1. 13
  1.  

  2. 25

    Copyleft licenses like AGPL and SSPL have a weaker result in terms of overall benefit to society than liberal licenses like MIT and Apache2

    People are motivated not only by a sense of purpose and the desire to contribute to something greater, but also by improving their lot in life and that of their families and heirs (read: capitalism)

    All hail MIT, because that will motivate the internal capitalist in our families and heirs to benefit society more since they can take the code out of the commons and use it to… amass capital. Or something.

    Or maybe, just maybe, the model of selling an intangible, infinitely copy-able good is just not a great business model and the enforcement we see with licenses, DRM etc. is an indication of this fact?

    Ultimately, in my philosophy, copyleft doesn’t represent real open source, despite what the OSI says. Copyleft is a restriction.

    An argument as old as the GPL itself. In similar vein: if democracy were democratic, why does it not offer the option of abolishing democracy even with a majority?

    1. 9

      Similarly, I like to argue that the thirteenth amendment to the US constitution doesn’t represent real freedom, despite what the abolitionists say. Banning slavery is a restriction.

      And it is! …on the masters. But the point is to protect the slaves from the masters. And the GPL is the same idea: it is supposed to protect the users of the software from misdeeds of the authors. For example, do you have a phone app that sends spurious push notifications to monopolize your time? If it was GPL, you’d have the right to hack that malicious code out (or have a developer do it and share their change with you, since not all users have the technical skills to do it themselves).

      I think a lot of commentators about the GPL forget this. It isn’t about money per se, it is about letting the users escape the author’s dark patterns.

      1. 2

        This argument is unclear. You can hack the part out if it was released under any OS license, such as MIT or Apache 2.0 The GPL adds a restriction irrelevant to your use case which is that you are required to release the modified source code if you distribute the modified software.

        1. 6

          adam_d_ruppe is assuming that the source code of the application would not be released unless it was built on, and with, GPL components. The assumption is correct: 99% of phone apps are closed source.

          1. 2

            Of the 1% that are open source, what proportion are copyleft and what proportion are permissive?

          2. 4

            MIT etc. let the publisher decide - they can distribute the program without the source (modified or no). Some of them forward this to the end user, but many don’t. The GPL says the publisher MUST let the end user have that access, by requiring they make the full source - including modifications - available to them too. That’s what the “viral” thing is all about - making sure those rights actually make it to the end user and aren’t stripped out by a middle man.

        2. 8

          The comparison is spot on! Also see: https://en.wikipedia.org/wiki/Paradox_of_tolerance

          “In order to maintain a tolerant society, the society must be intolerant of intolerance.” - Karl Popper

          1. 3

            I think you’re right that the selling model doesn’t work, which is why we have SaaS. The question is how to fairly compensate those who enable the sale of services, or goods built with open source, while retaining the benefits of open source.

            As I allude to above, I think that software is more like music than anything else, and we need something like collecting societies to share the wealth.

            1. 1

              “and use it to… amass capital. Or something.”

              Basically what they do.

              “why does it not offer the option of abolishing democracy “

              The 2nd Amendment in US is specifically included as a way to address government corruption, abolishing specific politicians or whole thing.

              1. 2

                The 2nd Amendment in US is specifically included as a way to address government corruption, abolishing specific politicians or whole thing.

                I understood the argument as “if we have democracy (roughly: rule of the [majority of the] people), why can’t such a majority decide that they prefer to live in an absolutist monarchy and thereby end democracy?”. The 2nd Amendment to the US Constitution is a feature specifically designed to prevent that from happening.

              2. 0

                not offer the option of abolishing democracy even with a majority

                It depends on the constitution, but it’s quite possible to turn a democracy into a defacto autocracy by changing laws and process through the democratic process. For example, it is possible to change the US constitution with sufficient majority in the house and senate or the state legislatures. The hope is that the distributed nature of the power makes such a hack very hard to do and requires a conspiracy of mind boggling proportion (or similarly sized stupidity and greed) but it can be done.

              3. 6

                I like this approach overall, but I think the discussion in general has suffered from a lack of clarity about the reasons for wanting source to be open. This article touches on that, from the perspective of a developer.

                Let’s try to delimit some use cases:

                1. Getting things done (or perhaps leeching).

                In this scenario, as an engineer I use open source because I can solve a problem and not have to engage with business types about paying for licences; and increasingly the business types are comfortable with this.

                1.1 Fix my own bugs, get bug fixes on my enhancements Like the above, but I don’t have to fix every bug myself, AND if I’m the first to really be bothered by a bug, I can fix it.

                These scenarios look like a shared resource, not created for innovation, but instead to amortize cost and reduce friction across their user community. This is largely the model of most successful open source today. Where the community of users is not vibrant enough to sustain innovation, or where the creators feel that they are not being rewarded for creating value for many companies, they may form exploitation companies, whose pressures then lead to the logic of visible source licences. (Cf confluent, redislabs)

                A related model is where a single sponsoring company is a major consumer of the technology, and maybe derives some benefit from wider use - react is a great example. Because the sponsor does not rely on this software to generate revenue directly, they can afford to let it stay open and not care about others making money from it.

                1. Curiousity, education, get around bad docs.

                This is the need to know exactly what I’m getting, instead of wondering why closed source software doesn’t work as advertised.

                Visible source licences meet this need.

                1. Freedom to build and innovate

                This is different from 1 in that it’s about putting software out in the world separate from (necessarily) running a business. I consider this the core need to which the FSF addresses itself. Not coincidentally RMS used to get his money from working at a university, and now from running the FSF.

                This doesn’t seem strongly compatible with a commercial ethic. Props to Influx for finding a way to partially square this circle.

                In many ways this model is the most pro-innovation, and if it could work would answer to the functions of the historical university: an endowed organization not needing to link activity to income, with no inherent need for growth, generating public goods.

                For this to work, probably large, successful companies would have to be held to a social or legal norm that they contribute to institutions that sustain such development.

                Compare with the large sponsor model (eg Facebook and React) - the results they can produce are similar on the level of individual projects, in terms of a dynamic community, at least.

                For my part, I’m increasingly convinced that in a pure market model, only visible source is possible. Some kind of reform to allow the use of software without catastrophic litigation, but with an expectation of payment is needed. Something like the mandatory licensing of music: https://www.thebalancecareers.com/what-is-a-compulsory-license-in-music-2460357

                This would likely then require the formation of collecting societies for software to handle the dispersed copyright of most open source.

                I submit that this would satisfy the needs of all three use cases above.

                The use of a licence that allows cure and protects users from a disaster (either open sourcing or being asked to disgorge all their profit), and the formation of a collecting society might allow us as a community to solve this without legislation.

                1. 4

                  Music and software have many of the same problems. Music has a complex ecosystem of copyright and licensing to help musicians get paid for their creativity and yet there’s still lots of middlemen, like record labels, making huge returns, somewhat akin to software’s VC. Read about ASCAP: https://www.petekeen.net/ascap-for-open-source-software

                  My latest thought is having all companies with an engineering team “tithe” 10% of their engineering payroll to license all OSS for commercial use but there’s a million questions to be dealt with, just like ASCAP.

                  1. 2

                    Agree with this (see my comment on this post). I think the details you suggest sound like a good starting point. Do you know of any projects that want to move to this model? I think we should try to do this.

                    1. 1

                      Why would they bother? They’ll just go back to writing their own internal libraries and toolchains. Paying 10% of your payroll to use code you don’t have control over? Are you mad?

                      1. 1

                        Why not? Access to all open source, royalty and litigation risk free?

                    2. [Comment removed by author]

                      1. 2

                        You are entitled to your opinion about when something is ‘liberal’, but the word ‘arbitrary’ in your comment serves no purpose (at best) or is just wrong (as it’s a very specific restriction with very clear arguments for why it is needed). When you remove it, it clearly raises the question! ‘restrictions like what?