1. 4
    1. 12

      I’m sorry, but as much as I dislike copyright, I don’t get it.

      Trust

      You cannot trust non public domain information products. You can only make due. By definition, non public domain information products have a hidden agenda.

      What hidden agenda would a software licensed under the BSD-2-Clause have “by definition”, that it wouldn’t have if it was in the public domain?

      Speed

      Public domain products are strictly faster to use than non public domain products. Not just faster, orders of magnitude faster.

      Cost to build

      Public domain products are far cheaper to build than non public domain products. Failure to embrace the public domain increases the cost to build any information product by at least an order of magnitude.

      Again, I’m sorry but without any attempt at a proof, it is more wishful thinking than actionable information. What sources do you use, what makes you believe that?

      1. 3

        I think the verbiage can be better around “hidden agenda”. Thanks for the question. Let me explain (and maybe I’ll come up with a tweak to improve the post).

        What hidden agenda would a software licensed under the BSD-2-Clause have “by definition”, that it wouldn’t have if it was in the public domain?

        Every human has hidden agendas. You can not see inside someone’s brain. I’m not saying most people walk around with evil hidden agendas—far from it. But most people don’t necessarily have your interests at the top of their mind either—why should they, everyone’s got their own problems.

        The “BSD-2-Clause” comes with conditions. If you don’t want to follow those conditions, the only way to legally do so would be to seek the permission of the author, who, as a human, of course has a hidden agenda.

        Therefore, by definition even a BSD-2-Clause licensed software has a hidden agenda. As I’m assuming that was your best counter, you can follow that the situation is worse as they license gets worse, and most things have worse licenses than BSD-2.

        [Not sure if I’ve made a good enough argument, will have to reread this tomorrow]

        Again, I’m sorry but without any attempt at a proof, it is more wishful thinking than actionable information. What sources do you use, what makes you believe that?

        It’s a fair point. It would be good to build up a big public dataset on this. I think that would make the argument stronger. I have an internal mental dataset from decades of experience but a good ole CSV would be 100x better. I’ve added a commit to that post with a todo and will start that soon.

        1. 5

          Why must their agenda be hidden, though? Isn’t it possible that people who choose, say, a “copyleft” license like AGPL have a fairly obvious agenda of not wanting modified versions of their software to be run by giant cloud companies without releasing the modified source? That is what such licenses are designed to prevent, so doesn’t it follow that someone using such a license likely just wants to prevent that? Nothing is hidden here, it’s just the right tool for the job.

          But if you’re still worried about agendas being hidden in cases like that, would it suffice for the author to have a statement of intent for the software including reasoning around their choice of license?

        2. 4

          the author, who, as a human, of course has a hidden agenda

          Of course, in that sense every human has a hidden agenda. It follows that the author of a public domain work also has one. My point is that, while I cannot see inside someone’s brain, I would guess that those hidden agendas are generally similar whether the software is in the public domain, or licensed under a permissive license.

          Edit: by the way, are you familiar with this recent article?

          1. 2

            It follows that the author of a public domain work also has one.

            For sure, but the key difference is that when the product is public domain the author’s hidden agenda is not transferred to the product. The product is free, with zero strings attached back to the author. Even if someone uses a very liberal “license”, say a string 1 micron wide, there is a big difference between 1 and 0.

            1. 8

              Even if someone uses a very liberal “license”, say a string 1 micron wide, there is a big difference between 1 and 0.

              The trouble with that reasoning is that for some of us[1], the 0 is not an option. For example in France, dedicating my work to the public domain is not a thing and using a very permissive license is the best alternative we’ve got.

              [1] Perhaps for most of us since, as argued in the article I linked to in a late edit, the situation might not be much better in the US.

              1. 1

                In the USA, federal agencies cannot hold copyright. For example, when NASA shares photography with us, the data is in the public domain. This might be a good example of a situation where the public domain is superior – NASA is generally considered one of the leaders in space exploration.

              2. 1

                For example in France, dedicating my work to the public domain is not a thing

                !!!! I didn’t expect things to be worse in Europe. Thanks for the information.

                1. 1

                  This article is mostly about books but there are lots of exceptions

                  https://medium.com/copyright-untangled/public-domain-why-it-is-not-that-simple-in-europe-1a049ce81499

                  I believe the real issue is that there’s no “positive” way to declare a work in the public domain in Europe, like in the US for works funded by the federal government. So everything is under copyright until 70 years after the author dies.

      2. 2

        Thank you @jmiven for the very helpful feedback.

        I just updated the paragraphs about Speed and Cost: https://github.com/breck7/breckyunits.com/commit/65bc0aa36eaaa9345350181c444ab0b41c54e434

    2. 10

      Hi @breck. I’m a developer, I use FOSS and I contribute to FOSS projects. I don’t want to care about licences, I just want to minimize friction for using FOSS, contributing to FOSS, accepting contributions to my projects. I’m probably a person that your message is meant to reach.

      I’m frustrated by licence proliferation and licence incompatibility, and by having to think about licences. Maybe you are too, I don’t know your motivation for writing this post. I’m not a lawyer, but I have paid attention. The issue of “should we dedicate FOSS code to the public domain” has been discussed to death, with lawyers involved, in many forums, and the answer turns out to be no, for good reasons.

      In some jurisdictions, if you dedicate code to the public domain, it has no legal effect, so the code is still copyrighted in your name, and nobody is legally allowed to use it without your permission. You say you are concerned about trust. Well, you can’t generally trust code that doesn’t have a copyright notice and a licence grant, because the owner can always pop up and claim their rights and threaten a lawsuit or try to take down your code. Even if there is something that looks like a FOSS licence grant, if the licence is not one of the popular licences whose meaning has been fully investigated by lawyers, then the owner can still pop up and say “gotcha” because you understood the licence differently than the owner. And this has happened.

      So as a community, we’ve collectively decided that the best way to achieve trust is to use a popular and well understood FOSS licence when you publish code. Not public domain.

      SQLite has dedicated their code to the public domain. That’s great, but they have had to take extreme measures to make this work for them, their customers, and their contributors. I don’t want to subject myself to that hassle, I don’t want to subject contributors to my FOSS projects to that hassle. My code wouldn’t be usable to businesses unless I did what SQLite does, and I don’t want the hassle. I do have businesses using my code, and I have no particular ill will towards them: it’s cool that they use my code as long as it doesn’t create bureaucracy for me. So public domain is a very high friction proposition, and therefore it isn’t fast and cheap, which is what you say you want.

      In order to be public domain, SQLite has imposed the following burdens on themselves:

      • they cannot use external library dependencies (other than the standard C library I guess)
      • they cannot accept contributions from external developers, unless those developers sign a legal document.
      • In order to enable businesses to use their code, they will sell a Warranty of Title to anyone who needs one.

      So much hassle. https://www.sqlite.org/copyright.html

      One thing a FOSS licence does is provide legal protection to the developer who is writing the code and releasing it to the public. All of the popular FOSS licences have a “no warranty” clause which provides the bulk of the legal protection. A few licences (eg, Apache 2, and the Universal Permissive Licence (UPL)) also have a patent clause, which further protects the developer. Say that somebody contributes to your project on a source forge like github or codeberg. But, unknown to you, that person owns a patent on the algorithm used by the code they contributed. Later on, they can sue you or your users for using that code. So the patent grant in the licences I mention protect you from that.

      A public domain grant does not provide you any of these legal protections.

      1. 2

        This was ann extremely informative post. Thank you. I figured that plublic domain would prove troublesome for using libraries which have say Apache or MIT licenses, but I did not consider the patent troll angle at all.

    3. 4

      SQLite’s meteoric success is not a fluke.

      Citation needed. SQLite feels like a textbook definition of a fluke to me; it was created by a mad-scientist coder who invented their own version control system. I don’t know of any other projects like it; it’s certainly not a success any moderately-skilled programmer could expect to replicate.

      1. 2

        I believe Daniel J Bernstein’s software is ceded to the public domain (as he sees it). I personally find his stuff much less useful than sqlite though.

        1. 2

          it’s like you read my mind; as I was writing this I almost write “other than someone like djb…”

          I can count on one hand the number of people I know to be competent enough at writing C for me to trust their stuff without treating it like a ticking time bomb.

    4. 4

      Public domain products are strictly faster to use than non public domain products. Not just faster, orders of magnitude faster.

      While I agree with the general gist of your argument, you’re building a tripod of support and this one seems unsupported. I think most people can figure out the trust and cost legs, but why is PD strictly faster?

      1. 2
        1. 3

          I appreciate the changes, though I’m not sure the updates will carry the day.

          When you sum the time savings across all possible use cases of all possible products, you’ll see the orders of magnitude speed up caused by public domain products.

          This assumes the time cost of public domain software is effectively zero … that is nonsensical, if for no other reason than, in many jurisdictions, it’s not entirely clear if it is even possible to declare a work to be public domain. If we assume a similarly expansive stance for this thought experiment then you’d have to admit that at least one of the all possible use cases for any public domain product would be one in which the legal system of one (or more) jurisdictions would have to decide on a challenge to the factual basis of the public domain claim. Such a challenge (for instance for SQLite), raised by a sufficiently well-resourced legal team (possibly including the legal team of the jurisdiction itself) could drag on for decades, and might never be decided in the favor of the public domain at all. In that edge case PD software could take an infinite amount of time to use, relative to something less than PD but permissively (if conditionally) licensed.

            1. 2

              Ahh, well, I admire the ambition… I dare say there may be some more pressing potential amendments, but good luck getting this one through.

              1. 1

                I dare say there may be some more pressing potential amendments

                On the surface, but when you look upstream you see how a lot of issues stem from restrictions on information rights.

                Edit: that’s my take anyway. But I could be wrong.

    5. 2

      There is a seed of something clever and insightful here. It needs clarity. One strategy that might help is to give up on a strict inequality. Sometimes it can be sufficient to simply provide an alternative for every proprietary strategy; the alternative doesn’t have to be strictly better in all ways, just equivalent.