3 ms·
I don't know, I don't feel I am in some camp. His license grants some permissions and forbids something else. I just don't feel the word "freedom" fits into wha
by greyman 10y ago
I don't know, I don't feel I am in some camp. His license grants some permissions and forbids something else. I just don't feel the word "freedom" fits into what he wants to achieve. Just let's call it some other way and I am fine with it. This is just grabbing a positive word, twist it, and use for his own agenda. And I am not willing to play such a game.
- thristian 10y agoLet's imagine a world exactly like our current world, except that I am allowed to punch anybody in the nose at any time, with no negative consequences for myself. Is that world "more free" than this one? My point is that "freedom" is a messy concept, and the freedom to do one thing (punch somebody in the nose, or to make proprietary software) is often at odds with the freedom to do another (go about your business without being punched, or to control the software that affects your life). That's not RMS twisting words, that's just an awkward philosophical truth. I guess you could say that the term "free software" is biased because it emphasises society-wide freedoms over personal freedoms, but I think it fits nicely alongside more widely held ideas like "only police should be allowed to fight crime" and "politicians should not write laws that directly benefit themselves". Both those ideas significantly limit personal freedom, yet they're very important parts of a free society.
- swhipple 10y agoTo add on to your point, this is called the 'paradox of freedom' -- that you cannot have both unlimited freedom and ensure freedom for the future. At the end of WWII, Karl Popper defended state interventionism with an argument citing the paradox of tolerance [1], and claimed ensuring a tolerant society in the name of tolerance, even when paradoxically intolerant. A parallel argument can be made in Stallman's case for copyleft to be claimed in the name of freedom. It's also worth noting that it is not only Stallman's FSF that uses this definition w.r.t intellectual property nowadays. Lessig's CC adopted the same terminology for their share-alike license (-SA), which has the goal of preserving freedom, while -NC or -ND are considered non-free because they contain restrictions not related to this purpose. The same is true in the greater free-culture community and academic contexts from my experience. [1] http://www.goodreads.com/quotes/25998 http://www.goodreads.com/quotes/25998
- TheOtherHobbes 10y agoThat's a general point, but I think it missed the criticism that the FSF's definition of software freedom fails on its own terms. In practice the "freedom" is exclusive to a highly educated and exclusive technocracy. The same freedoms are not available to the general public. This may seem pedantic, but IMO it undermines the argument about software freedom, because Stallman has never shown the slightest interest in making software that the general public can modify, or in designing, promoting, or doing anything at all to encourage the existence of such software. The idea that if you want to promote software freedom you need to make tools and systems that ordinary people can modify - preferably without learning Emacs - doesn't seem to have occurred to him. It seems that in Stallman's world the public are supposed to become MIT-grade hackers before they can earn their freedom. You could argue that software is hard, and that's just how it is. But I'm not so sure that's true. It's not obvious that the hardness of software isn't a side effect of the culture around it, and not something unavoidable in the process of development. The reality is that there has been almost no serious CS research into making accessible user-modifiable systems. FSF would be more convincing if there had been at least some token effort in that direction. But instead FSF has always promoted a rather old-fashioned and nostalgic view of computing, where everything happens on the command line, and development means hand-editing source code and running a build system to create a local binary. That was more or less the only model around in the 1960s and 70s, but things have moved on since then. Unfortunately the FSF mostly hasn't. Given that, the question is - what does software freedom mean to users now? If source code doesn't actually equal freedom - what does?
- swhipple 10y agoI agree that freedom and practicality are related, but disagree that it undermines the argument. If they were completely unrelated, source code availability would not be an issue; however the legal and practical freedom to do something is often not zero effort or little work. I'm wondering what you have in mind by user-modifiable systems. For example, KDE and Firefox are free and allow for quite a bit of user-customization without any programming knowledge. At some level you would naturally have to write code. I recently came across this interview and thought it was well done. Stallman responded to a similar question: https://youtu.be/NB8mCcLRxlg?t=59m55s https://youtu.be/NB8mCcLRxlg?t=59m55s
- theseoafs 10y agoI mean, obviously because some system gives you more freedoms than others doesn't mean it's the system that best preserves freedom. In the USA we don't have the freedom to enslave people but that doesn't make us less free. For the record I'm not an RMS-ite and I don't actually mean to compare proprietary software with slavery.
- pksadiq 10y ago> His license grants some permissions and forbids something else. I just don't feel the word "freedom" fits into what he wants to achieve. What do you think "freedom" is? We are here because we are not free. And that is called "freedom". :) Freedom may be considered as a set of values that can be exercised on every human. So that is very limited. Say like "I wish not to be harassed in the public." So that means, "I shouldn't harass someone else in the public." If you say you need the second, and not accept the first, you can't call it a "freedom" in a communal sense.