Strangeworld is done with Video Toaster and KDE kVideo. AFAIK, Strangeworld is the only active production using Debian right now. You'll find the Debian swirl hidden in at least one scene per show, so presumably they've got at least one Debian fanatic on board. Pretty damned cool!
You are completely mistaken. While the long version can be found on the scifi.com web site, the short version is that Strange World was produced, finished, and shelved in 1999. Only ten episodes were even made, and only three of those were aired before SciFi picked up the property. Nobody has touched Strange World in three years.
Whether Video Toaster or Debian Linux were used in the production of the series is open for discussion. But the question belongs entirely in the past tense.
Perhaps it's just me, but that makes no sense whatsoever. How do you ship a measure of speed? "Shipping" millions of instructions per second seems to me to be the same as "shipping" miles per hour.
It's just you.;-)
In the mainframe world, where virtualized hardware is the norm, systems are sold by the MIPS. In other realms of computing we talk about a 32-processor or a 128-processor system (or, if you buy Sun, a 106-processor system, for some reason). In mainframe land, you talk about a 12 MIPS system or a 45 MIPS system or whatever.
It makes a lot of sense, too, when you think about the fact that a four processor system, mainframe or otherwise, from five years ago is probably less powerful than a one-processor system now. It sort of normalizes it if you talk about system capacity in MIPS rather than in terms of "x processors of type y at z megahertz."
Don't compare the practice to shipping "miles per hour," but rather to horsepower. General Motors could, if they wanted to, say that they shipped umpteen million horsepower worth of engines last year, and it wouldn't be that hard to understand. It's just a different way of counting.
I'm curious as to the philosophical framework that leads you to that position.
Ask the people who wrote the law. I say that corporations are deserving of consideration because the law says they are. People smarter than I am-- and I dare say smarter than any of us are!-- have come to that conclusion over a period of many, many years.
And, just for the record, a 768-processor Origin 3000 from SGI works just like a 2-processor Origin 300. If you ignore little hardware differences that don't affect your apps, it also runs just like a 1-processor Origin 200 or Octane from five years ago.
I had the privilege of getting time on SGI's 768-processor system in Minnesota several months ago. It was... nice.
Okay. Let's talk about RSA. Although since you don't seem to draw any impacts from the RSA patent, I have to just guess at your meaning.
First of all, RSA was used as a key component in SSL. Everyone who banks or makes purchases using the Internet has access to SSL. To say that the patent on RSA hindered the adoption of SSL-- which you didn't, but I can only guess you meant by "the result of all of which certainly included the lack of security in software for any but the deep-pocketed"-- seems unlikely at best.
Furthermore, in 1997, Flinn and Jordan of Alston & Bird LLP published an article entitled, "Using the RSA Algorithm for Encryption and Digital Signatures: Can You Encrypt, Decrypt, Sign and Verify without Infringing the RSA Patent?" Read it here.
In this article the authors say, in part, "The existence of the patent, and RSA Data's aggressive litigation posture, have chilled the interest in both commercial and non-commercial implementations of public key encryption and digital signature technologies." This would seem to be evidence for your assertion. But read on: "Many have taken for granted the bald assertion that the 'RSA Algorithm is patented,' without examining the patent itself, or more particularly, the claims of the patent." In a footnote to this statement, the authors cite Schneier's Applied Cryptography as an example of one misinterpretation of the patent.
This article goes on-- at great length, not worth reproducing here-- to explain why the process of verifying an RSA-generated signature is not covered by the patent. Furthermore, the process of generating an RSA signature is also not covered by the patent, due to the existence of the identical Pohlig-Hellman algorithm invented in 1975, two years before RSA. Basically, the only thing that the patent covers is the process for generating RSA key pairs.
So basically, if you look at it closely, as those two patent attorneys did, you'll find that the RSA patent wasn't nearly as wide in scope as most people seemed to assume it was.
Maybe you can make the case that some kind of harm, direct or indirect, was done as a result of the RSA patent. But upon analysis it will seem pretty clear that that harm was the result not so much of the patent itself, but rather of the misinterpretation of the meaning of the patent by one party or another.
Why did you excise part of my post? When I said, "corporations deserve, and are legally entitled to, the same consideration as individuals," you quoted me as "corporations deserve... the same consideration as individuals."
I think you... are misrepresenting my... statements... by taking... them... out of context.
The law says the corporations are, in the legal sense and with exceptions, persons. That's it.
Imagine that in your flight sim for the USAF, that you had to go and check the patent status of every technique you used in that project.
We did. You wouldn't believe the things the Department of Defense makes you do when you're contracting for them. You have to deliver to them, as a condition of their acceptance of your product, the evidence of your research into patents and copyrights and so on that might have an impact on your work. I wasn't working on that, of course, but I got to hear about it. Everything we did fell into three categories: COTS (commercial off-the-shelf, which is exempt from background checks because the responsibility is the vendor's), non-commercial licensed (like licensing the onboard computer microcode from Motorola and Boeing), and in-house developed. Everything we did in house had to be gone over with a fine-toothed comb.
None of this crap about patenting processe. You don't patent "a way of propelling a ground vehicle", you patent "this engine right here", and "this specific drivetrain".
Hmm. Have you ever read a patent? The language is written by lawyers, which makes is hard to sift through, but also very, very precise. The title of a patent might be, "system and method of propelling a vehicle," which I'm sure sounds unacceptably vague to you. But the actual patent consists of a dozen or more pages of details: the what, how, and why. I read a patent this morning that describes an API; it was 70 pages long.
For example, Amazon's controversial patent (5,960,411) is entitled, "Method and system for placing a purchase order via a communications network." That's vague as all hell.
The abstract offers more detail in the form of a paragraph. The rest of the 19-page document describes Amazon's method in resounding detail, including where the cookies live and how single-item orders are consolidated into multi-item orders to reduce costs.
It's very thorough, and it's not obvious at all. I think the USPTO was entirely justified in giving Amazon a patent for it, in and of itself.
There are other factors, of course-- should Amazon have gotten a patent for a technology that others developed around the same time?-- but in general, the idea was specific and unique enough to warrant patenting.
So generally I think you might consider reading a few patents before you pass judgement on the whole system.
Yes, yes, RMS asserted that patents don't encourage what he called "incremental innovation." I don't deny that. I'm saying that reading the whole article left me with the impression that what he was calling "innovation" was nothing of the sort.
Back to the example of LZW compression. If you wanna use LZW, license it from Unisys. There's nothing wrong with that.
If you wanna improve on LZW, go right ahead! The algorithm is documented, fully and completely, in the patent application if nowhere else. If your improvement is significant enough to merit it, you'll be granted a patent on it yourself.
If you oppose paying Unisys a licensing fee-- which RMS clearly does, on (for lack of a better word) moral grounds-- then don't use LZW. Use another algorithm, or develop your own.
Through that process, you weren't prevented from innovating. You were prevented from making use of somebody's patented innovation without complying with that somebody's terms. It's not the same thing.
Now we're getting to the crux of the matter. If asked, I will argue in favor of the status quo. You argue against it. This is our topic. I debated in college; I'm comfortable with this.
In arguing to change the status quo, we must first establish that the status quo is worth changing. We can do that in two ways: one would be to demonstrate that a different circumstance would be markedly better than the status quo; the improved circumstance must be sufficiently better than the status quo that it merits making the change at all.
The other method is to demonstrate that the status quo is simple unacceptable in and of itself, by some arbitrarily assigned criteria. If we, for instance, all agreed that butterflies are wonderful, and we could demonstrate that the status quo implies the extinction of butterflies, then we have a prima facie reason to change the status quo.
I'm not hearing either of those arguments expressed very clearly, though. The "patents stifle innovation" argument doesn't really mean anything without a compelling "but for" case. Can you assert, in some reasonable way, that innovation would be better served in the absence of the patent system? That seems to be what RMS is arguing, but his assertion isn't really backed up by his examples.
On the other hand, is the patent system somehow morally wrong? Should we change or abolish it, not because there's a better way, but just because the status quo is so completely unacceptable?
I guess what I'm saying is that I don't see anybody advancing any good reason to consider another way of doing things.
Your logic is severely flawed. The development of a new car can't be reduced to a count of the number of parts. Vast amounts of effort from large teams of people, and huge amounts of money, go into the design of a new car before the first part is designed. The sheer number of pencil sketches-- and, more and more often, digital paint sketches-- that get produced, revised, reused, or discarded in the early design phase can't be underestimated.
Likewise, software is far more than the sum of its lines of code. The last big project I worked on-- a flight simulator for USAF-- was three years in development before the first line of code was written. There was a ton of work done on design and architecture, building concept demonstrators and mock-ups of the operator interface, and on and on. Practically every detail of the system was figured out on paper and in meetings. Once the time came to code it, it was almost just a matter of writing things down in Ada and C++.
There's no question that GM's designing of a new car and some guy's writing of a mail daemon are very different things. But when you start comparing apples to apples-- large-scale mechanical or civil engineering to large-scale software engineering-- I think you'll find that the two are more alike than they are different.
You seem to agree that the scientific model works well, so I guess I'm not understanding why you think corporations deserve distinguished status.
First of all, corporations deserve, and are legally entitled to, the same consideration as individuals. Corporations are, in the eyes of the law, persons, with standing distinct from their stockholders. For tax purposes corporations and individuals are treated differently, but that's a different matter entirely.
To promote the progress of science and useful arts, not to promote financial gain.
The end result of the patent system, as you so ably quoted, is to promote the progress of science and the useful arts. We promote progress by offering incentives. These incentives include, and have included for over 200 years, grants of rights of exclusivity.
One leads directly to the other. "Get rich by being inventive!" benefits both the inventor and society.
First of all, I disagree with your fundamental premise, which is that "society only benefits when a patented invention would not have been rediscovered and published during the term of the patent." This is just patently (pardon the pun) untrue. Whether somebody else might have come up with the invention in the next 20 years is irrelevant, both to the criteria for granting the patent and to the question of benefit to society.
But even more important is that you're asking the patent office to become expert in every field of human endeavour. Before one can judge whether an invention could be independently reproduced by another, you have to understand the invention, as well as its place in the greater scheme of things. I don't know anybody who could make that kind of judgement in a fair, fully informed, and impartial way, in any field.
The quote from Douglas Adams comes into play at this time: "It is a rare mind indeed that can render the hitherto non-existent blindingly obvious. The cry 'I could have thought of that' is a very popular and misleading one, for the fact is that they didn't, and a very significant and revealing fact it is too."
No need to patent them? Of course there's no need to patent. A patent is a right that the government grants if you choose to apply for it. It's not necessary. It's an incentive.
Look, the government makes a deal with you, the inventor. You can choose to get a patent on your invention, which will grant you the right to prevent others from making something like what you've made for a certain amount of time-- usually 20 years. But in order to get that exclusivity, you have to describe your invention in detail and make that description available to the whole world.
This serves two purposes: on the one hand, it gives your competitors a fair opportunity to keep from stepping on your patent, because you're telling them exactly what you own the exclusive rights to.
On the other hand, at the end of the term of your patent, the whole world knows exactly what you did and how you did it.
So at this point you have a choice. You can either try to keep your methods and ideas secret and see what happens, or you can take the patent and get to be the king for a while before turning over everything to anybody who wants to look it up.
Of course it's not necessary to patent your invention.
What? I really can't tell exactly what you're trying to say. It sounds as if you think patents are granted on individual components, like pistons. Nobody owns a patent on pistons; if anybody tried to apply for one it would get rejected because pistons have been used in many ways for thousands of years. Somebody may own a patent on a particular type of engine that incorporates a piston in its design, but that doesn't mean they have exclusive right to use pistons.
Basically I can't really figure out whether your post makes any sense, but is just obscured, or whether it's just nonsense from asshole to elbows.
I guess the range of experience is pretty broad. In my experience, the process of designing and building a big software application is pretty similar to designing and building a big physical thing, like a car. You have the same people doing the same sorts of jobs. The difference only comes in the manufacturing. When you release a car to manufacturing, huge factories full of people work to build them. When you release software to manufacturing, CDs get burned, usually in a big duplicating machine.
But the stuff that happens before manufacturing-- the designing, prototyping, testing, marketing, and so on-- is almost identical. The difference being that building a car involves making technical drawings which act as the schematic for manufacturing. Writing a big software app involves making source code that acts as the schematic for the compiler.
In short, if RMS is trying to apply his analogy to commercial software development-- as opposed to one- or two-person efforts-- then I think he's just plain wrong.
Your time frame is too short. Of all the software or software-related patents that have ever existed, the vast majority are still in effect. You can't say that software patents don't do their job of releasing knowledge into the public domain if you don't consider a sufficiently long timeline.
Again, you're confusing pure research with commercial efforts. The same argument you just used could be applied equally well to medicine: without the free exchange of ideas and information among the medical science community, we'd all be dying of smallpox and whatever the hell else.
And yet we still grant patents in the medical field. Medical instruments and devices are patented. So are drugs.
I'll say it again: there's a line between fundamental research and knowledge in the field of computer science, and commercial uses of that knowledge to create products, processes, and services. In medicine and other fields, that line is fairly well defined. (Although the recent discussions of the IP related to the Human Genome Project cast this in doubt.) In computer science and computer engineering, the line is not clear at all.
Examining the history of the past couple of decades I see scant evidence that the current patent law fosters this. I appears more likely that it suppresses it.
Obviously your timeframe is too short. Patents last twenty years. Look at the last 100 years and see if your opinion changes.
The rate at which innovation happens has increased in the past half-century. Perhaps it's appropriate to talk about reducing the duration of patent grants. Then again, maybe not. Wiser heads than ours picked the 20-year rule, so we shouldn't change it lightly.
Okay, so it's possible to use the patent system in a way that stifles innovation. That doesn't make patents inherently bad or wrong, like RMS seems to think they are.
I agree that some things shouldn't be patentable. Mathematical formulae, for instance, aren't patentable. But processes are. For instance, the process for making steel was patented when it was invented.
There's a line between algorithms (like bubble sort) and processes (like refining steel). But where is it, exactly? Does it lie in the complexity of the algorithm? Like maybe simple ones should be public domain while complex ones can be patented? That doesn't make any sense, though, because the simplest algorithms are often the most brilliant and innovative.
The problem here is that we're fundamentally talking about a but-for situation. The only way to judge the value of the patent system is to imagine a world in which it didn't exist, and that's hard.
In what way would we, as consumers, be better off if the spreadsheet had been patented?
Bah. The fact that you're asking the question means that you fundamentally misunderstand my point. The question to ask-- which is admittedly hard to answer!-- is "What innovations would we, as consumers, have access to if the spreadsheet had been patented, but that were not invented because no patent protection was available for them?"
I know, I know. It's impossible to answer that question. But that's kind of the point. If we don't offer patent protection for software innovations-- truly deserving ones, I mean-- then we'll never know what we missed.
I have yet to see a single piece of evidence that software patents truly discourage innovation. In fact, the converse is likely true. Today, it's not legal for me to use LZW compression in one of my programs without licensing it from Unisys. If I choose not to ignore that fact, I can do one of four things: give up, or license the algorithm, or use some other algorithm, or develop my own. If I don't want to license, and I don't want to give up, then I have to innovate! (Or I have to use somebody else's innovation that is freely available to me, but that's basically the same thing.)
If you read the article, you'll see that Stallman argues that patents don't encourage innovation. Any thoughts on this?
I admit that I didn't read the article slowly or carefully, but that's not how I understood it. (Marge: But you liked Rashomon! Homer: That's now how I remember it.)
It sounds to me like RMS is arguing that patents are an obstacle to the writing of software using well-known algorithms or principles. The example RMS gives of LZW compression, for instance, says that the programmer who wrote Compress "use[d] an idea he had found in a journal, just like software authors had always done."
That doesn't sound like innovation to me. It sounds like the programmer wrote a utility using an algorithm that was under consideration for a patent, and got burned by it. That has nothing at all to do with innovation, which is the process of starting from scratch-- or from a foundation of public knowledge-- to come up with a truly new idea. Not program, idea.
The larger question is this: should the LZW compression algorithm have been patented? This kind of goes back to what I said in a previous post about pure research versus commercial research. There's an argument to be made that fundamental algorithms-- like the stuff in TAOCP-- should be part of the field's collective body of knowledge, unencumbered by rights or ownerships. Maybe things like LZW compression should be included in that category.
Of course, Unisys disagrees with that assertion.
It's a tough question, and it's hard to know what to do.
Software doesn't need patents to protect it becuase the amount of effort to reverse engineer some code is often more than the energy needed to code something new in the firstplace.
But the code itself isn't what's patented. The patent goes for the "process, machine, manufacture, or composition of matter, or any new and useful improvement thereof." The code itself is covered under copyright, and possibly under various protections of trade secrets.
So you don't patent Adobe Photoshop. You could, possibly, patent software for editing digital images in such-and-such a way. It has to be fairly specific. For example, consider patent number 6,337,901, "Customer billing relationships software." (Read it here.) This patent doesn't cover the software itself. Rather it covers the design, methodology, and architecture of the software. In this particular instance, the meat of the patent describes the process that the software implements for generating its output.
So it's not about whether or not you can reverse-engineer Photoshop. It's about whether you should be given the right, for a limited time, to exclude others from writing software that does the same thing Photoshop does.
(This sounds silly because Photoshop is neither patented nor patentable.)
Maybe I should have qualified "important." I think it should have been clear from context. I meant "important" in the large-scale economic and social senses. New music and art is important, and that's why we directly subsidize artists in some cases, and ridiculously overpay them in others, depending on whether they're popular or not.
It's the same basic premise, though, implemented in two different public policies. Artistic development is encouraged through subsidy. Commercial development is encouraged-- among other ways-- through the granting of patents.
Fewer high-quality programmers who have access to each other's code will produce better code faster than a whole bunch of overpaid amatuers working in isolation.
If we were talking about pure research, or the sciences, you'd certainly be right. That's exactly the model that scientists follow: publish early and often, and read everything.
Maybe we should talk about the difference between academic programming and commercial programming. It's the same as the difference between biological research and drug manufacturing.
If computer programming is a science, then pure research in the field should be in the public domain. But the use of that pure knowledge to achieve commercial gains is something else entirely.
On the one hand, we have the closed, proprietary, patented, buttoned up shrink wrap license vendors. On the other hand we have free software. Compare the amount of resources each consumes relative to its output.
I don't mean to be a grouch, but I disagree. I deal with quite a bit of commercial software, and I have to say that a fairly small percentage of it is complete crap. Of all the freely available software out there, though, there's a lot that's poorly documented, buggy, maldesigned, or all three. A few outstanding exceptions-- like Perl or Apache-- do not an argument make.
Slashdot Mirror
Strangeworld is done with Video Toaster and KDE kVideo. AFAIK, Strangeworld is the only active production using Debian right now. You'll find the Debian swirl hidden in at least one scene per show, so presumably they've got at least one Debian fanatic on board. Pretty damned cool!
You are completely mistaken. While the long version can be found on the scifi.com web site, the short version is that Strange World was produced, finished, and shelved in 1999. Only ten episodes were even made, and only three of those were aired before SciFi picked up the property. Nobody has touched Strange World in three years.
Whether Video Toaster or Debian Linux were used in the production of the series is open for discussion. But the question belongs entirely in the past tense.
Perhaps it's just me, but that makes no sense whatsoever. How do you ship a measure of speed? "Shipping" millions of instructions per second seems to me to be the same as "shipping" miles per hour.
;-)
It's just you.
In the mainframe world, where virtualized hardware is the norm, systems are sold by the MIPS. In other realms of computing we talk about a 32-processor or a 128-processor system (or, if you buy Sun, a 106-processor system, for some reason). In mainframe land, you talk about a 12 MIPS system or a 45 MIPS system or whatever.
It makes a lot of sense, too, when you think about the fact that a four processor system, mainframe or otherwise, from five years ago is probably less powerful than a one-processor system now. It sort of normalizes it if you talk about system capacity in MIPS rather than in terms of "x processors of type y at z megahertz."
Don't compare the practice to shipping "miles per hour," but rather to horsepower. General Motors could, if they wanted to, say that they shipped umpteen million horsepower worth of engines last year, and it wouldn't be that hard to understand. It's just a different way of counting.
I'm curious as to the philosophical framework that leads you to that position.
Ask the people who wrote the law. I say that corporations are deserving of consideration because the law says they are. People smarter than I am-- and I dare say smarter than any of us are!-- have come to that conclusion over a period of many, many years.
And, just for the record, a 768-processor Origin 3000 from SGI works just like a 2-processor Origin 300. If you ignore little hardware differences that don't affect your apps, it also runs just like a 1-processor Origin 200 or Octane from five years ago.
I had the privilege of getting time on SGI's 768-processor system in Minnesota several months ago. It was... nice.
Okay. Let's talk about RSA. Although since you don't seem to draw any impacts from the RSA patent, I have to just guess at your meaning.
First of all, RSA was used as a key component in SSL. Everyone who banks or makes purchases using the Internet has access to SSL. To say that the patent on RSA hindered the adoption of SSL-- which you didn't, but I can only guess you meant by "the result of all of which certainly included the lack of security in software for any but the deep-pocketed"-- seems unlikely at best.
Furthermore, in 1997, Flinn and Jordan of Alston & Bird LLP published an article entitled, "Using the RSA Algorithm for Encryption and Digital Signatures: Can You Encrypt, Decrypt, Sign and Verify without Infringing the RSA Patent?" Read it here.
In this article the authors say, in part, "The existence of the patent, and RSA Data's aggressive litigation posture, have chilled the interest in both commercial and non-commercial implementations of public key encryption and digital signature technologies." This would seem to be evidence for your assertion. But read on: "Many have taken for granted the bald assertion that the 'RSA Algorithm is patented,' without examining the patent itself, or more particularly, the claims of the patent." In a footnote to this statement, the authors cite Schneier's Applied Cryptography as an example of one misinterpretation of the patent.
This article goes on-- at great length, not worth reproducing here-- to explain why the process of verifying an RSA-generated signature is not covered by the patent. Furthermore, the process of generating an RSA signature is also not covered by the patent, due to the existence of the identical Pohlig-Hellman algorithm invented in 1975, two years before RSA. Basically, the only thing that the patent covers is the process for generating RSA key pairs.
So basically, if you look at it closely, as those two patent attorneys did, you'll find that the RSA patent wasn't nearly as wide in scope as most people seemed to assume it was.
Maybe you can make the case that some kind of harm, direct or indirect, was done as a result of the RSA patent. But upon analysis it will seem pretty clear that that harm was the result not so much of the patent itself, but rather of the misinterpretation of the meaning of the patent by one party or another.
Why did you excise part of my post? When I said, "corporations deserve, and are legally entitled to, the same consideration as individuals," you quoted me as "corporations deserve... the same consideration as individuals."
I think you... are misrepresenting my... statements... by taking... them... out of context.
The law says the corporations are, in the legal sense and with exceptions, persons. That's it.
Imagine that in your flight sim for the USAF, that you had to go and check the patent status of every technique you used in that project.
We did. You wouldn't believe the things the Department of Defense makes you do when you're contracting for them. You have to deliver to them, as a condition of their acceptance of your product, the evidence of your research into patents and copyrights and so on that might have an impact on your work. I wasn't working on that, of course, but I got to hear about it. Everything we did fell into three categories: COTS (commercial off-the-shelf, which is exempt from background checks because the responsibility is the vendor's), non-commercial licensed (like licensing the onboard computer microcode from Motorola and Boeing), and in-house developed. Everything we did in house had to be gone over with a fine-toothed comb.
None of this crap about patenting processe. You don't patent "a way of propelling a ground vehicle", you patent "this engine right here", and "this specific drivetrain".
Hmm. Have you ever read a patent? The language is written by lawyers, which makes is hard to sift through, but also very, very precise. The title of a patent might be, "system and method of propelling a vehicle," which I'm sure sounds unacceptably vague to you. But the actual patent consists of a dozen or more pages of details: the what, how, and why. I read a patent this morning that describes an API; it was 70 pages long.
For example, Amazon's controversial patent (5,960,411) is entitled, "Method and system for placing a purchase order via a communications network." That's vague as all hell.
The abstract offers more detail in the form of a paragraph. The rest of the 19-page document describes Amazon's method in resounding detail, including where the cookies live and how single-item orders are consolidated into multi-item orders to reduce costs.
It's very thorough, and it's not obvious at all. I think the USPTO was entirely justified in giving Amazon a patent for it, in and of itself.
There are other factors, of course-- should Amazon have gotten a patent for a technology that others developed around the same time?-- but in general, the idea was specific and unique enough to warrant patenting.
So generally I think you might consider reading a few patents before you pass judgement on the whole system.
Yes, yes, RMS asserted that patents don't encourage what he called "incremental innovation." I don't deny that. I'm saying that reading the whole article left me with the impression that what he was calling "innovation" was nothing of the sort.
Back to the example of LZW compression. If you wanna use LZW, license it from Unisys. There's nothing wrong with that.
If you wanna improve on LZW, go right ahead! The algorithm is documented, fully and completely, in the patent application if nowhere else. If your improvement is significant enough to merit it, you'll be granted a patent on it yourself.
If you oppose paying Unisys a licensing fee-- which RMS clearly does, on (for lack of a better word) moral grounds-- then don't use LZW. Use another algorithm, or develop your own.
Through that process, you weren't prevented from innovating. You were prevented from making use of somebody's patented innovation without complying with that somebody's terms. It's not the same thing.
Now we're getting to the crux of the matter. If asked, I will argue in favor of the status quo. You argue against it. This is our topic. I debated in college; I'm comfortable with this.
In arguing to change the status quo, we must first establish that the status quo is worth changing. We can do that in two ways: one would be to demonstrate that a different circumstance would be markedly better than the status quo; the improved circumstance must be sufficiently better than the status quo that it merits making the change at all.
The other method is to demonstrate that the status quo is simple unacceptable in and of itself, by some arbitrarily assigned criteria. If we, for instance, all agreed that butterflies are wonderful, and we could demonstrate that the status quo implies the extinction of butterflies, then we have a prima facie reason to change the status quo.
I'm not hearing either of those arguments expressed very clearly, though. The "patents stifle innovation" argument doesn't really mean anything without a compelling "but for" case. Can you assert, in some reasonable way, that innovation would be better served in the absence of the patent system? That seems to be what RMS is arguing, but his assertion isn't really backed up by his examples.
On the other hand, is the patent system somehow morally wrong? Should we change or abolish it, not because there's a better way, but just because the status quo is so completely unacceptable?
I guess what I'm saying is that I don't see anybody advancing any good reason to consider another way of doing things.
Your logic is severely flawed. The development of a new car can't be reduced to a count of the number of parts. Vast amounts of effort from large teams of people, and huge amounts of money, go into the design of a new car before the first part is designed. The sheer number of pencil sketches-- and, more and more often, digital paint sketches-- that get produced, revised, reused, or discarded in the early design phase can't be underestimated.
Likewise, software is far more than the sum of its lines of code. The last big project I worked on-- a flight simulator for USAF-- was three years in development before the first line of code was written. There was a ton of work done on design and architecture, building concept demonstrators and mock-ups of the operator interface, and on and on. Practically every detail of the system was figured out on paper and in meetings. Once the time came to code it, it was almost just a matter of writing things down in Ada and C++.
There's no question that GM's designing of a new car and some guy's writing of a mail daemon are very different things. But when you start comparing apples to apples-- large-scale mechanical or civil engineering to large-scale software engineering-- I think you'll find that the two are more alike than they are different.
You seem to agree that the scientific model works well, so I guess I'm not understanding why you think corporations deserve distinguished status.
First of all, corporations deserve, and are legally entitled to, the same consideration as individuals. Corporations are, in the eyes of the law, persons, with standing distinct from their stockholders. For tax purposes corporations and individuals are treated differently, but that's a different matter entirely.
To promote the progress of science and useful arts, not to promote financial gain.
The end result of the patent system, as you so ably quoted, is to promote the progress of science and the useful arts. We promote progress by offering incentives. These incentives include, and have included for over 200 years, grants of rights of exclusivity.
One leads directly to the other. "Get rich by being inventive!" benefits both the inventor and society.
First of all, I disagree with your fundamental premise, which is that "society only benefits when a patented invention would not have been rediscovered and published during the term of the patent." This is just patently (pardon the pun) untrue. Whether somebody else might have come up with the invention in the next 20 years is irrelevant, both to the criteria for granting the patent and to the question of benefit to society.
But even more important is that you're asking the patent office to become expert in every field of human endeavour. Before one can judge whether an invention could be independently reproduced by another, you have to understand the invention, as well as its place in the greater scheme of things. I don't know anybody who could make that kind of judgement in a fair, fully informed, and impartial way, in any field.
The quote from Douglas Adams comes into play at this time: "It is a rare mind indeed that can render the hitherto non-existent blindingly obvious. The cry 'I could have thought of that' is a very popular and misleading one, for the fact is that they didn't, and a very significant and revealing fact it is too."
No need to patent them? Of course there's no need to patent. A patent is a right that the government grants if you choose to apply for it. It's not necessary. It's an incentive.
Look, the government makes a deal with you, the inventor. You can choose to get a patent on your invention, which will grant you the right to prevent others from making something like what you've made for a certain amount of time-- usually 20 years. But in order to get that exclusivity, you have to describe your invention in detail and make that description available to the whole world.
This serves two purposes: on the one hand, it gives your competitors a fair opportunity to keep from stepping on your patent, because you're telling them exactly what you own the exclusive rights to.
On the other hand, at the end of the term of your patent, the whole world knows exactly what you did and how you did it.
So at this point you have a choice. You can either try to keep your methods and ideas secret and see what happens, or you can take the patent and get to be the king for a while before turning over everything to anybody who wants to look it up.
Of course it's not necessary to patent your invention.
What? I really can't tell exactly what you're trying to say. It sounds as if you think patents are granted on individual components, like pistons. Nobody owns a patent on pistons; if anybody tried to apply for one it would get rejected because pistons have been used in many ways for thousands of years. Somebody may own a patent on a particular type of engine that incorporates a piston in its design, but that doesn't mean they have exclusive right to use pistons.
Basically I can't really figure out whether your post makes any sense, but is just obscured, or whether it's just nonsense from asshole to elbows.
I guess the range of experience is pretty broad. In my experience, the process of designing and building a big software application is pretty similar to designing and building a big physical thing, like a car. You have the same people doing the same sorts of jobs. The difference only comes in the manufacturing. When you release a car to manufacturing, huge factories full of people work to build them. When you release software to manufacturing, CDs get burned, usually in a big duplicating machine.
But the stuff that happens before manufacturing-- the designing, prototyping, testing, marketing, and so on-- is almost identical. The difference being that building a car involves making technical drawings which act as the schematic for manufacturing. Writing a big software app involves making source code that acts as the schematic for the compiler.
In short, if RMS is trying to apply his analogy to commercial software development-- as opposed to one- or two-person efforts-- then I think he's just plain wrong.
Your time frame is too short. Of all the software or software-related patents that have ever existed, the vast majority are still in effect. You can't say that software patents don't do their job of releasing knowledge into the public domain if you don't consider a sufficiently long timeline.
Again, you're confusing pure research with commercial efforts. The same argument you just used could be applied equally well to medicine: without the free exchange of ideas and information among the medical science community, we'd all be dying of smallpox and whatever the hell else.
And yet we still grant patents in the medical field. Medical instruments and devices are patented. So are drugs.
I'll say it again: there's a line between fundamental research and knowledge in the field of computer science, and commercial uses of that knowledge to create products, processes, and services. In medicine and other fields, that line is fairly well defined. (Although the recent discussions of the IP related to the Human Genome Project cast this in doubt.) In computer science and computer engineering, the line is not clear at all.
Examining the history of the past couple of decades I see scant evidence that the current patent law fosters this. I appears more likely that it suppresses it.
Obviously your timeframe is too short. Patents last twenty years. Look at the last 100 years and see if your opinion changes.
The rate at which innovation happens has increased in the past half-century. Perhaps it's appropriate to talk about reducing the duration of patent grants. Then again, maybe not. Wiser heads than ours picked the 20-year rule, so we shouldn't change it lightly.
Okay, so it's possible to use the patent system in a way that stifles innovation. That doesn't make patents inherently bad or wrong, like RMS seems to think they are.
I agree that some things shouldn't be patentable. Mathematical formulae, for instance, aren't patentable. But processes are. For instance, the process for making steel was patented when it was invented.
There's a line between algorithms (like bubble sort) and processes (like refining steel). But where is it, exactly? Does it lie in the complexity of the algorithm? Like maybe simple ones should be public domain while complex ones can be patented? That doesn't make any sense, though, because the simplest algorithms are often the most brilliant and innovative.
The problem here is that we're fundamentally talking about a but-for situation. The only way to judge the value of the patent system is to imagine a world in which it didn't exist, and that's hard.
In what way would we, as consumers, be better off if the spreadsheet had been patented?
Bah. The fact that you're asking the question means that you fundamentally misunderstand my point. The question to ask-- which is admittedly hard to answer!-- is "What innovations would we, as consumers, have access to if the spreadsheet had been patented, but that were not invented because no patent protection was available for them?"
I know, I know. It's impossible to answer that question. But that's kind of the point. If we don't offer patent protection for software innovations-- truly deserving ones, I mean-- then we'll never know what we missed.
I have yet to see a single piece of evidence that software patents truly discourage innovation. In fact, the converse is likely true. Today, it's not legal for me to use LZW compression in one of my programs without licensing it from Unisys. If I choose not to ignore that fact, I can do one of four things: give up, or license the algorithm, or use some other algorithm, or develop my own. If I don't want to license, and I don't want to give up, then I have to innovate! (Or I have to use somebody else's innovation that is freely available to me, but that's basically the same thing.)
If you read the article, you'll see that Stallman argues that patents don't encourage innovation. Any thoughts on this?
I admit that I didn't read the article slowly or carefully, but that's not how I understood it. (Marge: But you liked Rashomon! Homer: That's now how I remember it.)
It sounds to me like RMS is arguing that patents are an obstacle to the writing of software using well-known algorithms or principles. The example RMS gives of LZW compression, for instance, says that the programmer who wrote Compress "use[d] an idea he had found in a journal, just like software authors had always done."
That doesn't sound like innovation to me. It sounds like the programmer wrote a utility using an algorithm that was under consideration for a patent, and got burned by it. That has nothing at all to do with innovation, which is the process of starting from scratch-- or from a foundation of public knowledge-- to come up with a truly new idea. Not program, idea.
The larger question is this: should the LZW compression algorithm have been patented? This kind of goes back to what I said in a previous post about pure research versus commercial research. There's an argument to be made that fundamental algorithms-- like the stuff in TAOCP-- should be part of the field's collective body of knowledge, unencumbered by rights or ownerships. Maybe things like LZW compression should be included in that category.
Of course, Unisys disagrees with that assertion.
It's a tough question, and it's hard to know what to do.
Software doesn't need patents to protect it becuase the amount of effort to reverse engineer some code is often more than the energy needed to code something new in the firstplace.
But the code itself isn't what's patented. The patent goes for the "process, machine, manufacture, or composition of matter, or any new and useful improvement thereof." The code itself is covered under copyright, and possibly under various protections of trade secrets.
So you don't patent Adobe Photoshop. You could, possibly, patent software for editing digital images in such-and-such a way. It has to be fairly specific. For example, consider patent number 6,337,901, "Customer billing relationships software." (Read it here.) This patent doesn't cover the software itself. Rather it covers the design, methodology, and architecture of the software. In this particular instance, the meat of the patent describes the process that the software implements for generating its output.
So it's not about whether or not you can reverse-engineer Photoshop. It's about whether you should be given the right, for a limited time, to exclude others from writing software that does the same thing Photoshop does.
(This sounds silly because Photoshop is neither patented nor patentable.)
Maybe I should have qualified "important." I think it should have been clear from context. I meant "important" in the large-scale economic and social senses. New music and art is important, and that's why we directly subsidize artists in some cases, and ridiculously overpay them in others, depending on whether they're popular or not.
It's the same basic premise, though, implemented in two different public policies. Artistic development is encouraged through subsidy. Commercial development is encouraged-- among other ways-- through the granting of patents.
Fewer high-quality programmers who have access to each other's code will produce better code faster than a whole bunch of overpaid amatuers working in isolation.
If we were talking about pure research, or the sciences, you'd certainly be right. That's exactly the model that scientists follow: publish early and often, and read everything.
Maybe we should talk about the difference between academic programming and commercial programming. It's the same as the difference between biological research and drug manufacturing.
If computer programming is a science, then pure research in the field should be in the public domain. But the use of that pure knowledge to achieve commercial gains is something else entirely.
On the one hand, we have the closed, proprietary, patented, buttoned up shrink wrap license vendors. On the other hand we have free software. Compare the amount of resources each consumes relative to its output.
I don't mean to be a grouch, but I disagree. I deal with quite a bit of commercial software, and I have to say that a fairly small percentage of it is complete crap. Of all the freely available software out there, though, there's a lot that's poorly documented, buggy, maldesigned, or all three. A few outstanding exceptions-- like Perl or Apache-- do not an argument make.