Hacker News Reader

Comment by fpgeek | Hacker News Reader

The github issue is amazing reading: https://github.com/rxrz/exfat-nofuse/issues/5

Normally you expect the copyright issues to be going in the other direction at gpl-violations, but kudos to them for noticing the problem and dealing with it quickly.

View on HN

> Basically, I did microsoft's job, they even owe me money for the time I've spent fixing the code to work on modern kernels and "throwing the GPL tag on it".

Wow.

Oh it gets better. I had to stop reading the thread on Phoronix out of fear that my brain would pop out of my head and leave for Portugal.

> People with emotional insecurities are trying to make up something depressing and horrible, point a finger and make lots of noise.

> For me it was just sad to see people complaining more often than being grateful or happy about something.

> I hope they will patent and release a license on negative thinking.[0]

Actually, the entire Phoronix thread is an amazing display of incompetence.

> Were do you assume "this source code is not under GPLv2"? Source of this please? This implementation (code) was on github, that says something to you? [1]

My mind is boggled so hard right now. This better be a conspiracy to discredit FOSS, otherwise I'm losing yet another piece of trust in human sanity.

[0] http://phoronix.com/forums/showthread.php?81642-Native-Linux... [1] http://phoronix.com/forums/showthread.php?81642-Native-Linux...

Wow, again. Also found:

> Now, putting it publicly on github qualifies as distribution, and therefore this code is also under the GPLv2 license implicitly. The act has already been done, and even if it does contain copyrighted bits by MS, the released version is now forever GPL.

so, yeah.. by this logic I can take my company's code, put it on GitHub under a GPL license and now everyone is free to use it under GPL. Doesn't matter that I don't have the right to license it..

I can sort of see is logic. If a company has a private stash of GPL code (say, Google is running a special fork of Linux), and it gets leaked, even though the leak was illegal the code might still be distributable under GPL.

But ... this code was never (afaik) legitimately GPL. Maybe Samsung had based it on GPL code (in which case, it's GPL), but we'd need some proof.

I would love to hear Stallman's take on this.

Here's a couple of related answers at the FSF GPL FAQ: http://www.gnu.org/licenses/gpl-faq.html#InternalDistributio...

Unfortunately, the question isn't really directly answered.

The most relevant passage seems to be, "If the [unwittingly distributed] version in question is unpublished and considered by a company to be its trade secret, then publishing it may be a violation of trade secret law, depending on other circumstances. The GPL does not change that."

So I think a court would rule that the original leak does not qualify as "distribution," and thus the company is not required to distribute the source and everyone distributing the leaked code is under copyright violation.

I was confused for a minute, then I remembered that you don't actually have to accept the GPL as a user, only as an intentional distributor. So while the base version is GPL, the private version isn't - by default it's a hybrid of GPL and proprietary code.

Actually, you have to accept the license as a user to be able to use it. However, you don't give up your right to keep any change for yourself if you don't distribute it in a different format (e.g. binary). Whatever you do in your house is your business. That's probably what you meant (in spirit so to speak), though law wise it is very different.

>Activities other than copying, distribution and modification are not covered by this License; they are outside its scope.

Why do you say you have to accept it?

Because running software involves making copies of that software. This requires you to have permission from the copyright holder. The GPL provides you with that permission, without adding any further conditions if the copies are just for the purpose of running the software.

http://www.digital-law-online.info/lpdi1.0/treatise20.html

It is now well-accepted that copyright protects computer programs and other digital information, whether they are in readable source code form or are an executable program that is intended to be understood only by a computer. Copies are made whenever the program is transferred from floppy disk to hard disk or is read into the computer’s memory for execution, and those copies will infringe the copyright of the computer program if they are not permitted by the copyright owner or by copyright law.

1. Even if that's the current legal theory it's bullshit, and I can run code without having copies in ram if I need to.

2. 'if they are permitted by copyright law' Fair Use is part of copyright law, has that specific defense been challenged when it comes to software execution and caches in RAM?

Regarding 1, I really don't believe you can run code without copies, or copies of a derivitave work, being made in the process.

That said, I can't read a book without reflections in my eyes...

I could run it directly out of a ramdisk, or maybe transfer it to an embedded processor that can run it directly off the flash chip without even a processor cache. There's a whole variety of ridiculous things I could do to avoid the ridiculous restriction.

In practice I would trust fair use and try to use such as a defense if sued.

> I could run it directly out of a ramdisk

Hard to do when you were shipped a DVD.

Because that's the legal side of this. You can compile and run GPL software on your computer without agreeing to the licensing terms, but you may not do it. It is very unlikely that someone will check your HDs for offending uses, and besides, any use of GPLed software assumes an implicit acceptance of its license, so if such a check is happening, you would have to be there to say: "I have it but I am not agreeing to the terms". That sounds a bit silly doesn't it?

That's how the rules have been made, and not complying could be assimilated to piracy.

Licenses grant freedoms to do things that you would not otherwise be allowed to do. Here, the GPL means that someone with copyright grants a freedom (the freedom to change and redistribute). But these freedoms can only be granted by people who are in a position to say who gets to exercise those rights.

That code might come with a GPL, but the license it grants is worthless, because the people who put it there have no connection to the copyright holders.

I think Wisty was responding to the part where the crazy man claimed the driver code was originally GPL internally before being changed.

What chalst probably meant is that the fact that the code was orginally GPL doesn't make any illegal obtention of the code legal. If you keep code you wrote on your computer with a GPL attached, and someone breaks in and find it, he won't have the right to use it afterwards thanks to GPL.

The GPL defines the rights offered to "clients" by the "vendor", and how these rights are granted. Illegal obtention of the code is not one of the granting processes defined by the GPL.

Therefore, putting the GPL on that code is meaningless, whether it were GPLed initially or not.

Isn't the GPL unilateral? If something is actually under the GPL, it shouldn't matter how it was obtained.

I don't think so, no. The GPL talks about a lot of transference of rights that take place during the act of distribution. If one does not have those rights to start with (the code was stolen, for instance) then it wouldn't apply.

Actually it matters.

Original code is GPL. I make modifications. These modifications are mine. Someone stealing my modifications is performing IP theft.

Now, GPL stipulates when I have to release my modifications. I only have to release my patches when GPLv2: I distribute software to clients/users. In GPLv3 it is when clients/users use my software (aka: webservices).

That's a completely different issue from what I'm talking about. You're talking about modifications that aren't yet/ever released under the GPL. I'm asking what happens if it's already completely GPL, including modifications, and it gets leaked.

Nothing happens. The GPL is built on top of copyright. You cannot license code that you do not have copyright permission for under the GPL.

Wow, this conversation sure is going in circles.

>I think Wisty was responding to the part where the crazy man claimed the driver code was originally GPL internally before being changed.

As in, IF Samsung put version 1 under GPL, owning the copyright, on purpose, but they were keeping it hidden and someone snuck it out, would it be okay to use?

No, because the person who snuck it out didn't have copyright permission from the company. Employees at most corporations don't own the code they write.

If this were legal, then it would also be legal for an outsider to hack Samsung and release their code under the GPL.

For licensing purposes, employees at companies do not have GPL code they are working on distributed to them by the company; they are part of the company.

In the hypothetical, Samsung intentionally put it under GPL. It is truly and legitimately licensed GPL, they're just not giving it to the public at this moment.

I don't know how I can make it clearer.

Something doesn't truly become GPL until it is distributed, because the license is a contract between licensor and licensee. Only someone with copyright permission can distribute code. Samsung didn't grant copyright permission to its employees, nor did it distribute GPL code to its employees. (In the hypothetical.)

So if I write a piece of software, burn it onto a hundred disks, put the disks in a nicely labeled basket for people to take, go to lunch, and people grab some, at what point does the license kick in?

I find strange your statement that there is no license despite every file header (hypothetically) claiming a license, and this being authorized by the entire management chain.

In the example with you putting the disks in a basket, you've intentionally made them available for people that are not your employees to take, so you're distributing them. The license kicks in when one is taken. It's analogous to downloading software.

The point is that for a license to be binding it has to be offered willingly. If I sign a blank check and you steal it from me, filling out $1000 and your name, although you'll be able to cash it, you have no right to the money, and I can undo the action.

The reason internal code is special is because the employees are not considered individuals to whom the company has distributed the software. Instead they are a part of the company.

I identified some relevant questions in the GPL FAQ. I believe that the first one and third one are what you care about.

https://www.gnu.org/licenses/gpl-faq.html#StolenCopy

https://www.gnu.org/licenses/gpl-faq.html#GPLRequireSourcePo...

https://www.gnu.org/licenses/gpl-faq.html#InternalDistributi...

https://www.gnu.org/licenses/gpl-faq.html#DistributeSubsidia...

I hope that helps. The FSF might respond if you write to them, it's worth a shot if you're still not convinced.

I just find it very strange that you can't apply a license-for-distribution without actually distributing. That means if I want to make sure a piece of software is GPL, and nobody can take that away in the future, I have to perform an action such as uploading to github or burning a single CD and giving it to a homeless person. And somehow giving away that first single copy is what makes the GPL kick in for all other copies in the world.

That's still not quite correct. A license is a contract. It requires two parties. Every time somebody gives GPL software to somebody else, the contract is entered into and the license is granted. Giving away the first copy only establishes a license between you and the first recipient. That recipient in turn grants a license to whoever he gives it to, if and when he chooses to do so. In particular, I cannot demand that the first recipient give me a copy of the source code unless I have been granted license to the software by him.

If you want to be really sure a piece of software is GPL, it would probably be alright to include a file in your source tree on your harddrive saying that whoever finds it is free to make a copy of the project for themselves and consider that you have distributed it to them. It's about the clear intent to distribute.

> Giving away the first copy only establishes a license between you and the first recipient. That recipient in turn grants a license to whoever he gives it to, if and when he chooses to do so.

That's a logical way of looking at it, but it disagrees with the GPL FAQ. Note especially "If the version has been released elsewhere, then the thief probably does have the right to make copies and redistribute them under the GPL".

Plus, you can have a unilateral contract that's open to anyone willing to accept the terms. You don't have to take any action per person to 'establish' it.

The thief "probably" has the right to make copies because the diff between what he stole and what is readily available elsewhere for free is nil, and he can go and grab a copy with almost zero effort. But if A gives a copy only to B (it has still been "released elsewhere"), B keeps his copy to himself, and the thief steals a copy from A, the thief does not have any rights simply because B has a copy. That would be absurd! The FSF lawyers use the word "probably" because they believe this is what the court would find if a thief distributed something like an unmodified version of Linux taken off of a stolen hard drive.

However I was wrong about contracts vs. licenses. Obviously I am not a lawyer.

http://www.groklaw.net/articlebasic.php?story=20031214210634...

Complete and utter horseshit. What you're talking about in the last sentence is the Affero GPL.

The GPL states and has always stated: If you modify code that is licensed as GPL and want to share it, the modified version needs to be GPL as well ("share alike").

But: Nobody forces you to release anything. If you don't release software, you don't need to make a license for it. If you want, you can license it as "Adam Zochowski Super License 5000". Frame it and put it on your wall, nobody gives a damn.

The GPL only governs and can only govern what happens when you actually do release (GPLv3 uses the term "convey") software to other people.

Sidenote: This post again confirms Stallmans rule of thumb that anybody using the utterly corrupt term "Intellectual Property" is either a moron or trying to mislead you.

I wonder why people are voting you down. The GP is definitely confusing GPL with AGPL.

GP wording was perhaps a bit awkward, but I don't think he meant that modifications had to be released unilaterally: He just indicated the difference between gplv2 and v3 in the situations when the licenses force the release of the modifications (ie. when distributing the program in non source form).

I believe he means that, once the copyright on the code runs out (in 120 or whatever ridiculous number of years the copyright monopoly was able to extend further), that version, which is illegal right now (and not legally under the GPL) will automatically become GPL'd.

This is an interesting take on copyright and licensing.

Reply on news.ycombinator.com