Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
Software

The Future of GPLv3 Hangs In the Balance (sfconservancy.org) 66

Posted by msmash from the closer-look dept.
New submitter jms00 writes: A years-long legal battle has quietly escalated into what could become the defining moment for the future of GPLv3, with implications that could reshape software freedom as we know it.

At issue is whether licensors have the power to impose 'further restrictions' on open-source software, potentially undermining the explicit rights granted to users and developers under AGPLv3, GPLv3, and LGPLv3.

The outcome of this case, now before the U.S. Court of Appeals for the Ninth Circuit, could set a dangerous precedent, limiting the ability to remove proprietary restrictions from copyleft-licensed software.

With little public attention on the case, the Software Freedom Conservancy (SFC) has stepped up as a key voice in defense of user rights, filing a critical amicus brief to challenge the lower court's ruling and protect the principles of software freedom.

The Future of GPLv3 Hangs In the Balance More | Reply

The Future of GPLv3 Hangs In the Balance

Comments Filter:

  • Nawww (Score:5, Insightful)

    by Aighearach ( 97333 ) on Thursday February 13, 2025 @07:08AM (#65163173)

    The defining moment(s) of the future of GPL3 happened long ago, when it was rejected by most of the community in favor of either staying with GPL2 or switching to the Apache license.

    Kinda gross that instead of linking to the case or reporting on the case the link is to an amicus brief, which is completely irrelevant and unlikely to even be read by anybody on the court.

    Also you forgot to even say who the parties are, which is how cases are identified.

    • Re:Nawww (Score:4, Informative)

      by jmccue ( 834797 ) on Thursday February 13, 2025 @08:31AM (#65163289) Homepage

      The defining moment(s) of the future of GPL3 happened long ago, when it was rejected by most of the community in favor of either staying with GPL2

      The reason it was rejected was due to fortune 500 companies, they want to keep using free labor and keep their bits (enhancements) hidden from us. Just take a peak at Red Had/IBM and what they did to CentOS to see an example of not using the GPLv3. If Linux was on v3, CentOS would still be a thing.

      I worked at IBM, the owner of Red Hat, and when I was there, IBM forbade the use of any GPLv3 software on their systems, this includes people using a Linux RHEL Workstation instead of Windows (like I did).

      • Re: (Score:3)

        by tlhIngan ( 30335 )

        I worked at IBM, the owner of Red Hat, and when I was there, IBM forbade the use of any GPLv3 software on their systems, this includes people using a Linux RHEL Workstation instead of Windows (like I did).

        Most big companies did. They saw the GPLv3 and what happened with it. Many of the big ones suddenly implemented legal review of open source code - if you wanted to use open-source code in any product, that product needs to be vetted for its license.

        Internal only use of GPLv3 was restricted to make sure the

    • Re: (Score:1)

      by jms00 ( 10502445 )
      See my comment below - but I don't think you get how serious this is. The title should have been "all open source" not just GPLv3. This case isn’t just about GPLv3, I believe it threatens all open-source licenses. If upheld, it could create a legal loophole where licensors can attach restrictions to open-source software, overriding the rights explicitly granted in licenses like GPLv2, MIT, and Apache. The court’s misinterpretation of AGPLv3 could set a precedent allowing companies to undermine

  • Clickbait (Score:5, Informative)

    by vbdasc ( 146051 ) on Thursday February 13, 2025 @07:51AM (#65163209)

    I personally fail to see how this can affect the future of GPLv3. In this case, we have a developer who publishes software under a hacked GPLv3 (GPLv3 concatenated with some restrictions resulting in a license that is somewhat self-contradictory and hence on a precarious legal standing, but I digress). Whatever happens in the courtroom wouldn't affect the GPLv3 itself in any way.

    Even if the court decides that the non-free license the developer uses is not valid, this won't be a victory for free software. The developer will simply move to another non-free license which doesn't piggyback on GPLv3, eliminating the contradiction.

    • And you just know they're clickbait. You don't even need to read the fucking summary. The future of X hangs in the balance is one of them.

      It's because the headline doesn't really tell you anything useful. A proper headline written by a real journalist includes some actual information because the goal of a real journalist is to get information into people's heads.

      The goal of a clickbaiter is to keep you scrolling through ads as long as possible. So the less information in the headline and the less in
    • The developer is mixing their own code with AGPLv3 and licensing the whole thing under their own license. This isn't about whether you can distribute your own code that way, it's about whether you can remove clauses and redistribute AGPLv3 under the new license.

      • Re: (Score:2)

        by vbdasc ( 146051 )

        The developer is mixing their own code with AGPLv3 and licensing the whole thing under their own license[which is more restrictive than AGPLv3].

        Whoa, are you sure? I couldn't find such a claim in the linked article. If this is true, then it constitutes a clear and obvious violation of AGPL by the developer, and if the court stands with them, then it would indeed dismantle GPL and terminally cripple the free software movement.

        I hope it isn't true, for the sake of everyone on this planet.

        • Re: (Score:3)

          by unrtst ( 777550 )

          The developer is mixing their own code with AGPLv3 and licensing the whole thing under their own license[which is more restrictive than AGPLv3].

          Whoa, are you sure? I couldn't find such a claim in the linked article. If this is true, then it constitutes a clear and obvious violation of AGPL by the developer, and if the court stands with them, then it would indeed dismantle GPL and terminally cripple the free software movement.

          Would it not ALSO cripple all software licensing? If I can alter and amend a license on a copyrighted work and then redistribute it under that new license... WTF was the license doing in the first place? You can't add a restriction to something whose licensing condition is that you do not add additional restrictions without violating that license, and then you're back to plain copyright where you have zero license (that's kinda the point and strength of GPL).

          • It would cripple all open source licensing which is why SFC is involved! With commercial software, you have no right to further distribution. With open source licenses you do have a right to further distribution and the argument is being made that one can distribute a combined work under AGPLv3 but add additional restrictions as a *licensor* of other parts of the combined work. This would essentially turn all GPL code into LGPL.
        • From TFA: Neo4j further argued (and the lower court agreed) that since 103 did not mention licensors explicitly, then not only must licensors have the right to add “further restrictions” but this implicit right trumped the licensees' explicit right to remove such restrictions under 74.

          • Re: (Score:2)

            by vbdasc ( 146051 )

            But if Neo4j used some AGPL code, then they're licensees with respect to it. And 103 does mention licensees explicitly. They CAN'T add restrictions to the product that uses others' AGPL code. They can add their own proprietary code, but according to the AGPL, the whole result needs to be AGPL.

            I'll repeat my question: Are you sure that Neo4j uses someone else's AGPLv3 code? Because if they don't use such code, then they're within they right to use whatever license they choose to.

            • You don't have to take my word for it. Neo4j has an open source version and a paid commercial version. I don't know where the AGPLv3 comes in here since the open source version is GPLv3 not AGPLv3. And I don't know how they package their enterprise product. But I put the text below from their own repository for you. They are basically saying (as best I can tell, no being a lawyer) that if you execute a commercial license with them, you lose all of your GPL rights. https://github.com/neo4j/neo4j... [github.com] N

              • Re: (Score:3)

                by vbdasc ( 146051 )

                Thanks for the information.

                I wish only to comment that if the GPL code is theirs, then they can invalidate, add restrictions, add proprietary licenses etc. as they wish.

                Only if they use GPL code from someone else, then they become licensees and are disallowed to add restrictions according to the license.

                • Yes, they could dual license the code. That doesn't appear to be what they are doing. I presume that's because there are non-neo4j contributors to the code who haven't assigned copyright so they don't have permission to dual license. Instead, they are trying to achieve what is effectively dual licensing by using their license / contract related to the proprietary portions to have the customer waive their rights for the GPL portions. Again that's what I can tell since there is limited information availab

          • He used bloodhound to find this one

    • Re: (Score:1)

      by jms00 ( 10502445 )
      Imagine this: a large organization builds its core infrastructure around GPL-licensed open-source software, investing in staff training and becoming fully dependent on it. Then, the licensor imposes a per-minute fee, demanding massive payments to continue using the very software they thought was protected under open-source principles. If the court upholds this ruling, nothing would stop licensors from pulling this bait-and-switch type move. If this ruling is upheld, future disputes could reference it to ju

      • Re: (Score:2)

        by vbdasc ( 146051 )

        Imagine this: a large organization builds its core infrastructure around GPL-licensed open-source software, investing in staff training and becoming fully dependent on it. Then, the licensor imposes a per-minute fee, demanding massive payments to continue using the very software they thought was protected under open-source principles.

        When some particular version of software has been licensed under GPL, it STAYS under GPL. Licenses can't become more restrictive retroactively. I mean, when you download some GPL software, you download its license with it, and you use that license. You're not bound by any new restrictions imposed after the event of your download. That's why when some company wishes to transition their own free software to a non-free license, they do it with a new version.

        If you're alleging that Neo4j has retroactively chang

  • Controlling what some people can and can't do isn't freedom. AGPL doubles-down on insanity such that no one serious can ever use it. Code is free to use or it ain't.

    • Re: (Score:3)

      by vbdasc ( 146051 )

      AGPL doesn't control what people can and can't do. Nobody is forced to use AGPL, just like nobody is forced to enter in any particular contract.

    • Re: (Score:2)

      by Ichijo ( 607641 )

      Controlling what some people can and can't do isn't freedom.

      Freedom is a paradox in that you must create and enforce laws--taking freedom from those who would use it to take it from others--in order to maximize freedom for all.

  • It's way too toxic, not even Linux uses it.

    I'm more of a true freedom advocate, I like BSD, MIT or Apache much more than even GPLv2, but that one it's still tolerable.

  • This case isn’t just about GPLv3, I believe it threatens all open-source licenses. If upheld, it could create a legal loophole where licensors can attach restrictions to open-source software, overriding the rights explicitly granted in licenses like GPLv2, MIT, and Apache. The court’s misinterpretation of AGPLv3 could set a precedent allowing companies to undermine open-source freedoms in ways previously thought impossible. This isn’t just about one case, it’s about whether any ope

    • Re: (Score:2)

      by vbdasc ( 146051 )

      If upheld, it could create a legal loophole where licensors can attach restrictions to open-source software, overriding the rights explicitly granted in licenses like GPLv2, MIT, and Apache.

      This is not a problem by itself, IMHO. Licensors are the owners of the software, be it open-source or not. They are entitled to attach to their own software whatever license they want - even a closed-source license or a bastardized AGPLv3 license as in the case of question. And a bastardized AGPLv3 license is NOT AGPLv3. In particular, it might NOT grant users the same rights they have under the pristine AGPLv3. To be clear, using a bastardized open-source license is wrong. It creates ambiguities that need

      • Re: (Score:1)

        by jms00 ( 10502445 )
        The issue is, Neo4j added restrictions but kept the AGPL license name, FSF preamble, and FSF copyright notices, effectively tricking people who thought it was still open source. Had Neo4j done the right thing and renamed the license and not made it look like an official open source license , then that would have been the end of it. Hopefully that explains why this is such a mess. Do you see he issue now?

Slashdot Top Deals

[We] use bad software and bad machines for the wrong things. -- R.W. Hamming

Close