In a more serious and judgement-free style, there's unlicense https://unlicense.org/ But same - I'd like a license for some things that basically says "I don't care and it's not even technically mine anyway", without the ridicule.
There are various edge cases, even when you just don't care. Like for example the problem that you can't release your rights in Australia, so "public domain" is not really a valid option and you need SOME license.
I’ve never really understood the relevance of that last part. When I release something into the public domain, that’s that. No country other than the one I live in can alter my right to do that, and if another country chooses not to recognize public domain released outside their jurisdiction, that’s a problem for them and the people living there to fight over.
It's a contract (not in a legal sense) between you and potential users. Do you want others to be able to use your software, or not? System distributions care about this, corporate environments do it, anyone doing contract work for companies with presence in other countries should care.
Nobody's forcing project authors to release things at all, but if they do then wiring "public domain or Apache" costs them nothing. We live in a society and all that...
No, it deliberately is me saying "this is out there now for anyone to use without any restrictions and I do not claim any ownership on it anymore." Labeling it public domain or X means I am still slightly in the loop, and I do not want that.
If other countries do not wish to accept that, it is strictly their problem. Though even in such cases, someone would need to bring suit for copyright infringement, and I'd be the only one with standing to do so, and I'm not doing it.
I know nothing about this and hence wonder: if someone just writes "public domain" without any qualifiers, what trouble could someone in Australia get into for using that software? (Assume that for some reason the authorities care, which I suppose in practice they might not, for this.)
As I understand it, Australia (along with other countries) does not have the legal concept of voluntary abandonment of copyright. So if the author, or the author’s heirs, decide to sue you, that “public domain” dedication may have no legal substance. To avoid that, CC0 has built-in backups, such as an automatic license that takes effect to give you every possible licensable right.
I like CC0 and you should use it for this (and not worry about the explicit lack of patent grant because that's silly). However, I think practically speaking even "this is public domain" works fine. That's how sqlite is "licensed" and it's shipped in absolutely everything absolutely everywhere including places like Germany or Australia and nobody seems to care.
There are various edge cases, even when you just don't care. Like for example the problem that you can't release your rights in Australia, so "public domain" is not really a valid option and you need SOME license.
Did you mean Australia there, or somewhere else? As an Australian generally fairly knowledgeable about these sorts of things but definitely no lawyer, I understood public domain dedications to be considered acceptable in Australia, and essentially in all places derived from English law; and that it was laws of various European mainland countries that had incompatible ideas. Copyright abandonment or public domain dedication is not explicitly covered in the Copyright Act 1968 (most relevant are section 196 and Part IX), but I understand that to largely be the case in the USA too. In fact, the USA’s situation seems worse because you can back out of it all 35–40 years later!
Don’t Rely on OSI Approval. (Plenty of other good writings on that site too on licensing stuff, including at least one or two others on OSI, which is badly broken.)
But doesn't this license completely sidestep the issue of "license washing"
ie you vibecoded some slop but the code contains copyrighted material which
you have now released under this more permissive license. Who's responsible
for what follows?
I wonder if you could structure a license which that responsibility for accidentally included copyrighted problem becomes the profile of the the end user. That would seem to fit the vibes of the license here!
IANAL, but I don't think think that's possible for a copyright license. A copyright license grants the recipient rights (possibly under conditions) for reception and redistribution. If the licensor accidentally didn't have sufficient rights to the stuff they redistribute in the first place, then surely they also don't have sufficient rights to grant the recipient any rights?
...and another license that tries to be simple (and funny) and completely ignores patent protection.
So anyone malicious can contribute and then sue your users (and you).
Yes, it's unlikely but I decided to make it a pet peeve of mine to point it out since it really seems that people chose licenses with less care then when they go grocery shopping.
So anyone malicious can contribute and then sue your users (and you).
I think that sort of threat is best prevented not via copyright license but by a contributor agreement. You need one anyway if you're taking contributions from random people online and want to protect yourself from liability.
Also, neither of those things actually protect you from patent lawfare. If a third party has, unbeknownst to you, patented a thing that your FOSS software happens to implement before you published it, they can sue you and your users anyway. This is because patents are busted. The thing that is really protecting FOSS at large from such things isn't patent grants (BSD, MIT etc. don't have any and are hugely popular) but patent non-aggression agreements. Essentially, a bunch of big patent holding companies have come together and decide that patents are bullshit for FOSS, so it's mostly stopped being a problem.
This sums up the existential crisis the software craftsmen find themselves facing after 40+ years of caring. Some of our caring has been misplaced and overreached. But as Andrew Kellly surmised recently, much of the sucess of the millions of man hours we've poured into software creations boil down to somone caring.
Regardless of the enforcability or applicability of some thing like this, this 2 word declaration is the stark part that would make this "license" the meaninful attachment to any slop genned artifact.
There is an incredibly important point here: code (and other "creative works") produced by AI are not protected by copyright. (Thaler v. Perlmutter in the US)
This means I don't have to care what your license says: I can use it any way I want no matter what the license says. (Excluding patent, trademark, trade secret, and other laws.)
EDIT: Others have pointed out, that apparently others don't agree with my interpretation of this and related court cases, so perhaps the matter is in dispute. The above IS my own interpretation of it.
I don't know if this is intentional on your part, but in either case, please do not mislead others like this — at least not without making clear that this is your position, as well as if you're a lawyer or not (I'm not a lawyer but I've informally spoken to one about this). Thaler was a somewhat extreme case because he didn't claim to be the author of the work. There's Allen v. Perlmutter which is a bit more interesting and is still pending in the courts. And large bodies of software are likely to have more creative and engineering vision behind them still, as evidenced in things like (human-written) design documents. This is an evolving area of law.
I've also talked to an attorney. The important case is Naruto, which says that copyright law isn't extended to non-human animals; Thaler merely adapts the test in Naruto to say that copyright law isn't extended to non-human bots either. As you say, please don't mislead.
Unless your claim is that given Naruto, there is no credible claim that Allen vs Perlmutter can be decided in favor of Allen, you don’t have any grounds for criticizing @sunshowers’ comment.
That comment made no claim stronger than that Allen vs Perlmutter is not a forgone conclusion.
please do not mislead others like this — at least not without making clear that this is your position, as well as if you're a lawyer or not
I don't really understand this admonition; some people here are quite knowledgeable about the law, especially copyright, despite not being practicing lawyers. I don't think people need to attach disclaimers to every single comment.
TBH you'll probably know who the practicing lawyers are because professional responsibility obligations (at least in the US) would probably require them to disclose that they're not providing legal advice, etc.
I mean, when you read their comment did you actually think they were a lawyer and giving you legal advice?
If not, the lack of disclaimer was fine. I think everyone here knows that most of the people in this forum aren't lawyers so having everyone write IANAL over and over isn't necessary.
Looks like all the slop that inundated our senses, ended up yielding some positive things.
I feel like the fact that it is centered around generative AI, does change the tone fundamentally.
While such licenses used to pass a message of the world being a good deed, something charitable, this license is more in the lines of: "look, I don't care, I didn't even wrote it myself"
simonw | 15 hours ago
I wonder if a version of this could be made OSI-compliant. I'd actually consider using it for a few of my least valued projects.
viraptor | 15 hours ago
In a more serious and judgement-free style, there's unlicense https://unlicense.org/ But same - I'd like a license for some things that basically says "I don't care and it's not even technically mine anyway", without the ridicule.
This one would be classified as an informal license which gnu treats as "yeah you can, but please don't" https://www.gnu.org/licenses/license-list.en.html#informal
There are various edge cases, even when you just don't care. Like for example the problem that you can't release your rights in Australia, so "public domain" is not really a valid option and you need SOME license.
muvlon | 9 hours ago
Pandora | 14 hours ago
I’ve never really understood the relevance of that last part. When I release something into the public domain, that’s that. No country other than the one I live in can alter my right to do that, and if another country chooses not to recognize public domain released outside their jurisdiction, that’s a problem for them and the people living there to fight over.
viraptor | 14 hours ago
It's a contract (not in a legal sense) between you and potential users. Do you want others to be able to use your software, or not? System distributions care about this, corporate environments do it, anyone doing contract work for companies with presence in other countries should care.
Nobody's forcing project authors to release things at all, but if they do then wiring "public domain or Apache" costs them nothing. We live in a society and all that...
Pandora | 5 hours ago
No, it deliberately is me saying "this is out there now for anyone to use without any restrictions and I do not claim any ownership on it anymore." Labeling it public domain or X means I am still slightly in the loop, and I do not want that.
If other countries do not wish to accept that, it is strictly their problem. Though even in such cases, someone would need to bring suit for copyright infringement, and I'd be the only one with standing to do so, and I'm not doing it.
tomsmeding | 10 hours ago
I know nothing about this and hence wonder: if someone just writes "public domain" without any qualifiers, what trouble could someone in Australia get into for using that software? (Assume that for some reason the authorities care, which I suppose in practice they might not, for this.)
wrs | 4 hours ago
As I understand it, Australia (along with other countries) does not have the legal concept of voluntary abandonment of copyright. So if the author, or the author’s heirs, decide to sue you, that “public domain” dedication may have no legal substance. To avoid that, CC0 has built-in backups, such as an automatic license that takes effect to give you every possible licensable right.
muvlon | 3 hours ago
I like CC0 and you should use it for this (and not worry about the explicit lack of patent grant because that's silly). However, I think practically speaking even "this is public domain" works fine. That's how sqlite is "licensed" and it's shipped in absolutely everything absolutely everywhere including places like Germany or Australia and nobody seems to care.
wrs | 2 hours ago
Yes, but don't forget this (from their copyright.html):
jalcine | 11 hours ago
This is what I was thinking too; UNLICENSE seems like the "go wild" option (even more so than MIT)
chrismorgan | 11 hours ago
Did you mean Australia there, or somewhere else? As an Australian generally fairly knowledgeable about these sorts of things but definitely no lawyer, I understood public domain dedications to be considered acceptable in Australia, and essentially in all places derived from English law; and that it was laws of various European mainland countries that had incompatible ideas. Copyright abandonment or public domain dedication is not explicitly covered in the Copyright Act 1968 (most relevant are section 196 and Part IX), but I understand that to largely be the case in the USA too. In fact, the USA’s situation seems worse because you can back out of it all 35–40 years later!
asb | 13 hours ago
I like the MIT-0 license: https://opensource.org/license/mit-0
hoistbypetard | 14 hours ago
You might like the wtfpl.
pmc | 14 hours ago
I don't see why this wouldn't be compliant, but the real hurdle would be getting it OSI-approved.
chrismorgan | 11 hours ago
Don’t Rely on OSI Approval. (Plenty of other good writings on that site too on licensing stuff, including at least one or two others on OSI, which is badly broken.)
romforth | 12 hours ago
But doesn't this license completely sidestep the issue of "license washing" ie you vibecoded some slop but the code contains copyrighted material which you have now released under this more permissive license. Who's responsible for what follows?
simonw | 11 hours ago
I wonder if you could structure a license which that responsibility for accidentally included copyrighted problem becomes the profile of the the end user. That would seem to fit the vibes of the license here!
gspr | 9 hours ago
IANAL, but I don't think think that's possible for a copyright license. A copyright license grants the recipient rights (possibly under conditions) for reception and redistribution. If the licensor accidentally didn't have sufficient rights to the stuff they redistribute in the first place, then surely they also don't have sufficient rights to grant the recipient any rights?
luker | 14 hours ago
...and another license that tries to be simple (and funny) and completely ignores patent protection.
So anyone malicious can contribute and then sue your users (and you).
Yes, it's unlikely but I decided to make it a pet peeve of mine to point it out since it really seems that people chose licenses with less care then when they go grocery shopping.
kwas | 14 hours ago
Yeah, maybe. But "Software Patents" aren't a thing here (within EU, where author also lives).
viraptor | 14 hours ago
That's not entirely the case. It's harder than in the US to get a patent, but it's not impossible. For example the codec patent wars are still very much alive in the EU https://ipfray.com/interdigital-wins-third-munich-injunction-against-disney-court-confirms-bench-ruling-over-hevc-related-patent/
muvlon | 8 hours ago
I think that sort of threat is best prevented not via copyright license but by a contributor agreement. You need one anyway if you're taking contributions from random people online and want to protect yourself from liability.
Also, neither of those things actually protect you from patent lawfare. If a third party has, unbeknownst to you, patented a thing that your FOSS software happens to implement before you published it, they can sue you and your users anyway. This is because patents are busted. The thing that is really protecting FOSS at large from such things isn't patent grants (BSD, MIT etc. don't have any and are hugely popular) but patent non-aggression agreements. Essentially, a bunch of big patent holding companies have come together and decide that patents are bullshit for FOSS, so it's mostly stopped being a problem.
travisgriggs | 10 hours ago
This sums up the existential crisis the software craftsmen find themselves facing after 40+ years of caring. Some of our caring has been misplaced and overreached. But as Andrew Kellly surmised recently, much of the sucess of the millions of man hours we've poured into software creations boil down to somone caring.
Regardless of the enforcability or applicability of some thing like this, this 2 word declaration is the stark part that would make this "license" the meaninful attachment to any slop genned artifact.
saadia | 9 hours ago
Oof. Exactly. It still hurts.
edent | 14 hours ago
We already have this in the CRAPL https://matt.might.net/articles/crapl/
mcherm | 14 hours ago
There is an incredibly important point here: code (and other "creative works") produced by AI are not protected by copyright. (Thaler v. Perlmutter in the US)
This means I don't have to care what your license says: I can use it any way I want no matter what the license says. (Excluding patent, trademark, trade secret, and other laws.)
EDIT: Others have pointed out, that apparently others don't agree with my interpretation of this and related court cases, so perhaps the matter is in dispute. The above IS my own interpretation of it.
sunshowers | 13 hours ago
I don't know if this is intentional on your part, but in either case, please do not mislead others like this — at least not without making clear that this is your position, as well as if you're a lawyer or not (I'm not a lawyer but I've informally spoken to one about this). Thaler was a somewhat extreme case because he didn't claim to be the author of the work. There's Allen v. Perlmutter which is a bit more interesting and is still pending in the courts. And large bodies of software are likely to have more creative and engineering vision behind them still, as evidenced in things like (human-written) design documents. This is an evolving area of law.
Corbin | 12 hours ago
I've also talked to an attorney. The important case is Naruto, which says that copyright law isn't extended to non-human animals; Thaler merely adapts the test in Naruto to say that copyright law isn't extended to non-human bots either. As you say, please don't mislead.
hyperpape | 4 hours ago
Unless your claim is that given Naruto, there is no credible claim that Allen vs Perlmutter can be decided in favor of Allen, you don’t have any grounds for criticizing @sunshowers’ comment.
That comment made no claim stronger than that Allen vs Perlmutter is not a forgone conclusion.
legoktm | 6 hours ago
I don't really understand this admonition; some people here are quite knowledgeable about the law, especially copyright, despite not being practicing lawyers. I don't think people need to attach disclaimers to every single comment.
TBH you'll probably know who the practicing lawyers are because professional responsibility obligations (at least in the US) would probably require them to disclose that they're not providing legal advice, etc.
sunshowers | 5 hours ago
Disagree with this, especially when the statements are definitive (the way that comment was pre-edit).
mcherm: thanks for your edit :)
legoktm | 5 hours ago
I mean, when you read their comment did you actually think they were a lawyer and giving you legal advice?
If not, the lack of disclaimer was fine. I think everyone here knows that most of the people in this forum aren't lawyers so having everyone write IANAL over and over isn't necessary.
simonw | 11 hours ago
I don't think it's a good idea to assume that code written with AI assistance can't be copyrighted.
Especially since the legal departments at almost every large technology company in the USA appear to have come to a different conclusion.
saadia | 9 hours ago
Brilliant. Using this on my next slop proj.
mewse | 9 hours ago
This seems roughly equivalent to 2-clause BSD, right? Warranty clause, no attribution..
pm | 3 hours ago
Looks like all the slop that inundated our senses, ended up yielding some positive things. I feel like the fact that it is centered around generative AI, does change the tone fundamentally. While such licenses used to pass a message of the world being a good deed, something charitable, this license is more in the lines of: "look, I don't care, I didn't even wrote it myself"
ThatsInteresting | 11 hours ago
I laughed, grimaced, and closed the tab.