Slashdot Mirror
Report Says Patents Threaten Software Innovation
GORby_ writes "According to PriceWaterhouseCoopers, software patents are 'a particular threat to the European ICT Industry.' Quote from the report: 'There are particular threats to the European ICT industry such as the current discussion on the patent on software. The mild regime of IP protection in the past has led to a very innovative and competitive software industry with low entry barriers. A software patent, which serves to protect inventions of a non-technical nature, could kill the high innovation rate.' The full report (pdf) discusses Europe's ICT strategy."
There's a report that favors one side of a heated debate. Who thought we'd see the day.
Software patents are a reality, whether you and me want them or not. There is too much money on the pro side for software patents to go away, so stop dreaming. Better start thinking of ways to deal with the situation.
Have a look at the link in the story: www.pwc.com.nyud.net:8090...
This will have no impact on PWC's servers. Then again, I would have thought that they could afford to buy enough bandwidth to keep up.
flossie
Write now. Defend liberty
Although the preceeding comments have been largely on the order of "duh, /. users knew this for years!", we need to make sure this oppurtunity doesn't go to waste. If you've known it for years, then make sure other people know it as well. This is a good oppurtunity to spread the word. As the issue gets more press, it's going to be important to make sure it's given the gravity that it deserves.
Will this be enforced on existing software or is it just new software? All those applications that steal the outlook look and feel (or the components that mimic outlook) could mean Microsoft rake in a fortune.
Nothing costs nothing
The point is that they, unlike for example Richard Stallman, most surely have no axe to grind when they talk about software patents stifling innovation. When they complain about the effects of software patents, they are complaining only about their effect on the bottom line - and every informed analyst will know that. So their stance against software patents will carry a lot more weight than that of the people who've been crying out in the wilderness for all these years.
It's strange the friends we seem to be making these days ... First IBM, now PWC.
--
What short sigs we have -
One hundred and twenty chars!
Too short for haiku.
...that this is just the fault of the patent system, it is a flaw in the legal system, and it is a flaw in the community. Where patents are demonstratedly wrong, it should be easier for small companies or individuals to challenge that bad patent. Bad patents shouldn't just stand becasue one party has deeper pockets. And parties who are in thei right shouldn't be so cowardly as to run from these fights, if they want to change the system, they must work at it. It's all very well bitching about C&D and takedown notices, but you Americans meekly agree to do whatever they say anyway, grow some backbone.
I feel sorry for the USPTO. They obviously lack technical expertise, and can't afford the salaries to attain it. If they were getting feedback on what patents were downright bad from the court system, they could train and evolve to start granting more deserved patents.
Meine Schwester ist sehr, sehr reizvoll - Nietzsche
...is that the large software companies have patented so much of the fundamental building blocks of software engineering that even if you do come up with something truly "new and innovative" they can still get you on the sub-component functionality.
This means they can effectively hold to ransome any new software venture that turns out to be succesful, regardless of what they do.
There's however something special about this one. The magic word "PriceWaterhousCoopers" (try saying it quickly 10 times!) opens a lot of ears which remain deaf in other cases.
Donate free food here
This is great news. I know a lot of you think Duh, but you have to look at it from the other side. This report was not meant for the /. crowd, it was meant for the ministry of economic affairs in The Netherlands. The minister himself had to change it's vote after a vote in the parliament.
This give the recall of the vote a lot more cloud and hopefully will start to appeal to parliaments in other EU countries.
Disclaimer: This opinion was created without the use of any facts
Well, if profits are your motivation, then all patents, copyright laws, trademarks and registered designs are a good thing.
If, on the other hand, you're trying to write some decent software, patents are a very bad thing. What's more productive, coding, or running patent searches to see if you're allowed to use what you just wrote?
In fact, software patents and "real things" patents are very different beasts. Note the number of "RL activity... but on the Internet!" patents awarded. I couldn't patent, say, putting an index to a large document in the margin of each page. For some reason, though, a patent examiner awarded a patent on browser frames.
In any case, patent law was designed for the benefit of society, not the inventor. The aim was (and should be) to maximise the pool of useful innovations available, not maximise profits.
But, companies don't have ideas, people do. If we take away patents, some companies might get mad, but new software ideas will still appear. And FLOSS coders will be able to act on them.
A good number of times, patents don't do any good anyway, if you plan on producing products. Anything you do will infringe on somebody's patent, so they sue, and either you lose your profits, or cross-license, and gain competitors until the market is similar to one without patents. Except, of course, the non-patenters aren't in the market at all. And less competition = bad for the end users. And if you don't produce a product, merely patents, your company is a worthless leech on society.
There may be a few pieces of true innovation which deserve a good reward. This can be achieved, if desired, by selling the program at a reasonable price and being nice to your customers. Patents not required.
This may seem obvious, but legislators seem to forget (boy do they ever!) that the rationale for IP protections (patents, copyrights, trademarks, etc) has nothing to do with who deserves compensation for their work, and everything to do with guaranteeing that markets provide certain types of innovative goods for consumers. This is even codified in the US constitution:
"To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries;"
Of course, the "limited Times" phrase is there because protections prevent other types of innovation (based on immitation) from taking place. It is there so that (sadly) politicians can decide where to strike that balance for each market.
So the main question about software patents should be: to what extent has the absence of software patents negatively impacted the pace of software innovation? As far as I can tell, copyrights have been more than sufficient to strike the balance I talked about earlier, and patents would simply tilt the balance and harm innovation. European legislators should understand this and avoid going down the misguided path of the US congress in their counterproductive copyright extension route.
Imposing Libertarian views on everyone online since 1992.
"Well, how about CUSTOMS TARIFFS designed to bring the price of low wages countries products more in line with those in the high wages coutries???"
What a brilliant idea: let's increase the cost of goods to poor people in the West and shovel even more of their money to unproductive government workers through the new taxes.
I believe that if you were to look at...
Have you looked?
Until you do, your belief is nothing more than that. Your faith in the patent system is touching, but in this case it's misdirected.
Software is inherently more complex than just about anything else that individuals are able to build and expected to understand. Any non-trivial program involves hundreds of techniques that are potentially patentable under the current regime. Most programs people would consider trivial are going to contain dozens. This means that, first, determining if a program violates any patents requires first a detailed analysis by an expert in software patents to determine what potentially patented techniques are used, then hundreds to thousands of patent searches to determine if any of these techniques are patented. After this, you need to arrange licensing for the dozens of actual patented techniques that are used, and given the investment you have already made it would only be prudent to apply for patent protection on the remainder.
And that's for a program like, say, "Minesweeper".
Either that, or you just ignore the whole problem and hope for the best. Since even a company like Microsoft can't afford to hire an expert patent lawyer who is also a software developer and half a dozen paralegals for each programmer, this is all you can do.
So, since most people who develop potentially patentable techniques can't even tell if they're patentable or already patented, and a reasonably talented programmer would probably be doing so several times a week, software patents do not provide any useful protection for most inventors of potentially patentable software. They just provide a chilling effect on the development of publicly documented interfaces and protocols.
Of course large companies are in favor of software patents because patents protect the status quo. Large companies can file absurd software patents, e.g., one click shopping, to keep smaller companies from competing.
Once all software patents are enforced throughout the world, innovation will come to a standstill. It will be impossible for any new company to create anything. And large companies will rest on their laurels.
If someone says he and his monkey have nothing to hide, they almost certainly do.
Furthermore, patent filings, which express the "ideas" of a patent, are themselves neither patented nor even allowed to be copyrighted. After all, patents were intended to give incentive to disseminate information, and restricting patent filings in effect would undermine their claimed purpose. Similarly, source code expresses the "ideas" of a given software work and certainly should not be subject to patents in part for this very same goal.
Finally, since patents are about use, when they are applied to non-tangible things where use is expression, they in effect are a legal barrier to the very right to think. Hence, for thinking about, or using an idea, even in private, is at least theoretically a patent infingement. Certainly we do not patent books, and for many of the very same reasons we should not patent software.
All in all, I would love to see an active software patent repeal movement in this country.
Patenting of software is the same as patenting musical progressions (or riffs). Imagine where music would be if the riffs were patented. Imagine if authors could patent unique expressions that convey meaning more effectively than other expressions. Imagine if artists could patent color combinations that more effectively conveyed an image.
Hopefully, by thinking outside the cubicle for a moment, you may see the ramifications of this. The music industry would starve with patents on music like software patents. The writing industry would starve with patents on writing like software patents. All art and engineering would evolve much more slowly and be much^2 more expensive with the patent system that software enjoys.
InnerWeb
Freud might say that Intelligent Design is religion's ID.
Most software involves problem solving that is merely the means to an end, not an end in itself. Take for example the (patented) use of the XOR function for mouse pointers. The XOR function is not an invention that stands on its own as a result of long, painstaking and expensive research that eventually enabled the mouse pointer to be invented. It's the other way around: someone was programming a mouse pointer and decided to use the existing XOR function to solve the problem of restoring the graphics that the pointer passes over. This use of the XOR function does not qualify as an invention or research, it was simply a (very obvious) solution that needed to be solved to achieve the end result. It needed to be solved because these guys were the first to implement a mouse pointer, and as such they were able to patent this solution.
Simply being the first to solve a trivial and hitherto irrelevant problem should not be good enough to be awarded with a patent. Patents are designed to establish ownership of the result of expensive and difficult research, but most of the software patents protect things that were neither expensive or difficult to find, and are not worth 'protecting'
Ownership of software can be (and is) established through copyright; we don't need patents on top of that. Look at the current crop of lawsuits concerning software patents. Are they
a) An effort to defend the fruits of painstaking, expensive research that resulted in genuinely new ideas, or
b) An effort to lock out competitors, who would have spent the same 5 minutes to solve the problem under dispute without any effort, had they been there first rather than second.
Most cases are of type B. This is not surprising as another
Take Mom's Secret Recipe for apple pie. Copyright protects the entire recipe: I'm not allowed to publish it without Mom's permission. However if I decide to publish a slighty different recipe, using cinnamon rather than vanilla essence or something, I'll probably be in the clear.
Software patents are similar to patents on the various steps to make the pie: Mom might have patented the use of flour and butter to protect her particular recipe, but this means that I owe Mom royalties even if I am making apple pie to my own recipe, or even a completely different kind of pie.
If construction was anything like programming, an incorrectly fitted lock would bring down the entire building...
A software patent, which serves to protect inventions of a non-technical nature, could kill the high innovation rate.
What high innovation rate? Software is doing the same shit today that it was doing back in '95, we just have prettier interfaces now. I'd hardly call that innovation.
I keep hearing that the computer world is special because of the high turnover rate of products, but outside of the hardware world I really don't see it. Most people I know have been using the same basic software since at least 2000 but have upgraded their mobo/proc at least twice during that period of time.
The only "innovation" I have seen out of software is various bug fixes which shouldn't be there in the first place, but since the software writers are held to ridiculously low standards for quality control they can release the same piece of software 5 times and say: "Look at our high rate of innovation". Perhaps someone can point out all the great software innovations that have occured over the last 10 years. Since there is such a high rate of innovation it should be a trivial excersize for the typical slashdot reader.
"I have a porkchop, you have a porkchop. I have a veal, you have a veal".
I know I'm feeding the trolls, but this is a fundamental concept that few people can understand. However, I'm going to give it a go. Repeat slowly after me...
THERE IS NO SUCH THING AS INNOVATIVE SOFTWARE.
And there hasn't been in years. Modern software builds so much on previous work (much of it from before the days of software patents) that the amount of prior art renders the very idea of patenting it laughable. At a very high level, all software is is a means of making an existing job easier by automating that which lends itself to automation.
As computers progress, it becomes easier to automate more things. But ultimately, taking an existing business practise and appending the phrase "... on a computer" doesn't make it innovative.
You may be interested to read this.
What concerns me though is: if we do away with patents what will replace them?
/.ers seen or thought of a solution to this problem?
... if they did, you would probably be broke.
... a type of non-transferable copyright with manditory licensing attached, where the author is entitled to some percentage of any money made on their work (or derivative works), but cannot restrict how their work (or any derivative works) are used, with anti-plagerism statues requiring citation in perpetuity. Others have suggested shortening the length of copyrights back to their original 12 or 24 years. Reform is needed, and many have suggested all kinds of innovative approaches to replace or at least weaken the current monopoly entitlement schemes ... it is a subject that has b
You start off with a false premise: that something must replace patents (else there will be little or no innovation)
Patents do not need to be replaced with anything. The software world experienced much more innovation without them, and continued innovation in the United States only exists because they go largely unenforced.
Have any
There is no such problem. Your assumption is false.
I'm all for making software as "free" as possible, but I'm also of the mind that there would have to be some kind of IP structure in place.
I am assuming you are new to the software industry (apologies if this is not the case, but your statement indicates that you are unfamiliar with how the software economy worked in the 1970s, 1980s, and early 1990s prior to software patents, and prior to their being widespread). Having said that, you have some interesting thoughts despite the false premise from which you begin (you very correctly identify and express unease with the monopoly entitlement the government is granting on so many basic ideas and software implimentations, and the catastrophic consiquences to a robust and free market that follow).
Software has always enjoyed the protection of copyright, which has always been enough protection for companies large and small (c.f. Apple Computers, Microsoft & Joe's consulting) to make excellent profits. Patents are in fact antithetical to this, as they lock down basic ideas. Patents only came along much later (in the 1980s IIRC) If you write software in the United States, you violate patents. You can thank your lucky stars no one has decided to enforce them against you
Has anyone run into an IP scheme that would balance the creator/user relationship? Our present system is skewed and prone to monopolism, and a total absence of ownership would entail its own set of problems.
It depends on what you mean by "IP". If you're talking about trade secrets, the current laws work reasonably well. If you're talking about trademarks, the current scheme works pretty well modulo people abusing trademark law to silence critics using their name (this seems to get sorted out reasonably by the courts most of the time).
If you're talking about patents, the best reform is to eliminate patents. Government entitlement monopolies have been shown historically to not only NOT encourage innovation, but to actively stifle it. As an example, read up on the history of the airplane, the Wright Brother's patents, and America's desperation to catch up to advanced European (non-patent-encumbered) aviation technology during world war I. For those to lazy or uninterested to look it up, the short answer is that the US Government, in a tacit admission that the Wright Brother's patent on airplanes stifled innovation and improvements, effectively seized their patent (nationalized it), paid them a flat 1% royalty, and threw the technology open to all comers and competitors to develop modern airplanes. The amount of innovation that followed was truly phenominal.
If you are talking about copyrights, some have suggested a form of non-monopoly "authorright" as an alternative
The Future of Human Evolution: Autonomy
The problem here is not strictly to do with the law, or how patents are processed. It has to do with the nature of the industry. The patent review process depends on the ability to take concepts back to the prior art. For instance, if I propose to patent a modified rocket which can use air intake at high altitude, everyone who bothered to look into it would know that I was talking about the scram jet, which all kinds of rocket scientists were working on. You wouldn't know the details: that's their private research. But you'd know what they were doing.
... maybe ... caching the results of IO expensive queries to accelerate some database action, noone will spot it. Although everyone who has a database will have some kind of caching, MS, Oracle etc are hardly going to publish such details about their software because it would be almost equivalent to posting implementations.
However, software development is much closer to the conceptual domain. Knowing that someone is using a particular algorithm is almost as good as knowing their code. Many publications stop at description of the algorithm involved without bothering with any implementation, as any programmer with a little experience can go from one to the other. Thus if I propose to patent the technique of
There are other fields that behave like this: the field of novel writing simply couldn't have patents. If I proposed to patent "writing the ending first then structuring all the other chapters to introduce needed plot elements", we'd hardly have JK Rowling saying "look, I've been doing that for years". Or rather, she might say it afterwards, but there'd be no chance of it being public knowledge ahead of time. In that industry, it has to be copyright rather than patents, because people won't give away details of their works or the method of their composition, it would be too likely to compromise their own position.
Anyhow, you can see how it goes from there. The patent office starts out in good faith, and soon discovers they've approved a few patents for techniques that were already in common use, just in places noone published. Then it all starts to come apart, and everybody who beleives in principles like the consistency of law leaves, and the place fills up with empty minded officials who are happy to rubber stamp their way to a retired pension without thinking too hard. Disappointing, really.
I guess my point isn't really that patents are bad. They're a great idea. The last thing you want is people dedicating their lives to inventions that benefit all of society, and the profit of this stolen away by the big players. The tragedy is it nearly always happens anyhow.
The real problem is that the software industry doesn't have the openness needed for the prior art to be a useful examination. It's not impossible that open source could offset that (by filling out the prior art with a robust body of implementations), if only the relevant authorities would take OS seriously.
Ah well, that's all from me for now (goes back to working on thesis submission).
Maybe. We won't know for sure unless:
Someone comes up with a way to measure the contribution of "innovation" to the economy that doesn't depend on counting patents
That measure is applied to places and times with and without patents (US before/after software patents, US vs. EU vs Japan)
Without that kind of measure, we're reduced to dueling anecdotes - patent holder beats up megacorp and makes money, startup is crushed by megacorp with patent basket, etc.
To a Lisp hacker, XML is S-expressions in drag.
All economics is based on human nature because the actions of humans combine to make the economy. If you don't understand basic human nature, you can't understand basic economics.
Here are the basic principles of human nature you miss :
- insulate people from the effects of their inefficiency and THEY WILL NEVER LEARN TO BE MORE EFFICIENT.
- more money BEFORE good results NEVER LEADS TO BETTER RESULTS LATER, in salaries or government programs.
If you were able to race against a bunch of people who had to wear lead shoes (analogous to tarriffs) when you didn't, would you try as hard to compete ? Salaries are nice, but when I get the salary someone who outdid me should have gotten, that doesn't lead me to conclude I should try harder. It leads me to conclude that the politicians are useful idiots I need to cultivate.Assuming that government throwing money at the problem of competitiveness will help is absurd. Look at the effectiveness of nearly any government program. Why do public schools turn out morons when trillions of dollars have been spent on education ? Because more and more money is shovelled at it when it fails in order to "help it out" instead of waiting to shovel more money after it produces good results.
Let's look at the logic of Ford in your example: give money, get it back, lose a car. Net loss: one car, net gain: absolutely nothing. Ford succeeded by becoming so EFFICIENT it could offer LOW-PRICED cars and the masses (that didn't work for him) could buy lots of them.
You deride people who disagree as myopic. Let me spell it out for you :
- You aren't insulated -> you have to actually outdo your competitors -> you get more efficient to do this -> you can offer lower prices than your competitors -> the masses benefit from lower prices.
- You are insulated -> politicians just made it where you don't have to outdo your competitors -> you cultivate politicians -> you get rich and the masses have high priced crap.
Why in the hell do you think corporations spend so much time lobbying ?I have no problem having patents for new and innovative things. The problem is that, at least in the US, most of the patented stuff is obvious to an experienced worker in his field of expertiese, or it just reimplements something from real life on a computer.
E.g. nobody would ever try to patent sorting papers on you real life desktop into stacks, but doing so on your computer desktop all of a sudden becomes worhty a patent even though the task of fixing this is trivial.
If the things patented are trivial, then we can expect that any person with a university degree and/or some years of experience is extremely likely to produce patented solutions without knowing it just by coding what naturall comes to his mind. He could do patent searches, but this is quite hard since much in software design is about general ideas with sometimes overlapping contexts, chances are that he will not find what he is looking for as he expresses the same idea in different terms.
If we let a couple of monkeys hammer away at a keyboard they will sooner or later have written the complete works of Shakespeare. Now imagine that Shakesspeare had written all his stuff in a language containing the only words if, then, else, while. The time for that to happen would have been greately reduced. Now, replace the monkeys with humans that use this very limited language to express themselves, and you will very soon see sentences that look quite Shaksperian even if the never ever read a line from the original auther.
So if you like software patents, why not patents on litterature, and art. After all if I think that I'm the first person who write a crime story where sombody is killed by a knife why shouldn't I patent it. I'm quite sure that such patent have never been filed. As I privately have a slight suspicion that sombody actually may have written such a story before me, I am happy to licence the idea to other authers for a fee slightly less than it would cost to take the dispute to court. The only one who would not go to court would be sombody that positively would know that he could prove prior art, but as that person is most likely to be dead there is little risk.
Now it strikes me, why write a book in the first place. After all, the part where sombody is killed by the knife is a very minor part of a normal crime novel. Why spend all that time. Why risk to be sued by the auther holding the patent on having characters being transported by a car, or having a conversation, or kissing, or..., In fact by actually writing a book I open myself to all kinds of liabilities. Besides I'm not much of a writer anyway.
I think I just stick to just filing the patent on the knife usage. Then I can wait for some real auther to get a bestseller using my knife concept. After all people are known to have been killed by knifes as early as Julius Ceasar, so its bound to get into a novel soner or later. So I sit back and wait unitl somebody actually write such a story and get what is rightfully mine from him. That way the risk of being sued is much less, not to mention that it is much less work. I don't have any costs for marketing either.
Now, If software patents are so important for the software industry, the same thing would be valid for other ways of expression like litterature and art. So why should it not be possible to have patents in this field as well? After all there is big money in this business just as in software. Look at the movie and record industry, or that software industry such as Microsoft buy the rights to digital publishing of classic art.
The strange thing is that you seldom hear people argue that patents on art would be a good thing, even though my guess is that there are more artists in need of some extra income than there are software developers. Why should they not be able to get some money from their way of express themselves, when the software developer can?
Or is it just that patenting peoples way of expression themselves is a bad idea?
God is REAL! Unless explicitly declared INTEGER
Larry Ellison (CEO of Oracle) is against software patents because:
a) He is a really nice guy, a humanitarian, and wants to do the right thing for The Children.
b) He wants to foster innovation that could potentially allow smaller competitors challenge Oracle's dominance because industry-wide innovation is more important than Oracle's stock price.
c) He knows Oracle, being the second largest software company in the world and having the deepest inroads into corporate IT departments, can steal the ideas of smaller competitors and wipe them out easier when those pesky software patents are out of the way.
Surely the patent system is flawed, and it has problems, and it needs to be fixed, not eliminated altogether.
Before we call this a "duh" issue, can people please exercise their brain just a little to ponder why Larry Ellison, one of the most ruthless businessmen the world has ever seen, is against software patents?
I am listed as the inventor for several patents as well, though the patents are actually held (and were applied for) by the company I was woring for at the time.
I don't see the value of these patents. The patents involved were essentially an attempt to landgrab beyond the bounds of what was actually invented. A mix of copyright, and trade secrets (via closed source code) would have been entirely sufficient to protect the systems I designed from competitors.
What extra value do software patents provide beyond copyright and trade secret? It simply means if anyone else independently invents something similar, you can sue them - with copyright and trade secret you need to demonstrate that they actually stole it from you.
And before someone claims that patents cover the indea, while copyright covers the specific code, let me make it clear that that is not true. Have a look at the literary world, which has no patents, but does have copyright. To file a copyright suit for plaigarism you only need to find an author who has created a startlingly similar book - there's no need for any word for word copying. Want an example? J.K. Rowling sued DMitry Yemets over his "Tanya Grotter" books. The similarities were clear - both books were at wizard schools etc. - but the actual stories (plots for instance) were completely different. Equally there were rumblings of issues with "The Life of Pi" which was similar to a Brazilian book "Max and the Cats" in that they both featured someone adrift on a boat with a large cat after a shipwreck.
Jedidiah
Craft Beer Programming T-shirts
you need protection for your innovations. It's the only way for investors to assess just how valuable the software companies are. Microsoft for instance isn't valued based on the millions of lines of code it has - not just because the quality is hard to assess but also because it doesn't give any information about leverage on their competitors.
Also one shouldn't forget that about 50% of all patents are invalidated when taken to court, that coupled with the fact that when engineers unknowingly use patented technology, the companies aren't fined (but forced to license or decist) means that only that only the safe and truly valuable patents will be enforced.
Finally, if one looks at history, patents were created so that small companies and inventors cheaply could protect their inventions against big firms. This still holds, the problem with software patents lies not in the patents themselves but their broadness. By restricting what is software innovation and what isn't, one could get rid of most issues. For instance Amazon.com's one-click technology isn't really an invention but passed because the definition of software innovation is currently too vague.
Care to link to your patents?
Anyway, the problem with software patents is that with software, the line between "idea" and "implementation" is incredibly wide and fuzzy. This is because, once you have an idea to do *something* in code, there are very few different ways to actually go about doing it. Oh sure, there are different algorithms you could use for each step -- but more often than not software patents don't even get that detailed. All they seem to do is list what each step does, not how exactly it does it, and that's good enough for the USPO.
The other reason that the line between software implementations and ideas is so blurry is because there's never a physical invention. All you're doing is writing down instructions for a machine to follow. In essence, you're just stating your *idea* in a way that the machine understands.
For example, if you're to tell any semi-competent programmer who's never heard of Amazon.com that you wanted to make a web site that lets users buy an item with only one click, the great majority of them will come up with the same basic implementation -- store in a database all the information needed to make a purchase for that user, track which user is logged in, when the button is pressed cross reference the user into the database and automatically send the information through the purchasing mechanism. There really is no other reasonable way to go about implementing this kind of thing. Obviously, there's details depending on exactly how their database is set up, but it doesn't matter. The patent, as it's written and accepted by the USPO, effectively covers any attempt to do this sort of straightforward manouver.
Even if this programmer who's never heard of amazon comes up with the idea all by themselves they can't make a product that uses it because amazon has the "rights" to the idea.
Contrast this with an actual, physical invention -- say, one-button 4 wheel drive in an explorer. Now this is an idea that requires a lot of detail to hash out and get patented. And there's a lot of ways to go about doing it, none of which are very obvious. To get it patented, you have to have an electromechanical system pretty much designed that *works*, and that requires a lot of R&D.
To get the equivalent software patent accepted, all you need is the idea, and a rudimentary knowledge of how computer systems are set up. You practically don't have to do *any* R&D at all. And ideas are supposedly not patentable but for some reason high level software implementations are.
---
Ok, that answers why software patents may be bad philosophically, but you wanted to know why they're bad practically? Well, because it's so easy to "develop" a software patent, companies create thousands and thousands of patents as fast as they can to try to "stake out" their territory on the IP map. All that matters is that they were the first to be granted the patent, and no one else had used that particular technique yet. They may not have been the first to think of it, the first to R&D it (if they even did any R&D), or even the first to file for a patent! But they get a temporary monopoly on that new technology and that stifles innovation.
It also hurts the little guy -- smaller companies don't have the resources to push patents through with the same efficiency *or* to take on the big boys when patent disputes come into play.
Finally:
Capitalism vs Socialism
You think that having the government grant you a temporary monopoly in a certain area is "capitalism" and letting the market work for itself is "socialism"? Wow. Just wow.
The following sentence is true. The preceding sentence was false.