1. 31

Posting this as I was aware of rms’s objection to open source, but hadn’t done enough thinking/reading to clearly tease apart the definitions of ‘open source’ and ‘free’; a re-read of this was helpful.


  2. 18

    One thing that is so frequently forgotten these days is that “open source” was coined as a marketing synonym for “free software”. The two terms refer to the same software and licenses (except for minor obscure exceptions). Copyleft is not required for free software, and open source requires the same freedoms as free software. The only substantive difference is a matter of how you think of the software, whether you think it’s mostly about freedom or mostly about commercial advantage.

    1. 16

      And to expand on your point (and great blog post), “open source” was promoted by business interests to avoid RMS’s politics of “free software”.

      Huh. I just realized it’s been 20 years since I heard about it, 30 since it was coined, and “free software” is still a bad, confusing term. Maybe it’s time to stop digging in that hole and rebrand to “libre software” or one of the other alternatives.

      1. 7

        I had a meaner way of saying this, which I deleted…

        The popular misconception that free software is made by large companies and is free in the same sense as “free fries with a large Whopper” informs a lot of the attitude that often drives people away from making free software as their way of benefiting others.

        I do feel as though changing the terminology would help that. Looking at the larger context of why “free” is even a word that connotes “good” in the first place, which is a lot of the motivation for using it… it’s because political rhetoric has used that word as an undefined but good thing for centuries. Let’s pick a word that has a meaning that can be pinned down.

        Wow, I’m cynical in the morning.

        1. 6

          As long as it’s not the truly awful initialism “FLOSS”, I can get behind this.

        2. 8

          “open source” was coined as a marketing synonym for “free software”.

          I disagree. Open Source was created to promote toothless alternatives to licenses like those in the GPL family, and to ignore ethical questions about software distribution and control over computation in general. If the purpose of the Open Source labeling was just to dissociate themselves from Stallman or the FSF, then the community wouldn’t also be recommending weak/lax/permissive, non-copyleft software licenses –as opposed to the {A,L,}GPL– that fail to guarantee the very freedoms that make the work Free Software in the first place.

          Copyleft is not required for free software

          But it is required for Free Software to remain Free, and the de facto effect of this is that copyleft is the proper, logical way to promote and license Free Software specifically. If Free Software and Open Source really are the same, with a mere relabeling, then why bother with different licenses that leave the software and its users vulnerable to proprietary middlemen that turn once-Free Software into proprietary black boxes? That’s bad, right? Why sink millions of man-hours into permissively licensed projects, with the explicit goal of getting away from copyleft projects and licenses, even going so far as to rebuild entire toolchains from scratch to get away from the GPL? I mean, they’re the same thing right? Hell, why not just throw the source under the public domain? That’s Free Software too. And this whole Open Source schism was just a relabeling to sell Free Software, right?

          The reason is simple. The Open Source community are proponents of a development methodology –as opposed to an ethical software distribution framework which respects the freedoms of the user– that is preferential to the interests of business, typically to the detriment of the users' freedoms. The developers enjoy the four freedoms that Free Software provide, but often refuse to provide those freedoms to their users in turn, instead chaining them. The software isn’t Free/Open Source Software for me if the end result, as a user, is a binary blob that I must rely upon to do my computing and that strips me of my freedoms. It’s disingenuous to suggest that the Open Source movement is just a relabeling or marketing synonym, given its refusal to recognize the importance of copyleft in protecting the very Free Software tenants it claims to be identical to.

          This, to me, is the real “substantive difference.” It’s only obscure because of continued refusal to address the ethical components of software freedom. And it’s certainly not a minor difference. Inconvenient perhaps, but not minor. The linked essay goes into these issues at length and makes clear that the differences are not obscure or minor. They are fundamental.

          1. 3

            We already had the BSD license before GNU reimplemented [some of] Unix. You could ask why GNU reimplemented everything under a different license too, except now we have an answer to that: to pull a bait and switch by releasing the GPLv3. If a new toolchain prevents a GPLv4 it will be a small price for the effort.

            1. 5

              Well, there was the AT&T BSD lawsuit at the time which made it unclear if BSD really was free or not. I think the GPL was in part also a response to that sort of thing, so that there could be no doubt as to the freedom of the code.

              Also GPLv3 is not the bugbear people think it is. It is not some insidious plot of the FSF to trick people into using GPLv3. Can you articulate what your problem with GPLv3 is? I find GPLv3 to be widely misunderstood and thus feared.

              1. 0

                My own problem is much more basic than that. Back when I was younger and knew more, I studied the GPLv2 and used it to develop a working understanding of Title 17. That work was rewarding, but too far afield of my day job to give GNU a second study. By the time GPLv3 was published I was well enmeshed in MIT/X code and I’ve had no reason to wade through FSF license churn when those works have such an apparently short shelf life and attendant uncertainty. Linus' defection was a big help in letting me focus elsewhere.

                1. 3

                  Do you mean Title 17 of GPLv3? GPLv2 only has 12 sections.

                  If you are basing your understanding of GPLv3 on Linus not liking it, well… Linus himself doesn’t seem to understand GPLv3. I wanted to clear up confusion of what people think they dislike about GPLv3, but you seem to prefer to not want to attempt understanding.

                  1. 1

                    No I mean Title 17 of the United States code. I realize I sound like I prefer ignorance to understanding. What I dislike about the GPLv3 is the “I have altered the deal, pray I don’t alter it further” property. I was happy with the first deal and do not have any new licensing problems. I have been blessed to spend my time working on other things. If you think I have a problem the GPLv3 solves, I’m happy for you to tell me what it is.

            2. 1

              It’s only obscure because of continued refusal to address the ethical components of software freedom.

              I find all forms of intellectual property to be unethical (copyright + patents + trademarks + …) and therefore find copyleft itself to be unethical. The FSF doesn’t have a monopoly on the ethics of software freedom.

              1. 1

                What’s your problem with trademarks? I think they’re overall a pretty good thing. It’s important for consumers to be informed about what they are buying. I don’t want people who are not GNU to start using the GNU trademark and thus deceive the public.

                1. 2

                  It’s important for consumers to be informed about what they are buying.

                  I agree!

                  I don’t want people who are not GNU to start using the GNU trademark and thus deceive the public.

                  I agree!

                  What’s your problem with trademarks?

                  Because they are a form of intellectual property. There are lots of critiques of IP that resonate with me. I’m partial to “Information Feudalism.” I also disagree with IP philosophically. I don’t think creations of the intellect can be meaningfully owned, and therefore, I think laws granting monopolies on such things is unethical.

                  1. 1

                    I think the problem here is that you are thinking that trademarks are the same as software patents because people call them both “intellectual property”. The problem is that “intellectual property” has such vague and broad meaning that it makes people think that different things are similar. Trademarks are very different from patents and very different from copyright. So, without making reference to this nebulous “intellectual property”, what is the problem with trademarks?

                    I actually think some patents are really awesome. For example, Michael Jackson’s patent on the Smooth Criminal leaning trick. That’s how the patent system is supposed to work, a magician revealing his tricks! Without patent law, MJ may have taken the secret to his grave.

                    1. 2

                      I can assure you that I’m not confused. I am intentionally using the term intellectual property because my criticism is fundamental to it. I fundamentally disagree with the notion that ideas ought to be the foundation of legal monopolies. A direct consequence of this is that I find trademarks unethical (and anything else that is propped up by IP laws).

                      I don’t just dislike “software” patents. I dislike all patents. Especially the type of medical patents that IP rich countries use to strong arm developing nations. (cf. “Information Feudalism.”)

                      1. 2

                        But trademarks are not ideas. There is no originality or creativity required in trademarks. You can trademark a square if you want, but you can’t copyright or patent squares.

                        1. 2

                          Trademarks are creations of the intellect and depend on a legal monopoly given to them by the State. I don’t know what originality has to do with this.

                          1. 2

                            That’s the thing, trademarks are not creations of the intellect. They do not require any intellect at all. I can trademark “JordiGH’s thing” and it required absolutely no thinking to come up with this trademarkable name. Regional designations are similar, in that “made in Canada” is also not a creation of the intellect, yet is covered by laws that people call “intellectual property”.

                            1. 1

                              You are misunderstanding “creation of the intellect.” It is not meant to designate effort or ability, it is merely to designate it as something that only exists in the abstract. A trademark is itself some design or expression, it is not a tangible thing. A trademark has no scarcity. It has no physical presence. And by consequence, it has no inherent exclusionary properties; two or more people can use the same trademark simultaneously. (This is not to say anything about any potential confusion that may cause, which is itself dependent on a number of factors.)

                              It is these properties that cause trademarks to be classified under the umbrella term “intellectual property.” These properties are shared among copyright and patents, and they are specifically the things that I have a problem with being the basis of legal monopolies.

                              trademarks are not creations of the intellect

                              To be clear, you are implying something here that is at odds with the mainstream definition of intellectual property. The mainstream definition includes trademarks, among other things. The mainstream definition is also usually something like “legal monopoly granted to ideas” or “ownership of ideas” or “ownership over creations of intellect.” Trademark is included in all of these definitions.

                              1. 1

                                Last attempt, please read this.

                                1. 2

                                  I have. You are not the first FSF advocate I’ve run across. Although, I will say that most FSF advocates acknowledge my stance against IP and tend to agree with it, but that we part ways simply disagreeing with the means at which we express it. The GPL is purportedly using the system against itself, while I would rather not use more IP at all.

                                  I’ve demonstrated why trademark is IP. It specifically depends on legal monopolies granted to ideas. Trademarks are very clearly unlike property and fit perfectly in the framework of intellectual property.

                                  1. 2

                                    And for what it’s worth, that FAQ is a pretty bad response to the thoughts I’ve expressed here. It doesn’t actually address my specific criticism against IP. It seems to be addressing proponents of IP instead of detractors. It almost addresses it here:

                                    It leads people to focus on the meager commonality in form that these disparate laws have—that they create artificial privileges for certain parties—and to disregard the details which form their substance: the specific restrictions each law places on the public, and the consequences that result. This simplistic focus on the form encourages an “economistic” approach to all these issues.

                                    But this snippet is completely dismissive. The “meager commonality” is precisely the thing that I find unethical. So it would seem that your FAQ agrees that there does exist a commonality, but seemingly dismisses it as unimportant.

            3. 6

              If the problem with “open source” is that it entails readability but not writeability, per se, why not just switch to a term that makes it clear that both are required?

              1. 1

                The first name that came to mind was “deployable source,” which won’t do because it sounds too close to “deplorable source.” We could go with forkable source, but GitHub beat us to that idea. How about we jump on git’s bandwagon and call in cloneable software. It’s at least also a syscall, if you don’t mind the distinction between sbrk and free.

              2. 6

                I think open-source has won on the pragmatism side. In particular I’m thinking of LLVM vs GCC. GCC’s design was hobbled for ideological/license reasons; as a result academic work used LLVM (which used an open-source, permissive license) instead, which has lead to it overtaking GCC in technical quality.

                1. 5

                  That’s true. I was really distressed when I first saw the old post where RMS specifically declared, long before LLVM existed, that GCC should never support static analysis features because of the ideological concern. I found this recent summary of where that discussion has gone since: https://lwn.net/Articles/629259/

                  I don’t even disagree with the ideological goal that all software might someday be open; I just have advancing the state of the art as a higher priority. I’m offended by the idea of intentionally refusing to move forward because the wrong people might benefit from it.

                2. 2

                  Related to this topic, I can recommend this recent video discussion with rms here: http://lunduke.com/2015/12/16/need-for-compromise-in-free-software-with-richard-stallman/