The Divided House of GPL
Back in 2000 The Onion made fun of Libertarians and published this nugget:
Joking aside, a powerful idea attracts a wide range of people. Copyleft is such an idea. It turns copyright on its head by using authorship rights to enforce the public’s ability to distribute, modify and use the copyrighted work, rather than to curb it as is normally the case. Several copyleft licenses exist, the most prominent in software being the GPL, first released by Richard Stallman in 1989.
There are two main factions supporting the GPL: the pragmatic camp of Linus Torvalds and the ideological camp of Richard Stallman. The Linus camp sees copyleft as the enabler of a superior way to produce software, in which distributed and open development takes place because people are encouraged and protected by the license. The individual programmer is assured that their contributions must always remain a public good and cannot be coopted for private gain. Others may profit from the software, sell it, or support it, but the source code must be available, modifiable and distributable. This is a powerful motivator, the same force that makes people help the Wikipedia but not for-profit outfits.
For large-scale development involving multiple corporations, copyleft solves a type of free rider problem by ensuring that all participants must give back to the common pool of development. This protects investments and tends to boost returns, and a brief look at the Linux Kernel Mailing List shows that major tech companies are happy to play along. I bet you can do some game theory and prove some results for cooperation under GPL.
To the Linus camp the GPL is a means to foster this ecosystem, the end being better software. There are no moral imperatives or political reasons behind the whole thing, which surprises some people. Proprietary software is “alchemy” while open source is science. Here’s Linus:
In my book, what matters is what you do – whether you want to sell things is your personal choice, but even more importantly it is not a moral negative or positive. I’m a big believer in open source as creating good stuff, but I don’t think it’s a moral issue. It’s engineering.
So I think open source tends to become technically better over time (but it does take time), but I don’t think it’s a moral imperative. I do open source because it’s fun, and because I think it makes sense in the long run.
And here’s more:
Just to explain the fundamental issue: To me, the GPL really boils down to “I give out code, I want you to do the same.” The thing that makes me not want to use the GPLv3 in its current form is that it really tries to move more toward the “software freedom” goals. For example, the GPLv2 in no way limits your use of the software. If you’re a mad scientist, you can use GPLv2′d software for your evil plans to take over the world (“Sharks with lasers on their heads!!”), and the GPLv2 just says that you have to give source code back. And that’s OK by me. I like sharks with lasers. I just want the mad scientists of the world to pay me back in kind. I made source code available to them, they have to make their changes to it available to me. After that, they can fry me with their shark-mounted lasers all they want. This is where the GPLv3 diverges. It limits how you can use the software.
The Stallman camp, however, sees GPL-licensed software as the end itself. They claim that software should be free on moral grounds, citing several reasons. Hence it matters not whether the software or the process are superior. One must use free software regardless because it is the right thing to do, while proprietary software is inherently immoral. Here’s Stallman:
Supporters of open source (which I am not) promote a “development model” in which users participate in development, claiming that this typically makes software “better” — and when they say “better”, they mean that only in a technical sense. By using the term that way, implicitly, they say that only practical convenience matters — not your freedom.
I don’t say they are wrong, but they are missing the point. If you neglect the values of freedom and social solidarity, and appreciate only powerful reliable software, you are making a terrible mistake.
The fact that Torvalds says “open source” instead of “free software” shows where he is coming from. I wrote the GNU GPL to defend freedom for all users of all versions of a program. I developed version 3 to do that job better and protect against new threats. Torvalds says he rejects this goal; that’s probably why he doesn’t appreciate GPL version 3. I respect his right to express his views, even though I think they are foolish. However, if you don’t want to lose your freedom, you had better not follow him.
Discussions of copyleft often blur these two camps. For much development this is irrelevant – the license stands on its own irrespective of people’s motivations. But this schism explains periodical battles like the GPLv3 controversy, the endless flames when a proprietary source control tool was used for the kernel, and the GNU/Linux naming controversy. The distinction is also important when thinking about free/open source software and what to make of it.
October 16, 2008 at 1:40 am | Filed Under Culture, Linux
Subscribe to blog
Comments
15 Responses to “The Divided House of GPL”
Leave a Reply
Damn you, I was about to write a blog post on a very similar subject, and you beat me to it. The angle I was trying to highlight are the divide between people like Bryan on the Linux Action Show who are working on methods to “commercialize” the Linux desktop market with closed source, proprietary, software as a first entry for enterprises and the RMS/GNU/FSF-group and its followers who sees a philosophical and privacy/security problem with populating Linux desktops with closed source.
Bryan is trying to open up for a model for retail native high-profile software to be viable for Linux. Something that is a very small and to some degree controversial business today.
The divide in the industry is quite interesting to follow and I recommend those interested to listen to The Linux Link Tech Show (http://www.tllts.org/audio/tllts_266-10-08-08.mp3) where they did a lengthy interview with Bryan on this subject. The TLLTS podcasts are lengthy and quite slow paced but this interview is highly recommended for those curious about this issue.
Heh, nice summary of this controversy! Held a 2h presentation about copyleft licenses at my company yesterday (the slides are in Swedish) and was planning to write a blog post about it.
Thanks Burre for the tip about that TLLTS episode.
Claes:
Please point to those slides if they are public I’m from Sweden too and would be interested in how you portrait the issue. (I’m guessing this was pitched as a proposal for a business model at your company or something like it?)
I’m curious about success stories in the retail market for software on Linux. It is a quite uncharted territory. Most business on Linux is service based (except maby for some enterprise/virtualization software bundles).
Stallman is an extreme socialistic ideologue. He is a fanatic. Torvalds also believes in a socialistic model, but he is more pragmatic, and more wishy-washy when push comes to shove.
Of the two, Torvalds has had more a more significant impact on the world at large due to his superior technical abilities in writing kernel code. His ability to get along with others has also been important. Stallman has neither of these abilities to the that extent that Torvalds does.
But Torvalds permanently crippled Linux from ever becoming anything other than what it is: an elite OS for very smart geeks (many of whom are libertarians) and for corporate coders. He did this by deciding to release it under the GPL model. The GPL makes it nearly impossible to ever have Linux being a significant money-making product on its own. Torvalds did this by design. He baked this aspect into Linux on purpose, due to his own distaste for capitalism and money-making as an end in itself. In interviews, he implied that his dislike of Microsoft in the early 1990s had everything to do with how he decided to release his redo of the Unix operating system.
This is not to say that many companies don’t make money using Linux code because they do. But they do it in two main ways: by adding proprietary code to the Linux code, and by offering service contracts for their combined Linux-proprietary product. Red Hat is the best example of this usage, although IBM probably makes a lot more money off of Linux than Red Hat does.
But as for the adoption of Linux on the home desktop in America in a widespread manner, it will never come to pass despite the best efforts of starry-eyed socialists like Mark Shuttleworth. In this regard, Linux will always be a day late and a dollar short. It is not that you can’t do the basics of computing in Linux quite well. It is that you can’t do them easily unless you are a smart, dedicated geek who has a lot of time on his hands.
For the record, I’m pro-business in general and I lean Libertarian. I have NO PROBLEM whatsoever with proprietary software. I think the FSF’s morality arguments are flat out wrong.
@Burre: That was an interesting link, thanks for posting.
@Claes: is it online? Maybe Google can at least make it understandable?
@Rob:
It is true that Stallman is a fanatic, but is that always bad? Weren’t the founding fathers, in a very real sense, fanatics? What if you’re fanatical about the rule of law, or civil rights?
So I guess you need to decide whether the FSF’s moral arguments make sense to you. To me, they don’t.
I’m not sure what you mean by ‘socialist’. That became a non-word in current US political discourse. Bits of everyone’s ideas, from Locke to Adam Smith to Friedman, are “socialist” according to current usage.
Whether Linux will make money on its own, it’s hard to say. What do we attribute Red Hat’s profit to? Is that Linux “on its own”?
I disagree with your take on the GPL. It does not make it “impossible” by any means to have a decent, usable, flourishing desktop with good apps, many of which could be proprietary, at least from a legal / licensing point of view. There are many other problems, of course. See the Linux Haters Blog for some
I think though part of what you’re saying is that Linux won’t take off in the desktop because the ideological slant against proprietary software will keep it a marginal product. That might be true.
I didn’t say it was impossible to have decent, flourishing desktop using the Linux kernel. There are plenty of examples of those already, Ubuntu being the most conspicuous. I said it won’t happen in America, probably ever, unless we turn into another sort of nation in the next 30 years (not impossible given the way things are going now).
I believe the axiom from the Watergate movie “All the President’s Men” is the reason why Linux on the desktop will never be big here: follow the money. There is no money in Linux except under the conditions I spoke of before. In backroom server shops, where corporate coders can write and service the Linux and non-Linux code their bosses are selling, some money can be made via Linux. But in almost any other circumstance, Linux is a non-starter, money-wise. THIS IS BY DESIGN. It was baked into Linux by Torvalds. HE WANTED IT THIS WAY.
The GPL is anti-competitive. It is a legal embodiment of Richard Stallman himself. It is as though he is standing behind every legal transaction shouting his kindergarten teacher axiom “You MUST share with others.” Americans don’t like being told what to do. And the more dominating the personality they are, the less they like it.
When Steve Jobs was developing his NeXt computer operating system, he chose to use BSD as his starting point. The BSD license is almost the exact opposite of the GPL license. It places almost no commercial or sharing obligations on those who choose to use it. The last thing a guy like Jobs was going to do was put up with the likes of Stallman or Torvalds telling him what he could or couldn’t do with his company’s software.
And the issue of hardware drivers is also about money. Drivers are usually written by the hardware manufacturers. This costs them money. Linux exists in a financial Catch-22: manufacturers don’t want to spend money writing drivers for Linux because not enough people run one of the hundreds of Linux OSes. But, in part, people don’t want to run Linux OSes because there aren’t drivers written for some of their hardware. A lack of Linux drivers for printers is fairly common. And this problem isn’t going away either because new hardware is constantly coming on the market. So Linux is always in a game of catch-up with drivers.
Linux doesn’t have a seat at the table of corporate America BECAUSE IT ISN’T A CORPORATION. It is an operating system kernel that anyone can use for whatever reason they want. And this highlights a whole other set of problems having to do with Linux once it becomes something more than code kernel: it has no central authority representing it, no go-to set of people who are responsible for promoting it and representing it. The kernel itself does have this, but the hundreds of operating systems spun from it do not. So each one is left to figure out how to motivate its employees to do what employees are supposed to do: show up, get along and work through problems together. The issue of distro “forking” is an illustration of a situation where there is too much disagreement, not enough incentive to work out differences, and too much freedom without enough discipline or overarching authority. It is the equivalent of taking one’s marbles out of the game when the game isn’t going one’s way.
As for my use of the term socialism, yes, it is used very loosely to mean anti-capitalist in the sense that intellectual property is considered to be jointly-owned, or community-owned, not the sole product of one person, or set of people. Karl Marx saw socialism as the phase in between capitalism, or private ownership, and Communism, where all property was collectively owned, a sort of socialism on steroids. The socialism phase abolished private ownership of goods and services. Obviously, there can be mixed systems combining elements of both, such as China is currently experimenting with.
As for fanatics not always being bad, I disagree. By definition, fanaticism takes something to an extreme. Such conditions imply imbalance, lack of stability and sudden changes. I don’t believe the founding fathers of America were fanatics, in main. In fact, Madison’s checks and balances notion was a political idea used to tamp down extremist, or fanatical elements within the political process.
Rob,
Thanks for the thoughtful reply.
Regarding the “fanatism”, you are right. The usage of the word implies blind, extreme allegiance, inherently bad. So by definition, as you point out, I’m wrong. The founding fathers were perhaps “radicals” or “passionate”, but not fanatics. That’s not what you call people reading Montesquieu and advocating separation of powers. With that out of the way…
How exactly are you defining Linux above? If you’re thinking of Linux as a software collection, the kernel + GNU apps + X Windows + apps, and claiming that there is no money in selling that (say, in a box), then absolutely, there’s no money there.
So, for example, nobody will come along and take the kernel + X Windows, improve it, and sell it. By design, as you point out. Which means MacOS needs to happen on BSD, not Linux. Which means that, yes, one avenue for innovation and dissemination is shut off.
Is that a net loss for adoption? Maybe. Since this “innovative commercial Linux” will never show up, there will never be a greater pull on commercial software exerted by it. On the other hand, such a product might have introduced incompatibilities, so maybe it wouldn’t have helped. Has MacOS helped FreeBSD, net?
This point about competition is very interesting. Surely the GPL kills commercial competition. What’s the point of banging out that killer feature in a 3-week blitzkrieg if you’re forced to share it with your competitor in the end? This has profound consequences. Since I’m personally a free-market oriented guy (though not a fundamentalist, and not fanatical
, so this indeed is a troubling spot for me.
Now, there’s some technical competition going on. For example, both Rik van Riel and Andrea Arcangeli effectively wrote competing VM managers for the kernel back in the day, and the better one was picked. Companies compete to see what hypervisor code will get into the kernel, people compete to see what VFS should be like, and so on. There have also been separate trees, say the Alan Cox tree, that was certainly competition in a real way.
And of course, in the broad open source world there’s intense competition among apps. There are 985 blog engines, 500 editors, and so on, for any semi-popular type of app.
There’s an interesting way as well in which the GPL (BSD as well) fosters this non-commercial competition: since the source is available, the barrier to entry is drastically lower. It’s easy for someone to come along, tweak a project, and start competing. The cost of large pieces of software is prohibitive and stifles competition.
Ahh, here I disagree on what I believe are technical/factual grounds. The myriad distros have little to no effect on hardware drivers, because they depend solely on the kernel. So the same driver will run in any distro, provided it supports that kernel version.
There’s also the issue that manufacturers can get their drivers into the main tree, where for popular devices there’s a good chance the community will upkeep them as the kernel evolves.
The reason manufacturers don’t bother is that desktop Linux usage is puny. For video drivers the situation is more complex because X and other code are involved.
Again, I see facts very differently here. I believe Linux does have a seat at the table of corporate America. Look at the LKML, it’s @intel.com, @ibm.com, @google.com, @hp.com, all the biggies except for Microsoft. And that’s actual kernel development!
In more general business terms, I think Linux has penetrated mightly into corporations. For example, take a look at Gartner reports on server sales.
Where it lags is in commercial software production. But here I think the puny market share is reason #1, followed by users’ hostility towards proprietary software.
Rob, I understand your position but I respectfully disagree on a number of points:
* “GPL is socialism”: No, socialism is a political/economical ideology. GPL is a software/user rights based philosophy. It might be that GPL is a reflection of Stallman’s political beliefs, but that does not in any way require an economical system based on socialism for Linux to be widely adaptable and profitable. The two in this case are unrelated, and GPL can very much exist within a commercial free economy, it all boils down to what the merchandise is in both cases. In the retail market it is the physical (if one could call it that) product in the GPL society it is the service (administration, building, modifying, development). GPL is not anti-commercial, it is just different commercialism, with a different merchandise.
* “GPL is anti-competitive”: This could not be further from the truth. When Linus created Linux (or Freax as it first was called) it was from the standpoint of a young engineer with a limited budget operating in a software business dominated by tightly controlled monopolies, with a high barrier to entry for aspiring developers. The revolution Linus talks about is not about tearing down the profitability of software, it was about making it more accessible to developers and tearing down the established monopoly. As Gustavo mentioned, competition is very notable within the Linux community (between distros, between apps, between employees looking for well paid Linux jobs). GPL also creates competition within the traditional retail markets by changing the game rules and leveling the playing field. At my office all our servers are running Linux and pretty much all our operating software is open source. I, aside from developing apps, administer Linux based systems DAILY and I get paid for it. Open source gives us an competitive edge in a business that would otherwise be locked up by restrictions (both financial and by anti-competitive licenses). Linux and GPL helps expand the market, not the opposite. A company can freely choose commercial alternatives if they wish and feel that they are superior, GPL does not interfere with that. GPL clearly states that material produced by GPL software are not GPL licensed, that GPL programs can be run and installed side-by-side with other proprietary software. The restriction lies in mixing GPL and proprietary code and that free software remains free, which I don’t see the problem with. It is not like it is ok to mix and match different proprietary code bases with each other without infringing on licenses.
* “Money can not be made selling GPL software”: While I agree that selling codes or binaries is not a viable market it is an interesting thing to compare how GPL is sold as opposed to retail software. Few people can make a living developing Linux kernel code, but the same could be said about developing Windows kernel code. The main difference is distribution. With proprietary software being a reseller is a profitable job. Distribution, and more importantly in this case, distribution rights grants the reseller part of the profit. GPL shifts this balance and undermines the value and need of resellers. This is not anti-competitive nor is is socialism. It is however a clear cut utilization of the most basic of all rules of commercialism: supply and demand. Since, in my personal opinion, software is not a physical product (it is not the bytes or the disc you are selling, it is the cost of development) it makes no sense to treat it like one. And in a market where bytes are not the product it makes perfectly sense not to have resellers of a product that doesn’t exist, and this does not have anything to do with licensing philosophy. But by making software more open and accessible this “flaw” in todays notion about software becomes embarrassingly obvious. It is why DRM, as an example, does not work. It is a virtual enforcement of rules applied on non-physical product. It is not that non-physical products are not profitable, but if it operates under a virtually constructed business model it will always be haunted by the flaws of that model. A computer is designed to duplicate data (amongst other things) and a great computer can duplicate a great amount of data, why do we then stubbornly try to treat each copy as a separate product when the act of producing a copy costs close to nothing? It makes no sense. And until we start to make sense of it we will be leaking money into piracy and other illicit uses. Money lost that causes developers to loose their jobs, otherwise profitable companies to file for bankruptcy and sends a confusing message to users that doesn’t even makes sense inside the business. Just like an musician is an artist and not a CD-maker a developer is an engineer and not a byte-reseller.
We (most of us) live in a free and open economy and GPL is just one way of competing, and if the market is not able to adapt to those circumstances without restrictive and prohibitive laws that solely exists to uphold a virtual reseller model instead of letting the market find out how to best profit on these new game changing rules. If you want something that sounds a lot like socialism, a government regulated economy sounds very much like that. Economy is not a zero-sum game and as long as there are are users of software there will be a demand for developers, administrators, project managers, security consultants, service providets and other related positions in order to serve that market. GPL does not remove that demand, but it does alter the structure of the food chain. Well tough luck. This industry was created out the need for machines to assist us in our daily life and that is still the mission statement. A industry in constant change will have effects on the needs and means of fulfilling the needs of users. Just like the production industry changed the roles of factory workers (leaving hands-on production to machines and merely letting humans monitor the factory), the same thing will and is happening to the software industry. If you can freely download the software from the developers site, and if you can yourself easily redistribute it to others, how can you then justify a reseller market that only creates a need by actively restraining availability?
Do you think Microsoft, Adobe, Sun and others open source parts of their product out of good will? No they do it to survive. OSS have changed the rules of competition and they are forced to adapt their business to survive. Just like IBM had to when their model was challenged by Microsoft.
GPL did not destroy the business, it created a new one, and the survival of traditional businesses will depend on their ability to adapt, just like in evolution. They have to justify their place, compete with the newly founded rules or they will perish. That is how a free market operates and it is within that set of rules that the GPL exists. If you truly advocate a free market economy you could not possibly have any quarrels with GPL.
@Gustavo in #7: you are welcome re my reply. Thanks for the thoughtful blog and the ability to comment on what is said here.
Technical competition is useful, but, in and of itself, it doesn’t pay the monthly mortgage. That is one of the main issues here: THERE IS NO FREE LUNCH. If I get a free lunch, it is because somebody who wants to make a sales pitch to me is paying for it. So I pay for the lunch by listening their spiel and either agreeing to buy whatever it is they are selling, or not. As we progress further and further into the world of environmentally-conscious economics, we are going to find out that even taking a breath is not free because that air had to cleaned by a set of manufacturers who dirtied it while making whatever it is they make.
Linux is an intellectual commodity with strings attached to it. Those strings (the GPL) dictate how it must be dealt with. In this case, all source code must be given to whomever wants it enough to ask for a copy of this code. Any improved versions must follow these same rules. There is no choice about these decisions unless you want to be involved in a lawsuit with the GNU lawyers.
So we can look at these issues from two directions: 1) how does GNU and its work pay for itself?, or 2) how do proprietary software companies work and pay for themselves?
Of the two systems, the proprietary model is simpler and easier to understand. This is not a trivial distinction for some of the following reasons. Society needs to deal with all of its members, who exist on number of continuums, from the smartest to the dumbest, from the wealthiest to the poorest, and from those who most embrace the dominant modes in that society (those most involved with making things happen, I.E.,the movers and the shakers) to those who don’t embrace them at all (the outlaws, the fringe folks, the anarachists and the antinomians).
Software, by the nature of what it is, needs to be created by some of the smartest members of a society. Computers are extremely complex TOOLS which need careful care and feeding in order to continue to operate, and in order to continue expanding on and improving what it is they are asked to do. So these smartest people have something about them that 98% of ditch-diggers don’t have: brain-power. And, not only that, but a particular kind of brain-power: one that has an interest technical matters, and skillful prowess in careful manipulation of symbols. The coders may have this, but the ditch-digger may have other abilities, such as the ability to work 8 hours a day at an exhausting physical task without collapsing.
Under a more socialistic system, each of these people, the coder and the ditchdigger are considered to be worthwhile and useful. Neither is accorded any special status as a result of their unique abilities (supposedly). The phrase “From each according to his ability, to each according to his need” was used by Karl Marx to help explain how a socialistic system needed to be viewed. He saw a transition in what he thought was society’s inexorable march towards Communism, whereby the distinction between mental and physical work (labor) would be erased.
Under a more capitalistic system, values are assigned to each of these workers and what it is they do. This happens automatically, by virtue of the competitive marketplace, except when “artificial” processes are injected into the free market system. Labor unions and their demands for higher wages are an example of such an injected “artificial” process. Corporate welfare by a government, such as the recent set of actions by the U.S. Federal government are another example of the abortion of free market principles by virtue of having the taxes of ordinary citizens be “redistributed” to wealthy corporations because these companies failed to operate effectively in the marketplace.
Let us ignore these “artificial processes” for a moment (I understand that this is difficult to do) in order to proceed with the main argument. Under a more capitalistic system, the ditch digger doesn’t get paid as well as the software coder because his or her set of skills doesn’t make as much money for the company they are employed by. If they are employed by a government, the government’s table of wages generally reflects the prevailing market conditions. A ditch digger is not as useful to a government as a software coder is, so he gets a smaller wage. One reason is that ditch digging is a more common skill set, with less unique qualities about it. It is more common than the ability to code software. Another reason a ditch digger is paid less is because the ditch digger’s tools are simpler and require less money to buy and maintain.
But these reasons for wage disparities exist up and down the job continuum in a capitalistic system. @Burre in comment #8 speaks about the F/OSS (Free and/or Open Source Software) movement as tending to cut out the middleman in the process of creating and selling software. It only appears to be that way. In fact, many of these middlemen are simply not paid at all for what they do. An example would a private citizen who helps with Linux kernel coding entirely for his or her own satisfaction. This is an admittedly small, rare group of people, but it makes the point.
This relates to the beginning of my argument in this way: not everyone in society is smart enough to add to the Linux kernel coding, or even to write worthwhile code in a high-level programming language such as Visual Basic. But some of these people still are interested in computers and coding. Some of these people become middlemen. How? By becoming so-called “value-added” resellers. Many of these people become Microsoft Partners. Microsoft extends them certain abilities and advantages that make installing Microsoft software a “value-added” proposition for them. Their profits come from the fact that they (supposedly) have added value to what it is they are reselling. A small white-box computer OEM company which sells MS software with their boxes is a good example of these middlemen.
By this mechanism, less technically competent computer people are able to make a living via computers and the software they run. These middlemen don’t have the smarts, or the thinking capacity to create the code itself, but they are smart enough to, perhaps, add in a few pieces of code to knit together several pieces of pre-made software, or to add in another piece of software to the mix which already works with the existing code. In effect, this “spreads the wealth” of the coders’ and computer manufacturers’ expertise to those further downstream in society. In this case, the wealth referred to is a combination of intellectual wealth (being very smart) and infrastructure wealth (having a factory which can make computers or their parts).
A more socialistic system upends this set of social continuums by declaring that everyone is working equally for everyone else within society. In other words, the software coder is no more special or unique than the ditch digger.
Richard Stallman decided he wanted to inject his philosophy about free software and interpersonal cooperation into one of the most capitalistic (until recently) countrie in the world, America. So he attempted to encode his philosophy into a set of legal documents, each one getting more dictatorial and convoluted than the next one. He is attempting to create a mixed system, one slung between a more socialistic set of ideas and a more capitalistic set of ideas. There is nothing wrong with doing this, or with creating mixed ideological systems, except that in practice, odd, uncomfortable social compromise often have to be made as a result of doing so.
So what? So this: who decides how to parse a particular set of the above commpromises? In China, it is not too difficult, at least at the national level. Their Communist Party leaders decide what the overarching rules are about business and property ownership. But America already has a set of rules about these matters. So Stallman and his acolytes have to inject their set of requirements into every nook and cranny of the existing system in order to convey their principles. In practice, this means fighting things out in courts of law. But it only means this if someone decides to play on GNU’s turf. People such as Bill Gates and Steve Jobs, until relatively recently, decided that they didn’t want to play on the F/OSS/GNU turf. (I know that GNU does not equal F/OSS, as Stallman takes great pains to explain.)
At the heart of these matter are the issues surrounding intellectual property rights. GNU says that in matters of GNU software coding, anyone has the right and the ability to see and use the intellectual property created. In this case, that is a computer program, or a piece of a program, or a segment of code. The coder needs to be given original credit for it, but that is all the coder gets is bragging rights, nothing else. After that, the code is out of their hands as far its use goes.
Now look at the process of creating a fiction novel. It is a series of symbols put into a particular order by whomever created it, usually a single author, or perhaps, several authors plus their editors. Yet some novels are hugely popular and others barely scratch the surface of peoples’ imagination. Removing the issues of aggressive marketing for a moment, we can ask why? How come some novels are popular and others are not? This gets into a whole other set of considerations, such as the public’s taste, the nature of the novel’s plot, its applicability to the human condition, etc, etc.
But all of the same things are true about a software program as well. It is a unique set of symbols put into a particular order (here the order is even more critical, or even essential) than those in the process of creating a novel. In the end, both are particular sets of symbols in a particular order which the wider public either finds useful, or not. Yet very people in America question whether the proceeds from writing a successful novel should flow, in part, to the original author of that novel. Here, intellectual property law is well-established and agreed upon by most of the members of American society.
So why treat one set of ordered symbols one way, and another set of ordered symbols another way? Because Stallman says we should, and more socialistic systems of society say that we should do so. That is the main reason. All sorts of additional reasons have been cited, such as the usefulness of doing so, the power of widespread collaboration, etc. And some of these reasons are valid. But they and the systems that they engender are not without their problems.
The major problem is that there is no free lunch. The costs of designing and manufacturing computers and their software has to be born by some set of people somewhere. So it comes down to deciding who has to pay for these things. More socialistic systems say that everyone should help pay for them and everyone should be able to benefit from them. More capitalistic systems say that those who put the processes in motion in the first place (I.E., the authors, or authorities) should be the main and original ones to benefit, then those downstream from the original authors can benefit in a manner that suits them and they see fit to employ, within the laws of the country. So, if I don’t have the ability to write a best-selling novel, perhaps I have the wherewithal to build a bookstore which sells such books. I am downstream from the original act of creativity, but I still benefit from it, but not to the measure or degree that the author does.
These are different systems with different sets of beliefs about how society should conduct its business. They are naturally in conflict because of certain ideas behind them. These ideas are opposing ones. Conflict is inevitable in such situations and must be resolved. It is usually resolved by the authors, or the authorities within the society.
@Burre in #8: I think you are in error on this point in your 3rd paragraph: “Since, in my personal opinion, software is not a physical product (it is not the bytes or the disc you are selling, it is the cost of development) it makes no sense to treat it like one. And in a market where bytes are not the product it makes perfectly sense not to have resellers of a product that doesn’t exist, and this does not have anything to do with licensing philosophy. But by making software more open and accessible this “flaw” in todays notion about software becomes embarrassingly obvious. It is why DRM, as an example, does not work. It is a virtual enforcement of rules applied on non-physical product.”
It is the reification (making concrete) of intellectual property as a tangible commodity that is the hallmark of all of the laws surrounding intellectual property. Remove this cornerstone, and it all tumbles down. You seem to want this collapse. I don’t. If I write a best-selling novel, I want to be one of the primary beneficiaries of having done so. It is my property for some given length of time under copyright law. Any infringement on my property is dealt with in the same way as having someone coming onto to my physical property and stealing lawn furniture or anything else that is on it. These people, if caught, are prosecuted under existing laws. It is no different for software, movies, or other examples of intellectual property, UNLESS the author, or authors, decide otherwise.
@Rob: I think you have misunderstood GPL and my thoughts on it.
* GPL is not socialism, GPL is not communism and you can bet that GPL users DO care about the level of your skill and you can be sure that if you don’t cut it as a GPL developer you patches will not make it into the main branch. Unless you distribute them yourself of course, which you are free to do, but no one will care about your distribution of the code if it does not live up to expectations. It is a self-sanitizing system, just like in the current corporate world. It is not about equality in any other aspect than the freedom in distribution and modification. You talk about GPL as if it was a license made to limit developers. It forces you to contribute back to a community that have contributed to your application. Seems fair to me. If you want to write proprietary software you are free to do so, but then you are not allowed to benefit from free software (other than LGPL and such). In fact even if you write all your code proprietary, just the fact that other applications are open and transparent can help you with interconnectivity issues with their product. In many cases it is a more fair system as your worth is SOLELY judged upon what you are able to contribute and what services you are able to offer.
* GPL gives better access to code. This allows for helping many developers to improve their skills and by sharing information the computer industry as a whole grows faster. If you are worried about a uneducated mass locking up code will not help the situation. Today any teenager has a better access to computers, code, information etc than either Woz or Gates had in their day. Google, Yahoo and others are a fruit of a healthy software industry based in easily available software and code. I doubt any of them advocate the destruction of the free market and believes that open code is worth nothing. Price and value is not the same thing and not selling the code does not mean that you can not sell services based on it. GPL has the possibility to make “ditch-diggers” developers, if they wants to. And even if resellers will disappear their will instead be an increased need for other personnel to administer the services. Canonical for example have a lot of employees that isn’t developers/sysadmins.
* It is my belief that software as a market is slowly but steadily moving towards a service based model whether you or I agree with that development. It has to do with a change in the computer industry, innovative ways to profit from users and of course a growing interest in OSS. I believe that one contribution to this development is because of the problem with regarding software as a physical entity. Those who does does (and will) fight an endless battle with their user base over piracy and illicit usage while those moving to a service based model will only reap more and more benefits the more users they have. That is why, in the future I predict that only one of those models will have a real future. It will not kill the IT market, it will not remove the need for developers but it will revolutionize the industry. We can either hold on to old models and fight it tooth and nail, or we can prepare for it and benefit from it.
* I do not advocate a destruction of the free market economy. The software industry is a young and newly created industry
* GPL does not enforce a specific method of payment (or invalids commerce in any way). It is a philosophy on software distribution and in the way it is formed it invalidates some forms of commerce and open new ones. Treating binaries or source code in a GPL organization as a physical product ready for reselling does not make sense since distribution/reproduction of the software is both allowed and encouraged. Instead it focuses on making profit on development/maintenance etc. Switching from a retail corporation to a service based one. Many companies have testified that going GPL have opened up for more opportunities and given them the ability to expand and employ more developers.
* Openness increases adaptation, adaptation increases use, increased use causes increased demand for development and that puts developers in a very healthy position when trying to profit on their labor. If you want to contribute for free that is your personal choice, no one is forcing you and if you are skilled in your field your contribution is sure to be worth paying for. No free lunches.
@Burre in comment #10: I don’t think I have misunderstood your positions on software and the GPL model. I simply don’t agree with several of your major points:
a) software IS a physical entity. That is the only way it gets transferred from one person to another. This IS a significant point. Once software becomes something other than an abstraction, it needs a physical form in order to be useful.
b) you admit that the GPL DOES contribute to a decline in certain forms of software commerce, such as middle-man or value-added reseller business models. This means that it will never become a dominant mode of software distribution in the United States. Yes, other models of economic growth can and do develop, but I don’t see this happening to any appreciable degree in the U.S. There is no money in value-added resellers selling computers which run Linux. Most folks have no idea what software their computers run, and they don’t care either. They want the computer to do other work for them, such as accounting, or publishing, etc. These people, for the most part, are not asking for Linux by name. They ARE asking for Apple computers by name, or Dell computers or Hewlett-Packard computers.
c) software IS a patentable piece of intellectual property (IP) and will remain so in the future for most commercial purposes in the U.S. Software also allows for the ability to protect other patentable pieces of IP such as movies, TV shows, books in digital form, etc. I am speaking of digital rights management (DRM) here. DRM isn’t going away. Hollywood movie producers are one group who will not cave in on this issue the way some in the music industry have. Movies and music have different business models and different cost structures. The only people who want DRM to go away are the geeks who know how to circumvent it, but find doing so a pain-in-the-ass (PITA), and those who believe in a more socialist style of IP such as Richard Stallman, and to a lesser extent, Linus Torvalds.
d) The GPL-style, or open-source style of software development (GPL does NOT equal open-source, and is more restrictive than open-source licensing) is NOT going to gradually supplant the IP/DRM model in the future. Both will exist in varying degrees and amounts, depending on the country involved and its political economy. In America, the IP/DRM software mode will remain dominant into the foreseeable future.
e) As I said in my post #6: follow the (software) money trail. When you do so in America, you will find that Linux is barely a major player on its own. When it is employed by a large corporate IT firm such as IBM, Linux is useful and important. But left to its own devices, it barely registers. Red Hat is one of the most well-known American names in the Linux-software market. In 2007, Red Hat made approximately $80 million dollars profit. $80 million dollars is a lot of money for a private individual, but it is barely an accounting error for a company the size of Microsoft.
Linux has a lot of traction in the enterprise world as a useful set of software tools, but in these cases it is simply one set of tools amongst many. As a profitable entity in and of itself, it is not. It is my opinion that this is baked into the fabric of Linux by its creator, Linus Torvalds. As the twig is bent, so grows the tree.
Thank you for the response Rob.
a): I fully agree that software is regarded as physical entity today in a legal sense, I don’t dispute that. Although it is often disputed in court cases and there is often a debate regarding this when measuring damages and liability (i.e is it copyright infringement/theft/plagiarism/etc…). It is also sometimes a troublesome when separating services you provide with actual binaries in regards to warranties etc (since for many service providers the software is mainly a means to an end and not a product in itself). This is mostly due to poorly updates IP laws that in fact are based in “real” physical products. What I mainly was highlighting is that if software is moving towards OSS/GPL treating it like physical property will only cause more trouble than it solves and therefore I predict a change in this aspect in the future. Even retail businesses treat it more like licenses than physical property where the license grants you usage and the actual binary isn’t what you are paying for. For example I own licenses in both Adobe Flex 3 and Pixologic ZBrush 3. Both are bought as digital downloads and I can redownload them as many times I like and in some cases even move them between machines/OS:es (since they sell the license to use and not the binary).
b): I’m not sure I do admit to a decline, I certainly admit a huge change in the need of middle-man services (moving away from resellers towards administrators as, again just in my prediction, software will move towards services and away from retail). This is just a natural transition due to changed distribution and consumer interaction.
c): Software patents are only valid in the US afaik. Here in EU and I believe in other parts of the world they are not. So USA is the exception and not the rule in this case. I also do not agree that only geeks want to get rid of DRM. Low-end consumers are probably more annoyed with that than geeks. And if you talk about pirates I have still to see a DRM that have stopped piracy. The reason why FSF wants to get rid of DRM is simple, it is in direct conflict with open standards.
d): Well which model that prevails is to be seen. I strongly believe that OSS (whether it is GPL or something else), will slowly gain traction over the years and replace retail in time, unless it is stopped by anti competitive legislation. Remember that I talk about long term transition to openness, not in the next year or so. But we will see. As long as it is a model dictated by consumer demand and not by a virtual demand orchestrated by the business I don’t really care where it ends up. I’m not an evangelist, but I do believe that it is the superior choice for most applications and openness opens up more business opportunities than it prevents.
e): I often feel that the amount of money generated by OSS is hard to measure, and more so to compare to retail. My TV runs Linux (a typical Philips 42″ LCD) and I can get their source from their website if I want. Is that product considered when measuring money in the OSS industry. If Godaddy, Amazon, Googles back end is run by Linux how do that fit into the calculations? I often feel that those statistics focus on desktop PC or retail software when an increasing part of the consumer electronics market, mobile phones etc is moved over to Linux. I don’t doubt that it is a smaller part of the software industry, but I DO believe that it is growing fast and that it will equal or surpass proprietary software in time.
I am primarily speaking about the U.S. desktop consumer market in much of what I say.
For you to say that software patents aren’t valid outside of the U.S., this is not so. Windows is sold all over the world, and those who use it agree to abide by its terms of service when they first use a computer with it installed.
As for the true amount of revenue an operating system generates for the likes of GoDaddy or Google due to their backend operations running Linux, this is an apples versus oranges argument. These companies aren’t selling operating systems. GoDaddy is selling Web server space for Websites and Google is selling everything under the Sun except operating systems (so far) due to their search with relevant line-ads model of doing business. You may as well ask how much revenue a company running Windows servers generates. This approach isn’t relevant to what I was speaking about. Linux is one software development tool among many for a company like Google. But it isn’t what they are selling.
At the risk of repeating myself, Linux is a name and Windows is a brand. There is an enormous difference between these two facts. No single entity is representing Linux, other than the kernel development aspect of it. Torvalds purposedly set up Linux this way. Linux is not a corporation It is a set of kernel codes that can be used freely IF the GPL model is followed.
Windows is a brand of operating system by one of the largest companies in the world. If you steal a copy of Windows, you risk bringing the wrath of Microsoft upon you. Windows has a single entity looking after its welfare. Therein lies all the difference in the world.
As for Digital Rights Management, you are speaking of poorly-designed DRM software. If it is designed well, it should not impede with the legal use of the intellectual property IP) it is safeguarding. I agree that there are some issues in the “fair use” clause in American IP laws that current DRM software and hardware sometimes violate. These will have to be worked out on a case-by-case basis, the same way every other law in America is.
You spoke of America being the exception re software patent applicability. Well, yes, but America is an exceptional country in the world, in both the postive and negative senses of the word. Ignore us at your peril. And when in Rome, do as the Romans do. You may or may not be aware that during the current global financial meltdown, which is continuing apace, the flight of currency traders into dollars and Japanese yen has been striking. Not even the British pound was immune from frightened traders exchanging it for dollars. I am not simply boasting here. I am saying that the majority of the rest of the world, when the chips are truly down, still look to America and Japan as countries which have their shit together and are viewed as safe and reliable. Confidence is everything at a time like this.
The GPL has almost no future anywhere, IMO. It is simply too restrictive and is getting more and more so with each new iteration of it. The Free and/or Open-Source Software (F/OSS) movement, which is broader and more inclusive than the GPL, does have some traction and will continue to play well in undeveloped and underdeveloped countries. But as a country’s citizenry get more affluent and worldly, they will soon realize that Windows is another “language” that they need and want to know. At this point, they either pirate their Windows copy, or they abide by U.S. laws regarding its use.
I heard an anecdotal story that in India, many middle-class users buy computers with Linux pre-installed on it. One of the first things they do is download, or buy a pirated copy of Windows at street vendors, then uninstall the Linux OS, and run Windows on their box. I don’t know how true this story is, but I don’t find it difficult to believe. Computing is not a solitary act in its widest sense. It is a social act. If I have a piece of photo software on Windows that I rave about to my friend, he may want to run it as well. But if it only runs on Windows and he is in Linux, he won’t be able to do so. It is for these reasons that a certain number of people who try out Linux get rid of it after a few disappointments. It is also more difficult and complicated to repair than Windows IF the OS installation gets corrupted.
The Windows OSes, although they are waning in importance and relevance in the broad, longitudinal picture of software development, will stay dominant and central for at least 2 or 3 more decades wherever they are deployed. And if some smart, open-minded Gen X and Gen Y folks take over the helm of Microsoft after the Gates/Ballmer era is truly extinguished, Windows and the Office franchise could be streamlined, refreshed and re-invigorated for an entirely new generation of mostly American users. In the U.S., Linux as a desktop OS, is a nonstarter. And Linux as a stand-alone operating system has no hope of making any sigficant money here for anybody hoping to sell it as software.
@Rob
“Of the two, Torvalds has had more a more significant impact on the world at large due to his superior technical abilities in writing kernel code. His ability to get along with others has also been important. Stallman has neither of these abilities to the that extent that Torvalds does.”
I think you’re selling RMS short here. Stallman’s technical feats include emacs, GCC, and many of the linux userland apps (through the GNU project). Guess what compiler Linux uses? Guess what userland?
well, hi admin adn people nice forum indeed. how’s life? hope it’s introduce branch