Software Patents stifle innovation.
We don't want to fall into the same mess as the Americans with Software Patents.
There should be a world wide ban on Software Patents.
A date has been set for a hearing in the High Court to determine the legality of a patent office review of the level of protection it offers to software patents in the UK. On November 19, the High Court will hear four small UK companies argue that the patent office's refusal to accept patent claims covering disks and downloads …
With all respect to programmers - software is not an invention. Hard work, yes, but not a "novel idea or concept reduced to practice". Software is a set of instructions required to operate a device in order to achieve a specific result. Therefore it is pre-determined by the nature of device and by the desired result and as such is deterministic. In this way it is even less of an invention than a paperback novel. If the latter is no patentable the former should not be either.
Does that mean that..... if you have invented and built and are running a Quantum Computer which runs on Virtual Source Code, it is only possible for New machines using such Code to be as Clones to the Original and are therefore Liable to Require Purchase of a Licence, payable to the Originator.
Wow, what a crock and an easy way to make money for nothing.
Someone better informed correct me if I'm wrong but ... my understanding is that the EPO is happy to grant patents on software and, indeed, the usual suspects have sizeable arsenals of S/W patents to complement their US ones. It's just that EU member state law says that these valid software patents do not grant the holders a monopoly on their inventions - they are at present worthless.
This is why any change in the status of software patents in EU member states would have an immediate and catastrophic effect: all of a sudden, products which were legitimately being bought and sold would become illegal - the fact that they were infringing SW patents would change from being inconsequential to being illegal.
So if I understand it correctly, the UK Patent Office (I refuse to refer to it by its new name which includes the oxymoronic term "Intellectual Property") won't accept applications for software inventions because the resulting patent would have no value under current legislation. And the companies believe this is unfair, as they have to pay the higher fees to the European Patent Office, whose patents would also have no value under the law of EU member states. Their legitimate reason for filing software patents in the EU can only be to establish a priority date for future international filings - presumably in the USA.
So are the UK Patent office declining to accept initial filings for inventions which are based on software? Or are they saying they won't examine them? Or are they saying they won't grant patents on them? And if the patent is not granted in the country of first filing, does that still establish a priority date in the USA?
It all goes to show, when Law, Politics and Technology meet you end up with a complicated mess.
Thank God someone else is finally saying this out loud.
Writing software is more a form of carpentry than a form of artistry !!!
A programmers skill my improve the quality of his work but that is not innovation. And software companies, who poll their customers to discover what features their users want, can hardly call it unique or non-obvious if 95% of those polled thought of it.
The days of computer experts being thought of as wizards or geniuses is long gone. GET OVER IT !!! They need to stop expecting to earn 100's of millions off of a few weeks work!
Bill Gates did not get where he is because he was a great programmer or a great business man. He is a lawyer and he simply abused the inadequacies of the existing laws and applied slum-lord tactics to his marketing model.
There are two problems for the patent holders:
1. Software patents are expressly not enforcible in Europe.
2. European law forbids retroactive enforcement of a new law.
If the European parliament ever makes the grave mistake of legalising software patents, then any software patents (falsely) granted in the past will *not* suddenly become enforcible (which would constitute retroactive enforcement) but must be re-applied for. In the meantime, any software which only managed not to violate those patents by virtue of the patents themselves being a crock, can be cited as prior art in order to block the re-applications.
Beresford wrote that book "How to patent software under the European Patent Convention" which is akin to incitement to crime, since patenting software under that law is illegal.
Beresford has been angry with the UK High Court for ruling that his nice schemes to circumvent the spirit and letter of the law by... well, by convincing everyone that such things as spirit and letter don't really matter when there is profit to be made... that his nice schemes were not legit.
So he has rounded up this gang of clients to stand and make a 'fight' of it.
At stake is of course a lot more than the claimed case. This is about overturning UK high court case law to allow the flood gates to be opened once more.
It's quite probable that Beresford is not charging his clients a penny, rather he is quite possibly being sponsored by other businesses *cough* microbe *cough* soft *cough* who have a vested interest in seeing software patents brought back from near zombification in the UK.
And if Beresford fails, no matter, the patent industry is building up a new attack on the letter and spirit of the law, not to mention economic sense, by plotting to circumvent the UK High Court with a new, shiny, tightly controlled European patent court run and owned by the patent industry.
You really have to love these guys.
Copyright certainly *does* protect software. Allowing software to be patented stifles innovation, because it is usually possible to write software that is significantly (in copyright terms, "significant" means "at least 85%") different to achieve the same end result. Different software which achieves the same result may be more efficient, faster, compile to smaller executables, or be better than the "original" in other ways without infringing copyright.
Thus we see that the *purpose* of software patents is to stifle innovation.
Software patents are unwanted and unnecessary. The application for patents is absurd as implemented in the real world with software.
A patent should detail the result well enough for someone skilled in the art to reproduce THAT DEVICE. If it results in a different device, the patent is too broad or unspecific.
However (ignoring that this is a problem with patents as they are now for everything) if you do this for software, you've now given the source code away. Without the source code, you are not specific enough and with it you've already got it covered with copyright.
And if the way software patents are applied for, this would have been acceptable:
A method of controling rodent or other small pest species by means of a storage facility with a one-way method of ingress, use of which causes the device to attain a state whereby egress is blocked, either by the termination of said interloper or by impeded movement.
Now build a better mousetrap and NOT have my solicitors beat a path to your door.
Patents are now becoming a statement of the problem not the solution. "We need to stop people software tuning our device out of spec" not "do x y and z and the device is unable thereafter to change tuning by means of a blown fuse as detailed in spec A".
And how does making people have to apply for an EU/US patent rather than a UK one make it hard for them?
a) nobody else can block them by patent trolls
b) no licensing of patent this patent relies upon (whether you knew it or not)
c) though others can rip your patent off in the UK, you can rip others' patents and there a lot more others than you
d) if patenting abroad is a problem, you'd still have the issue that someone in italy will have THEIR patent in italy and now you
1) cant sell in italy without paying the fees
2) must sell in italy or fall afoul of trade restraint
the only ones that are DEFINITELY being hurt by no SW patent are the lawyers and the patent office. After all, each patent application pays the clerks and each attempt to get money for your patent has a solicitor drawing up the agreement.
Businesses? They only lose out.
..is a rare thing the problem with software patents in the states is that they've been granted for trivial and obvious things, as is the way in the states they leave that argument for a future challenge in court.
I'm willing to accept that there has been genuine innovation in software, Vladimirs point that it's dependent upon a machine isn't actually true, algorithms can be independent of a specific architecture or instruction set thus are not limited by a device.
Patents are supposed to encourage innovation, by allowing the inventor to profit from their idea. Without them, after years of working on something new, the inventor reveals his creation only for it to be copied and produced for peanuts in a sweatshop in China.
There's obviously problems with them though, some kind of validity test is required.
The both have a beginning middle and an end, and the best books ( like code ) can run infinitely until the paper wears thin.
Surely code could be copyrighted under the same laws that govern books ( and I assume e-books ).
Programmers merely become authors, of a predetermined "story" that follows the rule set of the authors creation.
Maybe software isn't an invention, but it is someones design. A set of "chapters" that enable the user to carry out whatever it is the author wants the user to go through.
On a side note - Bill Gates for Horror Writer of The Last 40 Years?
"A set of instructions" is by definition a large part of what a patent is. If your idea cannot be reduced to a set of instructions, it cannot be patented.
So if you implement your "set of instructions" on a programmable electronic device, it can't be patented.
But if you implement it with cogs and levers, or pipes and valves, it can be patented.
That makes no sense, except to software freeloaders -- people who expect programmers to work for free.
Working for free gets real stale real quick shortly after you graduate, and that is as true in China and India as it is in the USA or UK.
Declining and uncertain renumeration is largely why fewer and fewer intelligent and/or enterprising people are choosing programming and computer science as careers.
Complex software should be patentable!
The distinction should be whether the idea is trivial.
Any trivial idea, hardware or software, should NOT be patentable. And, of course, any previously discovered idea should NOT be patentable.
When you stake out a patent in software, why is it that you are allowed to freeload off someone else's innovation just because you thought of it first? As long as they thought of it independantly, he should be clear? Or do you want to patent the idea and let someone else come up with the implementation and then take all his profit?
Complex software isn't patentable because nobody writes a patent for complex software. The patents are written about a solution to a complex problem (or simple one but "on the internet") and NOTHING about how this solution was made.
What about "releasing their idea to the public"? If you can't release your idea as a trade secret WITHOUT patent protection, what is the public getting from the deal? We agree to a monopoly (which NEVER benefits the customer) in retrn for which you won't keep your idea secret. But if you could never keep your idea secret, why should we allow you a monopoly? We get only the downsides. If you won't release it without patent protection, and can't keep it a trade secret, keep it to yourself. See how much your "IP" is worth when you aren't selling it (as opposed to your home, a packet of crisps or a dirty mag: worth something even when you aren't selling).
Stop freeloading off everyone else on the planet and make your own way in life.
The set of instructions in a current patent describe how to build the patentable object. (Or in the case of manufacturing machinery how the patentable object builds other objects.)
The instructions in a computer program describe are the product itself, not how to make it. These instruction describe a procedure that, for the most part, could theoretically be carried out mentally, verbally or on paper. We do not have patents for business processes, algorithms or mathematical formulae carried out in "meatspace", so why should plugging them into a computer and having a shiny UI make them suddenly patentable.
Any patent law covering computers should also cover business process, algorithms and formulae. Then we can write of science as too much life hard work and go back to the hand-plough and pick-axe....
I'm sorry, but Vladimir is wrong if he believes that software should not be patentable simply because it is expressed in terms of machine code: that is only the tool set used to implement the idea. It is akin to saying that anything made out of metal is not patentable simply because you used a hacksaw, vice and sheet steel to make it, without any reference to exactly _what_ it is you have made.
It is true that most software patents are of dubious quality, being more concerned with protecting a particular implementation (or 'expression') than fundamental ideas. However, where truly novel _algorithms_ are invented there is no reason they shouldn't be patented as mechanisms, rather than covered by copyright that only protects that particular implementation of the idea. This is the way is has always been - witness, amongst others, the patenting of RSA in 1983. I don't have a problem with that.
Where there is a problem is in 'patent-creep', of the style seen in the States. This may be where patents are used to protect things that are better covered by copyright. It might also be that overly broad claims are permitted to stand, such that the patent covers not just what was invented but any other mechanism that achieves the same end no matter how obvious that end may be.
The protections of a patent are the furthest reaching of any in intellectual property law: if they are not to distort the market they should be reserved solely for their original intended purpose.
"The distinction should be whether the idea is trivial."
A fair point, but the problem as I see it is that it's quite difficult to see what is trivial and what is not when it comes to software.
Take the spreadsheet: the basic idea of it is very simple indeed. So is it "trivial"? I guess not, in hindsight.
What if I come up with an idea that takes 10,000 lines of code to make work? I guess that is non-trivial -- unless the next guy that does it manages it in 200 lines. Is it trivial now?
As a programmer myself, I certainly appreciate the idea of making moneyfrom code. But it's not that simple. I'd also like to live in a world where I could give away my work to a good cause, gratis. Software patents would put the OSS world in serious risk.
"That makes no sense, except to software freeloaders -- people who expect programmers to work for free."
You are confusing copyright and patents.
Copyright let you stop people taking your program and distributing/using it in a way that you don't want - often that means that you want to be paid for every copy. Nothing in copyright stops someone else writing a program that does the same thing as your program and selling/giving it to others... as long as they did not use any of your code in their program. You yourself probably benefited from other programmers' ideas when you wrote your program - that is part of the reason why software is advancing at the rate that it is: programmers stand on the shoulders of those who went before.
A patent is about an idea. If you were able to patent something you could stop someone from producing something that did the same as your idea/program.
Patents might be needed in the software world if there was a paucity of new software ideas, people not being willing to invest the time because they were not getting any reward. This is patently (ahem) not the case, programmers have shown themselves to be very inventive at producing new ideas.
In fact s/ware patents would be a bad idea since it would slow the rate of innovation.
Software patent terms would need to be shortened if they are ever to function correctly. A patent term of 2 or 3 years is more than enough to bring software products to market, and still short enough to wait out the trolls. I imagine most open source projects would be willing to wait for the related patents to expire. But a 15-20 year monopoly on software that is beyond obsolete in 5-7 years is just stupid.
This post has been deleted by its author
How about this idea - accept patents in the broadest sense possible while the law is in its present form? For example let a patent for 'a program that does stuff' be submitted to the EU patent office now.
In fifteen years time when the 'program that does stuff's' patent has expired, all conceivable future software programs will be covered by the prior art of an expired patent that was never enforced while it was valid.
Patents were originally created as a way to protect physical inventions (for 20 years) in return for publication of the technology's specifications and for use of the technology in the monopolist's products. Patents only exist because they foster innovation by allowing inventors (and investors) to reap the rewards of their toil.
Software patents, on the other hand cover inventive "concepts", not physical inventions. Applying patents designed to protect physical inventions to intangible, often broad and sweeping software concepts, benefits no one but those who have the resources to litigate; resulting in the stifling of innovation across the entire industry. This has been demonstrated time after time in the US; and until the advent of the General Public License and other Open Source initiatives, had caused the stagnation of innovation in many areas of IT. Abusive monopolists in the US are to this day targeting the Open Source industry with absurd patent litigation; hence the formulation of the Open Innovation Network as a consortium designed to fight such attacks.
Software is a very innovative area technology; and like other intangibles (such as art and music) deserves protection from theft. Copyright laws go a long way to protecting software publishers, but because they only protect the "form of material expression" from theft then this is considered by some to be insufficient. However it is important to realise that Copyright does cover the creation of derivative works. If that area of Copyright was strengthened, it could better satisfy those wishing to get protection from the likes of reverse engineering of sophisticated algorithms for example; and do it in a far less expensive way that is beneficial to the whole of society rather than the privileged few.
If you can't make money out of programming, you're either a moron, or lazy, or both.
I hold no patents or copyrights, and make plenty of money. Its not even hard - try cleaning or catering work for a while if you need to understand the concept of *hard* work.
And as for declining C.S. enrolments, it's just as likely caused by people seeing no future in starting software businesses without getting their arse sued off in patent litigation by big companies with armies of lawyers defending trivial patents.
> So if you implement your "set of instructions" on a programmable electronic device, it can't be patented. [...] That makes no sense, except to software freeloaders -- people who expect programmers to work for free.
Suppose I patent a method for buying goods or services with one click (it interfaces with a database to retrieve previously entered payment information). Now I want to charge every one who writes software that has the same effect $10 per end-user transaction. Now who's the freeloader?
And don't say you can't patent that because, in the US, Amazon did. And it held up in court. Twice. What's more, they didn't use it to get royalties but to stop a rival from writing anything of the sort.
If you think software patents are a good thing you clearly haven't seen the results thus far.
"This is why any change in the status of software patents in EU member states would have an immediate and catastrophic effect: all of a sudden, products which were legitimately being bought and sold would become illegal - the fact that they were infringing SW patents would change from being inconsequential to being illegal."
No, because -ex post facto- enforcement is prohibited under EU law (IMMSMC, that was what saved Pinochet: what he had done was not illegal under UK law at the time when he did it, therefore he could not be extradited on grounds of no dual criminality). A falsely-granted patent would *not* become enforcible simply because the law which made it bogus changed. It would remain false, and patent holders would have to re-apply for their patents -- which most probably would not be granted, due to the existence of substantial bodies of prior art.
Anonymous Coward wrote: "I'm sorry, but Vladimir is wrong if he believes that software should not be patentable simply because it is expressed in terms of machine code: that is only the tool set used to implement the idea. It is akin to saying that anything made out of metal is not patentable simply because you used a hacksaw, vice and sheet steel to make it, without any reference to exactly _what_ it is you have made."
No, using your ironmonger example let's say you've invented a hammer. The software for the hammer will be a set of instructions
- take hammer
- hold nail with sharp point to the board
- hit nail with the hammer
- if fingers hurt, adjust aim and repeat
- continue until nail is in the board
Clearly, this set of instructions is basically an operating manual and not a part of the invention.
Similarly, a software to run a washing machine (or iPod or A380) is a set of instructions how and in what sequence to operate the internal workings of the machine you invented to achieve a certain result. You can dispense with the software altogether and provide the user with large number of buttons instead + a manual which button to press and when.
If we agree that software is patentable then you would not be able to punch <2> <+> <2> <=> on a calculator without breaking someone's patent.
In the same way patenting algorithms is akin to patenting mathematical formulae - just imagine we needed to pay Pithagoras descendants every time we had to deal with a right-angle triangle...
The idea of patents is to provide a monopoly to the inventor as a reward for sharing their methods with the public. After a reasonable length of time, but while the invention still has value, the invention becomes part of the public domain. It's also only meant to protect *innovative* things that would not occur to a skilled practitioner of the art.
I could argue that 19 year old software has very little value, and should therefore not be protected by patent. It's also questionable as to whether ANY software is truly innovative, or simply an evolution of all the software that has gone before.
The same argument could be argued about copyright protection for software. Is 70 year old software really going to have ANY residual value for the public domain? If not, why are we, as a society, wasting resources protecting someones copyright for no benefit to society?
Divide current terms by 10 for software and it might work. 2 year patent terms or 7 year copyright protection for software makes some sort of sense. But the longer terms just don't cut it. But of course I live in a fools paradise. Money talks, and those with money will keep changing the rules to increase their money while keeping others out. It's the way of the world, and patents/copyrights have become just another tool by the ruling oligarchy to keep a firm control on things.