OK, Mark, I just now read for the first time your earlier copyright vs. patent posting that you linked here. Boy are you way off base! :-)

Given the excerpt that you provide, you make a decent case that software, if viewed as a series of instructions, should not be copyrightable. However, it does not therefore follow that it *must* be patentable.

Patents apply (at least in theory) to inventions. They do not apply to instructions, but to the product produced by said invention. If I patent a better mousetrap, I don't (again, in the original theory) patent whether I attach the left or right side of the hinge first; I patent that there are two hinges attached to a spring that launches a metal bar at the rodent's neck. At the same time, however, I do not patent the concept of a machine that kills mice.

Again, that is the original theory and design of the patent system. With that understanding, what would software (instructions) be? The code would be the "insert left hinge first" stuff, which is not what you're patenting. The purpose of the code would be the overly-broad "thing that kills mice", which would, again, not be patentable. So which part would fall into the specific implementation? None of it! There is no tangible invention, so there's nothing to patent.

The patent system went wrong, however, with the "business method patent" concept. That's when "ways of doing things" rather than the things themselves became patentable; that was via case law, btw, not an act of Congress as far as I am aware. Now the method, the instructions, the software, is somehow subject to patent. However, the courts have, in their infinite stupidity, established that a patent to *an* instruction to do X gives you a patent on *all* things that do X. The patent application is written that way (because patent lawyers are paid a lot of money to write them that way), and the courts gleefully enforce it that way. So once anyone has come up with the *concept* of killing mice, they can patent "a device that kills mice [using a computer]". Now, suddenly, no one else can kill mice at all. cf., the Microsoft Stacker case from the early 1990s.

So in theory software shouldn't be patentable, but in our new-and-not-improved patent system everything is patentable in the most general case. The result has been a chilling effect on innovation in the US. (cf., the various arguments made against software patents in Europe.)

So software should not be patentable. You argue that, as a set of instructions, it shouldn't be copyrightable either. So if software cannot be patented, and it cannot be copyrighted, and it's obviously not a trademark... then software must be in the public domain!

Guess what? That's exactly what the Free Software Foundation was founded for; software was being taken *out* of the public domain, and the GPL and company are an attempt to simulate putting it back.

This is exactly what I was looking for: some reasoning as to why software shouldn't be patentable. You're right; it has to be something. Just like a work of art or a mousetrap, software was created by someone, and as such, it must legally belong to someone. If a person cannot legally reap benefits from software in the same way that he can an invention or a book, then why should this person make software at all?

Part of the justification for the existence of copyrights and patents is that it gives the artist or inventor an incentive to produce work. A copyright guarantees that an artist can reap economic benefits from his work. This is good for society because art is good for society. This is good for the artist because he makes money. The same goes for an inventor, except the benefit to society is something "useful." Software is useful to society. But there also needs to be a reason for the software creator to continue to write software. People who write code for Linux do it as a hobby, but if we're talking about making Linux into the world's next major operating system, then it's going to be a full-time job, and people will want to be compensated for that job, because, even if we want to be idealistic and say that "information should be free," we live in a world that runs on money. The perfect world of the 24th century isn't here yet.

And if you'll look at 17 U.S.C. 102, a distinction is made between things that can be copyrighted and things that cannot. Ideas, procedures, processes, etc. cannot be copyrighted. In other words, utilitarian things cannot be copyrighted. This means that Lexmark's crazy toner chip shouldn't be copyrightable. It isn't "art." But over at 35 U.S.C. 101, we learn that "any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof" is patentable. This is how a "way of doing things" can be patentable; a "way of doing things" is a process. And you're right in that, as applied to software code, this can become stupid. Amazon should be able to have an exclusive license to its own particular code that does one-click shopping. Unfortunately, it now has an exclusive license on any and all code that does one-click shopping. That doesn't seem right.

I would suggest that (1) like authors and inventors, software writers should be granted a "limited" exclusive license to their software, so that they have an incentive to make that software in the first place. But I would further suggest, based on your observations about software, that (2) occupies a strange position between copyrights and patents. But this position is not the public domain. There's got to be a third way that preserves the rights of the public to use the software while at the same time respecting the rights of the person who made the software. This is one of the few times I disagree with the folks at EFF.

WHO CARES.

Mark, you make a few very correct points, and several very very wrong ones. :-) In no particular order:

1) I think you mean FSF, not EFF. EFF is not a copyright-handling body for the most part; they're more like an ACLU for the Internet. I happen to disagree with the FSF on some things, too, but not all of them.

2) "[S]oftware was created by someone, and as such, it must legally belong to someone." That is your first mistake. It's a common mistake, because for 30 years the copyright cartels have been actively and deliberately engraining that viewpoint into the public mindset. It is still untrue. Information is not a physical object. It does not obey the law of conservation of matter and energy. The concept of "property" as it applies to information *is completely and 100% artificial*. Specifically, in the US, it is a mechanism to "promote the progress of science and the useful arts". Any other "securing of exclusive rights" on information is unconstitutional. It's one of only two places that I know off of hand where the Constitution specifically gives a reason for Congress having a certain power (the other being the 2nd Amendment).

Your argument generally goes "All information must be exclusively owned by someone. Copyright doesn't apply. Therefore it must be patentable." The problem is that first statement is untrue.

3) The "Linux as hobby" myth sadly still persists. Wrong wrong wrong. The majority of code that goes into the Linux kernel these days is written by people who are paid full time to do so. Their day job is "improve the Linux kernel". They are paid a good salary to do so. They just don't happen to work for "Linux, Inc." They work for IBM, Red Hat, Novell, etc. Red Hat releases all of their code under the GPL. They are a multi-billion dollar company with positive cash flow. IBM, well, you know IBM. While many open source projects are coded by hobbyists, many are also coded by international corporations. The list of such is far too long to include here.

3) Why should someone write software? Because it gets the job done. Every company needs an accounting program. They can either each spend $1 million to license one that they can't control, or they can, through an open source process, co-develop one that everyone can use and that they can control the future development of. The latter saves them lots of green, and in no way whatsoever reduces their main business market in selling paint, manufacturing widgets, delivering carpets to your home next day, etc. Most software in actual use falls into that category. That's cold hard cash that the company saves by doing things open source. That's money in the bank. 24th century need not apply.

4) You argue that software authors should get "limited time exclusive rights". They already do. The problem is that "limited" is no longer limited. What's the lifetime of a patent? 20-something years. That's an eternity in the computing world. Under that, any concept of any kind in software since 1985 (that is, pre-Macintosh!) would still be patented. Copyright? 70 years after the death of the author, to be extended as soon as Disney decides they need to protect The Mouse again. In 70 years, the code is absolutely unusable. The hardware that runs it doesn't even exist anymore. Of course, unless you have the source code available it wouldn't do you any good anyway.

What's a reasonable length of time for copyrights or patents on software, even if we allowed patents? About a year on a patent, 5 at the most for copyright. The industry just moves too quickly for anything longer to not be damaging. (Of course, most printed books go out of print in a year, making them inaccessible to anyone yet the publisher and author are no longer making any profit, therefore it's not promoting the progress of science or the useful arts.)

For more on this topic, I highly recommend the following site, particularly the bonus article linked in the sidebar:

http://www.garfieldtech.com/copyright/

and the book "Free Culture" by Larry Lessig. It's even available as a free download at http://www.free-culture.cc/ .

I am not ant-copyright or anti-patent in general. I am anti the falsehood that information is innately "property" the way that a chair is "innately property". That is simply untrue, and until you understand the crucial difference, you will never understand copyrights or patents, or how they can be legitimately useful at "promoting the progress of science and the useful arts".

Man, I feel like I'm in STF again! Except without the threat of resignation and the personality conflicts.

(1) I should explain. First, as a member of EFF and a fan of Cory Doctrow, I'm quite aware of what EFF is. EFF is against software patents, too. Curiously, I use the same description when I tell people what EFF is: "It's like ACLU for the Internet."

(2) Maybe I should have made myself clearer. I have read Free Culture (and found it excellent), and I understand that information's natural state is to be free. I am also aware that, prior to the Statute of Anne, no one had an exclusive right to his work. In a state of nature, there would be no copyrights. A copyright is granted by the state (or some authority) to the author of a work. A copyright is artificial. A patent is artificial. They're both "artificial" in the sense that, if we were in a socialist utopia where money were no concern, there would be no copyrights or patents. No one would be granted the ability to "own" a concept. And the Second Amendment grants the people the right to keep and bear arms and provides a reason. None of the Bill of Rights give Congress any power <g>.

In the United States and other countries which have money and concepts of ownership, people are granted -- by the state -- an exclusive right to, say, books they write. Why does this happen? To "[promote] the progress of science and the useful arts." Okey-dokey. But more than that, why does granting someone a patent promote useful arts? Because it gives that person an economic incentive to invent something that will be added to the human catalogue of "useful arts." You've got a guy who can either write books or shovel coal for a living. Prior to the Statute of Anne, he would shovel coal for a living and either (a) not write, or (b) write occasionally as a hobby. No one could make a living as a writer because no one could own the rights to his work (the exception to "making a living," of course, is if you had a patron who paid you, or you were Shakespeare and could charge admission to your plays. But then again, anyone could perform your plays). Once the Statute of Anne comes into play, suddenly, if you want a copy of "The Lady of Shalott," you must go to Alfred, Lord Tennyson. The Statute protects his ability to make a living. If someone else starts selling copies of "The Lady of Shalott," Tennyson can take him to court. Prior to the Statute of Anne, there's no such protection. The Statute allowed writers to make a living writing by saying, "Okay, for twenty years (or so), no one else but you is allowed to sell what you wrote. This will encourage you to devote all your time to writing, because we believe that you're more valuable to society devoting 100% of your time to producing art than to shoveling coal 75% of the time and writing novels 25% of the time." So, while information's natural state is free, copyrights and patents are a necessary evil, much like the rest of the social contract. We surrender the complete freedom of information so that, if one day one of us produces information, that person will be able to make a living producing that information, which is good for society because more information is good. Art is good. And the Statute also recognizes the "society" element by creating the Public Domain. It recognizes that, okay, you've had your fun time making money from the art you produced. But after a while this art that you produced for society belongs to society. The Public Domain is good. I have no dispute with that.

(3) So, exclusivity of reproduction is an artificial creation of a state. Information's natural state is free. Fair enough. Software should be neither patented nor copyrighted because it's not quite an invention and it's not quite a work of art. Okay. So what is it? I argue that it must be "something" because, again, in a socialist utopia or a Hobbesian state of nature, no one would be granted exclusivity of reproduction for software just as no one would be granted exclusivity of reproduction for a book or a mousetrap. But we aren't living in a socialist utopia! Counselor Troi may be able to tell Mark Twain that in the 24th century, there's no money, but we're not quite there yet. We still live in a society where people need to reap economic benefit from whatever it is they do. And if they can't reap economic benefit doing X, they won't do X unless they're making money doing Y and do X as a hobby. Giving away software is no business model unless you're some sort of organization that gets its revenue from something other than the sale of your product (like the Mozilla Foundation). Those people at the Mozilla Foundation are getting paid! The people who work at Novell tweaking SUSE are getting paid. We can't "[promote] the progress of science and useful arts" unless people have an incentive to do so, and in 21st-century America, that incentive is monetary. Why can Larry Lessig give away copies of Free Culture? Because he works for Stanford University. Same goes for Cory Doctrow and any of his writings, except he works for EFF. Symantec has to charge people for Norton Antivirus because if it didn't, Symantec wouldn't be in business, and we wouldn't have Norton Antivirus (whether that's good or bad is debatable, of course).

Now, 120 years on a copyrighted work? That's ridiculous. It's no longer Walt Disney himself that's benefitting from the exclusive right; it's The Walt Disney Company, and Mickey Mouse should belong to the public. He's had his time belonging to someone. It's high time he came into the Public Domain. The Sonny Bono Copyright Term Extension Act is indeed ludicrous. The DMCA is indeed ludicrous. The DVD region system is ludicrous. DRM is ridiculous. But these last three things have to do with my rights as the owner of a physical thing and not with the author's rights. As indicated in your "Myth of Intellectual Property" essay, you rightly conclude that, once the DVD is mine, I can do whatever I want with it.

Again, I maintain that I'm not arguing that information's natural state is to be "owned." I maintain that a person who creates information is granted, by the state, an exclusive right to reproduce that information so that he will make more information. That's good for society. And in a society that runs on money and trade, that's how information gets created. In the society of the 24th century that doesn't care about money, it won't be that way. But until then, software is information, and in order for Steve Jobs to keep making software -- which, again, is good for everyone -- he's going to want to be paid for it. (Of course, Steve Jobs doesn't take a salary from Apple because he's already wealthy enough. And he's probably got stock options.)

You're always welcome back to STF if you'd like, Mark. In fact I dare say the Connie needs you. :-) You know you want to. You keep mentioning Star Trek!

You're slowly coming around to the correct point of view here :-), but you're still making a few errors.

1) The Statue of Anne and similar early copyright legislation was not intended for the author to make a living, but for the publisher. That is, the copy-er. The benefit to the author came in later. I don't disagree that it can be of benefit to have copyrights for authors, especially as someone whose primary income for the past several years has been the creation of copyrighted content. :-) However, let's remember the origins.

2) The financial incentive to create need not be for extorting profit from the itemized, productized sale of copies of a copyrighted work. As I explained above, billions of dollars of software, proprietary software and Free (Libre) Software, is written for incidental use. Pixar makes less than 5% of their profit from licensing their admittedly amazingly advanced graphics software to other studios. If they and the other studios got together and collaborated on it instead, Pixar would lose at most 5% of their annual revenue and gain hundreds of very expensive developers (at those other companies) in the process, a net savings. Their primary income source, making movies, would be unaffected. The same holds true for thousands of other applications.

Now, I'm not saying that because one can make money from software without productized sale item extortion (note: extortion in the "get something for" sense, not in the felony sense) that we should do away with copyright. I am pointing out that the "no one would write anything except as a hobby if it weren't for copyrights" argument is false, especially so for software, which has so much incidental benefit.

3) To get back to the original point, your argument still comes down to "Software must be protected by some sort of 'exclusive right'; copyright doesn't apply; therefore, it must be patentable." You are still making the error in assuming that (a) Software somehow "must" be protected by "something" and (b) Everything must be either a copyright or a patent.

(a) While you can argue that some sort of exclusive rights protection for software is beneficial, and I will not, in principle, disagree, it is not a given. There are a great many information-related concepts that have generally been outside the scope of exclusive rights. Take scientific facts about the world, for instance. That's certainly information, but it has generally been understood to be something to which one cannot have exclusive rights. One can have exclusive rights to a machine used to discover it, but e=mc^2 is not subject to exclusive rights. Similarly, databases of factual information were similarly uncopyrightable. Congress recently considered a bill to make them copyrightable, and there was a major outcry on the grounds that it was inane and did not "promote the progress of science and the useful arts". I do not recall what the eventual outcome was.

So your claim that "well it must be something" is false. Why does it have to be? Not everything in the realm of knowledge "must" be covered by copyrights or patents, only those things where doing so clearly "promote[s] the progress of science and the useful arts".

(b) Copyrights and patents are not the only God-given form of exclusive intellectual right. The Constitution does not specify the means by which Congress may "secure to artists and inventors exclusive rights..." Patents, Copyrights, and Trademarks are the current mechanisms. (Trade Secrets are by definition outside the scope of governmentally-granted exclusive rights, although under the DMCA they suddenly get patent-level protection. Joy.) Now, I don't believe that software falls under trademark law, to be sure. However, the point is that if you argue that i) Software should be protected by some exclusive right and ii) Copyright isn't it, then it does not follow that it should be patented, but that it should be covered by something that is not-copyright.

What should such protection entail? Let's suppose that software is protected by the Froboz system. What should the Froboz system of exclusive rights entail? A few thoughts:

a) Very short-term protection. 5 year maximum, I would say, with a "dead work" or "out of print" clause for it to go into the public domain earlier. (See the linked essays above for more details.)

b) Opt-in with deposit. All copyright used to be this way until the 1976 Copyright Act. I'd argue copyright should go back to that, but in the case of software it is most important due to the difference between "use version" and "editing version". Without the source code publicly available, having the binary in the public domain is worthless. Again, see the earlier linked essays.

c) Covers a specific implementation only, not the general concept. You can Froboz a program that compresses a disk, but you cannot Froboz all programs that compress a disk. Similarly, you can Froboz your code for one-click shopping, but not all code that accomplishes one-click shopping.

d) Interference with the work falling into the public domain once the Froboz has expired is forbidden. That is, DRM-and-friends are explicitly forbidden as they interfere with the eventual release into the public domain, as well as the deposit component.

e) Copyright-like fair use. I would actually suggest more lenient fair-use, to include all personal use even if it includes multiple personal copies. That would allow buying one copy of MS Office for both your desktop and your laptop, but not installing that copy of MS Office on the laptop of everyone in your dorm building.

Gee, this is starting to sound an awful lot like the copyright reform measures proposed in the linked essays. :-)

So how about it, Mark? Would you support Froboz for software?

I know I'm always welcome back at STF, but the question is, when will I have time for STF? Plus, I don't know any of the people in the club right now.

(1) Okay, fair enough. My reading of Free Culture was a year ago and I couldn't be bothered to go look up the information about the Statute of Anne.

(2) That's good for Pixar. Their business model has pretty much nothing to do with selling software, so they could give it away. But what about Microsoft? What about Apple? Their business model is dependent upon selling software. It's the same as Larry Lessig giving his books away online: his business model is being a university professor; being an author is almost a hobby at this point. He doesn't need the income from his books to survive.

(3) Man, now I know why Colin got so incensed with you in the chat room all the time. If you'll read the words I wrote, you'll note that I continually said in my second post, so that I wouldn't be misunderstood, that "Software should be neither patented nor copyrighted because it's not quite an invention and it's not quite a work of art." In that post, which clarified my original comment, at no point did I say that "Software can't be copyrighted; therefore, it must be patentable." While that was my previous assertion, you've convinced me that there must be a third way aside from patents and copyrights. My argument is no longer that software must be patented because it isn't copyrightable. Or perhaps I wasn't clear enough in my second post. I'll make myself clear now. Clearly, software isn't a work of art as defined by 17 U.S.C. 102, so it can't be copyrighted. And patent law, covered in Title 35 of the U.S.C., isn't appropriate for software. Therefore, there must be a third way aside from patents and copyrights.

Which brings us to your suggestion, which is great. It's what I was looking for, despite your assertions that it's not what I was looking for: a third way that makes a place for software. Yes, I like the Froboz system. I was never married to the idea of patents for software; it's just that I couldn't think of how else it would be classified.

Ah good, we finally agree that I'm right. Always a good way to end a discussion. ;-)

Other footnote, my point was not that all business models would survive, or that since there is one viable business model that we must abandon all others; it's that the "no one would [create | write software | innovate ] if they didn't get a monoploy on it in return" argument is bunk. Even if the law were rewritten to make software-only houses like Adobe impossible (MS' business model is to abuse monopoly power, and Apple is a hardware company primarily, so they're bad examples), software would still be a multi-billion dollar industry. I'm not proposing that we do that, mind you, just pointing out the failure of that line of thinking (one that you did not directly argue, but seemed to accept as true implicitly).

And on the Connie you'd know me, Jaret Hargreaves, maybe Irene McGary, and there's my brother who is now following in our footsteps as the CE. :-)