(originally posted
here∞)
Originally posted by pdampier:
As an inventor how would you prevent your idea from being coded up by someone else or worse copied by someone who then just gave the product away? They wouldn't be breaking your copyright.
Why not the old-fashioned way? More features, lower TCO, slicker marketing, higher expandability, better reliability, more talented salesmen...
The framework given to us (for free if you have a library card) by the computer scientists of 1950-1990 is an unmistakable cause of the tech revolution we've seen since then. What incentive do we have to restrict competition as you describe? Feel free to name an example of a software product or idea where we'd be better off had it been patented.
Excellently posted by drag:
Patents generally are good.
Software patents are not. I am glad if Microsoft wins this case.
This is my understanding of the issue, a pure opinion follows...
Patents were created for several reasons, most people figure they were created to 'protect ideas'. Which is false. That's their function, not their purpose.
The idea is to promote innovation thru giving a temporary monopoly over a idea. The reasoning is that:
1. Inventors can gain financially easier by selling patents. This promotes more innovation because the money that is earned is put back into more innovation because patents have a finite lifetime.. Unless you innovate more then when your patent runs out, so does your income.
2. Reduces the need for trade secrets. By helping to eliminate the need to obsofcate and hide new inventions it allows developers to impliment them in the open without fear of loosing competative edge. This allows other inventors and enginneers to closely examine and understand the operation of such innvention and hopefully use those ideas in their own products when the patent runs out. Also they can use patented ideas to help develop new ideas of their own.
3. Publish new ideas publicly. By making the patent form a public document and requiring a detailed and thourough description of the patented idea scientists, engineers, and inventors are allowed to easily go thru stacks and stacks of new ideas for their own research and hopefully inspire more innovation. Also advertises new ideas that may be bought up and increase revenue for the inventor which can be used to finance more innovation.
Now software patents defeat the whole purpose of having the patent in the first place.
The legal practicalities of it can potentionally render developing software prohibatively expensive.
Each peice of software is a excersice in solving innumerable problems. New ideas and such are created constantly as a matter of course. Plus those ideas can be implimented in innumerable different ways, in innumerable different circumstances and often without the programmer ever realising what he/she is doing.
Since programming isn't much more then applied practical mathmatics there are a hundred different ways to look at any problem and any solution.
If you allow each of these instances to be patentable then even simple programs can infringe on hundreds of different patents. Programming can be performed realively innexpensively. A individual or small group of people can whip up a very complex and usefull program using tools that cost next to nothing when compared to real engineering such as.. building a bridge or a building, or designing a wheel for a high performance automobile.
If you throw patents into the mix then in order to avoid stepping on anybodies intellectual toes you have to hire teams of lawyers expert in patent law to shift thru every loop, thru every interaction of software in the 'vertical software stack', thru every algorythm, for potentional patent violations.
A simple programmer can never ever hope to avoid patent violations.
This leaves writing software to only those that can afford legal fees in defending their software or have such patent portfolios of their own that they can threaten to sue in retalation to anybody that threaten to sue them and then enter into complex patent-licensing scemes.
Aside from the practical issues against software patents you also have the additional issues like software patents do nothing to discourage 'trade secrets'.
In fact software patents do the opisite of their intended purpose and ENCOURAGES trade secrets.
Because no programmer, no orginization or company that hires programmers can ever be assured that they are not violating patents, the primary means of defense against lawsuites, aside from threating patent lawsuites yourself, is to hide as much functionality of your program as possible. Unless you hide completely how your program functions it leaves you completely open to patent attacks.
I think Bill Gates had a great quote on software patents.
quote:
If people had understood how patents would be granted when most of today's ideas were invented and had taken out patents, the industry would be at a complete standstill today. ... The solution is patenting as much as we can. A future startup with no patents of its own will be forced to pay whatever price the giants choose to impose. That price might be high. Established companies have an interest in excluding future competitors.
Software developement is NOT like engineering mechanical devices or drugs. With engineering and such your using real-world technology, to solve real world problems. With software you can balance a hundred ton pyramid on the head of a pin and have it stand forever.
Software is more like writing, or a recepie.. Except that it's in mathmatics and layers of abstraction rather then English. Your work under software is protected by very strong and long lasting Copyright laws.. Nobody can legally take your code, your ideas, and use it themselves without your permission.
Right now patent law is a threat to the entire US software industry. Not only open source, but Microsoft, IBM, all American companies that are subject to US law can, and are, very screwed over right now in terms of software patents.
However on the other hand patents in general are a very good idea and one of the reasons that the US is were it is at today, economicly and technologically. Software patents are just a unfortunate legal mistake.
In my personaly opinion, of course. (I am no were near being a lawyer)
This is why I hope Microsoft wins this case, because even if we can't make software patents illegal (or just not a reality anymore) then maybe they can raise the legal requirements of successfull software patent litigation to the point were it might as well be a non-issue.
Very informative page from the League for Programming Freedom.
http://lpf.ai.mit.edu/Patents/patents.html∞
Their initial example of patent violating code:
for (x = 0; x < height; x++)
for (y = 0; y < width; y++)
screen [x + xpos][y + ypos] ^= arrow [x] [y];
much discussion, with a surprising amount of middle ground for an IP discussion if you ignore
PeterB's one-liners
Originally posted by domovoi:
Then we basically agree. The USPTO needs to get its act together so silly patents like the one granted to Eolas don't happen, but I don't see how software patents are universally detrimental.
I think we're heading in the same direction, but I'd phrase it a different way. The problem is not as large as "software patents are evil," but it's not as small as "we just need to retrain a couple of guys at the USPTO" either.
What I'm not sure everyone on "your" side groks is that the US patent system represents a <i>pragmatic</i> compromise, not an attempt to formalize morals. The question "what if someone steals my idea" is completely irrelevant to the language of the Constitution, while "how can we best facilitate the advancement of science" is specifically addressed in the Article through which patent law was created. (To be fair, you've raised both).
It's fair to say none of us has the exact answer to the latter question, but I hope you'll agree it requires more than mere USPTO scrutiny. Look again at the list of [pragmatic] grievances Drag elucidated on the 1st page. There are structural problems there that tweaking won't fix.
If you want specifics on my own opinions, I think there should be a shorter term length for software (3-4yr); some additional requirements on publishing functional code, especially where extensible interfaces are concerned; more specific regulation against wielding patents for monopolistic purposes (i.e. patenting something that you never intend to create); perhaps some provisions for holding complete source trees in escrow...
In sum: so long as we're letting the government intervene in a marketplace that would otherwise be highly competitive, there are some publicly useful things we can ask of it.
I'm going to restate that first clause because it's important to grasp -- as far as anyone has been able to argue, the patent-free bubbles within the software industry have flourished much more than those whose history is tarnished with lawsuits and infighting. Ideas have permeated into reality from all sources (large companies and dorm/garage rooms both) in more dramatic fashion than any industry in recent memory.
Yet despite the hyperbole, there's no reason to claim that the "old-fashioned" rules of competition I mentioned in my first reply no longer apply. Microsoft consistently makes zillions of dollars in rapidly-changing fields because they understand the market as well as any econ-textbook example. Linux took off not when it reached a kernel milestone, but when real companies decided to use real business practices to advance it. Patents are a smaller part of the equation than any of these factors...for now. My urgency comes not from the "evilness" of software patents, but from the myriad small, pragmatic effects that could combine to cripple an industry that's still not fully mature. We need to evaluate them as the compromise-laden legal fictions they are in order to find a solution that truly promotes progress.
I should clarify that some of my complaints -- thus some of my proposed solutions -- apply to other areas of technology besides software. More importantly, modern technology often makes the hardware/software boundary incongruous or downright irrelevant.
Good patent example: Dolby Labs patents the AC3 codec. Every time you buy a product that encodes or decodes 5.1 surround sound, they get a few bucks. Its algorithms are now well known by audio researchers and have fostered many expansions; the implementations themselves are ubiquitous, highly useful, and reached commodity prices quickly. Note that this analysis crosses the software boundary since many Dolby-licensed products are software-based (e.g. computer DVD packages, pro audio rendering programs, etc.), as are many of its offshoots.
Bad patent example: Creative Labs buys Sensaura in order to squash their patent for realtime PC-based surround encoding chips. Creative's monopolistic control of the soundcard arena is bolstered. Meanwhile, an entire market segment pioneered by nVidia (the most prominent Sensaura licensee) dies, leaving consumer hardware in roughly the same state it was in 1999 when they did the same thing to Aureal. Absent hardware innovation, related software is similarly stunted even though a "software patent" is not in question. (Coincidentally, when a company like id software tries to roll their own a software workaround, Creative sues them over an unrelated but
demonstrably baseless∞ software patent, giving them leverage to demand that Carmack base his sound engine on EAX instead.)
Originally posted by PeterB:
quote:
It's perfectly possible, and especially with software that it's VERY likely that somebody can come up a solution to a problem that is pretty much the same as yours and have it be completely unrelated to anything that your doing or done.
So they're not taking your ideas, so I don't really see what point you're trying to make....
Putting aside that the concept of "taking ideas" has no basis in patent law, I confess I miss your point too. Are you saying that patents are good when they protect you from unscrupulous competitors in particular, but bad when they inhibit unrelated (or simply equally creative) developers industry-wide? Well, duh. Please propose a patent system that differentiates between the two, because the present one certainly doesn't.
Originally posted by PaoloM:
quote:
Originally posted by resedit:
Can you patent a musical score? I wish the first rapper had patented rap and chosen not to license it ...
What do you mean, the succession of notes?
I think he's making a point along the lines of: "software patents, due to the nature of conceptual reusability, tend to apply much more broadly than patents in other domains. If it were feasible to create and patent abstract algorithms in non-computerized domains, we'd find counterintuitive situations there too."
At least I hope so, because I don't want nonsensical analogies on "my" side, and furthermore [some] hip-hop is great shit :p
Back to
InformationShouldBeFree
There are no comments on this page. [Add comment]