Unfortunately, anytime the value of something changes, someone wins and someone loses. I know hindsight is 20/20, but how far did you expect to get living in Europe with US dollars in the bank? Why would you not change those over to Euros when you decided to live there?
The real lesson here is the value of diversifying your investments. It's impossible to tell which currency is going to do better in the future. Therefore, it makes sense to hold assets denominated in all major currencies you have access to. Exchange rates are a zero-sum game, so if you get the proportions right, then every time one currency loses, well, you already hold an equivalent amount in assets of the currencies that won, so that, if you sell them and convert the proceeds to dollars, you get more dollars than you'd have gotten before the loss.
...ummm no, it means that people in a position too are trying to help others not get screwed by a vendor locked-in product that wants to charge you for a ringtone that you can make yourself. Instead of attacking developers who wish to enlighten a public entranced by Apple, perhaps they shouldn't base a revenue stream on vendor lockin and ripoff ringtones. If you ask me (flame on that noones asking), they should be the ones providing such a ringtone app. They are all about ease of use for the masses... oh wait, I forgot its easier for someone to pay them then do it themselves.
You forgot the part where you argue that it's OK to break the law when it's convenient for you.
TFA doesn't talk about "blaming" at all. All of the "blame" talk is in the Slashdot headline. This is an article about a bankruptcy court filing, citing documents where the company is required to provide a serious analysis of the reason it needs bankruptcy protection.
You're not offering offer a fair analysis of an economic situation if you reason about it axiomatically, from an impoverished set of axioms that assume that the parties to every exchange are perfectly rational, that what they value doesn't change by the act of purchasing, and that they possess perfect information. All you're doing is demonstrating that you have an unempirical adherence to the axiom that trade only happens because both parties wanted the trade to happen, and that whenever you see some situation that contradicts it, you will reject the existence or straightforward description of the situation.
You can always preserve your belief in a given claim by refusing to believe the things that would contradict it.
And they are routinely reported sensationalistically in the media, and most of you people who are reading this right now swallow it all hook and sinker.
It seems profoundly stupid to stress out the BSD license as the "most important" feature of this new software.
Yup. However, to the credit of the people developing it, they are not the ones stressing this. It's just the slashdot submission trolls and the gossip-greedy slashbots who are...
The problem with C isn't just the features (though a real module system would be cool, though). The big problem with C is that people write too much damn C code when they ought not to. Yes, writing to hardware registers and doing memory management needs pretty low level access to the hardware. Hardly anything else does, yet most C code out there is doing other stuff.
If I choose to distribute something under the GPL, the receipients only get permissions under the GPL.
Permission from who?
There already exists a legal document that specifies which permissions the recipients of a dual-licensed work get. It's called the license notice, and it says that the copyright holder gives anybody permission to make use of the work in question under either the terms of the GPL or the BSD.
Why do you think that your act of distributing the work has anything to do with what permissions the recipients get to the work? It doesn't. The permissions come from the license the copyright holder grants to everybody, not from you as a sublicensor.
So where does this leave , who used BSD-licensed code in their commercial software, which is distributed via a commercial license? Is this sublicensing? If so, how are you distinguishing between making a deritive work (allowed), and sublicensing (not allowed?) If not, what's the distinction you're drawing again?
The commercial vendor is distributing a derived work in a binary-only fashion. They can only control the license to the derived work, not to the original that it incorporates. Recipients of such a derived work, in order to be allowed to use it, must have a license to use both the original and the derived work. Since the BSD license gives everybody permission to use the original and any derived works as long as some very minimal conditions are met, the license to the original is never an issue.
Sublicensing makes sense only for proprietary software. I could license you to use the source code of a library that I've written in derivative products in a way that gives you permission to sublicense binary copies the of that library to the recipients of your derived works that incorporate my work, with the condition that you pay me a fee for any such sublicense that you grant. The recipients of your derived work still need permission to use both my library and your program; they obtain both permissions from you, and you obtain the permission to grant permission to my work from me.
No, what you are describing does not constitute a "relicensing" or sublicense of the BSD work in question. Neither Microsoft nor Apple have any right to give additional permissions to use the BSD work. They have a right to distribute it as part of a derived work, in binary-only form; their users have a right to use the resulting work, and if they obtain the sources for the original, to use them in the same manner as MS or Apple do.
I could take the FreeBSD distro and distribute it under any license I choose, it's my right under their license and they are free to do the same.
No, you can't. It's not your work. You can only distribute it under the licenses that the copyright holders grant you to distribute it.
Given the actual licenses involved here, this means that for most of the code in it, you can distribute only binaries if you so wish. This doesn't mean that you picked the license you wanted. The recipients of the binaries have all the rights that you have to those binaries; they just don't have the sources, and don't have any right to ask you to provide them. However, you must put in the documentation the notice that the source code for FreeBSD is licensed to everybody under the BSD license terms.
What Theo doesn't seem to realize is that as soon as a user chooses to comply with the GPL license over the BSD license in a dual-licensed software package, the text of the BSD license becomes modifiable under the terms of the GPL.
Oh, sure yeah. You are allowed to make it look like the original author, contrary to fact, licenses the code to other people under terms that you actually picked. Yeah. Nothing even remotely questionably legal with that.
If you cannot remove the BSD license, how can you release it under the GPL? The GPL is incompatible with portions of the BSD license (in particular, the fact that you can release modifications in binary form only).
You can't "release it under the GPL." You can make a derived work that incorporates the original work, and you can release the derived work under the GPL (or as a proprietary work, for that matter). The key thing is that your choice to use the GPL applies to your own work (your modifications and additions, and the derived work that results once your modifications and additions are applied to the original), not to the work of the original author of the BSD licensed work. Nobody can release your GPL-licensed work as binary-only, but your work contains parts that they may choose to release as binaries, as long as none of your own work is included in such release.
Sorry but I didn't see where you had to retain the license, in the license.
You do not have a right to sublicense a copyrighted work unless that right is explicitly granted to you by somebody who has the right to do so. In the case of BSD or GPL licensed works, nobody grants you the right to sublicense them. The license is a permission to do certain specified things. The only things you can do, beyond those that the license explicitly grants you permission to do, are those that the law says you can do (e.g., fair use). Sublicensing isn't one of them.
Now, obviously, slapping a copy of the GPL in the file is within your rights to "use, copy modify, and distribute" the software.
No, it is not, no more than taking a GPL-licensed file and changing its license notice to the BSD terms would be. You're knowingly misrepresenting the license you and your recipients have to the work in question. It may sound moot in the case of a work licensed as permissively as the BSD license does, but the BSD license terms aren't just about allowing you to use, copy, modify and distribute; they are also about allowing your recipients to do so, and because of that, you are required to tell your recipients of the rights they have to use the BSD code you took.
Most people belive that means you can either accept the BSD terms or accept the GPL terms (and from then on follow only the one chosen set of terms). Theo seems to be claiming that you somehow have to follow both sets of terms.
Nope, nobody is claiming the latter, except as a strawman, and everybody agrees on the former, for the cases where it applies (hint: there are disputed files that were only licensed under the BSD).
You can modify and redistribute dual-licensed code under either the terms of the GPL or the BSD. Insofar as the code you redistribute contains work that's covered under such a dual license, however, your recipients also have a dual license to those parts of the work you give them, because nobody has granted anybody a right to sublicense work. This means that:
You can't change the license to somebody else's work, unless they give you explicit permission to do so (which they don't, in this case).
Nobody else can dictate or change the license to your own work, for the same reason. However, for you to slap your own copyright and license notices on something, you really must have done some original and non-trivial work (a condition which the OpenBSD folks are claiming has been violated).
If you take a dual-licensed piece of work, modify the license notice to make it say what you would like, and give it to somebody else, you're misrepresenting the license that your recipient has to that work, which may be illegal.
The issue here is not what you can do with BSD licensed code, but what you can do with code licensed under both the BSD and the GPL licenses. Specifically, the question is whether you can accept one license and not the other; which is precisely the decision that precipitated the current debate.
You can always not accept a license. But, to paraphrase the GPL's friendly reminder, the only thing giving you a right to make any use of the covered work is the license in question.
To put it more bluntly, we need clarification on whether dual-licensing has any point in it at all.
Yes, it explicily allows the code to be used by anybody either under the terms of the GPL or the BSD licenses. If the BSD license terms truly make moot the alternative GPL terms, then this would be pointless; however, if a court at some point decides that the licenses are actually incompatible, then code that is dual-licensed is protected from the chaos than would ensue. Essentially, it's a legal safeguard.
(To disclaim something from the very start: IANAL.)
Forget the technical term "license," or the technical sense of "work" for a bit. What copyright law amounts to, essentially, is that when you do some creative work, you automatically have certain rights to control the use, modification and distribution of the result of your work. Except for certain exceptions (e.g., fair use), nobody can make use of your work without your permission.
This is what the term "license" means--you, as an author of a piece of creative work, give permission to other parties to use your work in certain specified ways, as long as they obey certain conditions.
If you want to use somebody else's work, you have to have permission to use it in the way in question. This means that if you want to run a program, you need permission to run the program; if you want to give a copy to somebody else, you need permission to do so; if you want to produce an original piece of work that contains, as an inseparable part of it, the work in question, and give this to somebody else, you need permission do that.
Also, when you give anything that incorporates the original work to somebody else, the recipient also needs to have permission to use it in whichever ways they do. There are two ways your recipients can obtain that permission:
The original author can give them permission to use their work.
The original author can give you permission to give others permission to use their work, under certain circumstances.
The second of these is called "sublicensing," and it's important to keep in mind, because neither the BSD nor the GPL licenses allow you to sublicense; they work by giving the same set of permissions to everybody.
Now, here's the deal with the "dual-licensed" code in the case in question (which shouldn't be taken as the general case for "dual-licensing," because the term is somewhat ambiguous). The author of the original work gives permission to everybody to make certain uses of the code, as long as they satisfy the terms of either the GPL or BSD. However, you're not given the right to sublicense the work; that means that you have no right to give others permission to use the original author's work, nor to dictate what terms they have to follow to use that work. The whole free software scheme works because the original author grants license to everybody, which means that your recipients have the original author's permission to use their code even if they receive it from another party like you.
Now, here's how I read the dual-license situation with the reproduction of the BSD license notice:
The dual license to use the code under either the BSD or the GPL terms does not require you to reproduce the BSD license terms in derived works that you give to other people.
However, if you do so, you are misrepresenting the license to those pieces of work; you are telling your recipients that they must abide by the terms of the GPL if they want to use the portions of your derived work that come from the original, when in fact, they may choose between the terms of the GPL and the BSD, because the original author licensed their work to everybody under the terms of either, and did not give anybody the right to sublicense as GPL-only. This misrepresentation may be illegal or not; but at the very least, you clearly shouldn't do that.
Another important thing is that for you to be able to assert copyright, you must have done some original, creative work. This is an important thing to keep in mind, because one of the arguments that Theo has made in this case is that there are some files that have had copyright and license notices added to them, but which do not show any original, creative work other than what's in OpenBSD. This may be true or not (I've heard the argument that making these files work in Linux, in and of itself, should qualify as original, creative work), but one thing for sure is that most people commenting about the situation are failing to even understand what arguments the participants are making.
The driver code in question was duel liscensed under the GPL and BSD liscenses. It contained text stating that it could be redistributed under the GPL or BSD liscense.
Nope. Some code involved in the dispute had a license notice that says that you may use, distribute or modify under either the terms of the GPL or the BSD. Other parts of code are BSD license only.
But even assuming that all of the code was dual licensed, removing the license notice is at least questionable. The people who receive that code from you, either in the original form or in the form of a derived work from it, have the right to the original work under the terms of either the GPL or BSD licenses, at their choice. When you strip out the BSD notice, you misrepresent the license to the original work, which is at least bad, if not flat out illegal.
You just need to make it clear that there is work in the files that you are distributing that is not yours, and that the holder of the copyright on that work license it under the terms in question. You don't need to distribute the original files.
Really, what the terms force you to do is to communicate to the people who receive your derived work's source that, whatever license you chose for your work, there is an original work that it's based on, and that it's available under the terms of the BSD license; this is why you are required to reproduce those terms. You don't have to tell your recipients which parts are yours and which are from the original work--in fact, if you modify somebody else's code extensively, it's difficult in general to decide which parts belong to who.
Has anyone ever stopped to realize that Google have actually produced nothing technically interesting since a mediocre research thesis (and I mean mediocre by standards of computer science / mathematics, not by layman's standards)?
I think you don't value Google's contribution to internet search fairly. If you just focus on CS and mathematics, you're just not being fair, IMO.
But still: yes, Google is immensely overvalued. They have exceedingly few lines of business that are actually profitable (selling ad space, the search appliance), specially considering all the stuff that they have developed that has fizzled and/or produced relatively little income. Yet despite this, people all over expect them as a matter of course to deliver revolutionary innovations in fields that they have absolutely no track record on.
Theo's email cited in TFA cites relevant US copyright law. I recommend you read it. (*resists urge to lash out against people who comment on this submission without actually reading through the linked materials, only to notice that parenthetical note yields to urge*)
The Berne Convention and the copyright law of other countries are substantially similar. The basic idea is that you can't slap your own copyright notice on something you don't own; you have to do some meaningful amount work before you can claim that you've got a derivative work.
Note that they only said; "works based on the material". Technically, a work that has even one character of code added is still 'based on the material'. What I mean is, if he's going to say it's illegal to do so, it's not derived *inherently* from the licence, but from the law of the country (in determining what 'based on' means). And that, of course, is dependent on the country. It may well be that it might be totally legit in some countries, thus.
Yes. And if you read what he says, he says that the law is being violated.
Slashdot Mirror
The real lesson here is the value of diversifying your investments. It's impossible to tell which currency is going to do better in the future. Therefore, it makes sense to hold assets denominated in all major currencies you have access to. Exchange rates are a zero-sum game, so if you get the proportions right, then every time one currency loses, well, you already hold an equivalent amount in assets of the currencies that won, so that, if you sell them and convert the proceeds to dollars, you get more dollars than you'd have gotten before the loss.
You forgot the part where you argue that it's OK to break the law when it's convenient for you.
You're talking as if philosophy of science were not an insanely controversial field that fails to agree on stuff seemingly as basic as that.
TFA doesn't talk about "blaming" at all. All of the "blame" talk is in the Slashdot headline. This is an article about a bankruptcy court filing, citing documents where the company is required to provide a serious analysis of the reason it needs bankruptcy protection.
You're not offering offer a fair analysis of an economic situation if you reason about it axiomatically, from an impoverished set of axioms that assume that the parties to every exchange are perfectly rational, that what they value doesn't change by the act of purchasing, and that they possess perfect information. All you're doing is demonstrating that you have an unempirical adherence to the axiom that trade only happens because both parties wanted the trade to happen, and that whenever you see some situation that contradicts it, you will reject the existence or straightforward description of the situation.
You can always preserve your belief in a given claim by refusing to believe the things that would contradict it.
Exactly. Incidentally, this is also the way that the no-global-warming myth started, too.
And they are routinely reported sensationalistically in the media, and most of you people who are reading this right now swallow it all hook and sinker.
Yup. However, to the credit of the people developing it, they are not the ones stressing this. It's just the slashdot submission trolls and the gossip-greedy slashbots who are...
The problem with C isn't just the features (though a real module system would be cool, though). The big problem with C is that people write too much damn C code when they ought not to. Yes, writing to hardware registers and doing memory management needs pretty low level access to the hardware. Hardly anything else does, yet most C code out there is doing other stuff.
Permission from who?
There already exists a legal document that specifies which permissions the recipients of a dual-licensed work get. It's called the license notice, and it says that the copyright holder gives anybody permission to make use of the work in question under either the terms of the GPL or the BSD.
Why do you think that your act of distributing the work has anything to do with what permissions the recipients get to the work? It doesn't. The permissions come from the license the copyright holder grants to everybody, not from you as a sublicensor.
The commercial vendor is distributing a derived work in a binary-only fashion. They can only control the license to the derived work, not to the original that it incorporates. Recipients of such a derived work, in order to be allowed to use it, must have a license to use both the original and the derived work. Since the BSD license gives everybody permission to use the original and any derived works as long as some very minimal conditions are met, the license to the original is never an issue.
Sublicensing makes sense only for proprietary software. I could license you to use the source code of a library that I've written in derivative products in a way that gives you permission to sublicense binary copies the of that library to the recipients of your derived works that incorporate my work, with the condition that you pay me a fee for any such sublicense that you grant. The recipients of your derived work still need permission to use both my library and your program; they obtain both permissions from you, and you obtain the permission to grant permission to my work from me.
...and in the same directory as the file you point us at, we find BSD-only files originally by Reyk Floeter, with other author's GPL and copyright notices attached to them.
There are multiple disputes in this case, remember:
No, what you are describing does not constitute a "relicensing" or sublicense of the BSD work in question. Neither Microsoft nor Apple have any right to give additional permissions to use the BSD work. They have a right to distribute it as part of a derived work, in binary-only form; their users have a right to use the resulting work, and if they obtain the sources for the original, to use them in the same manner as MS or Apple do.
No, you can't. It's not your work. You can only distribute it under the licenses that the copyright holders grant you to distribute it.
Given the actual licenses involved here, this means that for most of the code in it, you can distribute only binaries if you so wish. This doesn't mean that you picked the license you wanted. The recipients of the binaries have all the rights that you have to those binaries; they just don't have the sources, and don't have any right to ask you to provide them. However, you must put in the documentation the notice that the source code for FreeBSD is licensed to everybody under the BSD license terms.
Oh, sure yeah. You are allowed to make it look like the original author, contrary to fact, licenses the code to other people under terms that you actually picked. Yeah. Nothing even remotely questionably legal with that.
You can't "release it under the GPL." You can make a derived work that incorporates the original work, and you can release the derived work under the GPL (or as a proprietary work, for that matter). The key thing is that your choice to use the GPL applies to your own work (your modifications and additions, and the derived work that results once your modifications and additions are applied to the original), not to the work of the original author of the BSD licensed work. Nobody can release your GPL-licensed work as binary-only, but your work contains parts that they may choose to release as binaries, as long as none of your own work is included in such release.
You do not have a right to sublicense a copyrighted work unless that right is explicitly granted to you by somebody who has the right to do so. In the case of BSD or GPL licensed works, nobody grants you the right to sublicense them. The license is a permission to do certain specified things. The only things you can do, beyond those that the license explicitly grants you permission to do, are those that the law says you can do (e.g., fair use). Sublicensing isn't one of them.
No, it is not, no more than taking a GPL-licensed file and changing its license notice to the BSD terms would be. You're knowingly misrepresenting the license you and your recipients have to the work in question. It may sound moot in the case of a work licensed as permissively as the BSD license does, but the BSD license terms aren't just about allowing you to use, copy, modify and distribute; they are also about allowing your recipients to do so, and because of that, you are required to tell your recipients of the rights they have to use the BSD code you took.
Nope, nobody is claiming the latter, except as a strawman, and everybody agrees on the former, for the cases where it applies (hint: there are disputed files that were only licensed under the BSD).
You can modify and redistribute dual-licensed code under either the terms of the GPL or the BSD. Insofar as the code you redistribute contains work that's covered under such a dual license, however, your recipients also have a dual license to those parts of the work you give them, because nobody has granted anybody a right to sublicense work. This means that:
You can always not accept a license. But, to paraphrase the GPL's friendly reminder, the only thing giving you a right to make any use of the covered work is the license in question.
Yes, it explicily allows the code to be used by anybody either under the terms of the GPL or the BSD licenses. If the BSD license terms truly make moot the alternative GPL terms, then this would be pointless; however, if a court at some point decides that the licenses are actually incompatible, then code that is dual-licensed is protected from the chaos than would ensue. Essentially, it's a legal safeguard.
(To disclaim something from the very start: IANAL.)
Forget the technical term "license," or the technical sense of "work" for a bit. What copyright law amounts to, essentially, is that when you do some creative work, you automatically have certain rights to control the use, modification and distribution of the result of your work. Except for certain exceptions (e.g., fair use), nobody can make use of your work without your permission.
This is what the term "license" means--you, as an author of a piece of creative work, give permission to other parties to use your work in certain specified ways, as long as they obey certain conditions.
If you want to use somebody else's work, you have to have permission to use it in the way in question. This means that if you want to run a program, you need permission to run the program; if you want to give a copy to somebody else, you need permission to do so; if you want to produce an original piece of work that contains, as an inseparable part of it, the work in question, and give this to somebody else, you need permission do that.
Also, when you give anything that incorporates the original work to somebody else, the recipient also needs to have permission to use it in whichever ways they do. There are two ways your recipients can obtain that permission:
The second of these is called "sublicensing," and it's important to keep in mind, because neither the BSD nor the GPL licenses allow you to sublicense; they work by giving the same set of permissions to everybody.
Now, here's the deal with the "dual-licensed" code in the case in question (which shouldn't be taken as the general case for "dual-licensing," because the term is somewhat ambiguous). The author of the original work gives permission to everybody to make certain uses of the code, as long as they satisfy the terms of either the GPL or BSD. However, you're not given the right to sublicense the work; that means that you have no right to give others permission to use the original author's work, nor to dictate what terms they have to follow to use that work. The whole free software scheme works because the original author grants license to everybody, which means that your recipients have the original author's permission to use their code even if they receive it from another party like you.
Now, here's how I read the dual-license situation with the reproduction of the BSD license notice:
Another important thing is that for you to be able to assert copyright, you must have done some original, creative work. This is an important thing to keep in mind, because one of the arguments that Theo has made in this case is that there are some files that have had copyright and license notices added to them, but which do not show any original, creative work other than what's in OpenBSD. This may be true or not (I've heard the argument that making these files work in Linux, in and of itself, should qualify as original, creative work), but one thing for sure is that most people commenting about the situation are failing to even understand what arguments the participants are making.
Nope. Some code involved in the dispute had a license notice that says that you may use, distribute or modify under either the terms of the GPL or the BSD. Other parts of code are BSD license only.
But even assuming that all of the code was dual licensed, removing the license notice is at least questionable. The people who receive that code from you, either in the original form or in the form of a derived work from it, have the right to the original work under the terms of either the GPL or BSD licenses, at their choice. When you strip out the BSD notice, you misrepresent the license to the original work, which is at least bad, if not flat out illegal.
You just need to make it clear that there is work in the files that you are distributing that is not yours, and that the holder of the copyright on that work license it under the terms in question. You don't need to distribute the original files.
Really, what the terms force you to do is to communicate to the people who receive your derived work's source that, whatever license you chose for your work, there is an original work that it's based on, and that it's available under the terms of the BSD license; this is why you are required to reproduce those terms. You don't have to tell your recipients which parts are yours and which are from the original work--in fact, if you modify somebody else's code extensively, it's difficult in general to decide which parts belong to who.
I think you don't value Google's contribution to internet search fairly. If you just focus on CS and mathematics, you're just not being fair, IMO.
But still: yes, Google is immensely overvalued. They have exceedingly few lines of business that are actually profitable (selling ad space, the search appliance), specially considering all the stuff that they have developed that has fizzled and/or produced relatively little income. Yet despite this, people all over expect them as a matter of course to deliver revolutionary innovations in fields that they have absolutely no track record on.
Theo's email cited in TFA cites relevant US copyright law. I recommend you read it. (*resists urge to lash out against people who comment on this submission without actually reading through the linked materials, only to notice that parenthetical note yields to urge*)
The Berne Convention and the copyright law of other countries are substantially similar. The basic idea is that you can't slap your own copyright notice on something you don't own; you have to do some meaningful amount work before you can claim that you've got a derivative work.
Yes. And if you read what he says, he says that the law is being violated.