The competitor is encouraged to design another type of bridge, and guess what, that improves the whole bridgebuilding technology as designers consider and reject alternatives.
Sure you can label it "anti-competitive" if you want, but most countries disagree and have passed patent laws. At least in the U.S. it was considered so important to encourage "progress of the useful arts and sciences" that patents were written into the Consititution.
I'm still lookiing for a reason why software should be considered different from other patented technology.
"Problem with software patents, you can't search patent database to see if you are not infriging. If you do you might be considered contaminated, since you have seen other ideas."
I presume you are using "contaminated" in the context of situations in which copyright infringement was alleged and the plaintiff has to show the defendant copied the plaintiff's work. In those situations, the defendant is far better off if he can show that he's never seen the plaintiffs copyrighted work, i.e., he's not "contaminated."
However, your concern here is purely related to protection of software by copyright and your concern just emphasizes my point taht patent law makes more sense for things that are more like machines than books. Under patent law, there is no such thing as "contamination" since there's no need to prove copying. There's a written description of the patented invention - the claims - and if you're not doing the same thing specified in the claims, then you're not infringing. A copyright 1) has no written description of what is protected, 2) no one checks to make sure that it's worthy of protection (new, etc.) and it lasts for nearly forever.
"That reminds me like getting patent on "closed-room murder" for mystery novels."
And that's why patents make sense for software - no one could get a patent on a "closed room murder" since it's not something that's new and therefore it does not meet the standards of something that is deserving of a patent. YOu may have a concern about the poor quality of patents being issued - I have that concern too, but that's far different from saying that software patents don't make sense.
" "Didn't this take the same type of mental effort that it took to design the apple sorter?"
No."
Why not?
""Aren't I entitled to the same type of protection?"
No."
Why not?
""Doesn't it make sense to use patent protection designed for protecting useful things and not copyright designed for protecting expression of ideas, but not the idea itself?"
No."
Why not? How about some reasoned discussion here.
In case my point was not clear, a computer program, to my technical mind, is more like a machine than a book. Just because you build your software machine out of words and code phrases does not make it less of a device that performs a function just like any other patentable machines.
"... some clean-room engineering and produced a compatible BIOS, enabling competition. Had IBM patented the BIOS, there would have been no competition..."
You assume they could have "patented the BIOS," but that's not certain. Big mainframes used a similar startup procedure and that might well have prevented a patent from being granted on a BIOS.
In fact, when IBM challenged the alleged copying of their BIOS the court concluded that the copyright in the BIOS was not infringed, in part because the BIOS seemed more like part of a machine and the competing BIOS was simply performing the same functions. Copyright law for computer software has changed enough that the "clean room" procedure probably won't work for application software even if you write your own code and never see my source or object code. copyright protection now extends to what the program does (its machine-like function) and not just how you wrote the code, i.e., the "expression" you used and your style of code drafting.
"This has nothing to do with entitlement, and everything with consequences for the economy and innovation. Software patents are generally not beneficial to either, so it does not make sense to have them."
So why is it that it's beneficial for innovation and the economy to issue a patent on the apple sorter, but not the database sorter? If you want to argue that all patents are bad for innovation and the economy, fine, but almost all countries disagree and have patent laws.
1) I design a new type of machine to efficiently sort apples - I get a patent on my machine. You can build a machine to sort appples, that works differently, but not one that works like mine.
2) I write a book about how to sort apples and I describe my machine and how it works. I get a copyright on my book. You can write a book about how to sort apples and can describe my cmachine and how it works, but you can't use the saem words I used, nor paraphrase them.
My patent gives me some protection for my useful machine. My copyright on my book does not protect my machine or my method of sorting apples.
Now I write a new computer program to efficiently sort a database. Didn't this take the same type of mental effort that it took to design the apple sorter? Aren't I entitled to the same type of protection? Doesn't it make sense to use patent protection designed for protectiong useful things and not copyright designed for protecting expression of ideas, but not the idea itself?
At least with the patent, the duration is limited (20years) and there is someone checking to see that the patented item is really new and worthy of protection. Copyright doesn't have those checks, it's just a form you file and you can get over a 100 years. Teh reason is that it's not supposed to cover useful inventions, it's just supposed to cover expression, so anyone can write the same software as long as they don't copy the way you expressed your idea.
Slashdot Mirror
You are wrong on this. It is not necessary to actually make the object to get it patented. I am 100% certain of this.
A computer running software is a physical object, not a mental process.
Two identical computers running two different programs are different machines. They accept different inputs, process the data differently to produce different outputs. There's really no reason to require the computer part of the invention in the patent. If you look at patents on apple sorters, for example, you will find that they probably don't make any reference to a base or a frame or other supports that the gears and levers are mounted upon, yet all assple sorters will have such a base. Similarly, there is no reason to reference the underlying physical computer when patenting software.
I completely agree, and never intended to argue a natural "right" to a patent. The point of my first post was found in the title - software is more like a machine, than a book. The useful aspects of machines are protected by patents, so it makes more sense to use patents to protect software than it does to try to protect it with copyright, which is used for protecting the particular "expression" of an idea, but not the underlying idea itself.
I disagree with your phrasing, but not your underlying point. Society has concluded that a monopoly granted for a limited time is good if it stimulates R&D, or causes the inventor to disclose that which he would otherewise not disclose.
So we come back to the reasons for my original post.
1)Patents make more sense than copyright for protecting software since computers running software are more like machines than books.
2) The general opinion here seems to be that software patents are bad, so I'd like to know why people think that. Specifically, what do people think is different about software technology from all the other areas of technology covered by patents.
I don't believe in software patent "entitlement" any more than you. I'm not sure I even believe that software patents are good for society. I simply haven't made up my mind, and I'm looking for answers. I have read some of the studies that others here have pointed me to. If anyone reads t
But isn't it possible that if RSA *had* been patented in Europe, the "number of better algorithms" that were later developed in Europe might have been developed a few years earlier under the competitive pressure to design around the RSA patent? If better algorithms had been developed earlier, might Europeans be farther ahead now than they are?
A good point. For example, pharmaceuticals are not patentable in some countries. The argument is that medicine shouldn't be restricted by patents. It's probably worth noting that medical development, however, is mostly occurring in countries where pharmaceutical patents are available, and there's reasonably good evidence that the cost of pharmaceutical R&D would often not be acceptable if they couldn't be recovered during the exclusive atent period. So what, if anything, is it about software that you think is like pharmaceuticals?
I completely agree, but while it's easy to see when a patent restricts someone, it's much harder to see when the missing patent incentive casues us to lose out on a new or better pharmaceutical or software invention.
Again, we are in agreement.
And here we get to the crux. Why do you and others here think this, when it does not seem to be true for all other industries and technologies?
If this was true, I would join you, but I don't think it is true. Patents have served a major part in driving technolgy forward for hundreds of years. Why do you suddenly think it won't work any more?
But if the data sorter is trivial to make, then it's probably already either known (not new) or a simple variation on something that is already known (obvious) so it can't be patented. But let's suppose that a company decides that they could do better if they could come up with a data sorter that was better than anything out there, but it would cost them a few years of work to find out if it was even possible, and even if it was, after developing it, their competitors would immediately copy it. They have no incentive to try to develop it because you have removed that incentive. HOw do you know how many such improvements you have lost? I surely do not knkow either, but I can see historical evidence that providing the incentive for improvement pays off to society in the end.
There never was and is not now any requirement for the patented item to have taken immense amounts of time to have been developed. I'm not convinced that software is any easier or harder to develop than any other invention. In addition to the reduced incentive to develop software by eliminating patents, tehre's another effect. Patents are ther to encourage disclosure of improvements to the world. Suppose our data sorter company decides to build the new and improved data sorter that they really need. Are they going to tell the world about their breakthrough? No. They keep it as a trade secret, since if they tell, they lose their competitive advantage. Professionals in the data sorting field
"you agree software is not a device per se?" Yes I do, but it's directly transformable into an actual physical device simply by compiling it and loading it into a suitable computer. "How is this different from me taking your book describing the apple sorter reading it, following the instructions, and operating it on data (apples)? My point exactly! It's not at all different! If you want to protect your new design for an apple sorter, you don't write a book, send in the $30 and get a copyright, then sue someone 50 years later for copying the apple sorter design you described. That would be unfair, since no one checked to see if the design of your apple sorter in your book was really new. You have to write a patent application, send it in, have it examined to see if it's new, then you only have 20 years of protection. Right now, software is following the first model more closely than the second. It just does not make sense to me. "You seem to be asking why software should be a special case and I agree, but why should it be a special case not covered by copyright?" I think software should definitely be covered by copyright. Anyone who copies the source code exactly or the object code exactly, should be liable for copyright infringement. However, if the open source community wants to replicate the functions of that software by writing their own routines and without "copying" the "protected expression" of the software author, then they should be allowed to do that, unless there is a patent, because they are copying what the software *does* and how it *functions* and not copying the style or text or "expression" selected by the software author. "What you describe is basically a special 'state' of the computer - certain bits in memory are set to a predetermined pattern described by software" Agreed. "... but there is nothing inherently novel about this - all the states of the machine can, given sufficient computing power, be enumerated." True. But the number of possible states is so large that finding a specific one (by writing sotware to initiate that first state) is "novel". This reminds me of the following question - Walk into a digital camera store and ask the salesman "How many different pictures can this digital camera take?" The first reaction is that it can take a picture of anything, surely the number of things it can take a picture of are infinite, yet the number of possible pictures is not infinite. Each pixel can have a finite number of states (colors) and there are an easily enumerated number of pixels so the number of pictures it can take is easily defined by those two numbers, one to the power of the other. Yet, copyright protection is afforded to digital pictures, just like analog pictures, despite the finite number of possible pictures. " What you are really describing here is a discovery." You could make this argument for any technology. The guys who design lasers, and the guys who build microelectronic devices and the guys who build appple sorters all have the same uneasy feeling that a patent might be bad for them, since they might inadvertently infringe or be prevented from designing or building something. However, most of the successsful societies in this world have reached the opoposite conclusion and have set up a patent system. "Furthermore, this state you have discovered is one which is intrinsically linked to a function, i.e. the state describes the function. The two cannot be separated - the machine "reacts" to the state it is in along a set path, which performs whatever database sorting you described." True, but not much different from selecting from bolts and levers and gears to produce a machine. "You cannot create a 'new' device from software" Surely you can. No one has ever created a machine comprising a computer with the starting state defined by your software. "you cannot extend the bounds of computing beyond the limits of the hardware - you have simply discovered a pre-existing state" I think the point is that it was *not* pre-existing. It was a pre-existing
"No, the computer is the machine, and the software is the instruction that the machine follows, as the poster said."
... and define the charge pattern that is created by the software. There's no point in doing it taht way.
I agree, but I can build a hardware version of a computer running software that is entirely hardware with only hardware parts (firmware or hard-coded ones/zeros) instead of the software. You're pointing to a difference that has no substance.
"Is it that difficult to see?"
Let me ask the same question - Do you really think that a computer running software is somehow not "real?" That I can't build an only hardware version?
"You seem to be of the opinion books (and the content therein) can't be patented, right? "
Yes - sort of. The contents can be patented, but writing a book and obtaining a copyright on it does not give you patent-like rights in any invention you describe in the book. You have to prove that your invention is entitled to such protection, i.e., that it's new, etc.
"Now, imagine computers get smart enough to understand normal written words...then suddenly, ordinary text could be patented too?"
You can already do this. If I write a book about a new invention, copyrighting the book gives me no rights in the invention. However, I can write a patent and get rights in the same invention described in the book. This is my point. Software is an invention. If we are going to protect it, it needs to be good enough to deserve protection, and copyright isn't the right vehicle for that protection.
"Software is *not* a machine, it does *not* produce electrical charges: the electrical charges are provided by the powersupply of the computer. "
This is still a difference without meaning. The patent would then just claim a computer running a memory have ing a charge pattern as follows
"The only thing it does, is instructing these charges how to behave. Since instructions are mental processes and not physical objects, it falls outside the scope of patents."
Baloney. My apple sorter has electrical relays and mechanical arms. "The only thing it (my apple
sorter) does" is tell the parts what to do to achieve the sorting desired by my mental processes used to design my apple sorter. Therefore the apple sorter should also be outside of patents. That's ridiculous. Software is simply a set of instructions that controls how the computer operates. A computer operating according to those instructions is as real as an apple sorter and as much of a machine as an apple sorter.
"Except that software patents are given for the idea of sorting a database" We agree they should not be, and the patent law as I understand it agres with both of us. "Further than that, proving non-obviousness is nigh on impossible. So the patent office doesn't try." The patent office has to show that it's obvious, but they have to do that with everything else too. Why should it be harder for software? "Yet to any experienced software engineer, many database sorting mechanisms are obvious." Then the patent office should reach the same conclusion. This problem occurs in all areas of technology, not just software, why do you think software should be handled differently? "Add to that, if you hold a patent for an apple sorting machine, that gives you the ability to monopolise efficient apple sorting (until someone devises a competitive alternative). If you hold a patent for database sorting, nobody else can sort databases. You've just crippled every industry in the country that uses databases; that's the IT industry, finance, telecomms, travel, manufacturing, oil, and pretty much everything else. And nobody will devise a competitive alternative because you've patented the idea, not the implementation. So nobody will extend the sorting to optimise it, simplify it, take it into new areas - in other words, innovation is being prevented." If the stakes are that high, the incentive to design around will be higher. Patent law specifically states that patents are not granted on ideas, but only on specific inventionsthat implement the idea. "So no, I don't agree that you are entitled to the same protection. It just isn't needed. Look how far the software industry has come in the last 50 years, without software patents, and look what it's provided, and please, tell me exactly what benefit there is to society from introducing them?" The same benefit that all other forms have technology have received. Increased incentives to improve, increased incentives to try other options that wouldn't be funded or tried if not for a blocking patent.
"All you have actually done is write a detailed technical manual describing a method of sorting a database ... unless you can think of a way in which a computer program is something other than a detailed set of instructions."
There are two ways to look at this, which is why the copyright and patent laws have gotten so tangled up. First, you can see software as an actual physical machine. When software is loaded into memory, it produces an actual device with electrical charges that causes the processsor and display and printer to perform in a predetermined way (at lest it should :-) That's a machine and should be patentable if it's something that is new and meets patent standards. Software code is simply the way we describe that machine.
You can also see software as a written manual. It has the author's style and expression just like any other written book or manual. Different source code can compile to the same object code. The compilation process removes much of the author's expression IMHO, but not all. I would limit copyright protection more than it is now, which would lead to more freedom to write competing programs.
"The problem lies in that you'd get a very general patent that actually COULD cover any kind of database sorting." This is a legitimate concern, but why are you more worried that the software patent would cover all types of database sorting than you are that the example patent would cover all types of object sorting, and not just fruit or apples specifically? Sure, you're worried about overly broad patents, but that's not a concern that's limited to software. It sounds like an unjustified scare to me sinc eit applies to everything patentable. The decisiosn was made long ago that the rewards for issuing patents are greater than the downside risks. Patents are all about incentives to build the better database sorter and incentives to design around the last patented sorter to get an even better one. Why give up a system that has worked and driven technology for the betterment of all?
You need to cross a river. You construct and patent a "bridge". VOILA! Now no-one can build another bridge of *any* design "" That's not how patents work. There are lots of patents issued on specific designs for bridges. Those patents ensure that the guy who spent the time and money to figure out a good bridge design can pay for his R&D. The competitor is encouraged to design another type of bridge, and guess what, that improves the whole bridgebuilding technology as designers consider and reject alternatives. Sure you can label it "anti-competitive" if you want, but most countries disagree and have passed patent laws. At least in the U.S. it was considered so important to encourage "progress of the useful arts and sciences" that patents were written into the Consititution. I'm still lookiing for a reason why software should be considered different from other patented technology.
"Problem with software patents, you can't search patent database to see if you are not infriging. If you do you might be considered contaminated, since you have seen other ideas." I presume you are using "contaminated" in the context of situations in which copyright infringement was alleged and the plaintiff has to show the defendant copied the plaintiff's work. In those situations, the defendant is far better off if he can show that he's never seen the plaintiffs copyrighted work, i.e., he's not "contaminated." However, your concern here is purely related to protection of software by copyright and your concern just emphasizes my point taht patent law makes more sense for things that are more like machines than books. Under patent law, there is no such thing as "contamination" since there's no need to prove copying. There's a written description of the patented invention - the claims - and if you're not doing the same thing specified in the claims, then you're not infringing. A copyright 1) has no written description of what is protected, 2) no one checks to make sure that it's worthy of protection (new, etc.) and it lasts for nearly forever. "That reminds me like getting patent on "closed-room murder" for mystery novels." And that's why patents make sense for software - no one could get a patent on a "closed room murder" since it's not something that's new and therefore it does not meet the standards of something that is deserving of a patent. YOu may have a concern about the poor quality of patents being issued - I have that concern too, but that's far different from saying that software patents don't make sense.
" "Didn't this take the same type of mental effort that it took to design the apple sorter?" No." Why not? ""Aren't I entitled to the same type of protection?" No." Why not? ""Doesn't it make sense to use patent protection designed for protecting useful things and not copyright designed for protecting expression of ideas, but not the idea itself?" No." Why not? How about some reasoned discussion here. In case my point was not clear, a computer program, to my technical mind, is more like a machine than a book. Just because you build your software machine out of words and code phrases does not make it less of a device that performs a function just like any other patentable machines.
"... some clean-room engineering and produced a compatible BIOS, enabling competition. Had IBM patented the BIOS, there would have been no competition ..."
You assume they could have "patented the BIOS," but that's not certain. Big mainframes used a similar startup procedure and that might well have prevented a patent from being granted on a BIOS.
In fact, when IBM challenged the alleged copying of their BIOS the court concluded that the copyright in the BIOS was not infringed, in part because the BIOS seemed more like part of a machine and the competing BIOS was simply performing the same functions. Copyright law for computer software has changed enough that the "clean room" procedure probably won't work for application software even if you write your own code and never see my source or object code. copyright protection now extends to what the program does (its machine-like function) and not just how you wrote the code, i.e., the "expression" you used and your style of code drafting.
"This has nothing to do with entitlement, and everything with consequences for the economy and innovation. Software patents are generally not beneficial to either, so it does not make sense to have them."
So why is it that it's beneficial for innovation and the economy to issue a patent on the apple sorter, but not the database sorter? If you want to argue that all patents are bad for innovation and the economy, fine, but almost all countries disagree and have patent laws.
1) I design a new type of machine to efficiently sort apples - I get a patent on my machine. You can build a machine to sort appples, that works differently, but not one that works like mine.
2) I write a book about how to sort apples and I describe my machine and how it works. I get a copyright on my book. You can write a book about how to sort apples and can describe my cmachine and how it works, but you can't use the saem words I used, nor paraphrase them.
My patent gives me some protection for my useful machine. My copyright on my book does not protect my machine or my method of sorting apples.
Now I write a new computer program to efficiently sort a database. Didn't this take the same type of mental effort that it took to design the apple sorter? Aren't I entitled to the same type of protection? Doesn't it make sense to use patent protection designed for protectiong useful things and not copyright designed for protecting expression of ideas, but not the idea itself?
At least with the patent, the duration is limited (20years) and there is someone checking to see that the patented item is really new and worthy of protection. Copyright doesn't have those checks, it's just a form you file and you can get over a 100 years. Teh reason is that it's not supposed to cover useful inventions, it's just supposed to cover expression, so anyone can write the same software as long as they don't copy the way you expressed your idea.