Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Linux's GPLv2 licence is routinely violated (2015)

devever.net

Post reply on HN
Print view View original

121–130 of 202 posts

Re: Linux's GPLv2 licence is routinely violated (2015)

#121

Post by Borealid »

ISO-morphism wrote: post #98
Borealid wrote: post #39

Earlier quoted context omitted.

I think there's some confusion in the above post. The relevant "derived work" definition for the purpose of linking isn't one stemming from copyright law, it's the one in the GPL itself. This is because the GPL defines what licensees are permitted to do with the licensed work, and it imposes restrictions on those rights. Copyright law by default lets you make very little use of the software - the GPL broadens that ou…

I think there's a misunderstanding here on what "use" means. > You could ignore the GPL and static-link a piece of software to a GPLed library, but if you did so, you'd legally need a different right to use the GPLed library, because you hadn't complied with its license terms and so only have your minimal rights to use someone else's intellectual property. Before we get into anything particular to the GPL, Galoob v.…

Everything you said is correct. I was not sufficiently careful in my... uh... use... of the word "use".

I meant "use" in the sense of "allow to form part of your whole work which you then distribute". Anything that happens on your own local machine and never leaves isn't the intended target of the GPL.

Re: Linux's GPLv2 licence is routinely violated (2015)

#122

Post by michaelmrose »

zarzavat wrote: post #118
Borealid wrote: post #76

Earlier quoted context omitted.

The relevant clauses read as follows: b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License. [...] These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can…

Courts are supposed to balance the bargaining power of the parties when considering the enforceability of terms. So for the GPL which is a boilerplate clickwrap contract where the parties have not necessarily even communicated before, a court may take a narrow interpretation of such terms. So the unsatisfying answer (to programmers) may be: it depends who you are.

It's not at least in the US a contract at all its a license. As the text says you don't have to agree to it but have no right to distribute said software if you don't agree and no you can't ask a judge for better terms.

Re: Linux's GPLv2 licence is routinely violated (2015)

#123

Post by tkw01536 »

jarpadat wrote: post #94
Borealid wrote: post #73

Earlier quoted context omitted.

I find it difficult to imagine a situation in which you could use the library without copying the library. How would you obtain the code to execute it? Someone must distribute it to you, and/or you must copy it. Those are the governed actions. EDIT: note that the GPL broadly doesn't impose restrictions on interacting with a program over a network ("using" a server), which is why the AGPL exists.

So, one answer (not necessarily the right one!) to how you can use software without copying it is that "the license says you can". In other words the license definitely views "running" and "copying" as distinct, regardless of you, me, or federal law. The other, maybe more familiar idea to lawyers, and maybe more plausible to you, is the one in MAI v. Peak, that running is copying by definition. (The argument is based…

In its’ very basic form a license has two distinct purposes:

a) limit whatever rights are granted by copyright law b) grant new rights that go beyond copyright law

> Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted

I don’t see a contradiction here.

The first part states that anything being granted by this license (purpose b) only applies to copying and distributing.

The second part simply states that no new restrictions (in the sense of purpose a) are made.

To be more explicit, the paragraph should be read as:

> The act of running the Program is not additionally restricted

Re: Linux's GPLv2 licence is routinely violated (2015)

#124

Post by simion314 »

assttoasstmgr wrote: post #20
nix23 wrote: post #12

Earlier quoted context omitted.

GPL is a nice idea...but you need lawyers lots of money and time. I hate lawyers and i don't have time for that bs nor do it have/want money to spend for lawyers, i was a diehard gpl evangelist in my 20`s, now it's BSD/MIT/ISC (and not diehard anymore...more like IDGAS)...and i really don't care for religion anymore...just quality and real freedom.

My company does whatever it can to comply with GPL requirements and it is nothing short of a goddamn nightmare. Nothing like sitting in a conference room for hours with lawyers while we poured through the licenses for 250 different packages; we would have been better off just burning the money. My conclusions are that most Linux distributions are an amalgam of software packages with licenses that are fundamentally in…

>Ultimately, we concluded we are moving to BSD on future projects where possible. Not because of some overwhelming need to keep source closed from customers, rather just to avoid the mess.

It makes sense that developers or slave owners want different freedoms then users or slaves. If your code was BSD or GPL do you still need a layer ? If you are not a slave owner you are not affected by the fact there is a law that does not allow owning slaves.

Re: Linux's GPLv2 licence is routinely violated (2015)

#125

Post by gonzo »

georgyo wrote: post #89
snickerbockers wrote: post #84

Is there any legal significance to "clean-room" reverse-engineering? I hear this term a lot from engineers who seem to think that they're entitled to claim ownership over distinct works created by people who read their source code (and in extreme cases their documentation or even disassembled binaries), but that's not how any other form of intellectual property works.

IANAL This is the difference between patents and copyright. A mechanism that is patented can not be implemented by anyone. It doesn't matter if they did or did not read your source, they cannot implement it. It doesn't even matter if they knew the patent existed or that even that it had been done before. Copyright is a lot more forgiving. It only covers the exact implementation. Doing a clean room implemation means y…

That’s not how patents work.

Re: Linux's GPLv2 licence is routinely violated (2015)

#126

Post by pjmlp »

michaelmrose wrote: post #122
zarzavat wrote: post #118

Earlier quoted context omitted.

Courts are supposed to balance the bargaining power of the parties when considering the enforceability of terms. So for the GPL which is a boilerplate clickwrap contract where the parties have not necessarily even communicated before, a court may take a narrow interpretation of such terms. So the unsatisfying answer (to programmers) may be: it depends who you are.

It's not at least in the US a contract at all its a license. As the text says you don't have to agree to it but have no right to distribute said software if you don't agree and no you can't ask a judge for better terms.

In Europe unless the law has changed, in most jurisdictions the typical US style licenses are considered void unless the consumer of such software has been given and agreed to the license before acquisition.

That is why most ship wrapped software started putting a reference to the license on the box.

Re: Linux's GPLv2 licence is routinely violated (2015)

#127

Post by throwawaylinux »

pjmlp wrote: post #126
michaelmrose wrote: post #122

Earlier quoted context omitted.

It's not at least in the US a contract at all its a license. As the text says you don't have to agree to it but have no right to distribute said software if you don't agree and no you can't ask a judge for better terms.

In Europe unless the law has changed, in most jurisdictions the typical US style licenses are considered void unless the consumer of such software has been given and agreed to the license before acquisition. That is why most ship wrapped software started putting a reference to the license on the box.

If the license is void then copyright law applies.

Re: Linux's GPLv2 licence is routinely violated (2015)

#128

Post by tsimionescu »

Borealid wrote: post #39
tsimionescu wrote: post #25

> The userspace ABI is stable and intended not to cause anything targeting it to become a derived work. But once again, this seems to be a way in which the kernel project seems to think it has the power to interpret the GPL. In their minds, targeting the userspace ABI doesn't make a derived work, but writing a module does, unless it only targets GPL_ONLY symbols, in which case for some reason it isn't. I think this w…

I think there's some confusion in the above post. The relevant "derived work" definition for the purpose of linking isn't one stemming from copyright law, it's the one in the GPL itself. This is because the GPL defines what licensees are permitted to do with the licensed work, and it imposes restrictions on those rights. Copyright law by default lets you make very little use of the software - the GPL broadens that ou…

Let's imagine the following scenario. I am creating and distributing a program that, when run on a user's system, will dynamically link with a copy of libc. I am not distributing this copy, but my program will fail to run if no such copy exists. Furthermore, I tell my customers that the only lib C I officially support is FictitiousLibC, released under the full GPL (not the LGPL like the real GNU Libc).

The question is now: am I breaking any laws by delivering my program this way? According to the FSF, which believes an explicit linking exception to the GPL is needed - the LGPL - for such uses, then YES, I am breaking the copyright of FictitiousLibC, and so would need a license from FicititiousLibC to be allowed to distribute my code. But, if copyright law actually says "no, your program is NOT a derived work of FictitiousLibC", then the license is not needed. I don't know if this has ever been tested in court, but nothing would stop Congress from explicitly legislating this way and basically removing the difference between GPL and LGPL.

For an example of where this sort of logic applies is building programs to be run on an OS. If I build a program that only runs on Windows, but don't include any of Microsoft's code in my program, it is well established that my program is not a derived work of the Windows OS, so I can freely distribute it under any terms I want (not that Windows is special, the same is true of Linux or iOS or any other OS). I am not claiming that running on an OS is the same as dynamically linking with a library, just that in general something can be related to a copyrighted work without becoming a derived work by virtue of copyright law, not license terms.

Edit: note that if I want to distribute a bundle of my program + FictitiousLibC, that may infringe on FLC's license even if my program alone wouldn't, as FLC's license's terms of what constitutes a derived work could, in that case, go beyond copyright law definitions, and I would obviously need that license to be allowed to re-distribute FLC.

Re: Linux's GPLv2 licence is routinely violated (2015)

#129

Post by belorn »

tsimionescu wrote: post #128
Borealid wrote: post #39

Earlier quoted context omitted.

I think there's some confusion in the above post. The relevant "derived work" definition for the purpose of linking isn't one stemming from copyright law, it's the one in the GPL itself. This is because the GPL defines what licensees are permitted to do with the licensed work, and it imposes restrictions on those rights. Copyright law by default lets you make very little use of the software - the GPL broadens that ou…

Let's imagine the following scenario. I am creating and distributing a program that, when run on a user's system, will dynamically link with a copy of libc. I am not distributing this copy, but my program will fail to run if no such copy exists. Furthermore, I tell my customers that the only lib C I officially support is FictitiousLibC, released under the full GPL (not the LGPL like the real GNU Libc). The question i…

Let's remove the license and just look at what copyright says in order to understand your example. You have some copyrighted work and you want to make copies, and you do not have a copyright license from the author of said work. It this legal? Answer is mostly going to be no, especially if you are going to sell the copies. Commercially exploiting someones else's copyrighted work without a license is very often illegal.

Now let's say you do have a license, and the license is GPL. The license says that you have permission as long you do not combine the library using a linker with other programs that is not GPL. Do you have permission to copy the work if you disregard the conditions of the license? Likely answer: no.

Copyright do have a few exceptions. If you bought a copyrighted work like say Microsoft Word, there is an exception that says that you can also make a copy in memory in order to run it. Copyright law in some countries needed to be updated to do this, while in other places it occurred through precedents, but it was something that needed to occur. There might also exist a similar exception for running the program under Microsoft Windows, including using libraries called DLLs. Exceptions to the exclusivity of copyright is sometimes needed in order for the world to make sense.

The legal question then that remain is if the situation with FictitiousLibC has such an exception. That is less certain. There might be a copyright law exception, or there might not exist. What do you think?

Re: Linux's GPLv2 licence is routinely violated (2015)

#130

Post by phendrenad2 »

josephcsible wrote: post #117
jeremyjh wrote: post #110

Earlier quoted context omitted.

Ok then we just won't use Nvidia cards with Linux.

My point is that there's no reason that Nvidia can't release the source code of their Linux drivers. They just choose not to.

If there's no reason, why would they choose not to? The simplest explanation is that there's some reason they choose not to.
Post reply on HN