I'm not sure I can be any more explicit than my previous comment, but I'll try.
You can't just grab a permissive license, like MIT, ignore it, and go, "Hey, this is GPL now."
That's entirely incorrect. You cannot ever ignore the terms of a license, just because you want to use another license. You must under all circumstances, honour the terms of the license for which you were granted the right to reproduce someone-else's work.
MIT is compatible with GPL. In the sense that it's possible to honour both the terms of the MIT and the GPL simultaneously. You must however, still honour the MIT license in its entirety i.e. Display the MIT license notice in its entirety; it's right there in the license text.
Any new project isn't GPL licensed, it's GPL and MIT licensed. People need to understand this. If I grab someone's code off Github that's supposedly GPL, but it includes MIT code; I can't just chose to honour the GPL and ignore the MIT. That's copyright infringement, plain and simple.
> But anyway, they did get consent from all contributors: there's only three of them: "I have spoken to Sebastian and Daniel about this and we all want Lerna to do the same."
Github lists 169 contributors, not three. We're talking about contributors, not maintainers i.e. We're interested in the authors of the copyrighted works, not some open source nomenclature that has no legal bearing.
> You can't just grab a permissive license, like MIT, ignore it, and go, "Hey, this is GPL now."
Yes you can, that's sublicensing, which the MIT license allows. It's also why you can include MIT-licensed code in non-free applications. The MIT license does not forbid any of this. And you're not ignoring it when you say the whole work is now covered by the terms of the GPL as the text of the GPL says; you're actually exercising a right granted by the MIT license. I can totally grab MIT-licensed code and redistribute it to you under non-MIT terms[1]. I don't have any obligation to give you any freedom nor do I have an obligation to make sure you can impose GPL-incompatible restrictions on it. I can sublicense as I wish.
You're right that you still have to display the MIT license but that's not really the point. It's a very minor restriction that is not relevant to the point of this discussion. By licensing under the MIT license, the original contributors to Lerna agreed to let their code have additional restrictions imposed upon it such as denying licensing to certain companies. Lerna is still displaying the original MIT license anyway.
--
[1] Although in this day and age of widespread internet distribution, you could just go and grab the original from the original distributor, which may be enough for you if I haven't modified it significantly.
What you've just described is actually referred to as relicensing i.e. removing the old license, and replacing it, this is not the same as sublicensing. Without expressly having been given permission to do so, this is copyright infringement.
Sublicensing is when you, as a licensee, in accordance with your license, provide a license to another party. The MIT permits this as long as you adhere to the MIT license's conditions, namely:
[...] 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.
The point is that based on this, as laid out in the MIT license, all sublicensees are bound by, at best, the exact same conditions that you're bound by. The MIT license allows you to place additional restrictions, but this is inherently a "conjunctive and" situation i.e. The software is MIT and GPL licensed, meaning the sublicensee is bound by the conditions of both licenses.
> You're right that you still have to display the MIT license
You've acknowledged my point here.
Not only do "they" need to display the MIT license, so do sub-licensees, and their sub-licensees, and so forth, in aeternum.
> but that's not really the point.
That's entirely the point.
As I've said before, this is being bound by the terms of the MIT license and the GPL, not simply the GPL.
With respect to Lerna, we have a near parallel situation. Lerna can take MIT licensed software from their contributors, and sublicense a derivative of this software with an attached MIT+additional terms license. However, Lerna must continue to to adhere to the conditions of the MIT license i.e. Display the MIT license in full. As in the GPL situation, the software is now bound by multiple licenses, not simply MIT+additional terms, but rather MIT and MIT+additional terms.
> Lerna is still displaying the original MIT license anyway.
I will concede this is a fair point, in the sense that their new license incorporates most the text of the original MIT license. Given the surrounding context whereby additional restrictions are added, and the fact that they're collectively represented as the one license, I don't know how that would play out in court.
They've actually changed the copyright notice, thus they're not meeting the terms of the MIT license that was granted by their contributors.
Granted, that's being super pedantic and I doubt that alone would be reasonable basis for any sensible court awarding restitution.
Nonetheless, my entire point is that the Lerna team have claimed in that Github issue that they've relicensed the project under MIT+additional terms. They can't do this, they can only sublicense contributions of contributors - the MIT does not allow relicensing.
They can absolutely relicense their own contributions; as the copyright holder they inherently have that right to do so. However, with respect to contributor's works, they can only sublicense, as contributors have expressly permitted this, as long as sublicensees also adhere to the conditions set forth in the MIT license. So the contributor's portions, and therefore Lerna as a whole, are at best MIT and MIT+additional terms, not simply the latter.
If you're suggesting that:
MIT *and* MIT+additional terms == MIT+additional terms
Ignoring the aforementioned copyright notice change and concerns about contextual representation of the MIT license, then I'd say that's a fairly reasonable assertion. Thus, sublicensing contributions as MIT+additional terms would be essentially equivalent to relicensing as MIT+additional terms.
Nonetheless, as the MIT license makes no mention of relicensing, I would suggest that proclaiming you've relicensed is not particularly wise; which is precisely what the Lerna team are proclaiming (see their last comment before locking the issue). They have in fact sublicensed.
You can't just grab a permissive license, like MIT, ignore it, and go, "Hey, this is GPL now."
That's entirely incorrect. You cannot ever ignore the terms of a license, just because you want to use another license. You must under all circumstances, honour the terms of the license for which you were granted the right to reproduce someone-else's work.
MIT is compatible with GPL. In the sense that it's possible to honour both the terms of the MIT and the GPL simultaneously. You must however, still honour the MIT license in its entirety i.e. Display the MIT license notice in its entirety; it's right there in the license text.
Any new project isn't GPL licensed, it's GPL and MIT licensed. People need to understand this. If I grab someone's code off Github that's supposedly GPL, but it includes MIT code; I can't just chose to honour the GPL and ignore the MIT. That's copyright infringement, plain and simple.
> But anyway, they did get consent from all contributors: there's only three of them: "I have spoken to Sebastian and Daniel about this and we all want Lerna to do the same."
Github lists 169 contributors, not three. We're talking about contributors, not maintainers i.e. We're interested in the authors of the copyrighted works, not some open source nomenclature that has no legal bearing.