The cost of using GPL licences is that any changes you make to it everybody gets to have and anything you create with it has to be licensed under that licence. It was designed to create a viral effect so that free software remained free and that people could make money from those who didn't like free.
The MIT licence says "I have tenure in some organisation that is getting its money from something else and which can pay me to spend time on this".
Does GPL ever work this way? In practice, GPL discourages corporate (=widespread) adoption, driving down the quality of open-source software, and creating the market for paid closed-source software (which is often worse than open-source software, but packaged better). I'd love to see cases where a GPL-dev successfully negotiates a reasonable "cost-plus" arrangement with a company for a non-free license.
Your analysis of the MIT license is spot-on, btw.
So at least in my experience, yes dual-licensing can work
On the contrary, GPL tends to be associated with software from the era that predates the new "social coding" phenomenon. The tendency of folks whose first contact with OSS was GitHub and who are most likely to choose the MIT License usually give us brittle, highly niche devops boondoggle that is the byproduct of their responsibilities at their dayjob and that gets abandoned a couple years after being pushed out to GitHub (when the creator realizes the futility of trying to convince other programmers churning out corporate boondoggle that they or their company should pay the creator for their contribution to the mudpie).
isn't that a contradiction? How can it be worse, but packaged better, if the packaging is what people are paying for?
> GPL discourages corporate (=widespread) adoption
and i believe this is because most GPL software doesn't provide enough value over the cost. But for MIT style licenses, there is zero cost, and thus, adoption must be high by the laws of supply/demand!
In other words, the excess value provided by the software under an MIT license is extracted and kept by the corporations using it. GPL licenses forces some sort of non-monetary compensation in the form of contributions, and thus, the corp cannot extract and retain the full value of the software (and hence, they correctly decide to make a cost/benefit analysis, and choose the most profitable decision).
No, this is false.
- You can use a GPL library without making changes and contributing anything upstream
- You can use a GPL library internally and make changes without contributing anything upstream
- You can use a GPL library, make changes and distribute it to 3rd parties and, only in that case, you simply have to share the changes with the 3rd parties. [Not with upstream]
which is fine - your usage of GPL software doesn't affect anyone else. If you decide to charge for it, that's OK too - since if the market exists for such software, the price would equalize to the break-even point of the cost of production.
> You can use a GPL library internally and make changes without contributing anything upstream
If it's "internal", aka, not visible to the outside world, then that's fine too. There's no effect from anyone else's perspective.
> You can use a GPL library, make changes ... share the changes with the 3rd parties
This is the point i was trying to make - in this case, where you make changes, you have to share it. Even tho it's just the 3rd party, this 3rd party has the right to distribute these changes. And anyone that has access to the software is also the 3r party.
So, in other words, if you have visible effects to the outside world with your (changed) GPL software, you are effectively bound to contribute those changes for free, or relicense with the owner to hide those changes.
it works very well for Qt
I have never worked at an organization that does not use GPL software in some way. It discourages shipping GPL components/dependencies in binaries which is the whole point of the license.
Here's another opinion I've been mulling - the emphasis on FOSS is actively damaging to our community and we need to take steps to end it. We're highly trained professionals who often sacraficed a lot to get to where we are. Having this culture of giving away our professional work for free - often to people and organizations who can readily afford to pay for it - borders on masochism.
Instead of constantly singing the praises of open source, we should be encouraging a deep level of self-esteem in younger engineers by encouraging them to "know their worth".
There are of course clear and obvious benefits to FOSS - like no licensing friction and better security - but we have to find a sustainable way to reward the developers. Little donations here and there aren't going to do it either.
i would differ in my opinion, because it's not the 'singing praises' that's the root cause, but that the creator of said FOSS library expecting money in some form for having provided value up front (either contributions or donations etc).
That's an expectation that will never live up to reality.
> we have to find a sustainable way to reward the developers
this exists already - charging for the software you develop. Or, develop a business model around the software - aka, the software is free, but is just the bait for consulting business.
To be specific - I think a good argument could be made for free development in the service of the underprivileged or other altruistic causes. However, a large portion or most development ends up benefitting large public corporations who make large profits. But because, as a culture, we encourage this work, we have basically set ourselves up to be exploited.
I recognize that this is an unpopular opinion and is difficult to make well, in a thread, without being downvoted. But I think there is always value at questioning our underlying cultural assumptions.
All this "it's not real OSS" bullshit needs to end. People who make libraries like this gotta eat.
From tfa:
Most Fortune 500s are using Faker in some capacity. The scope of the Faker project is not small... Who pays for Faker development? No one... Most of these donations are from fellow developers, and not enterprises or corporations.
I don't think there's a single OSS license that'd force corps to pony up. Too restrictive terms and they'll simply go looking elsewhere.
Google does OSS right. Open source to commoditize your competitor's advantages, not to strengthen your own.
I guess the lesson is don't be that guy whose business plan is
1. Write FOSS 2. ... 3. Profit!
There is no such license, and there can't be. Just as a license that prohibits use by the military can't be open source, neither can a license that prohibits use by, or requires payment from, corporations.
Dual licensing under GPL (or AGPL) and a commercial license is the closest you'll get, and it relies on the aversion many corporations have to the GPL, rather than adding restrictions per-se.
Redis and Elastic tried something Frankenstein by mixing Apache and custom license code in the same repo, which spooked them and the community for good measure, with Elastic, later, notoriously doubling-down on "open" with SSPL.
Of all the source-available licenses, I like BSL by MariaDB, better.
"If you can't see the image, Retool's new service is using images hosted from fakercloud.com"
There's a difference between using someone's software and using someone's servers.
But it seems very similar in concept to automated web scraping, which has been deemed legal in the US.
Writing restrictions into the license makes sense in theory. Actually enforcing that license against an uncooperative company will take far more time, money and effort than most people are willing to spend.
Certainly an individual dev could pull in a library with a non permissive license without the broader business being aware. Or, I do agree, there may be particularly corrupt companies that knowingly violate terms of a license... But that's going to be less common than not.
Having worked for big tech, I can assure you they are very strict with licensing.
In this case it's a fast and loose startup, so I agree that it may have happened regardless. But at a certain scale these kind of things will be caught.
But you're not going to get AWS cloning your project if your license doesn't permit it.
That being said, AFAIU violating the GPL of the Linux kernel is very common in the embedded world. A lot of that is made in places largely out of reach of Western IP laws, and by the time anybody gets around to do something, the fly by night company responsible has ceased existing, replaced by a new company currently working on a device five generations ahead, again committing GPL violations.
which means either the market they serve don't care, or don't have enforcement within reach of the western courts. But then, them "stealing" GPL nor not won't really affect the owner of the GPL software, since the world would've been the same to them whether the fly-by-night company violated them, or did not exist in the first place.
We need something like this in spirit but which uses a legal mechanism thats more reliable.
I'm not aware of widely used templates for this kind of license, but definitely needs to become more common.
There should be a non-profit-software Linux distribution that is only for non-leeches. But as it is, the OSS landscape favors corporations.
I'm beginning to wonder how much that software-needs-to-be-free propaganda has been spread in the 90s and early 00s by the beneficiaries (corporations and foundations).
Not so much an issue with them using faker.JS, though honestly companies should do more to pay for development of these kinds of libraries that they depend on, it isn’t the core issue here
Which by the way is a great service if you are in the market and would fully managed fake data api
If faker.js, an open source library, creates links to fakercloud.com - a for-profit offering, then someone who is merely using faker.js isn't falling afoul of the fakercloud.com TOS since they were never even presented with it in the first place.
https://github.com/Marak/faker.js/blob/c9764abd20a198e318367...
But if so, then definitely a problem. They would then have a valid legal case against the company, I assume.
Please nobody hear listen to this guy. I don't want to live in that world.
There are simply those who choose to accept it and those who don't!
If you don't want people to make money off your software, or use it a certain way, don't put it under a license that says it's fine to do so.
Recognizing this fact and operating with that as an assumption is important.
E.g., if the author is obviously asking for sponsorship to maintain the project, as a major user/customer, you should consider a donation. Otherwise, you should be stripped of the benefit of the doubt, regarding the question, whether you are a civil person or not. Likewise, you shouldn't engage in actions, which are prone to doom (economically or otherwise) the project you're profiting from.
https://qbix.com/QBUX/whitepaper.html#Monetizing-Open-Source
Don't use a license that's more permissive than you're willing to tolerate.
It's absurd to me for a project to publish something with a 100% open source license and then complain when somebody tries to profit off of it.
Doesn't really matter how egregious or large the perpetrator is. If you want exceptions, write it into the license. People want the goodwill that comes with open source, without accepting the consequences.
That being said, I do empathize with the plight of the author. Perhaps they were naive to this possibility.