Linux's GPLv2 licence is routinely violated (2015) • Hacker News Board

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

#191

Post by tsimionescu » Sun, Feb 20, 2022, 6:27 PM UTC

Gaelan wrote: post #154

tsimionescu wrote: post #128

Earlier quoted context omitted.

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…

> I am not claiming that running on an OS is the same as dynamically linking with a library In pretty much any case, it literally is: you're dynamically linking against the libc. This is a little fuzzier on Linux (because, as you mention, multiple libc implementations exist), but on Windows, macOS, and the BSDs, the only supported way to make syscalls is to make calls into a dynamically linked libc, provided by the o…

Well, theoretically you could make a (pretty useless) Windows program that doesn't do any syscalls itself. It would mostly be limited to heating up the processor, but it would be a valid program. In fact, you could probably actually do some basic syscalls without linking to NTDLL by using very old DOS syscalls for things like printing to the console, perhaps with a compatibility mode.

Also, being pedantic, in Windows libc is a wrapper over Win32, which is partly itself a wrapper over NTDLL. For example, msvc's malloc() simply calls HeapAlloc() or some variant thereof.

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

#192

Post by fsckboy » Mon, Feb 21, 2022, 2:47 AM UTC

rtpg wrote: post #67

Dylan16807 wrote: post #59

Earlier quoted context omitted.

> writing the software and integrating it into their products That's distribution, not use. (Except for when they're testing it or running local instances, and GPL doesn't get in the way of those things.) > Not to say GPL is bad, just that the VAT-style "only the last user counts" view of things is playing with semantics. By my understanding VAT doesn't work that way. Once a user pays VAT and starts using something,…

the VAT reference is that businesses don't pay VAT, only end consumers do. If you buy something and then integrate it into another product, then re-sell it you are exempt from VAT (details are more complicated, but it's _not_ like sales tax)

if you resell items, then you don't pay sales tax on them, so that part fits your description

VAT is not a sales tax because it's an income tax. That's what value added means, profit, income, the measure of the value you've added. Wether it gets hidden in the price or tacked on to a fake price makes little difference.

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

#193

Post by butlerm » Mon, Feb 21, 2022, 9:24 AM UTC

hnfong wrote: post #179

butlerm wrote: post #176

Earlier quoted context omitted.

Static linking creates a work that incorporates another work, in a way that does not appear to be mere aggregation and which usually requires at least a modicum of creativity. I imagine end users might statically link things together all day long without getting in much trouble - 17 USC 117 practically exempts that sort of thing. But it seems like it would be pretty dangerous ground for a vendor to distribute a stati…

Might be a bit hair splitting (probably appropriate to do so in this context?), but the GPLv2 states that "mere aggregation of another work... does not bring the other work under the scope of this License". But it technically does not imply the converse is true, i.e. if it is not mere aggregation it must be a derivative work. Since "derivative work" does not seem to be clearly defined in GPLv2, I would say it's hard…

I think that is correct. "Derivative work" is a legal term in copyright law that the framers of the GPL were wise not to try to redefine in the text of the license. They seem to like the term "derived work", but do not define that either, probably for the same reason.

According to the U.S. Copyright Office, copyright only subsists in derivative works as such to the extent they show some independent creativity, which suggests that a non-mere-aggregation that is a mechanical or near mechanical combination of works may not be a legally protected derivative work at all, i.e. may be the sort of work you can prepare without permission from the copyright holder under Section 106 or some other legal allowance. All assuming you own legitimate copies of or have a proper license for the component works, of course.

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

#194

Post by butlerm » Mon, Feb 21, 2022, 9:40 AM UTC

hnfong wrote: post #177

butlerm wrote: post #174

Earlier quoted context omitted.

The claims that have never been tested in court here aren't part of the license, but rather commentary as to what constitutes a derivative work under copyright law. If true, they would affect the legality of a number of practices that apparently do not require distributing or modifying licensed components or accepting the license at all. If loading and running a program with dynamically linked modules with incompatib…

> commentary as to what constitutes a derivative work under copyright law That's part of my point actually. The question of API copyrightability in the Java case has a lot to do with whether copying APIs/ABIs in the code (whether from headers or binary symbol tables) is copyright infringement. That said, it's unfortunately true that the conclusion is far from clear. (Sidenote -- given that the Supreme Court dodged th…

The Supreme Court conclusively established that Google's use of the APIs was fair use under copyright law, so they did not need to reach the underlying question. I believe it would be safe to conclude that API descriptions such as header files are protected by copyright even in their essentials, that fair use is necessary to reproduce them without a license, and that Congress would need to act to create a more reliable exemption for that sort of thing.

Court of Appeals decisions are only binding precedents within the corresponding regions, but they generally have persuasive force throughout the country until and unless the Supreme Court overrules them.

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

#195

Post by butlerm » Mon, Feb 21, 2022, 9:53 AM UTC

tsimionescu wrote: post #186

butlerm wrote: post #182

Earlier quoted context omitted.

Congress has changed the law since MAI v. Peak specifically to allow users to create copies and adaptations of software programs as an essential step in running or utilizing the program in question on a machine, or for limited archival purposes (i.e. backups). "Adaptation" is another name for "derivative work". So it would appear that it is not a copyright violation for an end user to load or dynamically link modules…

> Congress has changed the law since MAI v. Peak specifically to allow users to create copies and adaptations of software programs as an essential step in running or utilizing the program in question on a machine, or for limited archival purposes (i.e. backups). That's not entirely true, as far as I understand. The rule that a user may copy the work to RAM as a fair-use exemption already existed at the time of MAI v…

Thanks for the clarification on that. Either way, MAI v. Peak was kind of a disastrous decision that presumably could have easily been resolved under fair use for the service contractor. Congress included Section 107 for a reason, and no doubt part of the logic is that courts use it instead of creating a make work program for Congress on the subject.

That would seem to go double for whichever court decided that ephemeral copies in RAM are "fixed" in a "tangible medium of expression", in direct contradiction of the plain meaning of the term.

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

#196

Post by nix23 » Mon, Feb 21, 2022, 1:46 PM UTC

pabs3 wrote: post #105

nix23 wrote: post #26

Earlier quoted context omitted.

>BSD/MIT/ISC rules can be broken as well No one cares and no one has interests to break them for financial gain...not even lawyers (that's the good part)

My router violates the BSD license and they got financial gain from that.

How is your router violating the bsd-license?

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

#197

Post by georgyo » Mon, Feb 21, 2022, 3:10 PM UTC

coliveira wrote: post #158

georgyo wrote: post #89

Earlier quoted context omitted.

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…

For patents, you can do your implementation, but you need to pay for the privilege.

With copyrights, you can use the the exact work, but you need to pay for the privilege.

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

#198

Post by hnfong » Mon, Feb 21, 2022, 4:11 PM UTC

butlerm wrote: post #194

hnfong wrote: post #177

Earlier quoted context omitted.

> commentary as to what constitutes a derivative work under copyright law That's part of my point actually. The question of API copyrightability in the Java case has a lot to do with whether copying APIs/ABIs in the code (whether from headers or binary symbol tables) is copyright infringement. That said, it's unfortunately true that the conclusion is far from clear. (Sidenote -- given that the Supreme Court dodged th…

The Supreme Court conclusively established that Google's use of the APIs was fair use under copyright law, so they did not need to reach the underlying question. I believe it would be safe to conclude that API descriptions such as header files are protected by copyright even in their essentials, that fair use is necessary to reproduce them without a license, and that Congress would need to act to create a more reliab…

Pretty sure the way they teach law students is to first determine whether a work is covered by copyright before going on to fair use... :D

But then, it's the Supreme Court so nobody's gonna fail their papers...

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

#199

Post by rtpg » Tue, Feb 22, 2022, 12:40 AM UTC

fsckboy wrote: post #192

rtpg wrote: post #67

Earlier quoted context omitted.

the VAT reference is that businesses don't pay VAT, only end consumers do. If you buy something and then integrate it into another product, then re-sell it you are exempt from VAT (details are more complicated, but it's _not_ like sales tax)

if you resell items, then you don't pay sales tax on them, so that part fits your description VAT is not a sales tax because it's an income tax. That's what value added means, profit, income, the measure of the value you've added. Wether it gets hidden in the price or tacked on to a fake price makes little difference.

My understanding (based off a French accounting class from many many years ago) is that VAT also can be deducted for things like factory equipment or, say, desks for your office workers. It's not purely a "if it's for resale you don't pay sales tax" thing, but a "if this is used for business then you can get a deduction for your VAT" (with some exceptions).

I would not really describe VAT as an income tax. It's not a tax on an income, it's a tax on the value add. Income tax means something specific, and perhaps VAT is one degree away from that, but money is fungible! On a long enough time scale everything is inheritence tax and income tax I guess!

EDIT: rereading my original post at the start of this thread I now feel a bit silly saying the explanation is "VAT-like". Metaphors and similes are bad!

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

#200

Post by ladyanita22 » Tue, Feb 22, 2022, 4:09 PM UTC

kmeisthax wrote: post #50

Several of the points in this article are either flat-out wrong, or are correct but being read in the wrong context. > The licence is specifically GPLv2 and not “GPLv2 or later”, which is the default when applying the GPLv2 to software. Because Linus didn't and doesn't want the FSF relicensing Linux with new restrictions to fit their specific political goals[0], on the basis that adding any further restrictions to GP…

> No, quite the opposite. The Linux kernel project has claimed that kernel-mode symbols equivalent to the userspace ABI do not trip the GPL copyleft. That's why they have the GPL_ONLY DRM[2]; it's to distinguish between "running two programs in the same address space" and "merging a module into the same GPL program".

> Compaq didn't need to ship code that links with the IBM PC BIOS. They needed to ship code that replaced it entirely. There is a difference between replacing a program and linking with it.

Does the Linux license (with the GPL_only thing) allows you to replace open source modules with self-contained proprietary ones? For this I mean, let's say I want to build a proprietary network stack tailored for a given hardware I'm building. In such case, may I do it as long as I write it from scratch and run it in kernelspace as a module? If that's the case, why do kernel infringements occur? Why wouldn't everybody just reimplement in their own proprietary way whatever they have to? Why is closed source DRM/anticheat non-existent on Linux?