This post was originally posted in GitHub/@vrc/vrc#792#issuecomment-657129783.
Because GitHub/@krainboltgreene decided to flag it as out of topic, I figured it's better to publish it somewhere else, in the hope that it can be useful to others.
Because of this issue, even though it was corrected (but krainboltgreene still won't admit there was ever a problem), I discurage anybody from ever using VRC as there might be other problems, given the absolutely lack of understanding shown by those who changed the license.
Issue #804 and pull request #805 are absolutely crucial.
The issue was fixed, but I will still write a comment on this (both are blocked). The reason is that I think it can be important for a potential reader and can fix some misunderstandings.
First thing first, let's summarize what happened and why it's important.
Computer code is a form of literature not any less than poetry and, thus, it's protected by copyright law. Some rights are therefore exclusive to the original authors and cannot be used by others without a license from the author.
A license is not a set of restrictions. A license is a permission. In order for you to be able to modify, copy or distribute code of which somebody else is the author, such author has to give you a license to do so and the license has to be valid. If, for any reason, the license doesn't apply to you, then you have no such permission.
Now, if you accept contributions, from others, to your code, then you are not the author of the contributions: you actually need a license from contributors to use the contributions. Unless otherwise specified, it's usually assumed the license of the project applies to the contributions as well and it's a license from the contributor.
The contributions this project received where under MIT license which allows you to do pretty much anything, but only applies if:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
This is a condition: nothing sais you have to do what the condition sais. However, the license only applies to you if and only if you meet the condition. If you do not, then it's as if you were never provided any license to begin with.
This project removed the text of the MIT license. Therefore, the license from contributors no longer applied to the authors of the project (at least not for this GitHub repository). By continuing to distribute copies of the project and to modify the project, they were essentially doing so without the permission from some of the copyright holders. This is not much different than illegally distributing digital copies of a book, for instance. You may only distribute them with permission from the author.
This is important to understand: the issue was not in removing the text of the license, per se. The issue was distributing, modifying and making copies of derived works of the contributions without having any permission to do so.
After merging #805, the codebase again contains a copy of the license ("this permission") and the copyright notices. Thus, because the MIT license is usually seen as not terminating, the permission from the authors of the contributions to the authors of this project applies again.
Of course, it still doesn't justify what happened previously.
I will respond to some points raised in #804:
As far as I know, we haven't removed the original license text. It's still visible at that point in the history.
I guess this could be used as a laughable defense. There are two problems, one moral and one practical. The moral problem is that many are unlikely to check the past states of the repository. For this reason, the current state should always reflect all attribution requirements. How would you like it if I forked your staff and then removed any trace of the license and, while I am at there, any copyright notice and your name from the project (the MIT license treats the notices and the text of the license the same way and they are just as important)? After all, they are still present in old commits! The practical problem is that very often only the current version of the project will be distributed. You have absolutely no reason to download the whole git history if you are just using the project, nor to distribute it. Also, notice that the MIT license still absolutely applies to the current version of the project! It's about the old contributions and it's given from the authors of such contributions to anybody. The new license also applies but it doesn't revoke or change the MIT license which is given for the contribution (if it did, no license would apply for the contributions and the whole software would be unusable).
The real important question is: Can someone sue me for creating a new distribution (to ruby gems) with to a more restrictive license?
The answer to this is absolutely yes. Somebody can sue you. Somebody can sue for absolutely anything. The "real" question is: would it be legal for you to create a distribution with a more restrictive license? (Thus, would they be successful if they sued you for doing so?). The answer to that is no. If it was "yes" copyleft licenses would be useless. Read the license: you have a permission from the authors to make derived works. In order to use such derived works somebody needs BOTH a license from the original author AND a license from you: they do not need to be the same license. Nobody requires you to provide any license at all for derived works: if you don't, no license applies, but the MIT license will always apply for anybody receiving a copy of the original work (including any copy contained in your distribution).
However, changing the license is not the issue here. Indeed, you didn't change the license. You are simply providing a DIFFERENT and ADDITIONAL license to make use of some of your own rights to the software. The point is, if you are not the only copyright holder (and you aren't if you accept contributions), you then need a valid license (aka permission) from all other copyright holders to distribute derived works (such as modified versions of the program, including its current version). You do have such license (the MIT license), but it's only valid and given to you if you keep its text intact in every single copy of the work to which the license applies. If you don't do so, then no license is given to you and by continuing to distribute copies of the program you are in violation of the rights of other copyright holders to the project.
According to the MIT license allowed to sublicense, which in my opinion covers this case (based on the legal definition of sublicense).
You actually do not even need to sublicense in order to be able to apply a more restrictive license to your own contributions to the work. Sublicensing simply means that you provide the rights you received to others. However, what you cannot do and what is not sublicensing is providing yourself with more rights than the license from other copyright holders give you. In this case, the permission (including the right to sublicense!) only ever applies to you if and as long as you provide every copy of the software with the full text of the MIT license ("this permission") from start to bottom (the condition is absolutely part of the permission). If you don't do so, then to those contributions absolutely no right is licensed to you, not even the right to sublicense. In additon, sublicensing is a tricky issue (that's why other licenses do not allow it or explicitly prevent it, despite not being copyleft and thus allowing different licenses on separate contributions). As for the MIT license, however, it explicitely sais it's provided to anybody who receives a copy of the Software (which can be assumed to be whatever work the license applies to). Notice that this only applies to those rights licensed under the MIT license. If any author used a different license (including Hippocratic) it's cool. But as a user I need BOTH licenses (and thus to meet the conditions of both licenses) to be able to use the program. That's because I need a permission from BOTH copyright holders, or else I am breaching the rights of one of them by making copies of the software without having received any permission (or without the permission being given to me).
Contributors own the changes they wrote in their distribution, no doubt.
This is absolutely true. Notice they are not the only copyright holders of the contributions, though, but that's pedantic.
Is the source in the repository that someone submitted a pull request for and was eventually copied with their consent now mine?
Absolutely not. Some of the rights are yours. Unless everybody transferred every right to you (which is not the same as licensing) you are just one of the many copyright holders. In order to use the project I, you and anybody needs a license from EACH and EVERY copyright holder. It could be the same license from all, it could be not. And I only receive the MIT license (from those contributors who are providing it) to the contributions if I keep the text of the MIT license. If I don't they don't provide any license to me. Thus even if I follow the Hippocratic license under which most of the project is licensed I am still breaching the rights of some copyright holders.
Is the source I copy from github into rubygems mine?
Absolutely not, at least not exclusively yours.
Does this count as a sublicense?
It doesn't have to be and there is no reason to introduce this complexity. Even if it did, you can only sublicense rights that you have because they have been licensed to you. And the rights of the contributors have only been licensed to you if you put the full text of the MIT license in each and every copy of the software, including this one on GitHub.
but I do think this is a valid case where the answer to 3 is "yes".
If the answer is yes, then you breached the rights of others, because the permission to sublicense was ONLY given to you if you did follow the condition.
Read the mit license:
Permission is hereby granted, free of charge, to any person obtaining a copy of this software […] to deal in the Software without restriction, including without limitation the rights to […] sublicense […] subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
Thus if you do not follow the condition then absolutely no permission is given to you, not even the right to sublicense absolutely anything. It's as if the MIT license had no text for you, it's as if it was empty. And the condition you have to meet is to include the text of the MIT license in every copy of the software. Either this or the permission isn't ever given to you.
Will anyone take me to court to assert their rights?
Who cares? You should respect the rights of others even if they don't bring you to court. Most likely nobody will, because it just isn't worth the time. But one doesn't just do anything they can do without going to court, right? We should follow the law and respect the rights of others regardless of whether we go to court or not.
@krainboltgreene I have seen another ruby gem change license and they had attempted to contact all contributors to the project to get their approval: prawnpdf/prawn#625
That is fine and I have seen it done successfully a few times. If you get a license from each and every single copyright holder you can do anything (well, within the limits of the law).
It might be hard to know if all of those contributor's contributions are still in master but it's an easier way to cover yourself.
It may not even matter which one are in master: if a modified version of a contribution is in master, even modified to the point that no original line is present, the copyright of the original owner may still apply.
Every time you commit to Github, you're creating a distribution
That's a very good way to see it.
Even if we don't follow this interpretation the MIT license should still be included in the latest version of the repo and in any commit to which it applies, please see above.
In addition, the license talks about "every copy". So every copy of the software must have the license, regardless if it is within a Git history or not.
Anyway, @j00bar absolutely managed it! Your comment is very well written and very well explains the problem!
The idea that no contact was done or that github is a means of contact. There's no need to discuss this here nor is there a need to discuss how or who to contact.
Why are you talking about contracts? Contracts are a more complex issue. I shall remind you that if the MIT license doesn't apply to you, then you received NO LICENSE and NO PERMISSION.
This is actually a huge misconception about software licenses. The misconception is that they restrict you. However, the opposite is true: licenses are permissions: they give you more rights you would otherwise have.
The MIT license doesn't require you to do anything. It simply gives you an extra permission, which copyright law doens't give you, if and only if you meet a certain condition. You, receiver of the license, have every interest in defending the license and in saying "yes, the license is absolutely 100% valid and it applies to me". If you manage to prove the license isn't valid then, congrats: it no longer applies to you and you received no permission, so you cannot use the contributions to do absolutely anything regardless of what you do.
No contributors have come into the discussion yet (as far as I can tell), I'd love to see those that have contributed weigh in.
They do not need to: they already gave you a license.
They do not need to forbid you from distributing copies of their contributions: the law does so. They gave you an EXTRA permission to do so if you meet a certain condition. If you don't, that's fine: the license doesn't apply to you. So no permission was given, so you musn't distribute or copy or modify anything unless you are the sole owner.
The point is: unless all, each and every contributor joins the discussion and says "yes, I give you an extra license which is different from the MIT and allows you to do blah blah blah", then the only license you received is the MIT license which has a condition.
Frankly, while a lot of what you've said makes sense (except the contributor licensing to me)
As I said, you have every interest in claiming that contributors DID license to you. If they didn't, then you received no permission from them and you are not allowed to distribute anything. To distribute contributions you need an explicit permission (a license!) from all copyright holders to do so. Luckly for you contributors did license to you and you have every reason to defend this: if not, the license doesn't apply to you and default copyright law does, which is maximally restrictive.
if it was true the open source community would have a lot of questions to answer and things to do should someone ever sue. Are there any judgements you can reference to back up these claims?
It is true! Copyright law makes no exception for open source. You can see this in action in the GNU project: contributors explictely TRANSFER their copyright to the FSF: if they didn't, they would keep hold of copyright to the contributions.
Now, it's true: the open source community has questions to answer. What is also true is that most of the time it answers them very well.
Are there any judgements you can reference to back up these claims?
Again, you are the one who would have to prove, in court, that contributions were indeed licensed to you. Contributions are works, protected by copyright law like any other work and to merge them you absolutely do need them to be licensed to you. So if they weren't licensed to you than you would absolutely be breaching copyright law.
This is what everybody was saying: the moment you removed the text of the MIT license, the contributions were not licensed to you. Thus, default copyright applied to them and you were not licensed any right to do anything.
If you want to use a contribution you are the one who has to defend the claim that they are licensed to you, not anybody else.
You've not actually answered the question, but repeated what I already said: The code they wrote is theirs, but by merging I effectively wrote it to the repository. They may be identical source, but who owns the one in vcr?
Copyright law essentially protects creative works. So whoever authored a piece of software, a painting or anything else holds copyright to it for their whole life, X many years after and to some extent actually forever.
You making a copy of some bytes didn't make you the author.
Anybody using such copies still needs a license from the original copyright holder. And that includes you!
That doesn't make sense. If someone make a change where it shifts every line down by one and I merge it in, do they now own all the code and have simply licensed it to me?
Code has multiple owners! If more people worked together to a certain piece of software, all of them are copyright holders.
Shifting the code down by one line, by the way, is completely irrelevant for the purposes of copyright law and such a contribution would not cause the contributor to become a copyright holder.
If you add something relevant, however, you are the copyright holder to that thing you added (not to anything else, not even if it was shifted by one line. Copyright law couldn't care less about bytes, Git). If Bob modifies the contribution, to any extent, the final version of the contribution is owned both by Bob and the original contributor. Using the software requires a license from each and every copyright holder to all parts of the code.
It's not that each line of code has an owner, it doesn't make sense. Every function, class, modules is influenced by the rights of the copyright holders. Who was the person that committed bytes to git history, what the number line is and similar stuff are irrelevant for the purposes of the law.
This phrase in particular is interesting. Do you have any legal research that suggests commit history is hiding?
This is a realy worrying way of reasoning.
First of, you are the one who has to claim that you got a license, so do absolutely anything you can in order to be able to claim you meet its conditions.
In addition, it's plain immoral: you should still follow the license even if you believe you found a way to effectively not to and say "well actually" in court. The MIT license won't be found by most if it's only in Git history.
However, the MIT license says the text has to be included in each copy of the software. Keeping it somewere in Git history does not reasonably meet this requirement and it's not how the MIT license is interpreted by the vast majority of the open source community.
Try to keep the discussion clear of suggesting the vcr team is doing something unethical. We released a new version with a license change.
No, ethical considerations are relevant when you are using the works of others, including contributions to your own software.
However, most of the discussion, and this comment, is about copyright law.
It's not that what vcr has done is only unhethical. It does breach copyright law. In addition, it also happens to be highly unhethical and unjustifiable plagiarism.
What we're discussing here is if that act is "compliant" (for whatever definition) with the MIT license.
Yes and no. There is no such thing as chaning the license to something you do not own (the contributions). You didn't change the license. You simply applied a different license to your own stuff, which is absolutely allowed by the MIT license and nobody has tried to claim otherwise.
The problem is that you removed the original text of the license and thus it no longer applies to you and thus you have no license from contributors and thus default copyright applies (to the contribution) and it's owned by them, not you.
Worst case scenario I(?) get sued, which is not a big deal for me.
The worst case scenario is plagiarizing things without permission. Since the permission is only given if and only if the condition is met, this is what happened.
My non-lawyer understanding of copyright on this matter is that if a user contributes a substantive piece to the combined work (a contribution of a blank line isn't substantive and therefore isn't copyrightable - typo fixes, and trivial fixes are likewise non-substantive), then any combined work which continues to include their substantive piece is using that substantive piece under license - in your case, under the MIT license. According to the plain text of the license, you must include the license text with the software so long as it includes their substantive contributions. You don't own them - they are licensed to you.
This is incredibly well put.
Dear @j00bar, I admire your ability to explain things shortly, clearly better than me.
Will you get sued? I don't know - probably not. But is that the real litmus test? You're violating your contributors copyrights - that's wrong and completely against the ethos of the free/open-source community.
This is why we are also talking about ethical issues.
Here is the thing. I believe that you should follow the law regardless of getting sued. I see, expecially from the US, a mentality of doing anything you can get away with regardless of the law as long as you don't actually end up in court.
This is a terrible way to think, it goes in the face of honesty, of the foundation of law and of everything democracy is.
In this case it's even worse however because an ethical issue applies regardless of the law. It's one of the lucky cases in which morality and the law say the exact same thing.
Merging contributions from other people does not erase their copyright. Their copyright persists. You might have an extra copyright based on the merging itself (beyond what you yourself contributed -- think of the editor of an anthology).
Not only this. There is absolutely nothing you can do to erase somebody else's copyright, except convincing them to release it in the public domain or being transferred all rights (and even that might not work depending on your jurisdiction).
Copyright law isn't specific to GitHub. It applies to books, paintings and a lot more.
I'm asserting that there's a longstanding community practice around open source to the effect that merely having a copy of a license like the MIT license in the commit history is not good enough to meet the MIT license's requirements.
And for a good reason!
- Every copy of the software needs to include the MIT license, copies in old commits won't. So you have to see the whole history as one copy, which is not trivial.
- Usually only the latest commit is distributed by others. So even if you believe you comply with the license, you are still tricking others into breaking the license.
- If we consider every commit and release to be a distribution, then this is the only reasonable interpretation of the license.
- Regardless this is the most reasonable reading of the license, just as it's reasonable to think the license must be in a format that one can read and in a place that one can find, or else it's as if it wasn't included.
This is, in my view, the only reasonable interpretation of the license and even if it was not (which it most likely is. It's so predominant you wouldn't be wrong in saying it's the only interpretation used in practice), this is key:
there's a lot of code in there that is under a license with obligations they have to comply with that is not immediately visible.
Receivers of the work will not now they need to comply with the condition of the MIT license but they do.
And even if you ignored this crucial point (which will cause others not to follow the license) and point 1, 2, 3 and 4, and even if you ignored the fact that this is the de facto standard interpretation of the MIT license, even if you claimed that your interpretation, which doesn't meet the intention of the MIT license is better than the interpretation everybody else uses and even if you could claim the MIT license still applies to you (you need from contributors) and you meet the condition, it would still be highly unhethical. Even if you could claim all this, you'd still be breaching the intention of the license and distributing contributions effectively against the will of their authors (or at least without their effective permission). Please notice that the law is to be interpreted in a reasonable way, and so is the MIT license and its intention is anything but irrelevant. So from a legal prospective keeping the license only in Git history is absolutely not cool and does not meet the requirement. In fact, it would only meet the requirement according to an ad hoc interpretation of the license. But the interpretation of the whole community, for which there are good arguments I believe I expressed, has a lot more value than an ad hoc interpretation which appears to be given specifically so we can believe this particular repository complies with the license.
I'm not a contributor, but many developers at my company use vcr. The requirements of the MIT license from past contributors apply to anyone using vcr, not just the maintainers, so it's not theoretical.
This is crucial as well. Even without this point the MIT text would have to be included though.
So users and downstream distributors would also be violating the contributors' copyright? 😕
Essentially yes, but they might have no legal issue expecially considering they had no way of knowing this (they could have checked the whole Git history but there was nothing to imply they were supposed to). Their most likely behaviour would be to assume the Hippocratic license comes from all authors and that's what they would do so the violation would be accidental. Your violation was not accidental though.
Those past contributors aren't likely to care,
Or they might care a lot. I'd be pretty pissed if I was kind enough to provide a contribution under a license permissive like the MIT license and somehow somebody managed to breach my rights anyways. A lot of open source contributors care about this stuff!
One example. Think about when StackOverflow started claiming that past contributions were under CC BY 4.0 when they were actually under CC BY 3.0 (and some 2.0). They thought they could get away with it, but it created a huge controversy. Now they correctly tag each and every post with their respective license. Despite CC BY 3.0 and 4.0 being very similar in practice this was still very wrong. The Hippocratic license is nothing like the MIT license so it's even worse.
vcr currently looks as though all those past contributors must have agreed to a wholesale relicense from MIT to Hippocratic For an attorney that seems like an extreme stretch to make. I mean in this very thread, the only discussion so far about the issue brought up, zero people assume I did any due diligence (regardless of if true or even needed).
I'd like to mention how @richardfontana didn't even mention that he was an attorney. And rightfully so in my view.
The law can be read by everybody, not just attorneys (who are the experts, of course) and while some laws are very complex and so is copyright law, in practice the basics are well understood.
Anyhow, this is not a stretch at all. You did make an extreme stretch by suggesting having a notice somewhere in Git history complies with the MIT license. That's a very ad hoc way of reading the only one clause of the MIT license.
Anywhow Fontana is absolutely right on this and the reason is trivial: you are not the only copyright holder thus you need a license from each and every copyright holder.
I think that after showing your misunderstanding of copyright you should have tried to learn and addressing Fontana with such statement was inappropriate and disrespectful. It's not that an attorney is always right, but in this case he is and you should have spent more time to actually investigate the claim.
The MIT license permits anyone to take MIT-licensed code and even use it in a proprietary project without open-sourcing anything.
It does! That's the beauty of permissive licenses.
It only allows you to do anything if you meet the condition, though. If not it doesn't apply to you (you may still be able to do some things of course, but not anything for which you need a permission).
A good point has been made about removing the MIT license itself from a project is a no-no, but what about leaving the MIT license there and adding a Hippocratic clause?
This would absolutely meet the requirement of the MIT license and it's how every proprietary project (or even GPL-licensed for this matter) does.
Secondly, MIT permits sublicensing.
It does but sublicensing is not even needed.
All you have to do is to provide the Hippocratic license to new code: even if the user gets the rights of the MIT license to old portions (which they do) this isn't a problem: they still ought to comply with the Hippocratic license. Other licenses do no allow sublicensing but are permissive just like MIT is.
Anyone know what needs to be done for that to be triggered, such as renaming a project, or creating a "new" repo that simply imports and exports VCR, where that new "wrapper" repo is licensed using the Hippocratic License?
You can use whatever license you want in new projects as long as you meet the requirement of the license. Regardless of whether you decide to call it "sublicensing" or not.
Can a new version of a project not sublicense the old version, for example?
New code does not need to come with any license at all, as long as it includes the text of the MIT license in a reasonable way (of course making it clear it only applies to old code).
Seems pretty clear to me. I think it's an honest mistake,
It's a mistake that comes from a misunderstanding of copyright but I wouldn't say it was honest.
First of, it wasn't fixed immeditely despite the points raised. But in addition it was still unhetical because it removed that which is a simple attribution requirement. So even if there was no intention of breaching copyright law (which I am not certain about, given the following questions about "going to court" instead of about how to actually comply with the law even if one could get away with it), expecially considering the ad hoc interpretations of the license and the intention of not actually complying with its intention I would say that it was way too immoral to consider it as "honest".
There's an easy, clear solution: add the MIT notice back in, cut a new release. Nice.
@richardfontana When you re/double/sublicense/create a derivative work
Those are 3 different things.
When you use a work from 2 copyright holders you need a license from BOTH. They may be different licenses: if so, this is not dual licensing (what I assume you ment by "double licensing"). Dual licensing is when you get two licenses from the same owner: in this case you can chose which one to follow and you only need one license to apply to you (you then ignore the requirements of the other).
What happened here is that a derived work was created.
otherwise RG will probably do it for you because they don't want to distribute material that violates IP law. lol
LOL! Rights of other people are so funny! It's almost like people have rights!!
There's an easy, clear solution: add the MIT notice back in, cut a new release. That's a solution to a problem that does not exist.
So is it an honest mistake or willful misunderstanding? You were explained the problem!
Further, I remain unconvinced that the change we made isn't compliant with the MIT license requirements.
Did you read the only one requirement? There is only one!
That you include the text of the MIT licene ("this permission" is the MIT license) and the notice in every copy of the work. It doesn't say every new code must be under the same license, of course. But it does say every copy of the work must still contain a full copy, from start to bottom, of "the permission". Either the repository as update does or it doesn't. And it doesn't. Notice that whatever the license sais before the condition (aka no matter what rights it gives you) is completely irrelevant! Those rights are given to you if and only if you meet the requirement. So if you don't you can't sublicense or do absolutely anything else.
I firmly stand my ground after this discussion so far that this is a sublicense and therefore no consent is required from contributors.
That's not what the law says though, nor what the MIT license says. Again, to be able to sublicense you need to follow the condition of the license. It only allows you to sublicense if you still include the text of the old license.
You can only sublicense rights you were given and you were never given the right to remove the text of the MIT license and no, it's not what "sublicensing" means.
It was a bunch of github comments. Some of those users are banned, their accounts no longer linked to that comment, with no authorship or authority bestowed.
Being banned from GitHub is irrelevant for copyright law. If you cannot reach them and get extra permissions, you do not have extra permissions.
Which brings me to my final point: Even if we conceded that this was worthy of reaching out for consent, determining who to reach out to becomes a nightmare that frankly isn't worth anyone's time:
This is redicolous.
Just because you don't want to reach everyone for consent it doesn't mean you can breach rigths of others without consent. What it means is that you do not have any consent, it being hard for you to get it does not equal having it.
Luckily I think we're well within our rights to do what we did and I'm willing to deal with any challenge to this.
You are absolutely not within your rights to do what you did and you were clearly explained why multiple times.
Now, for the correction:
"Some portions of vcr are covered by the following copyright and license notice:" is no more correct in my mind than not doing this.
Then you do not understand the issue.
The MIT license says you have to include the notice and the permission. It DOES NOT say that the permission has to apply to every copy of the program. It does not say that additional contributions must be covered by the same permission. It only states you have to CTRL+C and CTRL+V the text of the license in every copy of the program which is what the PR does.
I agree that it would be better to update the year and ownership to the appropriate people and timelines.
It would not. You should keep that copyright notice because it only refers to contributions covered by the MIT license. You can have an additional copyright notice for new contributions.
I think the reason this pull request wasn't blocked is likely because you forgot to, so I don't know how likely my comment is to get removed. I think it's better if it doesn't get removed though as a reminder for future "ethical source" developers not to make the same mistakes and also to remove some misunderstanding about such a simple license as the MIT license.
As a final note, those who do not care to follow a license as simple as the Expat license should not expect others to comply with their own rights and the requirements of more complex licenses such as ethical ones, right?
The funny thing is that @CoralineAda got this right:
https://ethicalsource.dev/blog/adopting-an-ethical-license/