If, however, I want to include GPLed code in
my program, the GPL forces me to release my program under the GPL. It has *infected* my program.
Wrong. You have infected your program by including the GPL covered code. No one held a gun to your head and demanded that you include it, nor did the code sneak in through an open window and jump into your source repository. That was your choice. Calling it viral is ungracious at best.
[...] under non-free licenses like the GPL [...]
That you prefer the BSD terms to those of the GPL does not make the latter non-free in any meaningful sense and your claim to that effect is nothing more than religious zealotry. Learn to accept that not everyone will genuflect at your church and get over yourself.
Java is far more important in its security [...] features.
Security features? Most of the "security" features in Java are aimed at permitting more than one trust boundary in a single address space. This is a poorly understood problem that Java doesn't actually solve. But even if it did, a stack security system is much more complex and difficult to configure than Unix setuid(2) and therefore even less likey to be used properly in practice. Furthermore, putting different programs in separate address spaces has advantages for robustness.
As for the rest (mainly array bounds checking and garbage collection), equivalent or better alternatives are available in C. Better still, they can be used only where they're needed. Check out Apache resource pools and the glib object system for some examples. Obviously C++ has even more of these toys. And if we don't restrict our choices to these three there are hundreds of interpreted languages with equivalent or better features. Lisp comes to mind...
I find that I'm far more productive in Java than I ever was in C++ and C. I find that's even truer for less experienced programmers.
Exactly the problem. A rigid system like Java is indeed wonderful if the goal is to allow mediocre programmers to create mediocre applications. Cobol had the same sort of appeal in its day. There's nothing wrong with that, but those who aim for excellence don't need to be pampered. All those fluffy features just get in the way.
Free software depends on a few companies' ability to actually make money developing and using free software.
No, it doesn't. Software development does cost money, money which must come from somewhere, and profitable companies are one place money can come from, but not the only one. As a general rule, the greater the extent to which companies can make money through developing and using free software the more they will contribute to it. All of this we agree upon. However, this doesn't mean that every contribution should be accepted without thought for the consequences. Companies such as RedHat that manage to contribute without resorting to proprietary software are laudable, but those like Lindows are justifiably regarded with more skepticism.
Free software in general and Linux in particular have been around far longer than BitKeeper, and can thrive without it. While it's true that BitMover is entitled to choose the license they prefer for their product, their "contribution" to Linux is a dubious one for two reasons. First, whatever its technical merits, the philosophy it comes with is fundamentally defeatist. There are some problems, we are told, that free software just cannot address. Actually, there is no reason that someone with the same skills and motivation as Larry McVoy but with scruples couldn't do exactly what he's doing technically using a free software compatible business model. Second, the existance of high quality proprietary tools removes pressure to create good free ones, and is therefore a step backward for free software. Every kernel hacker who uses BitKeeper right now is someone who would otherwise have an incentive to contribute to Subversion otherwise, even if only in terms of testing, feedback and endorsement. A similar case can be made for free software friendly business models, which McVoy refuses even to discuss.
The argument that GPL code is 'better' is not persuasive if it doesn't fit in their business model. Not everyone has academic salaries or other means of support, so these are real concerns of people who support the idea of Free Source in a deep way.
Proprietary software companies are not the only employers in the world -- in fact they employ less than 10% of software developers! The rest are working for companies that do something else but need specific software for some internal purpose. The right way to address the concerns of wages for free software developers is to create a robust free software industry, but even in the meantime there is no need to make the free software movement to pander to the interests and business models of proprietary software companies by watering down licenses.
The no incentive to create a free alternative is bogus. If you think that, you are too concerned with the short term.
Please pay more attention. I didn't say there would be no incentive, but that there would be less. And there would indeed be less, because while there are people (including myself) to whom freedom is important, there are plenty who don't want to think about it. The lower the quality of proprietary tools, the more willing those people will be to assist in the development of free alternatives. I agree that the free software movement is likely to prevail, but I would rather see that happen sooner than later.
It's a take-back because the LGPL came into being because in many cases strict GPL made it difficult or at least questionable to even use GPL tools in the production of software for sale. Then much later RMS decides that too many people are using LGPL and he doesn't like it. That's a take back as far as I'm concerned.
The term "take-back" clearly implies that you were given something by someone who had a change of heart, and that as a result you are deprived of it. In the case of RMS frowning on the LGPL, we have the first and second, but not the third. No one has revoked the LGPL, nor have any projects I've heard of changed from LGPL to GPL licensing terms, so the term "take-back" can't possibly apply. Perhaps you can think of a project where the license was so changed, in which case I'll agree that (though in a somewhat crabbed sense, because you can still use the software versions prior to the license change) a take-back has occured.
I'll address the rest of your post in a separate reply.
From where I stand, RMS's stance on the LGPL is a take-back that is just as damaging, if not more so, as the EULA change being discussed. LGPL gives you a lot more choice in terms of integrating free and proprietary subsystems and components. Where free libraries have significantly extended functionality, he explicitly recomends GPL over LGPL.
Exactly how is this a "take-back" or in any way comparable to the BitMover EULA changes? The closest comparison would be a copyright holder re-releasing a particular library under the GPL rather than the LGPL after a certain version (if indeed that has ever happened), but even that is much less malignant because there is no direct attempt to curtail competition. The GPL and LGPL are concerned with the license of derivative works and are otherwise content neutral.
This is the one case where I would claim that it goes beyond style, and the message itself actually hurts the movement.
How so? It clearly hurts your employer, but the production of a clever proprietary tool cannot -- by definition -- help the free software movement. The higher the quality of such a tool, the less incentive there is for people to assist in the creation of a free alternative. Furthermore, you now have an excellent argument to make to your employer: "if we choose to license our tools under the GPL we can use this body of well-tested and freely available code to get a head start."
1) Almost every known root CA targets businesses as their primary customers.
So? People who run businesses are entitled to target any subset of potential customers they choose. Usually this means the people most willing to spend money will get the most attention. Nothing obligates a company to be generous toward those providing free services. I agree that this is an unfortunate situation, but it's not the fault of the certificate vendors.
The internet community should establish a trustworthy non-profit body to administer certificates that charges just enough to cover administrative costs. Until that happens we're stuck with a choice between self-signed certificates, self-certified certificates, or profit-oriented services.
2) 'Wildcard' certificates cost an absurd amount of money, usually $500 or more. Excuse me? The entire premise of the certification, is that Thawte (or VeriSign, or whoever) is certifying my trustworthiness as an organization.
Excuse me, but that is completely wrong. An end-entity certificate certifies that you are who you say you are, not that you are trustworthy.
Clearly a wildcard certificate is no more expensive to produce than a more specific one, but the fact remains that this is a market economy and there are reasonable alternatives. There is nothing fraudulent happening here.
3) VeriSign, the biggest fish in the pond, has demonstrated on more than one occasion that it is in fact not trustworthy.
True indeed, but again not a scam. Software is complex and security even more so. Being trustworthy is difficult, and while I see nothing praiseworthy about VeriSign, they should not be vilified for trying and failing. (There are plenty of unrelated things for which they truly deserve blame, but that's another story.)
From the start, the entire digital certificate business has been about politics and moneymaking, nothing more.
Hello? Politics and moneymaking are a legitimate part of society. We get nowhere by turning up our nose at these things. Accept them and get busy making things better.
You may have given somebody permission as far as your browser goes but that doesn't give you the right to change a link on a persons website...
While I share your sentiment I don't think this argument will wash. Installing the spyware on your computer doesn't change an actual website, just your view of it. A ruling that this violated the rights of the website owner would imply that many legitimate and useful things (such as Googles language translation service) would also require permission, which would make them impractical.
None of this reasoning makes what is being done by Kazaa and friends right, but it would be better to seek a remedy on the grounds that a clickwrap EULA is not enforcable, that the practices are deceptive and harmful the user or that this is abuse of the Amazon affiliate program terms.
Ummmm... GNU does NOT comprise 80% of a Linux distro. I refer you back to this article, Section 3. Adding up the 35 projects listed, GNU provides 26%.
Of course it doesn't; I didn't say that it did and that wasn't the point of the analogy. The goal of the GNU project has always been to create a complete free system, not to reinvent everything from scratch. Wherever a suitable piece of free software existed it was adopted and integrated. This is a poor way to win a contest over who writes the most lines of code (one has to spend time learning how someone else's code works and testing things that doesn't get counted), but an excellent way to get a complete free system put together.
On the other hand, the GNU project probably had more than 80% of the system put together (as opposed to written from scratch) by the time Linux came along. Clearly the Free Software Foundation has done more of the work in creating a complete free operating system than Linus Torvalds or Red Hat, and most of it at a time when it wasn't a glamorous thing to be doing.
What I'm getting at is that the people who drop in the last piece are not necessarily the ones who deserve most of the credit, and I think that presenting a simplified hypothetical situation to make the point is valid.
But it's obvious to me that I'm not going to convice you, and so far nothing that you've said is any more convincing than anything I've heard or read before. Would you agree that we disagree?
I will indeed, but under protest. I feel that you have given the arguments I've made only the most casual scrutiny, refuting points I wasn't making and ignoring the rest. Under these conditions I don't think anyone could convince you of anything.
If they're not antithetical to each other, then why would you feel the need to say this:
"Do you really care so little about your freedom that you can't be bothered to prepend two little syllables?"
I'm not sure why you find this difficult to understand, but I'll try again. The concepts that "Linux" and "GNU" are associated with are not antithetical. This means that using them together is not a contradiction in terms, but it does not follow from this that they mean the same thing. It also doesn't follow that because they are not in direct opposition using "Linux" alone gives a sufficient emphasis to the freedom the "GNU" stands for. Get it?
"Who should get to make the choice of what the operating system is called?" The people who actually put it together.
That is an awefully aribitrary standard. Obviously, the people who put things together can get away with calling it whatever they want. That doesn't make it right. For example, I can repackage a GNU/Linux system and call it a "NewBSD" if I want to, but I doubt you would find that defensible. Why not? Because it is misleading. This situation is essentially the same.
I'm glad that they were able to produce those independant projects in such a way as they would fit together. But they did not put them together. They couldn't. They didn't have a kernel.
Acually, these are not truly independent projects. While they are designed to be modular enough to be useful separately (this makes them easier to test and useful for people who are forced to use proprietary systems for some purposes), they were designed as part of a complete free system. Just because they were easy to take apart doesn't mean they were never put together. This is true even though the result was not yet a complete free system.
While it's true that the GNU project took too long to finish their kernel, and it is arguable that they ought to have abandoned it in favor of Linux when it appeared, these tactical considerations aren't really relevant. Would you be pleased if, after completing 80% of a large free software project and getting bogged down on the remaining 20%, I added the missing pieces and released the product under a different name? What if people began to credit me with organizing the entire project?
Clearly in a free software community there is nothing you can do to forbid this, but I think you would find yourself doing a little "coattailing" of your own.
If the terms GNU and Linux are so antithetical to each other[...]
They are not antithetical, and no one said they were. Narrow technical advantage and a just-for-fun mentality are things that have their place. But the question here is whether freedom has a place too. Some of us think it does, and that it should come first when talking about what differentiates free and proprietary systems.
That's like saying that I can turn around what murder represents for by calling it GNU/murder.
No, actually, it's not. Get a grip.
[...]Linux uses the GPL and *insists* on guaranteeing the same freedoms as every FSF project. How is it that Linux represents anything other than the FSF's definition of freedom?
No one is complaining about the license used for Linux (except perhaps BSD license advocates and corporations who would like to take it proprietary). The problem is that people are using the term "Linux" -- which is the only legitimate thing to call the kernel as it is what Linus Torvalds chose -- to refer to an entire operating system built from GNU components around the GNU vision of a complete free system. Doing this implicitly gives Linus Torvalds credit for an entire operating system he didn't create.
Who should get to make the choice of what the operating system is called? Consider, as an analogy, who should get to choose what the kernel we call "Linux" is called. Obviously that would be Linus Torvalds because even if many different people and organizations contributed, he put the process in motion and guided the design. Similarly, the FSF conceived and organized the GNU system, so they have a strong claim when it comes to naming it.
By what measure, other than "we were here first", can GNU make the claim that they're the principle developer?
How about the measure of system design and organization? Not only was GNU there first, they conceived the idea of developing a complete free system, which is the essential point of a modern "Linux" distribution. As a result, they did all the boring development work that a system requires, rather than choosing the thing that was most fun to work on. They are also responsible for most of the non-technical infrastructure that makes free software possible, most notably the GPL, but also more mundane acts like encouraging developers to contribute.
Things like lines of code and memory residence are easy to measure, but that doesn't make them good indications of who concieved a system. Kernighan and Ritchie wrote exactly none of the lines of code in modern Unix derivatives, but that doesn't mean we should forget about their contributions.
But the real point here is not that we should be splitting hairs to allocate credit. Instead, we should be thinking about the messages we want to send. The terms "Linux" and "Open Source" stand for a focus on narrow technicial advantage and a just-for-fun mentaility. "GNU" and "Free Software" stand for preserving meaningful liberty in a digital era. Do you really care so little about your freedom that you can't be bothered to prepend two little syllables?
The difference here between the magazine or television and the web is that the guy who runs the site gets money when people click/lead or whatever the pricing plan may be. If you cut ads out of a magazine, the magazine doesn't care.
This argument is complete nonsense. People and organizations pay for an advertisement because they believe that it will recieve attention. Magazine executives most certainly do care about what happens to their ads because they want to get paid for the next issue too.
What makes internet advertisements different is that eliminating them can be done efficiently. Paying someone minimum wage to spend an hour eliminating ads from a magazine will cost more than the cover price for each affected issue. On the web this can be done at nearly no cost for any number of impressions.
Nevertheless, this is about business models. A free market economy derives its strength from subjecting companies to selection pressures like this. That an activity makes it difficult for some people to make money does not make it wrong or even illegal. Freedom should trump commerce in cases like these.
No, sorry, this time you've got it wrong. BSD was distributed as a complete system, even if most of that was AT&T code and even if it was based on the original Unix system created by AT&T. See (for example) this: "Over the next year, Joy, acting in the capacity of distribution secretary, sent out about thirty free copies of the system." Not patches, the whole system.
The difference was that GNU was to be a completely free system from the begining, but this is not germane to the basic premise. Actually, all of this is hair splitting anyway. Even if it had been distributed as patches, BSD clearly never refered to a single component the way Linux did and does.
Of course, all the BSD's use pretty much the whole GNU system as well, and you don't see him whining about calling them GNU/BSD.
The S in BSD stands for SYSTEM. From the begining BSD was intended to be a complete operating system with all the tools and components necessary for that. Similarly, GNU was intended to be a complete operating system. Linux was never meant to be more than a kernel and that is still true.
I think it's quite a good kernel, and I use it on all of the GNU systems I currently maintain. But a kernel is not an operating system. There is no such thing as a "Linux" operating system.
The fact that BSD adopted GNU tools to fill in the pieces originally supplied by AT&T does not make it BSD/GNU -- not in the eyes of RMS or anyone else. Bash is a component, not a system. Similarly the fact that GNU adopted XFree86 does not make it GNU/XFree or anything like that. XFree86 is a component, not a system.
Actually the fact that people use GNU with a Linux kernel does not make it GNU/Linux either: it's still GNU. But calling it GNU/Linux is a reasonable compromise because it gives credit to Torvald's particularly important contribution and allows one to distinguish between GNU systems with different kernels.
Slashdot Mirror
If, however, I want to include GPLed code in my program, the GPL forces me to release my program under the GPL. It has *infected* my program.
Wrong. You have infected your program by including the GPL covered code. No one held a gun to your head and demanded that you include it, nor did the code sneak in through an open window and jump into your source repository. That was your choice. Calling it viral is ungracious at best.
[...] under non-free licenses like the GPL [...]
That you prefer the BSD terms to those of the GPL does not make the latter non-free in any meaningful sense and your claim to that effect is nothing more than religious zealotry. Learn to accept that not everyone will genuflect at your church and get over yourself.
Security features? Most of the "security" features in Java are aimed at permitting more than one trust boundary in a single address space. This is a poorly understood problem that Java doesn't actually solve. But even if it did, a stack security system is much more complex and difficult to configure than Unix setuid(2) and therefore even less likey to be used properly in practice. Furthermore, putting different programs in separate address spaces has advantages for robustness.
As for the rest (mainly array bounds checking and garbage collection), equivalent or better alternatives are available in C. Better still, they can be used only where they're needed. Check out Apache resource pools and the glib object system for some examples. Obviously C++ has even more of these toys. And if we don't restrict our choices to these three there are hundreds of interpreted languages with equivalent or better features. Lisp comes to mind...
I find that I'm far more productive in Java than I ever was in C++ and C. I find that's even truer for less experienced programmers.Exactly the problem. A rigid system like Java is indeed wonderful if the goal is to allow mediocre programmers to create mediocre applications. Cobol had the same sort of appeal in its day. There's nothing wrong with that, but those who aim for excellence don't need to be pampered. All those fluffy features just get in the way.
Real hackers do it without a net. ;-)
Paul Graham makes each of these points in the article. Read first and comment later. That way everybody wins.
Free software depends on a few companies' ability to actually make money developing and using free software.
No, it doesn't. Software development does cost money, money which must come from somewhere, and profitable companies are one place money can come from, but not the only one. As a general rule, the greater the extent to which companies can make money through developing and using free software the more they will contribute to it. All of this we agree upon. However, this doesn't mean that every contribution should be accepted without thought for the consequences. Companies such as RedHat that manage to contribute without resorting to proprietary software are laudable, but those like Lindows are justifiably regarded with more skepticism.
Free software in general and Linux in particular have been around far longer than BitKeeper, and can thrive without it. While it's true that BitMover is entitled to choose the license they prefer for their product, their "contribution" to Linux is a dubious one for two reasons. First, whatever its technical merits, the philosophy it comes with is fundamentally defeatist. There are some problems, we are told, that free software just cannot address. Actually, there is no reason that someone with the same skills and motivation as Larry McVoy but with scruples couldn't do exactly what he's doing technically using a free software compatible business model. Second, the existance of high quality proprietary tools removes pressure to create good free ones, and is therefore a step backward for free software. Every kernel hacker who uses BitKeeper right now is someone who would otherwise have an incentive to contribute to Subversion otherwise, even if only in terms of testing, feedback and endorsement. A similar case can be made for free software friendly business models, which McVoy refuses even to discuss.
The argument that GPL code is 'better' is not persuasive if it doesn't fit in their business model. Not everyone has academic salaries or other means of support, so these are real concerns of people who support the idea of Free Source in a deep way.
Proprietary software companies are not the only employers in the world -- in fact they employ less than 10% of software developers! The rest are working for companies that do something else but need specific software for some internal purpose. The right way to address the concerns of wages for free software developers is to create a robust free software industry, but even in the meantime there is no need to make the free software movement to pander to the interests and business models of proprietary software companies by watering down licenses.
The no incentive to create a free alternative is bogus. If you think that, you are too concerned with the short term.
Please pay more attention. I didn't say there would be no incentive, but that there would be less. And there would indeed be less, because while there are people (including myself) to whom freedom is important, there are plenty who don't want to think about it. The lower the quality of proprietary tools, the more willing those people will be to assist in the development of free alternatives. I agree that the free software movement is likely to prevail, but I would rather see that happen sooner than later.
It's a take-back because the LGPL came into being because in many cases strict GPL made it difficult or at least questionable to even use GPL tools in the production of software for sale. Then much later RMS decides that too many people are using LGPL and he doesn't like it. That's a take back as far as I'm concerned.
The term "take-back" clearly implies that you were given something by someone who had a change of heart, and that as a result you are deprived of it. In the case of RMS frowning on the LGPL, we have the first and second, but not the third. No one has revoked the LGPL, nor have any projects I've heard of changed from LGPL to GPL licensing terms, so the term "take-back" can't possibly apply. Perhaps you can think of a project where the license was so changed, in which case I'll agree that (though in a somewhat crabbed sense, because you can still use the software versions prior to the license change) a take-back has occured.
I'll address the rest of your post in a separate reply.
From where I stand, RMS's stance on the LGPL is a take-back that is just as damaging, if not more so, as the EULA change being discussed. LGPL gives you a lot more choice in terms of integrating free and proprietary subsystems and components. Where free libraries have significantly extended functionality, he explicitly recomends GPL over LGPL.
Exactly how is this a "take-back" or in any way comparable to the BitMover EULA changes? The closest comparison would be a copyright holder re-releasing a particular library under the GPL rather than the LGPL after a certain version (if indeed that has ever happened), but even that is much less malignant because there is no direct attempt to curtail competition. The GPL and LGPL are concerned with the license of derivative works and are otherwise content neutral.
This is the one case where I would claim that it goes beyond style, and the message itself actually hurts the movement.
How so? It clearly hurts your employer, but the production of a clever proprietary tool cannot -- by definition -- help the free software movement. The higher the quality of such a tool, the less incentive there is for people to assist in the creation of a free alternative. Furthermore, you now have an excellent argument to make to your employer: "if we choose to license our tools under the GPL we can use this body of well-tested and freely available code to get a head start."
1) Almost every known root CA targets businesses as their primary customers.
So? People who run businesses are entitled to target any subset of potential customers they choose. Usually this means the people most willing to spend money will get the most attention. Nothing obligates a company to be generous toward those providing free services. I agree that this is an unfortunate situation, but it's not the fault of the certificate vendors.
The internet community should establish a trustworthy non-profit body to administer certificates that charges just enough to cover administrative costs. Until that happens we're stuck with a choice between self-signed certificates, self-certified certificates, or profit-oriented services.
2) 'Wildcard' certificates cost an absurd amount of money, usually $500 or more. Excuse me? The entire premise of the certification, is that Thawte (or VeriSign, or whoever) is certifying my trustworthiness as an organization.
Excuse me, but that is completely wrong. An end-entity certificate certifies that you are who you say you are, not that you are trustworthy.
Clearly a wildcard certificate is no more expensive to produce than a more specific one, but the fact remains that this is a market economy and there are reasonable alternatives. There is nothing fraudulent happening here.
3) VeriSign, the biggest fish in the pond, has demonstrated on more than one occasion that it is in fact not trustworthy.
True indeed, but again not a scam. Software is complex and security even more so. Being trustworthy is difficult, and while I see nothing praiseworthy about VeriSign, they should not be vilified for trying and failing. (There are plenty of unrelated things for which they truly deserve blame, but that's another story.)
From the start, the entire digital certificate business has been about politics and moneymaking, nothing more.
Hello? Politics and moneymaking are a legitimate part of society. We get nowhere by turning up our nose at these things. Accept them and get busy making things better.
You may have given somebody permission as far as your browser goes but that doesn't give you the right to change a link on a persons website...
While I share your sentiment I don't think this argument will wash. Installing the spyware on your computer doesn't change an actual website, just your view of it. A ruling that this violated the rights of the website owner would imply that many legitimate and useful things (such as Googles language translation service) would also require permission, which would make them impractical.
None of this reasoning makes what is being done by Kazaa and friends right, but it would be better to seek a remedy on the grounds that a clickwrap EULA is not enforcable, that the practices are deceptive and harmful the user or that this is abuse of the Amazon affiliate program terms.
Ummmm... GNU does NOT comprise 80% of a Linux distro. I refer you back to this article, Section 3. Adding up the 35 projects listed, GNU provides 26%.
Of course it doesn't; I didn't say that it did and that wasn't the point of the analogy. The goal of the GNU project has always been to create a complete free system, not to reinvent everything from scratch. Wherever a suitable piece of free software existed it was adopted and integrated. This is a poor way to win a contest over who writes the most lines of code (one has to spend time learning how someone else's code works and testing things that doesn't get counted), but an excellent way to get a complete free system put together.
On the other hand, the GNU project probably had more than 80% of the system put together (as opposed to written from scratch) by the time Linux came along. Clearly the Free Software Foundation has done more of the work in creating a complete free operating system than Linus Torvalds or Red Hat, and most of it at a time when it wasn't a glamorous thing to be doing.
What I'm getting at is that the people who drop in the last piece are not necessarily the ones who deserve most of the credit, and I think that presenting a simplified hypothetical situation to make the point is valid.
But it's obvious to me that I'm not going to convice you, and so far nothing that you've said is any more convincing than anything I've heard or read before. Would you agree that we disagree?
I will indeed, but under protest. I feel that you have given the arguments I've made only the most casual scrutiny, refuting points I wasn't making and ignoring the rest. Under these conditions I don't think anyone could convince you of anything.
If they're not antithetical to each other, then why would you feel the need to say this: "Do you really care so little about your freedom that you can't be bothered to prepend two little syllables?"
I'm not sure why you find this difficult to understand, but I'll try again. The concepts that "Linux" and "GNU" are associated with are not antithetical. This means that using them together is not a contradiction in terms, but it does not follow from this that they mean the same thing. It also doesn't follow that because they are not in direct opposition using "Linux" alone gives a sufficient emphasis to the freedom the "GNU" stands for. Get it?
"Who should get to make the choice of what the operating system is called?" The people who actually put it together.
That is an awefully aribitrary standard. Obviously, the people who put things together can get away with calling it whatever they want. That doesn't make it right. For example, I can repackage a GNU/Linux system and call it a "NewBSD" if I want to, but I doubt you would find that defensible. Why not? Because it is misleading. This situation is essentially the same.
I'm glad that they were able to produce those independant projects in such a way as they would fit together. But they did not put them together. They couldn't. They didn't have a kernel.
Acually, these are not truly independent projects. While they are designed to be modular enough to be useful separately (this makes them easier to test and useful for people who are forced to use proprietary systems for some purposes), they were designed as part of a complete free system. Just because they were easy to take apart doesn't mean they were never put together. This is true even though the result was not yet a complete free system.
While it's true that the GNU project took too long to finish their kernel, and it is arguable that they ought to have abandoned it in favor of Linux when it appeared, these tactical considerations aren't really relevant. Would you be pleased if, after completing 80% of a large free software project and getting bogged down on the remaining 20%, I added the missing pieces and released the product under a different name? What if people began to credit me with organizing the entire project?
Clearly in a free software community there is nothing you can do to forbid this, but I think you would find yourself doing a little "coattailing" of your own.
If the terms GNU and Linux are so antithetical to each other[...]
They are not antithetical, and no one said they were. Narrow technical advantage and a just-for-fun mentality are things that have their place. But the question here is whether freedom has a place too. Some of us think it does, and that it should come first when talking about what differentiates free and proprietary systems.
That's like saying that I can turn around what murder represents for by calling it GNU/murder.
No, actually, it's not. Get a grip.
[...]Linux uses the GPL and *insists* on guaranteeing the same freedoms as every FSF project. How is it that Linux represents anything other than the FSF's definition of freedom?
No one is complaining about the license used for Linux (except perhaps BSD license advocates and corporations who would like to take it proprietary). The problem is that people are using the term "Linux" -- which is the only legitimate thing to call the kernel as it is what Linus Torvalds chose -- to refer to an entire operating system built from GNU components around the GNU vision of a complete free system. Doing this implicitly gives Linus Torvalds credit for an entire operating system he didn't create.
Who should get to make the choice of what the operating system is called? Consider, as an analogy, who should get to choose what the kernel we call "Linux" is called. Obviously that would be Linus Torvalds because even if many different people and organizations contributed, he put the process in motion and guided the design. Similarly, the FSF conceived and organized the GNU system, so they have a strong claim when it comes to naming it.
How about the measure of system design and organization? Not only was GNU there first, they conceived the idea of developing a complete free system, which is the essential point of a modern "Linux" distribution. As a result, they did all the boring development work that a system requires, rather than choosing the thing that was most fun to work on. They are also responsible for most of the non-technical infrastructure that makes free software possible, most notably the GPL, but also more mundane acts like encouraging developers to contribute.
Things like lines of code and memory residence are easy to measure, but that doesn't make them good indications of who concieved a system. Kernighan and Ritchie wrote exactly none of the lines of code in modern Unix derivatives, but that doesn't mean we should forget about their contributions.
But the real point here is not that we should be splitting hairs to allocate credit. Instead, we should be thinking about the messages we want to send. The terms "Linux" and "Open Source" stand for a focus on narrow technicial advantage and a just-for-fun mentaility. "GNU" and "Free Software" stand for preserving meaningful liberty in a digital era. Do you really care so little about your freedom that you can't be bothered to prepend two little syllables?
This argument is complete nonsense. People and organizations pay for an advertisement because they believe that it will recieve attention. Magazine executives most certainly do care about what happens to their ads because they want to get paid for the next issue too.
What makes internet advertisements different is that eliminating them can be done efficiently. Paying someone minimum wage to spend an hour eliminating ads from a magazine will cost more than the cover price for each affected issue. On the web this can be done at nearly no cost for any number of impressions.
Nevertheless, this is about business models. A free market economy derives its strength from subjecting companies to selection pressures like this. That an activity makes it difficult for some people to make money does not make it wrong or even illegal. Freedom should trump commerce in cases like these.
No, sorry, this time you've got it wrong. BSD was distributed as a complete system, even if most of that was AT&T code and even if it was based on the original Unix system created by AT&T. See (for example) this:
"Over the next year, Joy, acting in the capacity of distribution secretary, sent out about thirty free copies of the system." Not patches, the whole system.
The difference was that GNU was to be a completely free system from the begining, but this is not germane to the basic premise. Actually, all of this is hair splitting anyway. Even if it had been distributed as patches, BSD clearly never refered to a single component the way Linux did and does.
Well alright then. But it *is* a system and it's *not* a standard, so... :-P
Of course, all the BSD's use pretty much the whole GNU system as well, and you don't see him whining about calling them GNU/BSD.
The S in BSD stands for SYSTEM. From the begining BSD was intended to be a complete operating system with all the tools and components necessary for that. Similarly, GNU was intended to be a complete operating system. Linux was never meant to be more than a kernel and that is still true.
I think it's quite a good kernel, and I use it on all of the GNU systems I currently maintain. But a kernel is not an operating system. There is no such thing as a "Linux" operating system.
The fact that BSD adopted GNU tools to fill in the pieces originally supplied by AT&T does not make it BSD/GNU -- not in the eyes of RMS or anyone else. Bash is a component, not a system. Similarly the fact that GNU adopted XFree86 does not make it GNU/XFree or anything like that. XFree86 is a component, not a system.
Actually the fact that people use GNU with a Linux kernel does not make it GNU/Linux either: it's still GNU. But calling it GNU/Linux is a reasonable compromise because it gives credit to Torvald's particularly important contribution and allows one to distinguish between GNU systems with different kernels.