Why can't Linux be more like OS X? If the open-source development model is so great, why did it take Apple a couple of years to do what Linux has failed to do for almost 15 years?
The open source development model isn't all that great; it's good at certain things (e.g., superior implementations of well-understood stuff that has been invented before), but not so good at others (e.g., inventing new stuff).
However, it took "Apple" more than a couple of years to do what you're attributing to them. OS X is NextStep 4, in effect; it's been under development for over 20 years by now.
I really hope that Theo does sue over this, even if the matter is said to be resolved.
There are two major misconceptions in this sentence:
Nobody who actually is involved in the dispute has said that it is resolved, AFAIK.
Theo is not the copyright holder, so he can't sue anyway.
I'd also like to see the OpenBSD devs counter-sued over the BCM driver affair. I know that Theo thinks there's a world of difference and that it can be dismissed as a deliberate attack by those "inhuman" Linux devs, but judges tend not to suffer from paranoid schizophrenia and may not see things the same way.
Sued how, exactly? You're misremembering what happened, or just failed to understand it at all.
The BCM driver folks notified the OpenBSD folks that some of the code submitted to their CVS tree was infringing. The OpenBSD folks admitted their error and removed the code within 24 hours. There was no controversy over the existence of a legal problem, or over its nature. There was controversy over the way the BCM folks handled the process, and over Theo's way of complaining about this. Theo's complaint is reasonable: the issue should have been raised in a less formal and more private manner; nothing was gained by voicing it so loudly and publically. Theo's way of voicing this reasonable complaint was, unsurprisingly, very undiplomatic.
Yes, users don't understand computers very well, that's true. Computer companies, however, should build this into the design, and minimize the amount of understanding and knowledge that users need to deploy to use computers. Blaming the users for the failures of the software industry is lazy, dishonest and self-serving.
Theo said it was "unethical" for someone to take BSD code and license a modified version under the GPL. And the thing is, it clearly isn't. It clearly isn't because that's essentially what the BSD people said they were perfectly ok with - they licensed it under the BSD license because they didn't care whether or not it stayed open, [...]
When you license your work under the BSD license, you surrender your right to sue other people in certain, specified circumstances; you don't surrender your right to disagree in some of those circumstances, or your right to protest when you disagree. There is nothing inconsistent about disagreeing with somebody's use of your BSD licensed code. You have just decided beforehand to surrender the use of the courts to pursue most of those disagreements.
I know what you're driving at, vaguely, that BSD licensed code encourages propagation of standards a nnd interoperability, but for most that doesn't come close to outweighing the idea that they don't want their effort to be able to be taken and sold for profit by others.
That's not exactly what I have in mind. The case I have in mind is along the following lines: suppose I think it's desirable for others to take my BSD-licensed software and incorporate it into propritetary works, as long as I judge the resulting work to be truly innovative. I think FOSS licensing makes sense for software that is, in effect, a superior, community-driven reimplentation of something that already exists, but that it's good that people can license truly innovative software proprietarily, so as to create incentives for such innovation. On the other hand, I think people who take my software and incorporate it, without doing any truly innovative work, are being selfish.
Now, of course, the problem is that I see no good way at all of writing a free software license that allows people to incorporate my work into proprietary products only they are "innovative." I can't see myself ever being able to nail down what "innovative" means without the benefit of hindsight. I'd be setting myself up for legal nightmares if somebody tried to test my attempts in court. Also, an unclear and complex legal document will most likely prevent interested parties from using my work in ways I intended, because of the legal unclarity of the status of their derived work. I could have such parties contact me for special licenses, but that's a hassle both for them and me. Many of them might never be aware that they I'd be interested in doing so. More importantly, it becomes really hard to manage in a large FOSS project with many authors.
In the light of these issues, I judge that it's better to just allow people to do what they want, because it's guaranteed to empower the "good guys" (in my hazy, not-well-defined conception) to do the things I want them to be able to do. The downside, of course, is that I'll get "bad guys" doing things that I think are bad. But I will sure as hell reserve the right to protest when they do so. I've renounced the right to sue them, not the right to tell them they're being bad.
I know what you're driving at, vaguely, that BSD licensed code encourages propagation of standards a nnd interoperability, but for most that doesn't come close to outweighing the idea that they don't want their effort to be able to be taken and sold for profit by others.
That's not exactly what I have in mind. The case I have in mind is along the following lines: suppose I think it's desirable for others to take my BSD-licensed software and incorporate it into propritetary works, as long as I judge the resulting work to be truly innovative. I think FOSS licensing makes sense for software that is, in effect, a superior, community-driven reimplentation of something that already exists, but that it's good that people can license truly innovative software proprietarily, so as to create incentives for such innovation. On the other hand, I think people who take my software and incorporate it, without doing any truly innovative work, are being selfish.
Now, of course, the problem is that I see no good way at all of writing a free software license that allows people to incorporate my work into proprietary products only they are "innovative." I can't see myself ever being able to nail down what "innovative" means without the benefit of hindsight. I'd be setting myself up for legal nightmares if somebody tried to test my attempts in court. Also, an unclear and complex legal document will most likely prevent interested parties from using my work in ways I intended, because of the legal unclarity of the status of their derived work. I could have such parties contact me for special licenses, but that's a hassle both for them and me. In the light of these issues, I judge that it's better to just allow people to do what they want, because it's guaranteed to empower the "good guys" (in my hazy, not-well-defined conception) to do the things I want them to be able to do. The downside, of course, is that I'll get "bad guys" doing things that I think are bad. But I will sure as hell reserve the right to protest when they do so. I've renounced the right to sue them, not the right to tell them they're being bad.
And pretending straight BSD code cannot be incorporated into GPL projects so long as the BSD license is retained is part of that. BSD code can be incorporated into GPL projects. Why some still persist in claiming otherwise, I still do not understand.
I think the only correct thing about any of that is "I still do not understand." Nobody's claimed that BSD licensed work can't be incorporated into GPL licensed ones. In fact, the "morals" part of the argument that Theo first made clearly presupposes so.
If you're going to release your code to the public under a license specifically allowing reuse and extension without releasing the source for the extensions, then get annoyed when people do exactly that - You're doing it wrong!
Unless you judge that the disadvantages of doing what you recommend outweigh the advantages of using the more permissive license. In that case, then, using the permissive license and getting annoyed when people do something bad is exactly the right thing to do.
Why do you insist in seeing this in such absolute terms? Why do you insist in assuming that the people you're criticizing haven't thought about their alternatives carefully? Is it because it allows you the convenience of assuming that you're always right?
I for one am quite pleased, and I rather surprised that the matter was able to be resolved successfully in a manner that didn't completely screw over the BSD folks.
Has the matter been actually resolved successfully? All I see is a set of pronouncements by the SFLC, some of whose staff apparently work as legal counsel for the parties in one side of the dispute.
It seems kind of petty to whine about someone stealing your code if you're releasing it under the BSD license though. By using the BSD license instead of the GPL you're choosing to let people take from you without giving back.
If I release my code under a BSD-style license, the fact that I allow you to use my code in nearly any way you want doesn't mean that I allow you to steal the credit for my work. This is really what this issue boils down to, which people repeatedly seem to miss.
The OpenBSD side's legal allegations are that first that their code was taken without attribution (stripping the original author's copyright notices), that somebody who doesn't own their code acted as if they did (by changing the license notice to GPL only), and then that even when the original notices were more or less restored, there was a claim to owning part of the work in question (by adding additional copyright and license notices), when there were (by their allegation) no original contributions, and thus, no part of the released work that was owned by anybody other than the BSD folk.
The SLFC's document about originality requirements spends nearly all of its time citing USA court decisions, whereas any action would be brought in Germany, not the USA. Yes, the very last section of the document (section 7) handwaves away this critical issue, by saying that we can use American copyright law as a guideline as to whether requirements of E.C. copyright law are met.
I wouldn't go as far as calling it "glossed over," but there is sure as hell a lot of talk about the issue being "resolved" without waiting for Theo and Reyk to chime in.
The real meat of the whole thing is in this analysis, where the SFLC argues in some detail that the changes made by the Linux Wireless folks do qualify for a derived work of their authorship. Do Theo and Reyk agree with this? We don't know. The licensing concession might be enough to get them to settle, independently of the authorship dispute, or then again it might not.
Certainly calling this "resolved" is very, very premature. You don't get to call this "resolved" until the other guy agrees not to sue.
Before you could communicate that to me, you had to think of it first.
Says who? That's not an empirical claim. It's a philosophical claim, and a controversial one. E.g., Fodor supports this ("language of thought"), while (late) Wittgenstein was very much against it. When you get down to concrete and realistic psychological models, how do you propose to distinguish the parts that constitute "thinking" from those that constitute "speaking"? How are you going to reconcile this with pervasive parallel processing in the brain, or with clear evidence that people start to say their sentences before they've managed to formulate them completely?
Are you suggesting that one of those functions is primary?
No, I'm suggesting that there do not exist principled criteria by which we can take a set of agreed-upon functions of language, and decide which of them is "primary." That's value-talk, not fact-talk. (And in fact, I don't even agree that there are criteria that we can use to individuate "functions" of language in a discrete manner. Which doesn't mean that language serves no functions, just that there is no well-defined set of functions that language serves.)
Reading quickly, the real meat of this is code analysis document, where the SFLC seems to argue, contrary to Theo and Reyk, that the files with the added copyright notices do qualify for derived work. (Still reading it.)
These are both in TFA, but it seems that most people here will find them more interesting than what the writeup actually says. Of course, important caveats: if this is really important to you, consult an unbiased lawyer.
Ok, they are indeed announcing that supposed changes by the Linux Wireless folks involved in this dispute will be released under a dual GPL or ISC license. But the last time I heard about this dispute, Reyk and Theo's most pressing claim was that the Linux Wireless developers in question illegally put their own copyright and license notices on work that they did not own; i.e., their position is that the Linux Wireless folks, in more than one instance, hadn't done enough original work for their release to qualify as a derived work of Reyk's code.
Still true, as you'd quickly learned if you researched it. But instead, of course, you prefer to hypothesize whatever fits your preconceptions, and don't bother to check.
You and GP both share a bad assumption that's common among non-linguists when they discuss language. To put it very roughly, it's the assumption that a language is a bag of words.
One of the important things in learning linguistics is to understand that, contrary to what lettered people in your culture would have you believe, words are boring little details; what's important about the language isn't whether it has words for such and such concept, but rather, how it organizes the encoding of situations into utterances; what aspects of the situation, given the system of choices that the grammar forces upon you, must obligatorily be encoded by the sentence, which may optionally be so, and which require circumlocution to encode.
Good examples are evidentiality and spatial relations. Evidentiality is when a language's grammar requires you to encode, as part of your sentence, the nature of the evidence that supports it: whether you saw the situation described with your own eyes, whether somebody told you it, whether you inferred it from other things, etc. Spatial relations have to do with whether you encode the spatial relation between two objects referred to in a sentence in, for example, terms that are relative to the orientation of your own body ("sit on the chair to my left"), or, as another example, in absolute directional terms ("sit on the northern chair").
Yes, the claim that other languages are special because they have words for concepts that yours don't is a bit of a red herring (when you try to give it substance, you're really going to end up talking about anthropology, not about linguistics). But the reason it's a red herring is because words are not very important.
"The limits of my language mean the limits of my world."
Wittgenstein was wrong, then. Or else how do you think a child learns the word for "soft", if the concept is beyond him before learning the word (and associating it with the concept) ? In fact, how is anyone able to learn anything ?
Heh, that Wittgenstein quote is totally out of context. Wittgenstein says that in the Tractatus Logico-Philosophicus, which is his early work, that he later repudiated (and I'm not sure that it means in context what GP thinks it does; I don't understand the Tractatus very well). In his latter work, he makes exactly the same argument you're making here: "only someone who already knows how to do something with it can significantly ask a name" (Philosophical Investigations, 31).
The difference now is that there are forces that speed up the extinction of non-self-sustaining types of cultures.
Aren't you assuming here that whether a culture is "self-sustaining" is an independent fact about the culture itself? This statement, as I read it, is profoundly circular, because the cultures in question are quite likely "non-self-sustaining" precisely because of the forces in question.
Slashdot Mirror
The open source development model isn't all that great; it's good at certain things (e.g., superior implementations of well-understood stuff that has been invented before), but not so good at others (e.g., inventing new stuff).
However, it took "Apple" more than a couple of years to do what you're attributing to them. OS X is NextStep 4, in effect; it's been under development for over 20 years by now.
If you read the comment you're replying to, you can answer your own questions. (Hint: the answer is "no.")
Because they might try to act on their beliefs, that's why.
It's not horrible. It's false. RTFA. Note who's said what, and who hasn't said what.
There are two major misconceptions in this sentence:
Sued how, exactly? You're misremembering what happened, or just failed to understand it at all.
The BCM driver folks notified the OpenBSD folks that some of the code submitted to their CVS tree was infringing. The OpenBSD folks admitted their error and removed the code within 24 hours. There was no controversy over the existence of a legal problem, or over its nature. There was controversy over the way the BCM folks handled the process, and over Theo's way of complaining about this. Theo's complaint is reasonable: the issue should have been raised in a less formal and more private manner; nothing was gained by voicing it so loudly and publically. Theo's way of voicing this reasonable complaint was, unsurprisingly, very undiplomatic.
Yes, users don't understand computers very well, that's true. Computer companies, however, should build this into the design, and minimize the amount of understanding and knowledge that users need to deploy to use computers. Blaming the users for the failures of the software industry is lazy, dishonest and self-serving.
When you license your work under the BSD license, you surrender your right to sue other people in certain, specified circumstances; you don't surrender your right to disagree in some of those circumstances, or your right to protest when you disagree. There is nothing inconsistent about disagreeing with somebody's use of your BSD licensed code. You have just decided beforehand to surrender the use of the courts to pursue most of those disagreements.
That's not exactly what I have in mind. The case I have in mind is along the following lines: suppose I think it's desirable for others to take my BSD-licensed software and incorporate it into propritetary works, as long as I judge the resulting work to be truly innovative. I think FOSS licensing makes sense for software that is, in effect, a superior, community-driven reimplentation of something that already exists, but that it's good that people can license truly innovative software proprietarily, so as to create incentives for such innovation. On the other hand, I think people who take my software and incorporate it, without doing any truly innovative work, are being selfish.
Now, of course, the problem is that I see no good way at all of writing a free software license that allows people to incorporate my work into proprietary products only they are "innovative." I can't see myself ever being able to nail down what "innovative" means without the benefit of hindsight. I'd be setting myself up for legal nightmares if somebody tried to test my attempts in court. Also, an unclear and complex legal document will most likely prevent interested parties from using my work in ways I intended, because of the legal unclarity of the status of their derived work. I could have such parties contact me for special licenses, but that's a hassle both for them and me. Many of them might never be aware that they I'd be interested in doing so. More importantly, it becomes really hard to manage in a large FOSS project with many authors.
In the light of these issues, I judge that it's better to just allow people to do what they want, because it's guaranteed to empower the "good guys" (in my hazy, not-well-defined conception) to do the things I want them to be able to do. The downside, of course, is that I'll get "bad guys" doing things that I think are bad. But I will sure as hell reserve the right to protest when they do so. I've renounced the right to sue them, not the right to tell them they're being bad.
I think the only correct thing about any of that is "I still do not understand." Nobody's claimed that BSD licensed work can't be incorporated into GPL licensed ones. In fact, the "morals" part of the argument that Theo first made clearly presupposes so.
Unless you judge that the disadvantages of doing what you recommend outweigh the advantages of using the more permissive license. In that case, then, using the permissive license and getting annoyed when people do something bad is exactly the right thing to do.
Why do you insist in seeing this in such absolute terms? Why do you insist in assuming that the people you're criticizing haven't thought about their alternatives carefully? Is it because it allows you the convenience of assuming that you're always right?
Has it? All I see is SFLC pronouncements.
Has the matter been actually resolved successfully? All I see is a set of pronouncements by the SFLC, some of whose staff apparently work as legal counsel for the parties in one side of the dispute.
If I release my code under a BSD-style license, the fact that I allow you to use my code in nearly any way you want doesn't mean that I allow you to steal the credit for my work. This is really what this issue boils down to, which people repeatedly seem to miss.
The OpenBSD side's legal allegations are that first that their code was taken without attribution (stripping the original author's copyright notices), that somebody who doesn't own their code acted as if they did (by changing the license notice to GPL only), and then that even when the original notices were more or less restored, there was a claim to owning part of the work in question (by adding additional copyright and license notices), when there were (by their allegation) no original contributions, and thus, no part of the released work that was owned by anybody other than the BSD folk.
The SLFC's document about originality requirements spends nearly all of its time citing USA court decisions, whereas any action would be brought in Germany, not the USA. Yes, the very last section of the document (section 7) handwaves away this critical issue, by saying that we can use American copyright law as a guideline as to whether requirements of E.C. copyright law are met.
IANAL, but isn't this a pretty bad idea?
I wouldn't go as far as calling it "glossed over," but there is sure as hell a lot of talk about the issue being "resolved" without waiting for Theo and Reyk to chime in.
The real meat of the whole thing is in this analysis, where the SFLC argues in some detail that the changes made by the Linux Wireless folks do qualify for a derived work of their authorship. Do Theo and Reyk agree with this? We don't know. The licensing concession might be enough to get them to settle, independently of the authorship dispute, or then again it might not.
Certainly calling this "resolved" is very, very premature. You don't get to call this "resolved" until the other guy agrees not to sue.
Says who? That's not an empirical claim. It's a philosophical claim, and a controversial one. E.g., Fodor supports this ("language of thought"), while (late) Wittgenstein was very much against it. When you get down to concrete and realistic psychological models, how do you propose to distinguish the parts that constitute "thinking" from those that constitute "speaking"? How are you going to reconcile this with pervasive parallel processing in the brain, or with clear evidence that people start to say their sentences before they've managed to formulate them completely?
No, I'm suggesting that there do not exist principled criteria by which we can take a set of agreed-upon functions of language, and decide which of them is "primary." That's value-talk, not fact-talk. (And in fact, I don't even agree that there are criteria that we can use to individuate "functions" of language in a discrete manner. Which doesn't mean that language serves no functions, just that there is no well-defined set of functions that language serves.)
Reading quickly, the real meat of this is code analysis document, where the SFLC seems to argue, contrary to Theo and Reyk, that the files with the added copyright notices do qualify for derived work. (Still reading it.)
I mention in an earlier post an SFLC document about originality requirements. They've also put together a set of guidelines for using permissively-licensed software in a GPL project.
These are both in TFA, but it seems that most people here will find them more interesting than what the writeup actually says. Of course, important caveats: if this is really important to you, consult an unbiased lawyer.
Ok, they are indeed announcing that supposed changes by the Linux Wireless folks involved in this dispute will be released under a dual GPL or ISC license. But the last time I heard about this dispute, Reyk and Theo's most pressing claim was that the Linux Wireless developers in question illegally put their own copyright and license notices on work that they did not own; i.e., their position is that the Linux Wireless folks, in more than one instance, hadn't done enough original work for their release to qualify as a derived work of Reyk's code.
I don't see anything in TFA that directly addresses this. There is a link to a new document about originality requirements under the law (which I haven't read yet, I'll admit), but I would hope that this issue was addressed explicitly.
Still true, as you'd quickly learned if you researched it. But instead, of course, you prefer to hypothesize whatever fits your preconceptions, and don't bother to check.
c) The descendants of recent immigrants to the USA are adopting English and losing their native language at a faster rate than in the past.
You and GP both share a bad assumption that's common among non-linguists when they discuss language. To put it very roughly, it's the assumption that a language is a bag of words.
One of the important things in learning linguistics is to understand that, contrary to what lettered people in your culture would have you believe, words are boring little details; what's important about the language isn't whether it has words for such and such concept, but rather, how it organizes the encoding of situations into utterances; what aspects of the situation, given the system of choices that the grammar forces upon you, must obligatorily be encoded by the sentence, which may optionally be so, and which require circumlocution to encode.
Good examples are evidentiality and spatial relations. Evidentiality is when a language's grammar requires you to encode, as part of your sentence, the nature of the evidence that supports it: whether you saw the situation described with your own eyes, whether somebody told you it, whether you inferred it from other things, etc. Spatial relations have to do with whether you encode the spatial relation between two objects referred to in a sentence in, for example, terms that are relative to the orientation of your own body ("sit on the chair to my left"), or, as another example, in absolute directional terms ("sit on the northern chair").
Yes, the claim that other languages are special because they have words for concepts that yours don't is a bit of a red herring (when you try to give it substance, you're really going to end up talking about anthropology, not about linguistics). But the reason it's a red herring is because words are not very important.
Heh, that Wittgenstein quote is totally out of context. Wittgenstein says that in the Tractatus Logico-Philosophicus, which is his early work, that he later repudiated (and I'm not sure that it means in context what GP thinks it does; I don't understand the Tractatus very well). In his latter work, he makes exactly the same argument you're making here: "only someone who already knows how to do something with it can significantly ask a name" (Philosophical Investigations, 31).
Aren't you assuming here that whether a culture is "self-sustaining" is an independent fact about the culture itself? This statement, as I read it, is profoundly circular, because the cultures in question are quite likely "non-self-sustaining" precisely because of the forces in question.