3 ms·

> That definition was largely a post hoc rationalization designed to alleviate the cognitive dissonance between the increasingly libertarian, permissive bent ta

> That definition was largely a post hoc rationalization designed to alleviate the cognitive dissonance between the increasingly libertarian, permissive bent taking over open source ideology via OSI/ESR and the overt prescriptivism of GPL. This is... a view of that debate, and to be fair it's not necessarily unfounded. But let's assume it's 100% accurate. Would you say that debate was a good thing for the FOSS movement as a whole, or was it a painful years-long argument that fractured the community and forced otherwise productive developers to waste hundreds of hours arguing with each other? I don't understand the point-of-view of someone who looks at the GPL, says, "that exception was an arbitrary compromise" and then concludes, "let's add more arbitrary compromises." > I don't apologize for writing copyleft licenses that eschew the compromises of the earlier forms. I'm not asking you to. I just don't see why that has to involve attacking a tribe that is resistant to fundamental changes of this sort, and that is perfectly happy with its current growth and direction. I don't see what any of that has to do with your ability to build or promote new licenses, and I don't see why the narrow nature of the FOSS tribe means that suddenly we shouldn't be free to create our own terms that accurately and narrowly describe what we stand for. Say what you want, brand dilution is a real thing, and you are causing it. And what possible benefit could there be to having this argument? Call your licenses Copyleft++, or Shared Source, or Liberty Licenses, or Social Good Licenses, or Share Alike. You say that you're not trying to co-opt a movement, but I don't understand what benefit you could ever gain from going to war with the existing FOSS community other than to seize on the existing expectations of users who feel like they already understand and can trust what FOSS software is and how it acts in the real world. > When a better mousetrap comes along, the designer of the last model doesn't get to pull the words "mouse trap" out of the newcomer's mouth. No, of course not. But if you build a bear trap and start advertising it as a mouse trap, I'm going to call you out on that. You can say the distinctions that we have are arbitrary, you can say that the community is small, you can say that we're dumb for holding on to them, you can say that most people want a new movement that's more receptive to their specific goals, you can say we're ineffective. Fine. But the distinctions still exist, and they allow us as a community to precisely describe ourselves, and we have used them for years, and you don't have the right to arbitrarily take them away. If you think you can build a better movement, and you obviously think you can, then go build a better movement. I have no objection at all to different communities approaching user freedom from different angles and philosophies. I welcome that, the more the merrier. I would genuinely, sincerely love to see more alternatives to the FOSS movement spring up, because I admit that the FOSS movement can't address everyone's needs on its own. It's especially fantastic to see lawyers getting involved, because there's a huge, real need for more variety of pre-built licenses that devs can safely adopt, and there are a very small number of people who are both advocating for user freedom and equipped to build those licenses. It's really, really good. Just stop saying that it's FOSS.

kemitchell 7y ago
Enjoying this conversation! > I don't understand the point-of-view of someone who looks at the GPL, says, "that exception was an arbitrary compromise" and then concludes, "let's add more arbitrary compromises." I don't think we should carve more ugly exceptions into a bad rule. I think we should have a better rule that better serves the underlying purpose. Specifically: license conditions that require developers using open source to build software to contribute their own work alike ought to be welcomed, no which kind of "use" as we slice and dice that term triggers the obligation. In FSF-speak: copyleft can regulate exercise of freedom 0 and freedom 1 as well as freedom 2 and freedom 3, to the same end, adaptable for both activist and pragmatic business use. > I just don't see why that has to involve attacking a tribe that is resistant to fundamental changes of this sort, and that is perfectly happy with its current growth and direction. Because that resistance to change creates and perpetuates restraints that prevent developers from addressing pressing, practical problems. Embedded library developers have recognized license choices at hand to keep their work free, or to dual license. Service library developers do not. Distributed-application developers have established license choices to keep their work free or to dual license. Developer toolmakers do not. And so on. Intentionally or unintentionally, consciously or unconsciously, the old-guard posture picks winners and losers. No picking losers, then claiming to represent them. > going to war with the existing FOSS community I'm not at war. But there is no singular FOSS community to go to war with. And again, doctrinal specifics---like being able to rattle off the four freedoms from memory---aren't characteristic of any broad set. They're common knowledge only among a particular cohort of developers who came into open practice in particular ways at particular times. I suspect that includes both of us, but not nearly everyone. It took me way too long to internalize that. For example, I published a noncommercial software license a while back and took great care not to present it as "open source". Many of its early adopters did anyway. When I reached out to them to warn of the feedback I thought they should expect, the answer was often "nobody I code with cares about any of that". I'd send them the Open Source Definition or links to fsf.org. No awe. Lots of good questions I couldn't answer. That was my experience with the maximal-copyleft license and OSI, too. After a number of weeks, I had to report back that approval didn't seem likely, but that I didn't feel I could make a strong case why in any but political terms. The developers I thought I was serving by taking the debate for the team responded, almost unanimously: "We don't care. I know open source, and it's open source. Why are you wasting your time on that process in the first place?" > But the distinctions still exist, and they allow us as a community to precisely describe ourselves, and we have used them for years, and you don't have the right to arbitrarily take them away. Sure I do, but not arbitrarily. I believed as you do, that FSF and OSI definitions represent rigorous, legal-flavored standards. Having actually dove into them now, the only precision to take away is false precision. Without linking you to a bunch of long blog posts: The OSD is a fork of DFSG, which OSI holds out as a technical standard, but Debian never did. When the text actually met many eyeballs under OSI, the org ended up making a bunch of hasty changes and commentaries to stem the tide of hackers reading it to exclude GPL. Twenty years later, I still saw comments on OSI's mailing list to the effect that the nondiscrimination criteria prohibit copyleft---more confounding than helpful. In the intervening time, OSI approved a number of controversial and just plain bad licenses. They approved stronger-than-GPL copyleft terms like OSL and RPL. If the organizations' work in today's climate is to preserve a winners' history of open licensing circa 2000-2010, they should build that monument. Monuments don't move. Movements have to. If "the movement" was just a tribe, you're right. It belongs to its time and place ... long past. But if "the movement" was really more than that, a set of goals beyond mere market penetration and a very niche kind of fame, the mantle belongs to those advancing those goals. Developers unhappy with where software freedom has run aground don't deserve to be marker for scorn or shorn of their "open source" identity. If that happens, they won't bend over backwards to honor open source as predecessor. They'll turn on it. New tribe.

kemitchell 7y ago

Enjoying this conversation! > I don't understand the point-of-view of someone who looks at the GPL, says, "that exception was an arbitrary compromise" and then concludes, "let's add more arbitrary compromises." I don't think we should carve more ugly exceptions into a bad rule. I think we should have a better rule that better serves the underlying purpose. Specifically: license conditions that require developers using open source to build software to contribute their own work alike ought to be welcomed, no which kind of "use" as we slice and dice that term triggers the obligation. In FSF-speak: copyleft can regulate exercise of freedom 0 and freedom 1 as well as freedom 2 and freedom 3, to the same end, adaptable for both activist and pragmatic business use. > I just don't see why that has to involve attacking a tribe that is resistant to fundamental changes of this sort, and that is perfectly happy with its current growth and direction. Because that resistance to change creates and perpetuates restraints that prevent developers from addressing pressing, practical problems. Embedded library developers have recognized license choices at hand to keep their work free, or to dual license. Service library developers do not. Distributed-application developers have established license choices to keep their work free or to dual license. Developer toolmakers do not. And so on. Intentionally or unintentionally, consciously or unconsciously, the old-guard posture picks winners and losers. No picking losers, then claiming to represent them. > going to war with the existing FOSS community I'm not at war. But there is no singular FOSS community to go to war with. And again, doctrinal specifics---like being able to rattle off the four freedoms from memory---aren't characteristic of any broad set. They're common knowledge only among a particular cohort of developers who came into open practice in particular ways at particular times. I suspect that includes both of us, but not nearly everyone. It took me way too long to internalize that. For example, I published a noncommercial software license a while back and took great care not to present it as "open source". Many of its early adopters did anyway. When I reached out to them to warn of the feedback I thought they should expect, the answer was often "nobody I code with cares about any of that". I'd send them the Open Source Definition or links to fsf.org. No awe. Lots of good questions I couldn't answer. That was my experience with the maximal-copyleft license and OSI, too. After a number of weeks, I had to report back that approval didn't seem likely, but that I didn't feel I could make a strong case why in any but political terms. The developers I thought I was serving by taking the debate for the team responded, almost unanimously: "We don't care. I know open source, and it's open source. Why are you wasting your time on that process in the first place?" > But the distinctions still exist, and they allow us as a community to precisely describe ourselves, and we have used them for years, and you don't have the right to arbitrarily take them away. Sure I do, but not arbitrarily. I believed as you do, that FSF and OSI definitions represent rigorous, legal-flavored standards. Having actually dove into them now, the only precision to take away is false precision. Without linking you to a bunch of long blog posts: The OSD is a fork of DFSG, which OSI holds out as a technical standard, but Debian never did. When the text actually met many eyeballs under OSI, the org ended up making a bunch of hasty changes and commentaries to stem the tide of hackers reading it to exclude GPL. Twenty years later, I still saw comments on OSI's mailing list to the effect that the nondiscrimination criteria prohibit copyleft---more confounding than helpful. In the intervening time, OSI approved a number of controversial and just plain bad licenses. They approved stronger-than-GPL copyleft terms like OSL and RPL. If the organizations' work in today's climate is to preserve a winners' history of open licensing circa 2000-2010, they should build that monument. Monuments don't move. Movements have to. If "the movement" was just a tribe, you're right. It belongs to its time and place ... long past. But if "the movement" was really more than that, a set of goals beyond mere market penetration and a very niche kind of fame, the mantle belongs to those advancing those goals. Developers unhappy with where software freedom has run aground don't deserve to be marker for scorn or shorn of their "open source" identity. If that happens, they won't bend over backwards to honor open source as predecessor. They'll turn on it. New tribe.