Everything is a compromise. GPL versus MIT is a compromise either way. Both are more free than the other. So don't tell me there's such a thing as fully free and that nothing else matters.
> I'm certainly for fundamental freedoms like the freedom of the press, even if I'm not myself a journalist.
Freedom of the press has more restrictions. In particular violating an NDA in your press activities is somewhat similar in how it affects few people and is time-limited.
GPL and MIT fully fulfill the exigences of the free software definition, there's no compromise here.
Choosing between GPL and MIT may involve compromises, but we are talking about something else, you are moving the goalposts here.
(note: I kept updating my previous comment, probably while you were writing this, sorry for this)
(edit: ah, you noticed)
> Freedom of the press has more restrictions. In particular violating an NDA in your press activities is somewhat similar in how it affects few people and is time-limited.
Violating an NDA is a matter of law and contracts that can equally apply to software, it's not at the level of the fundamental right or at the level of the license.
If you are speaking about protecting one's source, then it's orthogonal (and probably also an important mechanism for) freedom of the press.
MIT gives less freedom to the user. GPL gives less freedom to the dev. I'm not moving goalposts, those are both very important and you can't maximize both.
> it's not at the level of the fundamental right or at the level of the license.
I would say the ability to run a competing business within two years is less fundamental than the ability to talk about what you saw at your former job.
Also what are you saying about "not at the level"? It overrides the fundamental right, but people accept that.
No. Someone can use some MIT licensed software for any purpose and can strictly do anything with it that the GPL would allow.
Just as free.
However, I do agree that permissive licenses like the MIT allows someone to build proprietary software with the code, and the users won't enjoy software freedom with that result. That would be the reason why I wouldn't license my software under permissive license except for specific cases.
One could argue the MIT actually gives more freedom to the user: they can, with a MIT source code, build and distribute a proprietary derivative. Of course, this doesn't mean this additional freedom is desirable.
> Also what are you saying about "not at the level"? It overrides the fundamental right, but people accept that.
Yeah, just like laws limit your freedom to run free software for any purpose. You, for instance, can't legally take /bin/ls with you to go kill someone even if its license itself doesn't explicitly forbid it. I'm pretty fine with this.
I have some feeling we do agree that freedom is never absolute and kinda argue past each others.
> the users won't enjoy software freedom with that result
That's the part I was talking about.
So I don't know why you started your post with "No."
> I'm pretty fine with this.
Cool. And I'm fine with some other restrictions. So we're agreeing that being fine with a restriction doesn't necessarily mean anyone has "lost the direction" of free software? And instead we can talk about the specific dangers of a restriction without reflexive rejection?
>> the users won't enjoy software freedom with that result
> That's the part I was talking about.
> So I don't know why you started your post with "No."
You are mixing two different users. FreeBSD user A enjoys software freedom. macOS user B doesn't. A and B are different people. And notably, the release of macOS didn't make FreeBSD disappear, so A still fully enjoys their freedom even though B doesn't with the downstream software (but could with the upstream software as well).
If you are mixing user A and user B, your mental model of this stuff is not clear and reasoning with it is going to be difficult and frustrating. And it makes you reach incorrect conclusions such as "permissive licenses are less free for the users than copyleft licenses".
> So we're agreeing that being fine with a restriction doesn't necessarily mean anyone has "lost the direction" of free software?
Laws are supposed to protect individual and collective freedom and interests. They do this by putting limits on what you could do that would prevent others from enjoying their own freedom.
(of course, I know we are far from this ideal; also laws aren't perfect and issues should be fought against - it's actually, in practice, a constant fight because different (groups of) people have different interests and priorities - but that's another concern and mixing concerns makes things hard to reason about)
License restrictions that compromise on software freedom are simply not comparable to whatever it takes to make a society work. They are not making a tradeoff for your own good as a user. They are benefiting the software provider at the cost of the user's freedom.
I'm not mixing up users A and B. Let me try phrasing it better. MIT does not give software freedom to all users of the code. A huge number of users get excluded in practice, far more users than this clause in the liquibase license.
> Laws are supposed to protect individual and collective freedom and interests. They do this by putting limits on what you could do that would prevent others from enjoying their own freedom.
I'm not sure where you're going with this when my example was NDAs. NDAs aren't there to protect freedoms, they only remove freedom. Laws enforce them, but in the same way that laws enforce license restrictions.
> Okay, I would agree more with this phrasing. That's the issue I have with permissive licenses.
So do you also see what I meant about both permissive and copyleft licenses giving up some freedoms? You can't have it all.
A restraint on running a particular type of business isn't great for software freedom, but if it's narrow and temporary I don't think it ruins the entire license. The gap in freedom is avoidable but also it's much smaller than the gaps you can't avoid.
For ongoing freedom, If I was choosing between plain MIT and a GPL clone that bans competiting companies but reverts to pure GPL after a couple years, I would pick the latter.
> So do you also see what I meant about both permissive and copyleft licenses giving up some freedoms? You can't have it all.
Not a fan of the framing but I get your point (of view).
> For ongoing freedom, If I was choosing between plain MIT and a GPL clone that bans competiting companies but reverts to pure GPL after a couple years, I would pick the latter.
Reasonable. I would find the 2 years delay unbearable and pick MIT, but I wouldn't enjoy doing so. It's fortunate I don't have to pick :-)
> I'm certainly for fundamental freedoms like the freedom of the press, even if I'm not myself a journalist.
Freedom of the press has more restrictions. In particular violating an NDA in your press activities is somewhat similar in how it affects few people and is time-limited.