Copyright vs. Copyleft (2007)
gnu.org
gnu.org
I was wondering how FOSS teams could get their fare share of revenue from cloud providers.
It just seems reasonable that AWS would pay royalties to the elasticsearch, postgres, rabbitmq teams. For examples.
Am noob, just started foraging for details. I didn't quickly find anything, so started reading about how the music industry handles royalties, because that's copyright. Stuff like sampling, covers, airplay.
Next, I'll probably read about royalties for patents.
Thoughts?
https://www.mongodb.com/licensing/server-side-public-license
I've heard of this in a couple of other places, I'm not aware of a common approach just yet. And I'm not sure how it's played out.
No. The goal of copyleft is to ensure that the users of software are free to do what they want with it. Royalties would encumber that freedom.
The only freedom not allowed is giving the software to someone else without granting them the same freedoms.
As I said, I'm pondering fairness. Something practical, a la Hintjens (zeromq). Like maybe dual licensing.
Edit: Hintjens has some advice about using trademarks to maintain control. I'll go back and check if his strategy talks about getting paid.
IMHO it's working ok. It definitely could be better. I think GPLv3 or AGPLv3 with a dual license is currently the way to go. That offers the free software folks the freedom and increases awareness of the code, and the commercial license can be negotiated on a case by case basis. The challenge is to get projects to require CLA before they grow and lose the control needed for dual licensing.
I did reread some of the seminal documents, refresh my memory. Free vs open, libre vs gratis.
I now lean towards dual and maybe three-fold licensing. Free for non commercial use. Open for others. Something something for cloud service providers.
I can't yet guess what that something something will look like. My hunch is it'll be analogous to music royalties and publishers rights.
Mostly, I want some boilerplate deals and heuristics, so that people can frame their arguments, instead of reinventing the wheel each round.
There's e.g. the polyform project https://polyformproject.org/ but that's explicitly source-available proprietary software, not open source.
Timescale DB have a license [0] that is basically open source with the one added restriction that you can't use their open source to offer a hosted SAAS version of their product. Basically to allow them to open source the software for those that want to host it for their own use, but preventing the big cloud providers from taking all their business. It's not really 'copyleft' though because it takes away a users freedom (albeit a very small and limited one).
The GNU Affero GPL [1], adds the requirement that using the open source software in a server environment you must also provide the ability for your users to download the source code. It doesn't address the problems through royalties because that's kind of against the GNU philosophy, but it does stop the cloud providers from just taking the software and modifying it and using for cloud infrastructure without providing the modifications back to the community.
I'm sure there are other examples of this kind of this.
[0] - https://blog.timescale.com/blog/building-open-source-busines...
Hopefully companies go back to it. Hopefully I can have a positive impact and maybe that will cause others to try it.
I have 2 toasters at my house, one pulls 600W while on, the other 1400W. Both toast 2 slices of bread and are otherwise the same. If someone accidentally installed the 1400W elements in the 600W toaster something would melt and possibly start a fire. I’m all for promoting the repair of consumer devices but toasters are one device that I don’t think the general public has the knowledge and skill to repair safely.
That line of thinking has been removed from global consciousnesses because it was less profitable.
Code released under such a license would not qualify as open source, but it could allow for individuals to maintain the software themselves. A license like the BSL could even be structured to transition the code to open source after some amount of time.
Of course, copyleft licenses such as the GPL _can_ be used to protect the authors' rights to compensation for commercial use of their work - but the framing matters. I think that's the fundamental reason open source licenses such as BSD and MIT have caught on - they centre on the need of the developer to get productive work done, not on the far more abstract principle of the home PC user being able to recompile their own software stack from source.
MIT and the like, while being good open source licences, allow anyone to break the chain and not release their code to the public. So it is good in terms of being open source and free/libre now, but does do anything for the future, and does not do enough to bring about the vision of a fully free world.
Personally, I'd prefer the GPL licenses unless there's an good reason not to.
The idea that there is some wonderful "community" on GitHub and all this code just flows back into a beautiful, better place simply isn't how programming works. No one would do that.
Therefore GPL helps both users and devs.
What is the range of opinion regarding the MIT license? From what I’ve seen, people are either fine with it, or mildly disgruntled. Exceptions are rare. And of course, corporations love having access to software libraries under this license.
How about the GPL? Individual developers seem to be either ok with it, or enthusiastically in support. That’s cool—I’m sure we can all relate to those feelings. But a lot of, ahem, for-profit institutions simply will not get involved in GPL-licensed software. We’ve seen repeatedly the effects of their willingness to walk away, to invest in alternatives, etc.
It seems obvious, if depressing, how things will move over time.
I think this is painting the picture in a too one-sided way. Here is another way to see it: While for-profits might develop their own alternatives, they will fail to make those GPL licensed in most cases, because that is what they wanted to avoid: Giving back to the society, that allows them to exist. So they will have proprietary code. The next company will do the same and so on, all while slowly the GPL licensed alternative creeps up on them, until finally it has caught up. It gets ever more expensive to not use GPL software, as free software slowly gains ground.
I contributed hundreds of hours of work into libraries. They are used by the same companies that sell me locked-down devices that I cannot modify or even inspect. They even invade the my privacy and I cannot disable that.
At least GPL tries to mitigate this problem. MIT turns volunteering into unpaid labor for some billionaires.
Calling it "far more abstract principle" is quite absurd.
These companies try hard to push the narrative that MIT is "practical" and GPL is "political" - guess why.
Um... no? The idea behind MIT/ISC and the like is to let other people use your software, which they aren't automatically allowed to do because copyright is broken by default.
There's a lot of software that I would like to use in my day job that I can't touch because it's GPL. This is by design.
> These companies try hard to push the narrative that MIT is "practical" and GPL is "political" - guess why.
I have no sympathy for FAANG, but that happens to be correct. GPL is explicitly political in its intent.
God forbid people use a license that makes it easier to turn a profit so they don't have to beg to stay afloat.
The “rolling release” model, where the latest version's proprietary but six months ago's version is libre, simply cannot be done with something like MIT unless you're willing to risk somebody else taking it all for themselves.
Can you name an IC that profits from their software with the GPL?
By contrast I know of a very well-known IC that profits (handsomely) from their software with an even looser license than the MIT - SQLite.
This is exactly what happened even when they used AGPL, something that was thought to scare cloud vendors, but instead it wasn't scary enough, so these companies had to invent new proprietary licenses.
Plausible used to be MIT, and exactly this happened to them.
But my day job's fine with me using that software. Why? It's not using the law to restrict people's ability to use their computers when that software is installed.
Good: the license is working as intended.
> I have no sympathy for FAANG, but that happens to be correct. GPL is explicitly political in its intent.
Don't be cheeky. The sentence was [ push the narrative that MIT is "practical" and GPL is "political" ]
The MIT is equally political - just siding with the freeloaders.
It's not different from soldiers taking weapons from the enemy they are fighting. A pretty common tactic.
Fighting fire with fire. Where's the hypocrisy?
What does the endgame look like for an abolitionist?
However, I'm under no illusion that it'll prevent my unpaid work from being used by billionaires.
Setting aside those people/companies that are willing to violate the AGPL, even the ones that respect it are free to use it as long as they redistribute the source to modifications they make.
Anyone (including billionaires) could do that, and not every business model relies on keeping their source code secret.
There's plenty of evidence that AGPL is reasonably effective at that. SaaS providers are clearly afraid of it.
> not every business model relies on keeping their source code secret.
I was never talking about secrecy and business model.
It's about allowing end users (including myself) to have some reasonable chances to inspect / update / reflash their devices.
The GPL has a different usage definition: run and execute. E.g. think of an image manipulation program under the GPL. If you design a game, you run the program to create assets, but none of the created assets contain a trace of the original program any more. The game you ship doesn't either. The GPL doesn't affect you. However, if you tried copying one of the algorithms implemented in the program into your own proprietary competitor, you infringe on the copyright now.
Anyway, I certainly think the MIT license is political - I also believe in the political position of free education for all people, even if they're billionaires (or children of billionaires), and in the political position that I should be taxed to provide these services for people richer than me. I donate to charities that help people who do not repay me. And so forth.
If some billionaires want to profit off of unpaid labor, don't think you can stop it by using the GPL, which relies on the legal system and copyright laws that made them billionaires in the first place. If you sue them for infringing the GPL, you might have a great strategy to win their game, but you're still playing their game in their courts. There's a saying from a totally different context that applies here too, I think: the master's tools will never dismantle the master's house. We've had the GPL for decades now and the result is there's a lot of billionaires whose fortunes were made via GPL'd software like Linux.
If you want to get rid of billionaires, or of freeloading billionaires, or of locked-down devices, or of invasions of privacy, take action against that thing. Do not fall into the trap of "Something must be done, and the GPL is something."
That is reducing what GPL is about quite a lot.
For example it is equally important, if not more, that modifications in most cases have to be contributed back to the community, which means getting companies to take part in the action, instead of simply using work of others, profitting from it and never contributing anything back.
This is wrong. There is no requirement to contribute back to the community, just the requirement that changes (or rather, the source with your changes) be made available to your users. This means they could mail CDs of their code to their paying customers only, if they so choose. A far cry from "contributing back to the community", closer to something like "a pain in the ass."
All you have to do to comply is make your own source code available on request to users who accept your license, along with installation instructions to build the running product. That's not contributing anything to the other dev's community.
I am not so sure about what you say about releasing it to only paying customers. That might be true. In that case one of the original developers could buy 1 license or the community could pay for 1 license and this way get the source code. Then the developers of the modification would get some pay and the original developers would get the source code.
The GPL only requires you to make the source code accessible to those people who also get the binary from you, not everyone, but most entities just allow downloads for anybody. It's more expensive to implement checking than it's to just allow downloads for everyone. xyzzy_plugh is splitting hairs here.
"Contributing back to the community" sorta implies you're opening a PR, following coding guidelines, writing tests even perhaps. What happens in practice is corps just hack everything together and rush stuff out the door -- no communities want those contributions. It's rare that these sort of releases contain anything novel or useful. Take a look at the OSS releases for Amazon's products (Echo, FireTV) and be prepared to be underwhelmed.
The theoretical possibility of charging money for distributing GPL software has not materialized for any company. You can get RHEL for free if you want, for example, and build sell your own devices with it.
If the world stabilized on a few software products that did what people wanted to do, like and office suite, that would be great. No more work needs to be done on it. Just a little maintenance going forward. Like roads. They cost a lot to build, but little to allow someone to drive on them.
GPL is not necessarily for developers to get productive work done. It is for people to ensure maximum benevolence with their work for everyone and take the highest moral position. I believe what needs market penetration is this ideal, rather than the GPL itself. GPL is for people who already hold this kind of ideal. They are looking for exactly this.
I don't say you shouldn't ever make money from your work or never use any copyright/MIT/BSD licenses. Most people are not filthy rich and need the money. What I am saying is GPL is not for this, and does not have the requirement to be convenient or framed in a different way than the heroic selfless endeavor it is, in any sense.
Try getting some productive work done without the source. Like fixing a printer of a proprietary driver, for instance. No, the GPL is about exactly about getting practical work done. That's why it emphasizes so much on the 4 essential freedoms, and why 3 of those 4 freedoms are essentially the same thing "the freedom to tinker with the source code and share that information with other developers who can tinker with it some more and do the same".
This seems a naive take when considering that it's Corporations who hold the most Intellectual Property[1] (Patents and Copyrights) and are the 'customers' or beneficiaries of Intellectual Property laws and systems.
It's Corporations who focus on IP acquisition (and thus monopolization) strategies (e.g. Google buying smaller startups to add their Intellectual Property to it's repertoire).
...in other words, the likelyhood of average Joe being helped by the Intellectual Property system to get access to the fruits of his labor is not a common situation at all. Most new innovations are the result of access and engagement with existing Corporate Intellectual Property (plus an expensive education).
"5% royalties due on profits derived from this product."
So if something is given away free, nothing would be due. But if something were used for profit, part of that profit would have to be paid upstream. And that would be profit for that node which, if used the same license, would have 5% of that trickle upstream.
This could be an automatic process and handled through an open registry. Anyone could take anything, register, and then pay their "taxes". If you wanted to donate anything you made, make that an option also.
Of course there are an infinite amount of "what if" scenarios, but that's the gist of it based on the following philosophy:
No one should need permission to copy anything or profit from anything. But they should share their earnings if they "opt in" to capitalism.
If that counts as a victory, however, it was a pyrrhic one. In the process of gaining mainstream popularity, the social movement of “free software”—which rejected the very idea of treating software as intellectual property—morphed into the more palatable notion of “open source” as a development methodology, in which free and proprietary software could happily co-exist. The corporations that latched onto the movement discovered a useful technique for developing software, but jettisoned the critique of property rights that formed its ideological foundation.
Yet it was precisely the weakness of that foundation that made the free software movement vulnerable to co-optation in the first place. The movement’s greatest limitation was its political naivete. Even as it attacked the idea of software as property, it failed to connect its message to a wider analysis that acknowledged the role of property rights within a capitalist framework. Free software pioneers like Stallman tended to approach the issue from an individualized perspective, drawn from the 1970s-1980s hacker culture that many of them came from: if you could change how enough hackers wrote and used software, you could change the world. This highly personalized model of social change proposed an individual solution to a structural problem, which necessarily neglected the wider social context."
+
"the neoliberal consensus of the last few decades has meant that the benefits of technological development have largely flowed to corporations, under the aegis of a strong intellectual property regime. As the free software movement came up against these prevailing economic forces, its more contentious aspects were watered down or discarded. The result was “open source”: a more collaborative method of writing software that bore few traces of its subversive origins."
— Wendy Liu, https://logicmag.io/failure/freedom-isnt-free/
Civilization has multiple centers on Earth and always has, not simply the US East Coast. The GPL has carried on for three+ decades, and look what has exploded in that time. What comes next? A legal framework must engage its epoch, and times change, of course. Massive forces are at work right now, with acceleration.
Free as in Freedom
The same could be said about your analysis. She, at the least, engages the reader to think about Capitalist Property Rights. It's also a small snippet of a larger text/essay, so I would ask you to please consider reading the whole thing before you make up your mind about her positions and overall vision.
> Civilization has multiple centers on Earth and always has, not simply the US East Coast. The GPL has carried on for three+ decades, and look what has exploded in that time. What comes next? A legal framework must engage its epoch, and times change, of course. Massive forces are at work right now, with acceleration.
> Free as in Freedom
I honestly have no idea what you are talking about here, and you provide no useful rebuttal to the points she brings up.
The difference is:
The old system produces social status for daily use by ordinary people. You work at a day job and your parents, your partner, your children, your peers all know about that.
The hacker culture doesn't have a method to do that. It can only produce fame, which is only usable for a few individuals.
The old system needs adjustments. More and more people see that property alone nowadays makes a suboptimal signal of status.
There are now more people who care about these things and more quality Free software available than ever before in human history! It's just that the complacent open-source idea, as well as proprietary software (biggest offenders are locked smartphones, etc), have spread even faster :(
I don't think the movement has weak foundations nor that it is politically naive. In my eyes it goes directly to the problem it cares about, which is human freedom in the context of using a computer. It doesn't oppose commerce, but it is aware of the current socio-political order's (i.e. capitalism's) tendency to turn anything into a sellable commodity.
Even if the second developer isn't evil, he may be willing to release his mods to the first person's code, but there can be several technical and business reasons that he cannot release his build process to his customers.
BTW, nothing in GPL requires that the second guy communicate with the first one in any way. There is no "community contribution." It only says he has to make a copy of his source available to his customers on request. That's a pretty huge delta from the idea of folks sitting around in a community and improving software.
To actually use the second guy's code, you'd have to diff it from the first! There is no requirement that the code have any comments or documentation. It cannot be intentionally obfuscated, but it can be obtuse by design. Again, none of this implies any community of people working to improve some theoretical software.
And that's because the second person is likely building very different software from the first! The first person and his community probably don't need or want most of the second guy's code. And if there's a community of devs for the second guy's product, well that's kind of his problem.
The real burdens that GPL places on the second dev are pretty bad problems. Releasing the build process and the surrounding code which is not related to the first guy's product is a huge problem. You don't have to be evil to see that.
There are also many, many cases where the second person is making something that improves the first product but doesn't compete with it. It's in a different technical niche or customer application and in that way, they're not related. (Same software in a TV and a fridge, for example.) Does the TV company care about the fridge people's mods? No, not really. Unless there are some bugs or real improvements in the TV software that the fridge people could make. Since this does happen if the TV code base gets big enough, eventually you see paid software companies like Google and Red Hat who go ahead and fix errors and make improvements in the free software they use. But this only happens when there is a certain critical mass of users or an overlap between the use cases of the base code.
Nothing in the fridge people's proprietary software would be of any help to the TV users! There is no great "GitHub in the sky" where all the mods to the TV code for a fridge or some other thing come together to make the "great, perfect TV software!" That just isn't how software development works.
The idea is that first guy is irrelevant, your user is the person that has the right to demand the code so he can change the software. If you want to sell me some software without giving me the sources then don't use GPL. If the GPL code is some trivial small thing that is not related with your product then find a BSD version or pay for a proprietary version of that library (and have fun not having the freedom to check it's code and fix/improve it)
Let's say you've written a library and published it under GPL and it gets used in a cellphone or an automobile. The "community" of users for that paid product is not the same as the users for your dev library. In fact, most commercial software innards don't have a "community" at all. I think that's a complete fiction. What they have is customers who pay.
When the automaker changes that code, do you really want it back? Are you really going to publish their automobile version or cellphone version of your library back into your "community?" If its a simple bug fix, you might, but again this big company isn't republishing your software just to fix your bugs. They've modified it for their application. Do you want or need their mods? Are their devs in the same "community" as yours? I don't think so.
GPL tries to reach into the surrounding and supporting code that the commercial entity writes and somehow force that company to "contribute" that code back to a non-existent community of people. What right does the first developer have over any code that the second dev wrote? That idea that the first author could reach into my code or my product and have anything to say about it is antithetical to the values of open source software right from the start.
Non-permissive licenses use the tools of the enemy to try to do something that sounds good on the surface but in practice is immoral and largely unenforceable. I would argue that, far from encouraging liberal software use and development, restrictive FOSS licenses like GPL and its cousins prevent a great deal of software from ever being commercialized and also prevent great products from reaching users because no one bothers to commercialize them.
If you love our industry, please help your fellow devs by giving them the maximum ability to use your work in their own work so that they can create businesses, pay the bills, hire other devs, and so on and so forth.
Well, maybe that "commercial entity" could have picked a different library to begin with, and your rant has no bearing on the GPL at all?
> When the automaker changes that code, do you really want it back?
Well, if you buy that car, yes?! Because that's the very stipulation you set forth when you published the library.
> If you love our industry,
What a load of bollocks.
Unfortunately, improvements above the kernel stack will keep closed, since the Android userland don't adopt GPL as much.
edit: for example, OpenWRT was created through a GPL enforcement lawsuit! Router companies weren't complying with GPL and were closing their kernel (which had in-house drivers). The companies were sued and had to open their code. You would think that their customers weren't interested in this, but actually.. many use openwrt to this day (arguably, there's even an environmental angle to this - openwrt increases the longevity of the router by having it do more stuff with the same hardware, reducing e-waste)
The license was never about "the community". In this case it's about the owner of the car having the ability to modify it.
Most people dont want to modify the software they use, but those that use these licenses want the people who use their code to have that option.
>> If you love our industry, please help your fellow devs by giving them the maximum ability to use your work in their own work so that they can create businesses, pay the bills, hire other devs, and so on and so forth.
That's quite literally asking to use someone else's work for your own profit. Someone else asked about royalties and I think I see room for a license that would prevent such exploitation without compensation. That would be a commercial license.
The point of Copyleft (and the Free software movement in general) is to ensure that the user of a computer (and the software in it), a human being, is free -- meaning she can do with it what she pleases. Use if for whatever, study what it does, change it, share the original or modified version -- those are the four freedoms https://www.gnu.org/philosophy/free-sw.en.html . Humans can't be free, if they're not in control of the computers/software that they use.
The Free software community considers it unethical, if a cellphone owner or an automobile owner can't do with her device (including the software it runs) what she wants, even though she bought it with her hard earned money.