The freedoms he is talking about are about your customers being allowed to do as they want with the software they bought from you. You releasing non-free software takes away those freedoms by definition.
Ok... so after buying this non-free software, what can I suddenly not do that I could do before buying it?
He has been angry at Linus for years for not pursuing NVidia to release source for their graphics module. It arguably is a derivative work of the GPL kernel.
I am highly skeptical of this. Is a research article a "derivative work" of all the earlier papers it cites?
My understanding of the GPL is this: if I want a program's source code to be freely available as well as any modifications to said source, I use the GPL. People are not allowed to extend my stuff without GPLing the result. They also cannot include it in their projects without GPLing the whole thing. Thats all I understand about it.
It becomes slightly more complicated if you consider that not everything that uses your program is necessarily a derivative work (The obvious case is including it as a separate executable, and communicating with it over stdin/stdout.).
Now, GPL 2 vs. GPL 3: what exactly changes from a programmer's point of view? Googling this is hopeless, its full of blogs FUDing around, and I dont think the FSF is an unbiased source. So, anyone here with a clue?
GPLv2 says that dependencies have to be redistributable and modifiable. GPLv3 says that dependencies must be provided under the complete terms of GPLv3. Since one of the terms of GPLv3 is that "additional permissions" can be removed, I would expect this to be problematic WRT dependencies under more permissive licenses that don't let themselves be removed.
GPLv3 talks more about patents. If you happen to have any relevant patents, GPLv3 makes more demands about licensing them to everyone for free.
There are a few things you can't use GPLv3 code for, although you probably only care if you want to make something that's locked down.
I think there are still some (at least one--me) who agree with Stallman's original "four freedoms", but believe the GPLv2 is sufficient for protecting those freedoms, and the GPLv3 adds too much complexity to be worth it.
It doesn't just add complexity, it adds restrictions that I think can semi-reasonably be considered to abridge some of those freedoms.
I personally contribute to GPLv2 projects (and write my own), but I will not contribute to GPLv3 projects. I don't know if I'm the only one in the world--certainly on Slashdot I appear to be the minority--but I do think there are others.
You probably mean the "accidental" GPL users, chief amongst them Linus, who never really bothered to understand the ideology behind the GPL and simply used it out of "convenience".
The ideology expressed in the GPL is not quite the same as the ideology of the FSF.
Speaking of Linus, for an example of the consequences of his short-sighted, "technocratic" approach, witness the the Bitkeeper fiasco, amongst many other such examples.
So what exactly was lost in this "fiasco"? If you pick a crooked landlord because the rent is $100/mo less, and they make up a reason to keep your $500 deposit at the end of the year, what overall have you "lost" here by not choosing the other landlord?
Again, that depends on if you actually subscribe to GPL ideology, or are merely using GPL because it is "convenient" or for some other such mis-guided reason.
How is it "misguided" to use a license that you like the effects of, simply because the license's authors happen to be raving lunatics?
As to how many people are in this camp, I cannot even try to estimate. I would venture however to say that many of them do instinctively understand that GPL protects their work from being simply appropriated by some business for commercial use and that is what keeps them away from BSD.
I'm starting to think that any such instincts are somewhat misguided. If you don't actively look for your code, infringements probably won't get noticed. If you actively maintain your code, companies have an incentive to minimize the size of their patchset.
Anyone who beleives in the ideology behind the GNU project would have no problems at all adopting the GPLv3. It adds additional copyleft restrictions to promote the freedom to hack - in addition to making a few important clarifications. If you feel uneasy with GPLv3, ask yourself if your ideals match those of the GNU project. If they don't, there are other copyleft and non-copyleft licenses available - including the BSD licence.
That ideology is approximately "not sharing is wrong". Their licenses are approximately "we refuse to share with you, unless you share with others". It is very possible to agree with the former and have serious problems with the latter.
We need a new license -- the Linux license -- that has GPLv2 features and perhaps the patent provision but not the anti-tivoization clause.
Eh, I actually don't think the patent provisions are very good. They seem stupidly specific to one particular case that annoyed people, rather than cleanly holding a particular position. Which is really rather odd, since RMS and the FSF like to diss "pragmatism" in favor of "idealism".
I've read and I understand the differences between the three versions of the license, and I really don't see how that is going to really affect me. I've been using Red Hat/Fedora and Gentoo since 2000, and I can't think of a single instance of a software license ever really affecting me. Maybe its because I'm not a software developer, but does the regular user really care about any of this?
Not directly. Indirectly, however, it will affect what is available to you.
Parts of GPLv3 were specifically designed to attack Tivo -- so if you have a TV it should eventually affect what choices you have for that sort of product. It will probably (indirectly) affect you even if you don't have a TV, since it affects anyone who wants to sell you something that's locked down (even things that *have* to be locked down, say to guarantee standards compliance for something).
There are some things that GPLv3 software simply cannot be used for. So if lots of people move to GPLv3, there are some things that become harder to make, and therefore less available to you.
Better read that section again, because it's referring to its own code. If you release code as GPL3 with additional permissions, it includes an offer to "relicense" itself without those permissions. It will have no effect on any other code in the project.
GPLv3 demands that the "corresponding source" (basically, source code of all dependencies) be made available under GPLv3. How would doing this not turn those dependencies into the "covered works" that section 7 talks about?
/me wonders how the BSD licence not allowing itself to be removed would work with a GPLv3 project trying to use BSD code (since FSF thinks this is OK). One of the terms of GPLv3 is that additional permissions can be removed...
I think the problem this is trying to solve isn't *network* bandwidth, it's *upload* bandwidth. As in, how to get more people to share a given file so that people can download it better. The way to solve the *network* bandwidth problem is to find a way to make ISPs not become legal targets by caching data (which I would assume means that it has to be impossible to find out what exactly it is that they're caching).
GPL does not put restrictions on use of code. It puts restrictions on distribution of code, created to ensure the lack of restrictions on use of code.
There are many ways to use code which involve distributing it. Such as using it as a component of some other system you produce. This form of use is very much restricted by GPLv3.
One things is for sure - they all rely on proprietary Microsoft produts (.Net, sharepoint, SQL server, etc) to run. They're not particularly useful to the Open Source community, just the Microsoft community. (In Debian, they would sit in the non-free repositorty).
I thought "free, but depends on something non-free" was what contrib/ was for?
So why the hypocrisy with respect to pornography and other sexual or erotic descriptions.
Religious institutions. You're much less likely to get excommunicated from your church for beating your wife vs. cheating on her. I never understood the logic myself. Why are we so much more lenient in censoring violence--an act that inflicts pain and can end human lives vs. sex--acts that bring pleasure and can create human life?
Anything fun is sinful, because it tempts you to take time away from meditating upon the Word of the Lord.
Or maybe the monks/priests/whoever who made these policies in the first place were just jealous or something.
I don't understand how this is considered censorship. Censorship is when the government tries and silence speech. Walmart, as large as it is, isn't the government.
No.
Censorship is when one person/group/whatever prevents another from saying something, or from seeing/hearing something; the government does not have to be involved. The issue with government and censorship is that our government is generally not permitted to engage in censorship, and that much harder for someone who isn't the government to censor effectively.
You're overstating the GPL's claimed definition. GPLv2 simply references copyright law's definition of derived works; only GPLv3 specifically references shared-library linkages as inferring derivative status, and even then only when the 3rd-party code is written to the interface provided by the GPLed code ("shared libraries and dynamically linked subprograms that the work is specifically designed to require", such that use of a common, standardized interface not specific to some GPLed work is explicitly acceptable).
Hm, right. I guess I must have been thinking of the LGPL, which says "linking a "work that uses the Library" with the Library creates an executable that is a derivative of the Library".
It pretty much depends on what you understand by "derivated work". This is by far the biggest error in the GPLv2 - it's just not very clear on what it means.
It's not so much that the GPL isn't clear, as that GPL tried to use an insanely broad definition, and the limits that copyright law imposes on this definition aren't clear.
Maybe someone can help me understand something I'm unclear on: How does one go about creating proprietary software that runs on Linux (the kernel) without "violating the GPL?" Based on the blurb, and many similar stories I've read here over the years, it almost seems like the FOSS community expects every piece of software that is written to run on a Linux platform to be Open Source. If that's the case, why?
I have seen claims that the only reason Linux userland isn't subjected to GPL by the kernel is that Linus declared a linking exception for things that only talk to the kernel through the standard userland interfaces. Or in other words, some people consider anything written to run on the Linux kernel to be derived from the kernel.
Unfortunately, the FSF has seen fit to encourage such nonsense, with the "anything that links to this is derived" claims they put in the GPL.
So if I'm reading this right, anything that uses Linux as a bootloader is "derived from" Linux (because it depends on Linux, because nobody bothered to implement another bootloader) and must be GPLed? That seems very, very bogus, and I will be very annoyed if such nonsense is upheld. Just because something looked at your code funny once, does not automatically make that thing derived from your code.
Unlike copyright the patent concept is easy to defend. The benefit for progress of engineering and technological culture can be logically demonstrated - and unlike copyright the limits on duration are not totally insane.
I don't care whether it can be logically demonstrated. It's been around for quite a while, so the question isn't "logically, should this work", but rather "has this been observed to *actually* work". And I kinda get the impression that the answer to that is "it depends" and "somewhat, but not as well as we'd think".
Like all other drugs, caffeine loses its effect unless you keep increasing the dose. The stimulating effect of caffeine is vastly overestimated and doesn't last if you keep "using". If you don't believe this, don't consume caffeine in any form for half a year and then see what effect a single cup of coffee has on you after you've been weaned of caffeine. I would suggest that caffeine causes more accidents by making people think they can stay awake with coffee than it prevents by keeping people awake a little longer.
I haven't noticed that it does much at all, actually. For a while I had basically random coffee consumption of 0-4 cups/day... no noticable effect other than from the sugar to make it drinkable (same as from hot chocolate or a candy bar), and no withdrawl effects going from a couple days at 4 cups to 0 cups when I decided it was all very silly (maybe I just hadn't been drinking it long enough?). (And no, it wasn't decaf.)
Slashdot Mirror
...is that legally enforceable? Is it even legal at all?
Ok... so after buying this non-free software, what can I suddenly not do that I could do before buying it?
I am highly skeptical of this. Is a research article a "derivative work" of all the earlier papers it cites?
Yeah, it's not a gun... just a licence designed to force people to conform.
It becomes slightly more complicated if you consider that not everything that uses your program is necessarily a derivative work (The obvious case is including it as a separate executable, and communicating with it over stdin/stdout.).
Now, GPL 2 vs. GPL 3: what exactly changes from a programmer's point of view? Googling this is hopeless, its full of blogs FUDing around, and I dont think the FSF is an unbiased source. So, anyone here with a clue?GPLv2 says that dependencies have to be redistributable and modifiable. GPLv3 says that dependencies must be provided under the complete terms of GPLv3. Since one of the terms of GPLv3 is that "additional permissions" can be removed, I would expect this to be problematic WRT dependencies under more permissive licenses that don't let themselves be removed.
GPLv3 talks more about patents. If you happen to have any relevant patents, GPLv3 makes more demands about licensing them to everyone for free.
There are a few things you can't use GPLv3 code for, although you probably only care if you want to make something that's locked down.
It doesn't just add complexity, it adds restrictions that I think can semi-reasonably be considered to abridge some of those freedoms.
I personally contribute to GPLv2 projects (and write my own), but I will not contribute to GPLv3 projects. I don't know if I'm the only one in the world--certainly on Slashdot I appear to be the minority--but I do think there are others.Yeah, there's at least a few more.
The ideology expressed in the GPL is not quite the same as the ideology of the FSF.
So what exactly was lost in this "fiasco"? If you pick a crooked landlord because the rent is $100/mo less, and they make up a reason to keep your $500 deposit at the end of the year, what overall have you "lost" here by not choosing the other landlord?
How is it "misguided" to use a license that you like the effects of, simply because the license's authors happen to be raving lunatics?
I'm starting to think that any such instincts are somewhat misguided. If you don't actively look for your code, infringements probably won't get noticed. If you actively maintain your code, companies have an incentive to minimize the size of their patchset.
That ideology is approximately "not sharing is wrong". Their licenses are approximately "we refuse to share with you, unless you share with others". It is very possible to agree with the former and have serious problems with the latter.
Eh, I actually don't think the patent provisions are very good. They seem stupidly specific to one particular case that annoyed people, rather than cleanly holding a particular position. Which is really rather odd, since RMS and the FSF like to diss "pragmatism" in favor of "idealism".
Not directly. Indirectly, however, it will affect what is available to you.
Parts of GPLv3 were specifically designed to attack Tivo -- so if you have a TV it should eventually affect what choices you have for that sort of product. It will probably (indirectly) affect you even if you don't have a TV, since it affects anyone who wants to sell you something that's locked down (even things that *have* to be locked down, say to guarantee standards compliance for something).
There are some things that GPLv3 software simply cannot be used for. So if lots of people move to GPLv3, there are some things that become harder to make, and therefore less available to you.
You probably want the ap.google.com version instead.
GPLv3 demands that the "corresponding source" (basically, source code of all dependencies) be made available under GPLv3. How would doing this not turn those dependencies into the "covered works" that section 7 talks about?
/me wonders how the BSD licence not allowing itself to be removed would work with a GPLv3 project trying to use BSD code (since FSF thinks this is OK). One of the terms of GPLv3 is that additional permissions can be removed...
I think the problem this is trying to solve isn't *network* bandwidth, it's *upload* bandwidth. As in, how to get more people to share a given file so that people can download it better. The way to solve the *network* bandwidth problem is to find a way to make ISPs not become legal targets by caching data (which I would assume means that it has to be impossible to find out what exactly it is that they're caching).
There are many ways to use code which involve distributing it. Such as using it as a component of some other system you produce. This form of use is very much restricted by GPLv3.
I thought "free, but depends on something non-free" was what contrib/ was for?
Religious institutions. You're much less likely to get excommunicated from your church for beating your wife vs. cheating on her. I never understood the logic myself. Why are we so much more lenient in censoring violence--an act that inflicts pain and can end human lives vs. sex--acts that bring pleasure and can create human life?
Anything fun is sinful, because it tempts you to take time away from meditating upon the Word of the Lord.
Or maybe the monks/priests/whoever who made these policies in the first place were just jealous or something.
No.
Censorship is when one person/group/whatever prevents another from saying something, or from seeing/hearing something; the government does not have to be involved. The issue with government and censorship is that our government is generally not permitted to engage in censorship, and that much harder for someone who isn't the government to censor effectively.
Hm, right. I guess I must have been thinking of the LGPL, which says "linking a "work that uses the Library" with the Library creates an executable that is a derivative of the Library".
That is simply incorrect.
It is also the basis for the whole article, never mind how silly it is.
It's not so much that the GPL isn't clear, as that GPL tried to use an insanely broad definition, and the limits that copyright law imposes on this definition aren't clear.
I have seen claims that the only reason Linux userland isn't subjected to GPL by the kernel is that Linus declared a linking exception for things that only talk to the kernel through the standard userland interfaces. Or in other words, some people consider anything written to run on the Linux kernel to be derived from the kernel.
Unfortunately, the FSF has seen fit to encourage such nonsense, with the "anything that links to this is derived" claims they put in the GPL.
So if I'm reading this right, anything that uses Linux as a bootloader is "derived from" Linux (because it depends on Linux, because nobody bothered to implement another bootloader) and must be GPLed? That seems very, very bogus, and I will be very annoyed if such nonsense is upheld. Just because something looked at your code funny once, does not automatically make that thing derived from your code.
I don't care whether it can be logically demonstrated. It's been around for quite a while, so the question isn't "logically, should this work", but rather "has this been observed to *actually* work". And I kinda get the impression that the answer to that is "it depends" and "somewhat, but not as well as we'd think".
I haven't noticed that it does much at all, actually. For a while I had basically random coffee consumption of 0-4 cups/day... no noticable effect other than from the sugar to make it drinkable (same as from hot chocolate or a candy bar), and no withdrawl effects going from a couple days at 4 cups to 0 cups when I decided it was all very silly (maybe I just hadn't been drinking it long enough?). (And no, it wasn't decaf.)