1. 36
  1.  

  2. 22

    I understand your arguments in favor of copyleft, but I will still use a permissive license for my new library project, even though AFAIK it’s unique (a cross-platform abstraction over accessibility APIs, independent of a large monolithic codebase like Chromium or Qt, meant to be used from multiple languages). The reason is that I don’t want anything to hinder acceptance and usage of my work, because I want more applications, including applications using niche GUI toolkits, to be accessible (e.g. for blind people with screen readers). This goal matters more to me than the likely outcome that I’ll be helping out proprietary software projects. They’re not going to choose copyleft anyway, but at least they’ll be accessible to more users, and I might be able to avert a scenario where someone would otherwise lose their job (or be unable to get one) due to inaccessible applications.

    I’m not linking to my new project, because I don’t want to distract too much from the topic being discussed. I just wanted to explain why convincing people to copyleft unique libraries is a hard sell.

    1. 13

      Accessibility is a really interesting edge case, because it’s sort of similar to the libvorbis example: people are more likely to not use the lib than switch to copyleft, and the primary goal (improving accessibility / avoiding the MP3 patents) was deemed strategically more important than copyleft (by you / xiph). In the libvorbis case, it’s due to there being a lot of sound libraries already; for accessibility, people are more likely to just give up on the users with extra needs.

      Were I in your position, I genuinely don’t know what I’d do either. It’s hard being the first mover, especially when your decisions impact under-served people like in this case. Kudos to you for thinking hard about the implications of your licence choices.

      ETA: If I were to try and put my finger on why this feels like such an interesting case, I think it’s because good accessibility APIs should be “commodity”, but are currently “unique”. (If you take as given certain principles like “everyone should be able to use a computer”, the necessity of good accessiblity software almost immediately follows.) Whereas an implementation of a novel algorithm or a major contribution to a new language ecosystem doesn’t seem to interact along that moral dimension quite as much.

      1. 3

        Thanks for the thoughtful reply. I hope you’re right that there are still cases where choosing copyleft for a new library is a rational move and a net benefit.

      2. 8

        I will still use a permissive license for my new library project […] The reason is that I don’t want anything to hinder acceptance and usage of my work, because I want more applications, including applications using niche GUI toolkits, to be accessible (e.g. for blind people with screen readers). This goal matters more to me than the likely outcome that I’ll be helping out proprietary software projects

        The FSF and the GNU project share your view: sometimes GPL is not the best choice, especially for libraries used for interoperability, or reimplementations of public APIs.

        From https://www.gnu.org/licenses/why-not-lgpl.html (emphasis mine):

        Using the ordinary GPL is not advantageous for every library. There are reasons that can make it better to use the Lesser GPL in certain cases. The most common case is when a free library’s features are readily available for proprietary software through other libraries. In that case, the library cannot give free software any particular advantage, so it is better to use the Lesser GPL for that library.

        This is why we used the Lesser GPL for the GNU C library. After all, there are plenty of other C libraries; using the GPL for ours would have driven proprietary software developers to use another—no problem for them, only for us.

        From https://www.gnu.org/licenses/license-recommendations.html (emphasis mine):

        Some libraries implement free data formats that are competing against restricted data formats, such as Ogg Vorbis (which competes against MP3 audio) and WebM (which competes against MPEG-4 video). The success of the free format requires allowing many proprietary application programs to link in the code to handle the format. For instance, we wanted nonfree media players, especially appliances, to include the code for Ogg Vorbis as well as MP3.

        In these special situations, if you are aiming to convince proprietary application developers to use the library for the free format, you would need to make that easy by licensing the library under a weak license, such as the Apache License 2.0.

        1. 2

          Thanks for posting the extra detail. Interop situations (like glibc) are an extra special case of what I was calling “commodity” in the OP.

        2. 3

          I agree with your specific points, but I want to contrast this with the idea of copylefting an entire language ecosystem at once, which I think is still desirable even in the face of unique libraries which crave adoption. Because it is hard to predict which libraries will be exploitable by corporations, we need to apply copyleft licensing uniformly to every piece of work within an ecosystem, as with the Creative Commons, in order to keep everybody protected. It is something like herd immunity but for software.

        3. 19

          The original sin of Free Software is that it’s fundamentally based around “POV: you work, maybe live, at the MIT AI lab, surrounded by other hackers”. It’s freedom for hacking on what you have and hacking first, anything else is a side effect. The benefits felt by end-users feel abstract and nebulous.

          1. 10

            I think you haven’t gone quite far enough:

            The original sin of Free Software is the underlying assumption that the hackers and the end-users are the same.

            It’s a foolish errand trying to ensure the freedom of users who don’t care to take responsibility and an active interest in the tools they use.

            1. 8

              It’s very easy to explain Free Software to business folks because they understand the concept of a second source: you don’t want anything your business depends on for success to be limited to a single vendor because then you have a huge switching expense if they put their prices up to more than you’re willing to pay. With open source, you can always get another company to provide the features / bug fixes / support that you need.

              It’s much harder for consumer markets because second sourcing isn’t generally something that folks think about for consumer devices, in part because they don’t tend to build things on top of single-vendor devices and most things don’t have lock-in effects. If you buy a Samsung fridge, that doesn’t prevent you from using a Whirlpool dishwasher in the same kitchen and if the fridge breaks you can replace it with a Panasonic one without needing to buy different food.

              1. 3

                Relying on free software to safeguard privacy, for example, is maybe necessary but hardly sufficient.

                For that you need legislation (with teeth).

                The ready availability of free, high-quality software that can be distributed with zero marginal cost has probably done more to erode the public’s privacy than anything else - in the form of highly scalable ad delivery and analysis networks.

              2. 4

                At this point in history, more education about computers and what basic hacking is/can be is as important if not more important for freedom than licensing.

              3. 4

                I wouldn’t say that it’s the “original sin”, just that the FSF and the GNU project haven’t really reconsidered their original assumptions on how people relate to software.

              4. 11

                The meat of this article is pretty good, unfortunately it is halfway down after reading through what anyone outside the movement could only see at an anti-government borderline-covid-denying rant. I honestly feel that, your personal feelings on how the pandemic should be handled notwithstanding, the point would be clearer and more forceful if given the main spotlight of the article.

                1. 11

                  The meat of this article is pretty good, unfortunately it is halfway down after reading through what anyone outside the movement could only see at an anti-government borderline-covid-denying rant.

                  That “Unfortunately” represents an extremely politicized opinion. There are lots of good reasons to be opposed to (your particular) government, including actions that (your particular) government has undertaken in the name of managing the coronavirus pandemic. Similarly it is reasonable to be suspicious of people using “borderline-covid-denying” as a way of dismissing an argument - “borderline covid-denying” implies that the OP did not in fact deny the existence of the covid pandemic, and there are lots of statements that are semantically close to denying the existence of the covid pandemic - like “it’s bad if the government forces people to install closed-source software on their cell phones in the name of fighting pandemics” that constitute extremely reasonable political arguments.

                  1. 4

                    Again, I did not dismiss the argument or (necessarily) even disagree with the rant. This was meant as constructive criticism of the article to say that focussing on the point/meat would make the point punch harder with all audiences.

                  2. 8

                    The opening section is a bit of a waffle; I have trimmed it down. The “papers, please” section in particular was true but off-topic. The techno-authoritarianism remark remains true, but is hard to elaborate in an on-topic way, so it’s been removed too. (Aside: Population-level vaccine targets provide aligned incentives without compromising people’s medical privacy or bodily autonomy.)

                    It is possible to build privacy-preserving check-in, and to legislate protections for the data; it is a travesty that Australia has pretty much done neither. (Small exceptions: Western Australia legislated protections after it was revealed that the police accessed contact tracing and border pass data, and the federal government’s failed COVIDSafe app has strong legal protections also.)

                    I think you are being extremely rude and uncharitable to lump me in with some “movement” of nutters, and I’d like you to apologise. Like most Australians, I am waiting patiently for our vaccine rollout to crawl down the age brackets, increasingly frustrated that additional restrictions and surveillance are imposed on citizens because of the government’s failures.

                    (Quote of the removed original text below, for reference):

                    Over the past year, there have been incredible restrictions on people’s freedoms of movement, speech, and association, in the name of protecting public health. On a recent trip, I was required to carry travel papers for 14 days, and to show them on request to certain classes of authorised officers — something I never thought I’d be required to do in a first-world liberal democracy.

                    In Australia, we are now required by law to pump our personal information into one or more “contact-tracing” platforms in order to visit the supermarket or have something like a normal social life. This all-but-forces people to own a smartphone, and to interact with an enormous pile of proprietary software just to participate in society. Despite earlier promises to only be used for contact tracing, these databases have been used by police. Because of course they have.

                    Elsewhere, there are talks of “vaccine passports” on smartphones, which accelerates techno-authoritarianism and should therefore be resisted. (Search for “raise the more generalized version of this objection”.)

                    1. 9

                      I think you are being extremely rude and uncharitable to lump me in with some “movement” of nutters, and I’d like you to apologise. Like most Australians, I am waiting patiently for our vaccine rollout to crawl down the age brackets, increasingly frustrated that additional restrictions and surveillance are imposed on citizens because of the government’s failures.

                      1. He meant the Free Software Movement, the one you’re advocating in the piece.
                      2. That removed original text is pretty easily construed as an anti-government borderline-covid-denying rant.
                      1. 5

                        I disagree with your claim #1: as it’s embedded in the sentence “anyone outside the movement could only see at an anti-government borderline-covid-denying rant”, it reads pretty clearly to me like it applies to the various covid-is-a-conspiracy “movements”.

                        I can back everything in #2 with links to mainstream (i.e. not Murdoch/fringe/conspiracy-tier) news articles or to the relevant health orders, but to do so would take us further off-topic. I still do not understand this reading that you and singpolyma seem to see in the original text. (The “borderline-covid-denying” part, in particular. Whatever; it’s off-topic and removed from the main post, anyway.)

                        1. 5

                          If you have been paying even a modicum of attention to news, you know that there are and for a long time have been people who, no matter what crisis or emergency occurs, immediately denounce it as either a complete fiction or at best having been blown out of proportion as a cover story for the “truth”, which always is that the (insert preferred evil conspiratorial one-world-government entity here) wants an excuse to finally confiscate all our guns, terminate our freedoms, and institute worldwide dictatorship, all of which is only held at bay by badly-formatted postings from lone heroes on the internet.

                          What people are telling you is that the quoted section of your post sounds like one of those paranoid rants, in that it can be read to imply that contract tracing and vaccination status are sham reasons being used to impose the real agenda of “techno-authoritarianism”.

                          1. 2

                            What people are telling you is that the quoted section of your post sounds like one of those paranoid rants, in that it can be read to imply that contract tracing and vaccination status are sham reasons being used to impose the real agenda of “techno-authoritarianism”.

                            They’re worse than sham reasons to impose the real agenda of techno-authoritarianism - they’re superficially good reasons to impose the real agenda of techno-authoritarianism. Techno-authoritarianism is advanced more by a broad social atmosphere where opposition to software authoritarianism is widely thought of as low-status and uncomfortably adjacent to people with heretical politics, than it is by genuine conspiracies.

                            1. 3

                              Your formulation appears to be a literally irrefutable claim, in that anything and everything could be described, via argumentative gymnastics, as advancing or supporting “techno-authoritarianism”.

                          2. -2

                            Ignore them, they’re robots.

                          3. 2

                            Thank you for reading me correctly and providing this good summary of my comment

                            I would like to further clarify that I in no way meant to assume the rant was in bad faith (as someone in the Free Software movement, I could see the intent) but only that it would be easy to read it that way and thus distract from the meat of the article.

                          4. 3

                            I don’t think those three paragraphs make you out as a crank. I think it is fine and good to ask for serious privacy protections in contact-tracing, vaccine passport, and identity systems.

                            I would like to point out that many liberal democracies require their citizens to carry ID with them at all times. I believe this is the law in Germany and France, where I understand it is mostly used for harassing ethnic minorities.

                            FWIW, I think this history could be used to strengthen your argument: “liberal” democracies are very comfortable with authoritarianism for some of the population and these illiberal mechanisms are rarely dismantled, so they will accumulate if not resisted.

                            1. -2

                              a first-world liberal democracy.

                              In Australia

                              I think king Murdoch would disagree.

                            2. 5

                              Being anti-government is not immoral per se. It depends on the government. For example, you would probably agree that it’s moral to be anti-government in North Korea. You don’t have to deny Covid, the respiratory disease, to see that some governments have massively overstepped their boundaries. For example, there is mounting evidence that dusk-to-dawn curfews have no impact on the pandemic situation whatsoever, yet they were broadly imposed, often violating the constitution of the respective country. I would appreciate if people like you would stop with the moralistic shaming and inspect the subject matter with more care and thought. Not everyone who criticises the measures or aspects thereof is a science-denying anarcho-terrorist.

                            3. 6

                              I get the point of these arguments and I feel for your views, but the overall market has just changed from creation to consumption. The average person really doesn’t care what their computers run as long as they’re able to work with the restrictions that are put on them (use these apps, talk to these people, socialize here, etc). I get the philosophical point about free software being a moral imperative (this is something I whole-heartedly agree with) but overall that doesn’t pay the bills. You can’t exactly make rent purely on doing the right things or giving things away with nothing in return. It’s not economical, unfortunately. I really wish this wasn’t the case but the “starving artist” trope exists for a reason.

                              1. 2

                                That’s not entirely true. Drew DeVault is a good example of using, building and making a living out of free software. He’s proved that it can, in fact, pay the bills.

                                1. 2

                                  His software caters almost exclusively to other developers, so, far from “the average person” as a market.

                              2. 2

                                I think most of the problems mentioned in the article are not solvable with better licenses. As some free software projects now also have started to show ads on the console, for example, it proves that even if you want to opt out of that stuff, licenses won’t help and thinking that they do is naive: the producer decides what goes in a software. You can decide to strip it out, but then you’ll be playing catch-up forever with the one who produces the software. Also note the recent audacity debacle.

                                As the article correctly states, the market won’t save you as it’s a race to the bottom: price seems to be more important than privacy to most consumers, so that’s the strongest driving factor.

                                I think we could take our examples from animal and human rights and maybe climate change activists. There, the market also strongly favours bad behaviour over good behavior because they yield better prices (in the case of human and animal rights that’s probably the reason these practices started in the first place, as with tracking and showing ads). Naming, shaming and educating the public as well as lobbying politicians to create new laws seems to have helped a little. Not sure how that translates to “free software”, specifically, but I’m not sure I care as much about the software’s specific license as the net effect. Of course free software is better at empowering technical users than non-free software, but in the larger scheme of things that’s unlikely to get nontechnical users excited about.