> Yes you can, that's sublicensing.No, that's copyright infringement.
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.
Nonetheless, if we're being pedantic:
- Copyright (c) 2015-2017 Sebastian McKenzie
+ Copyright (c) 2015-present Lerna Contributors
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.