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

111–120 of 202 posts

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

#111

Post by jeremyjh »

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.

I think it is just as much about the corporation protecting itself; they don't trust their engineers not to just copy parts of code.

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

#112

Post by yuubi »

snnn wrote: post #101

Firmware of TP-link routers is licensed under GPL. But would anyone of you get the source code? They say they provide machine-readable source code. I don't quite understand what it means.

GPL 2 says "machine-readable", which I read as meaning on a tape or ftp site or something in a useful form, not handwritten on paper or similarly useless form. Also, I doubt obfuscated code would comply with the "preferred form" part.

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

#113

Post by jeremyjh »

phendrenad2 wrote: post #68
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…

> Linus is a copyright holder...So...the court will listen to Linus to determine where the boundary is This seems completely insane. If I become a programmer on Monday, and download the Linux source on Tuesday, and read the license file on Wednesday, and publish code on Thursday, and Linus sues me on Friday, am I going to be penalized by the judge because I didn't read random LKML posts by Linus? Surely the license i…

The insane words are ones that you made up. The documentation is literally in the source code. If courts need to understand precisely what it means as it applies to a particular case, yes they can call Linus as an expert witness.

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

#114

Post by rizky2410 »

charcircuit wrote: post #69
fsckboy wrote: post #23

> It is rather ironic that in some ways, one has more freedom in how one licences a kernel driver for Windows than the freedom one supposedly has in writing a kernel module for Linux. it is not ironic at all: freedom in Free Software is the freedom of USERS of the software to be able to see and modify the source. This is a freedom granted by the writers of the software, and copyleft means that this right is ensured t…

>In this same way, the GPL and like licenses are the most permissive licenses from the perspective of every eventual user No, take for example the real world use case of distributing ZFS on Linux in binary form. Users are unable to distribute it due to conflicting copyleft licenses. Copyleft licenses limits people by creating their own exclusive ecosystems which can't be mixed.

Ubuntu supports ZFS since 20.04.

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

#115

Post by kmeisthax »

phendrenad2 wrote: post #68
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…

> Linus is a copyright holder...So...the court will listen to Linus to determine where the boundary is This seems completely insane. If I become a programmer on Monday, and download the Linux source on Tuesday, and read the license file on Wednesday, and publish code on Thursday, and Linus sues me on Friday, am I going to be penalized by the judge because I didn't read random LKML posts by Linus? Surely the license i…

In this particular case, we're talking about a license clarification that gives you more rights than you otherwise could. So, if you hadn't read Linus's interpretation of his own license, you would think that you always had to GPL your LKMs. So there's no point in which Linus could actually sue you for not reading his LKML rants - they strictly increase what you are allowed to do under the license.

If this was the opposite case - say, Linus saying that the GPL required something that the text of the license said it didn't - then a judge would probably be inclined to not listen to Linus.

There's two fallacies engineers commonly get tricked by when talking about the law:

- "Code is law"[0]: the belief that legal protections, contractual clauses, and so on can be replaced with technical protection measures and "smart contracts".

- "Law is code": the belief that judges are merely human interpreters for an English-language-shaped virtual machine that contracts and laws are written in.

These fallacies are dual to one another; the root cause is a fundamental misunderstanding of how laws work. Judges are not merely tasked with executing a computer program in their heads; they are allowed and expected to actively comprehend and rewrite the law based on the facts at hand. "Legislating from the bench" is a deliberate feature, especially in "common law" jurisdictions where prior rulings are allowed to be treated as good as actual passed-by-Congress laws.

A good example of how this works would be Oracle v. Google. Strictly speaking, as much as we hate Oracle, they had the most meritorious argument, and they were trying to enforce the GPL. If you take copyright as a given on software, then that copyright can be applied to the thinnest sliver of creativity present in that software. Yes, even lists of function names and signatures alone should be copyrightable, and the GPL copyleft should trip if you copied them. However, the entire software industry did not actually consider these to be copyrightable, and acted as if they weren't. In fact, even Oracle's own cloud division was busy doing to Amazon what they were suing Google for doing to them. When SCOTUS saw this, they actively decided to change the law in favor of Google because it was the interpretation that most actual copyright holders in the industry had been operating on.

[0] With apologies to Cory Doctorow

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

#116

Post by kmeisthax »

shagie wrote: post #100
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…

Some more on Linus and his feelings on GPLv3 from 2014 - https://youtu.be/PaKIZ7gJlRU

I remember watching this years ago and getting really pissed at Linus. At the time I didn't actually get what he was getting at, and just assumed he was just standing in the way of Stallman grabbing half the computer industry by the balls/ovaries and demanding they never lock down a CPU ever again.

I rewatched this about a month ago and it has aged like a fine wine. As much as I personally like GPLv3, Linus was in the right to reject it. GPL has a no-further-restrictions clause for a reason, and the license upgrade mechanism should not be used to circumvent the intent of that clause just to make a stronger copyleft. In fact, I'm starting to sour on the concept of supra-GPL-strength copylefts in general, because going any stronger than GPLv3 requires interfering with freedom zero in some way.[0] This is supposed to be a software commons, after all.

[0] AGPL is designed to force users to retain what may be an antifeature to them. SSPL tries to force users to embed said antifeature into anything remotely related to the program. "Ethical Source" licenses try to enforce moral restrictions with economic rights.

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

#117

Post by josephcsible »

jeremyjh wrote: post #110
josephcsible wrote: post #81

> You have use cases for binary kernel modules which need to be accommodated I can't think of a single reason that binary-only kernel modules would ever need to be distributed. There are some cases when out-of-tree modules make sense, but in all of these cases, they could be distributed with source code, in compliance with the GPL.

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.

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

#118

Post by zarzavat »

Borealid wrote: post #76
Dylan16807 wrote: post #64

Earlier quoted context omitted.

I don't see anywhere in the GPL that defines deriving, except the part where it mentions "derivative work under copyright law".

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.

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

#119

Post by kuang_eleven »

snnn wrote: post #101

Firmware of TP-link routers is licensed under GPL. But would anyone of you get the source code? They say they provide machine-readable source code. I don't quite understand what it means.

Like this page?

https://www.tp-link.com/us/support/gpl-code/

It's not even slightly hidden, pretty easy to get the code.

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

#120

Post by tzs »

e-dt wrote: post #102
ISO-morphism wrote: post #98

Earlier quoted context omitted.

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.…

To those clicking on link [1] and finding it broken: you need to add a full stop at the end, as Hacker News's software has seen fit to exclude it from the link.

One can work around HN’s mangling of dots in URLs by writing them as %2e. E.g., https://en.wikipedia.org/wiki/Lewis_Galoob_Toys,_Inc._v._Nin....
Post reply on HN